Money Trends / Liquidity

Special Report

Highlights China and Brazil are two extremes in regard to investment and savings - the former saves and invests a lot, the latter very little. The key difference between Brazil and China is neither the existing amount of deposits nor their propensity to save. Rather, it is their real economies' capacity to produce goods and services. Regardless of how capital expenditures are financed, when inputs for capital spending are procured domestically it is recorded as national "savings," but when they are imported there is no change in the level of national "savings." In China, policymakers are currently being forced to walk a very thin line between inflation and deflation. Brazilian consumers do not need to save more for companies to get financing for their investments. Instead, businesses - along with facilitation from the government - should build the supply side. Banks can finance the latter by originating loans "out of thin air." However, the natural consequence of this adjustment in Brazil will be considerable currency deprecation. Feature The Fallacy This is the fifth report in our series on money, credit, savings and investment. Its objective is to show that financing of investments is not constrained by national and foreign savings. This report argues against a postulate in mainstream economic literature which holds that in order to invest, nations with low savings rates need to either reduce consumption and boost national savings or to borrow foreign savings. Some examples of this economic thesis can be seen here: As Lindner neatly summarizes: "Many economists hold the position that "saving finances investment." They argue that saving - a reduction of consumption relative to income - is necessary for the provision of loans and the financing of investment." (Lindner 2015).1 Linder also provides other examples suggesting that this thesis is well entrenched in the economic theory and analysis. For example, he cites Gregory Mankiw's influential introductory macroeconomics textbook that upholds: "Saving is the supply of loans - individuals lend their saving to investors, or they deposit their saving in a bank that makes the loan for them. [. . . ] At the equilibrium interest rate, saving equals investment, and the supply of loans equals the demand." (1997, p. 63) (Lindner 2015).2 This mainstream economic thesis - that financing is constrained by savings - is intuitive, and not surprisingly many investors take it for granted. Yet this is a false proposition. This thesis is correct for barter economies but is not pertinent to modern economies with their own banking systems and national currencies. Further, Lindner (2015)3 argues: "The fallacies loanable funds theory commits might be explainable by the mis-application of some ideas and concepts of neoclassical growth models - especially the Ramsey (1928), Solow (1956) and Diamond (1965) models - to the sphere of money and finance... The Ramsey and Solow models are models of real investment only. Financial markets, financial assets and financial saving do not play any role in those models. There is only one good which, for simplicity, will be called "corn". Corn has three functions: it can be consumed, invested and used as a means of payment since wages and interest payments are made with it..." Clearly, modern economies with their fiat money systems are much more complicated than a barter economies with no banks and money. The Veracity: Financing Is Different From "Savings" This and previous reports4 clarify and elaborate on the following aspects of banking, money creation and financing as well as savings and investments: 1. Attributing the lack of investment in many emerging market (EM) economies to their low savings is a major fallacy. Borio (2015)5 argues: "Crucially, the provision of financing does not require someone to abstain from consuming. It is purely a financial transaction and hence distinct from saving... The equality of saving and investment is an accounting identity that always holds ex post and reveals nothing about financing patterns. In ex post terms, being simply the outcome of expenditures, saving does not represent a constraint on how much agents are able to spend ex ante. If we step back from comparative statics and consider the underlying dynamics, it is only once expenditures take place that income and investment, and hence saving, are generated." 2. Banks do not need deposits or "savings" to lend. They create money/deposits when they originate loans or buy assets from non-banks. To settle payments with their peers as well as the central bank, they require reserves at the central bank. Reserves at the central bank - not client deposits - constitute true liquidity for banks. For a more detailed discussion on loan origination and money creation in absence of new deposits entering into the banking system, please refer to Appendix 1 and 2 on pages 14 and 18. Certainly, there are several factors such as regulations and shareholder preferences that can curtail banks' ability to expand their balance sheets. However, households' or nations' "savings" do not constrain banks' ability to originate new loans/create deposits. 3. In an economy where banks exist, "savings" and financing are very different things. Many investors use the term "savings" to refer to bank deposits. Yet, in macroeconomics, national and household "savings" are not related to deposits or money in the banking system at all. Chart I-1 demonstrates that there is no relationship between the savings and changes in the amount of money in the banking system. Chart I-1Savings And New Money ##br##Creation Do Not Correlate The confusion between national "savings" and financing creation is dealt with nicely again by Fabian Lindner. Having modelled it, Lindner argues: "... the aggregate economy's saving is equal to the newly produced tangible assets and inventories. That total saving is equal to just the increase in tangible assets ... (because) all changes in net financial assets in the economy add up to zero... Thus, for every economic agent increasing her net financial assets, there is a corresponding decrease in net financial assets of all other economic agents in the economy (Lindner 2015).6 Put in more general terms: An economic agent can only save financially if other agents dis-save financially by the same amount... That is why in the entire economy (that is the world economy or a closed economy) only the increase in tangible assets, thus investment, is saving...." In another paper, Lindner asserts: "Investment is the production of any non-financial asset in an economy and thus is always directly and unambiguously savings: it increases the economy's net worth... The economy as a whole cannot change its net financial wealth since it always equals zero. The aggregate economy can only save in the form of non-financial assets...The only way an economy can save is by increasing its non-financial wealth, i.e., its physical capital stock" (Lindner 2012).7 On the whole, deposits are a monetary concept; they represent money savings. Deposits are created by banks "out of thin air," as illustrated in Appendix 1 on page 14. Meanwhile, "savings" are a net addition to capital stock. Not surprisingly, there is no relationship between "real savings" and money savings, as illustrated in Chart I-1. In a nutshell, "savings" is an addition to the capital stock of a nation, which is the same as investment. Hence, the Savings = Investment identity for a closed economy is nothing other than a tautology as it de-facto means Investment = Investment. That is why in this report we use "savings" in quotations whenever we refer to it in the traditional sense of economic theory. 4. Households' (or businesses') propensity to save alters the velocity of money, not the amount of deposits/money in the banking system. A decision by a household to spend more rather than save does not change the amount of deposits in the banking system and does not affect the banking system's ability to provide more financing. When households or companies decide to spend their deposits, the velocity of money rises. Conversely, when households and companies decide to save (retain) their deposits, the velocity of money drops. The amount of deposits in the banking system stays constant. In turn, the amount of deposits and hence broad money supply in any banking system equals the cumulative net money creation by banks and the central bank over the course of their history. This has nothing to do with household and national "savings," which form the country's capital stock. 5. In a country with its own national currency, the true macro constraint on commercial banks' ability to expand financing infinitely are inflation and currency depreciation - not "savings." This is of course apart from demand for loans, regulations and shareholder preferences that can limit commercial banks' capacity to expand their balance sheets. Bottom Line: In an economy with banks, one does not need to save in the form of a deposit in a bank for the latter to lend money to another entity. Tales Of Brazil And China Chart I-2Two Extremes Of Investment ##br##And Savings: China And Brazil We use China and Brazil solely for illustrative purposes. One can use any country with a low savings rate instead of Brazil or a high savings rate economy such as Korea, Taiwan or Singapore in place of China. China has enjoyed a very high national savings rate and has been investing substantially (Chart I-2). In contrast, both the national savings rate and the investment-to-GDP ratio in Brazil have been depressed. It is very tempting to argue that Brazil has been experiencing very low investment because it saves so little. The narrative goes like this: Brazil's national savings rate is low because households save so little and the public sector dis-saves a lot - i.e., the government runs enormous fiscal deficits. This constrains the pool of available "savings" to finance private capital expenditures. This typical analysis concludes that Brazil needs to boost its "savings" - i.e., reduce its spending. This will allegedly enlarge the pool of available "savings" for investment and allow the country to invest, and consequently boost productivity and its potential growth rate. This narrative is misplaced in our view, because as we have shown in the past and in this report, banks do not need households, businesses or the government to save in order to provide financing. Banks can provide financing by simply expanding the money multiplier, among other things (see a more detailed discussion about the money multiplier below). So what is the true difference between Brazil and China? How has the latter achieved such high savings and investment rates, while the former has failed to finance its capital spending? Why have Brazilian banks not expanded their balance sheets more rapidly to finance investment (Chart I-3)? Chart I-3Snapshot Of Bank Assets-To-GDP Ratios Let's consider a hypothetical example. For simplicity and illustrative purposes, we assume there are two economies of equal size and have the same level of investment: savings and net exports. In short, they have identical starting points. We refer to these economies as Brazil and China. Now, commercial banks in both countries provide new financing of $50 - or equal to 5% of their respective GDP - to businesses for infrastructure building. This is new purchasing power created by commercial banks "out of thin air" in both economies. We assume that the only difference between these two countries is that in China, 100% of inputs for infrastructure (materials, machinery/equipment and so on) are produced/purchased domestically. In contrast, in Brazil, 100% of the inputs for infrastructure construction are imported, because this economy lacks production capacity. Table I-1 illustrates this hypothetical numerical example. As this infrastructure project is implemented, Brazil's imports will surge, and its net exports will deteriorate. Chart I-4 shows that this indeed is the case in Brazil - when capital spending expands, its current accounts deficit widens, entailing that Brazil imports a considerable portion of inputs for its investments. Table I-1A Hypothetical Example Of Investment - Saving Dynamics Chart I-4Foreign Content Of Brazil's ##br##Capital Spending Is High If there is no matching rise in foreign investor demand for Brazilian assets, the nation's currency will depreciate. Consequently, to support the plunging currency, Brazilian interest rates would have to rise. As a result, higher borrowing costs short-circuit the credit cycle. In China, because inputs for infrastructure are sourced and procured locally, there is no impact on its exchange rate or interest rates. If there is excess capacity in China to produce these inputs for infrastructure building, this new purchasing power will not lift inflation. A caveat is in order: Similar dynamics in trade balance deterioration, currency depreciation and inflation will prevail if there is a rise in consumer spending instead of capital expenditures. Importantly, the outcome will be the same in both economies if investment spending is done using existing money savings (deposits), not new credit. This example illustrates that a similar amount of capital expenditures financing via money creation "out of thin air" in both economies has increased national savings in China from $250 to $300, yet Brazilian savings stayed at $250 (Table I-1). In terms of savings rate, China will record a rise in its national savings rate from 25% to 28.6% of GDP (Table I-1). In Brazil, however, the national savings rate will remain at 25% of GDP, even though its banks, like Chinese ones, originated money "out of thin air" to finance infrastructure spending. The starting-point difference between China and Brazil is neither their banking systems' ability to expand their balance sheets nor the existing amount of deposits and assets. Rather, it is their real economies' capacity to produce goods and services. Therefore, we conclude: Regardless of how capital expenditures are financed - via new borrowing from banks or non-banks or using the investing company's own financial resources - when inputs for capital spending are procured domestically it is recorded as an increase in national "savings" level, but when they are imported there is no change in the level of national "savings." Over the decades, China, Korea, Taiwan, Singapore and Japan have all aggressively expanded their capacity to produce goods and services. They funded this capacity build-up via both money creation "out of thin air" and by attracting foreign capital. In the meantime, their large exports shielded their currencies from abrupt depreciation - as and when local bank financing was used to acquire foreign inputs. In the past decade, in China, loans - which banks have originated to build infrastructure - were largely spent on domestic inputs: cement, steel, chemicals, machinery and equipment all produced in the mainland. Even though some of that money/loans was used to purchase foreign inputs (commodities and equipment), China had large U.S. dollar revenues from exports that acted as an offset in its balance of payments. In short, Brazil and other low "savings" rate nations do not need to raise interest rates to curtail consumption and boost savings in order to release funds for financing capital expenditures. Chart I-5 demonstrates that there has been no positive relationship between real interest rates and the national savings rate in Brazil. Remarkably, real interest rates in this nation were often very high but that still did not lead to high "savings." Chart I-5Real Interest Rates And Savings Are Not Positively Correlated As They Are Supposed To Be What Brazil and other low "savings" rate economies need is to build efficient and competitive productive capacity - i.e., they need changes in the supply side of their economies. Only then can their banks expand their balance sheets and provide financing similar to how banks in high "savings" countries do. However, to shield the exchange rate from depreciation, these nations need to boost their exports first. This can be done by depreciating the currency and developing their global competitiveness. This is in effect what China has done in the past 25-30 years. Bottom Line: The key difference between Brazil and China is not their propensity to consume versus save, but their ability to produce goods and services domestically. So long as a nation builds and maintains excess productive capacity, its banks can originate loans "out of thin air" and finance capital and consumer expenditures. Money Multiplier Versus "Savings" Redundancy of the mainstream economic view that a pool of "savings" represents a constraint on financing investments becomes apparent when one applies the money multiplier concept, which is in fact accepted by mainstream economic theory. The money multiplier is the ratio of broad money relative to excess reserves. A rise in the money multiplier will lead to more money creation and financing in an economy per one unit of excess reserves (liquidity provided by the central bank), everything else held constant. In brief, money supply/the amount of deposits in the banking system will change regardless of the level of national or household "savings." Let's assume two countries with the same level of income per capita and GDP have identical national savings and investment rates as well as money supply and excess reserves. In short, they have indistinguishable macro parameters. Now suppose their banking systems in the past year had different money multipliers. The monetary authorities in both countries maintain the banking system's excess reserves at 10 units. If the money multiplier were to remain constant, say at 15, the money supply/deposits in both banking systems would remain at 150 units (10x15). Let's assume the money multiplier increased to 20 in Country A while held constant at 15 in Country B. In such a case, broad money supply would have risen to 200 units (10x20) in Country A and would stay at 150 (10x15) units in Country B. This entails that banks in Country A increased their funding yet those in Country B did not. That is despite the fact that the savings rates (and amount of savings) were identical before the change in the money multipliers occurred. This is one way to prove that a nation does not need to cut consumption for its banks to provide financing. The reason why the money multipliers could vary in these two countries with otherwise similar macro-economic parameters is due to animal spirits: In Country A, banks may have felt increasingly confident to lend more per one unit of their excess reserves, and there was demand for credit from borrowers. In the meantime, the money multiplier remained the same in Country B. In China, the money multiplier - the ratio of broad money to excess reserves - has risen dramatically since 2013 (Chart I-6). Interestingly, the amount of excess reserves at the People's Bank of China has been broadly the same over the past five years, yet broad money has grown by an enormous 75% (Chart I-6, middle and bottom panel). The exponential money/credit creation in China since 2009 has to a large extent been due to the rising money multiplier - wild animal spirits among bankers and borrowers - rather than high national "savings." Bottom Line: In any country, banks can provide more financing simply by expanding the money multiplier. This can happen regardless of the country's savings rate. Investment Relevance Why is this analysis pertinent to investors? First, this issue is critical to assess whether China's excessive credit expansion is an outcome of the nation's high savings - like many economists and investors claim - or due to the enormous amount of money/deposits and credit originated by the mainland's banks "out of thin air." If it is the former, investors have no need to worry about China's money and credit dynamics. If it is the latter, we are facing a typical banking and money/credit bubble. This report corroborates that it is the latter. Chart I-7 shows that China's broad money has grown 4-fold since January 2009 and has reached RMB 200 trillion, or the equivalent of $30 trillion. Chart I-6China: Money Multiplier Has Risen A Lot Chart I-7A Money Bubble In China? Does this enormous quantity of RMBs pose an inflation and/or currency depreciation risk? Or will the ongoing policy tightening cause another deflationary slump in China? It is clear that Chinese policymakers are currently being forced to walk a very thin line: On the one hand, the immense amount of money created "out of thin air" could stoke inflation or currency depreciation. It may not take much of a rise in the velocity of money for inflation to become a problem. On the other, tightening policy amid high leverage in an economy that is addicted to money and credit could push it into a growth slump and deflation. There is always a chance that policymakers will get it right and manage it perfectly so that neither inflation/currency depreciation nor a growth slump transpire. We would assign a 25-30% probability to this benign outcome. Hence, in our opinion there are 70-75% odds of either inflation or deflation in China in the next 12-24 months. Given these odds, we have been and remain reluctant to chase the rally in EM and China-related plays. In particular, the Chinese authorities have been tightening liquidity and banking/shadow banking regulation as well as projecting the ongoing anti-corruption campaign into the financial industry. This poses a meaningful risk given the existing macro imbalances. Second, this analysis re-shapes how investors should think about economic development and understand how nations with low savings can grow without relying on foreign funding. This provides us with a framework to assess the developmental path and the sustainability of growth in various developing economies. These include but are not limited to nations with low national savings rates such as Brazil, South Africa, Turkey, Russia, Colombia and many others. Finally, this analysis leads us to argue that Brazil does not need to maintain high real interest rates as a way to force consumers to cut spending and boost savings. In fact, this is the wrong prescription for Brazil. The most optimal macro adjustment path for Brazil is to reduce interest rates much further and encourage banks to finance private investment. Brazil needs to build an efficient supply side, and banks can provide funding by originating loans "out of thin air." Brazilian consumers do not need to save more for companies to get financing for their projects and invest. The natural causality of this adjustment will be considerable currency deprecation. However, Brazil is currently suffering from low inflation and high real interest rates (Chart I-8). Hence, reflationary policies are the right policy prescription. Chart I-8Brazil Needs To Reduce ##br##Interest Rates Much Further Foreign investors are therefore at risk due to potential currency depreciation. The new leaders to be elected in the October presidential elections may well adopt such a macro policy mix. Markets will front run this by pushing the real down and this will be negative for foreign investors. However, there will be a buying opportunity after the currency finds a floor. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com Andrija Vesic, Research Analyst andrijav@bcaresearch.com Appendix 1: Loan Origination, Deposits/Money Creation And Settlement The amount of deposits is not a constraint on a banking system's ability to make loans and buy assets from non-banks. Figure I-1 and I-2 present stylized cases of how commercial banks can originate new loans without requiring a new deposit or extra excess reserves entering the banking system. Specifically Figure I-1 illustrates how commercial banks can originate loans with the subsequent net settlements among themselves taking place via inter-bank borrowing/lending. In this stylized example, the banking system is comprised of three commercial banks. These commercial banks hold all deposits in the system. Cash does not exist and all payments are done via wire transfers. Figure I-1Money Creation By Banks With Net Settlement Among Banks Via Inter-Bank Lending/Borrowing Figure I-2Money Creation By Banks With Net Settlement Between Banks & Central Bank 1. Loan Origination/Money Creation In the morning, Bank 1 originates a new loan worth $100 for Client 1. This transaction creates a new asset and, for the balance sheet to balance, Bank 1 should also increase the liabilities side of its balance sheet. Therefore, it simultaneously credits Client 1's chequing account by $100. Bank 1 does not transfer other depositors' money to Client 1's chequing account; it creates a new $100 deposit. The rest of the bank's depositors still have their full deposits, which they can draw on. In a nutshell, both assets and liabilities of Bank 1 rose by $100 - this was done "out of thin air" by just pressing the enter button on the computer. That also means that a $100 of new money was created by Bank 1 which increases the overall money stock in the banking system. Meanwhile, Bank 2 lends $200 to Client 2 and Bank 3 lends $300 to Client 3. Let's assume these were the only lending transactions during that day. In aggregate, the three banks originated $600 of new loans, and consequent new deposits/money "out of thin air." 2. Money Transfer / Payments Debtors do not borrow money and leave it sitting idle. They borrow money to pay their suppliers and others they owe. Even though Clients 1, 2 and 3 wire their payments to their respective suppliers on the same day, the total amount of deposits in the banking system does not change: Deposits simply move from one bank to another or from one bank client to another. In Figure I-1, Client 1 wires its $100 from Bank 1 to Supplier B that has an account at Bank 2; Client 2 pays its $200 invoice to Supplier C which in turn has an account at Bank 3; and finally Client 3 transfers $300 to Supplier A, who holds an account at Bank 1. The amount of money/deposits in the overall banking system has not changed as a result of these wire transfers. 3. Multilateral Net Settlement At the end of the day, banks should settle with other banks. Many countries employ a multilateral net settlement system typically operated by the central bank. In a multilateral net settlement system, at the end of the day, each bank pays (receives from) the system only the net amount they are due to pay to (receive from) other banks combined. Importantly, banks settle their payments with other banks using their excess reserves (herein called reserves) at the central bank, not the deposits of their clients. This entails that banks do not need deposits to pay their dues to other banks or the central bank. Figure I-1 illustrates the impacts on the banks' reserves under the multilateral net settlement system: Bank 1's reserves at the central bank change as follows: -$100 (Client 1's wire transfer out) + $300 (this is the amount that Supplier A with an account in Bank 1 gets from Client 3) = $200. The impact on Bank 2's reserves is as follows: -$200 (Client 2's wire transfer out) + $100 (this is the amount that Supplier B with an account in Bank 2 gets from Client 1) = -$100. The net change in Bank 3's reserves is: -$300 (Client 3's wire transfer out) + $200 (this is the amount that Supplier C with an account in Bank 3 gets from Client 2) = -$100. If we assume that all banks had no excess reserves before this day, then how do they settle their accounts? There are various alternatives, but we highlight two: Figure I-1 demonstrates the case of interbank lending. As a result of the settlements, Bank 1 has $200 in extra reserves, while Bank 2 and Bank 3 each have a $100 deficit in reserves. As such, Bank 1 lends $100 to each of Bank 2 and Bank 3. Why does it lend to other banks rather than keeping these reserves at the central bank? Because interbank rates are typically slightly above the central bank's rate - the rate Bank 1 would get if it were to lend the $200 to the central bank. Figure I-2 portends the same transactions with the sole difference being the reserves flow. Unlike Figure I-1, here banks do not lend to/borrow from each other. Banks lend excess reserves to the central bank as well as borrow deficient reserves from the central bank. This is done to settle their payments with other banks. Bank 1 lends its free reserves of $200 to the central bank. Bank 2 and Bank 3 each borrow $100 reserves from the central bank to settle with the system at the end of the day. As a result, the aggregate amount of reserves at the central bank does not change. On the whole, banks created $600 of new deposits/money/loans during the day without requiring savings from households, companies, the government or foreigners. Thereby, the money supply was expanded and new financing in the amount of $600 was provided "out of thin air." Appendix 2: Deposits Versus Liquidity Below are additional questions that we seek to answer to provide further elaboration on the issues of banks creating money and the difference between deposits and liquidity: 1. Why would central banks provide reserves to banks? When a central bank targets interest rates, which is nowadays the most common policy framework in both advanced and developing countries, it must provide liquidity to banks: the latter is required to preclude interbank rates from deviating from the policy rate. Under an interest rate targeting regime, the central bank does not have complete control over banks' reserves nor broad money supply. A central bank can control either the quantity of money or the price of money (interest rates), but not both simultaneously. The following two quotes from the New York Federal Reserve Chairman William Dudley and the European Central Bank confirm that central banks nowadays provide banks with reserves on demand - i.e., the amount of reserves is determined by demand from banks. "The Federal Reserve has committed itself to supply sufficient reserves to keep the fed funds rate at its target. If banks want to expand credit and that drives up the demand for reserves, the Fed automatically meets that demand in its conduct of monetary policy. In terms of the ability to expand credit rapidly, it makes no difference whether the banks have lots of excess reserves or not." (Dudley, 2009) European Central Bank (2012), May 2012 Monthly Bulletin: "The Eurosystem ... always provides the banking system with the liquidity required to meet the aggregate reserve requirement. In fact, the ECB's reserve requirements are backward-looking, i.e. they depend on the stock of deposits (and other liabilities of credit institutions) subject to reserve requirements as it stood in the previous period, and thus after banks have extended the credit demanded by their customers." 2. Why do banks compete for deposits if they create deposits themselves? The true reason banks compete for deposits is not that they require more deposits, but because they require more reserves. When a bank attracts a deposit from another bank, the latter must transmit to the former reserves equal to the amount of the deposit transferred. When a bank is experiencing a liquidity shortage, more deposits are of no help. Banks can always create more deposits themselves, but they cannot create reserves at the central bank. The true liquidity for banks is their reserves at the central bank - not deposits. Reserves are solely created by central banks "out of thin air." A central bank may decide not to provide funding to certain banks in some cases when the authorities deem these banks insolvent and/or in breach of regulations. Otherwise, if the central bank wants to keep policy rates stable, it must provide all liquidity (reserves) banks require. 3. Why do banks attract deposits if the central bank provides liquidity on demand? The primary reason why banks seek to attract deposits instead of borrowing from the central bank is due to the cost of funding and duration of liabilities as well as regulatory requirements. Deposits may be cheaper and have longer duration than short-term funding from the central bank. 1 Lindner, F. (2015), "Does Savings Increase the Supply of Credit? A Critique of Loanable Funds Theory", Macroeconomic Policy Institute, World Economic Review 4, 2015. 2 Lindner, F. (2015), "Did Scarce Global Savings Finance the US Real Estate Bubble? The Global Saving Glut thesis from a stock flow Consistent Perspective", Macroeconomic Policy Institute, Working Paper 155, July 2015. 3 Lindner, F. (2015), "Does Savings Increase the Supply of Credit? A Critique of Loanable Funds Theory", Macroeconomic Policy Institute, World Economic Review 4, 2015. 4 Please refer to the Emerging Markets Strategy Special Reports from October 26, 2016, November 23, 2016, January 18, 2017 and December 20, 2017; available on ems.bcaresearch.com 5 Borio, C. and Disyatat, P. (2015), "Capital flows and the current account: Taking financing (more) seriously", BIS Working Papers, No. 525, October 2015. 6 Lindner, F. (2015), "Did Scarce Global Savings Finance the US Real Estate Bubble? The Global Saving Glut thesis from a stock flow Consistent Perspective", Macroeconomic Policy Institute, Working Paper 155, July 2015. 7 Lindner, F. (2012), "Savings does not finance Investment: Accounting as an indispensable guide to economic theory", Macroeconomic Policy Institute, Working Paper 100, October 2012. Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights Several economic and financial market indicators point to a budding downtrend in Chinese capital spending and its industrial sector. The recent underperformance of global mining, chemicals and machinery/industrials corroborate that capital spending in China is starting to slump. Shipments-to-inventory ratios for Korea and Taiwan also point to a relapse in Asian manufacturing. This is occurring as our global growth sentiment proxy sits on par with previous peaks, and investor positioning in EM and commodities is overextended. Stay put on EM. Markets with currency pegs to the U.S. dollar, such as the Gulf states and Hong Kong, will face tightening local liquidity. Share prices in these markets have probably topped out. Feature On the surface, EM equities, currencies and local bond and credit markets are still trading well. However, there are several economic indicators and financial variables that herald negative surprises for global and Chinese growth. In particular: China's NBS manufacturing PMI new orders and backlogs of orders have relapsed in the past several months. Chart I-1 illustrates the annual change in new orders and backlogs of orders to adjust for seasonality. The measure leads industrial profits, and presently foreshadows a slowdown going forward. Furthermore, the average of NBS manufacturing PMI, new orders, and backlog orders also points to a potential relapse in industrial metals prices in general as well as mainland steel and iron ore prices (Chart I-2). The message from Charts I-1 and I-2 is that the recent weakness in iron ore and steel prices could mark the beginning of a downtrend in Chinese capital spending. While supply cuts could limit downside in steel prices, it would be surprising if demand weakness does not affect steel prices at all.1 Chart I-1China: Slowdown Has Further To Run Chart I-2Industrial Metals Prices Have Topped Out Although China's money and credit have been flagging potential economic weakness for a while, the recent manufacturing PMI data from the National Bureau of Statistics finally confirmed an impending deceleration in industrial activity and ensuing corporate profit disappointment. Our credit and fiscal spending impulses continue to point to negative growth surprises in capital spending. The latter is corroborated by the weakening Komatsu's Komtrax index, which measures the average hours of machine work per unit in China (Chart I-3). In both Korea and Taiwan, the overall manufacturing shipments-to-inventory ratios have dropped, heralding material weakness in both countries' export volumes (Chart I-4). Chart I-3Signs Of Weakness In Chinese Construction Chart I-4Asia Exports Are Slowing Notably, global cyclical equity sectors that are leveraged to China's capital spending such as materials, industrials and energy have all recently underperformed the global benchmark (Chart I-5). Some of their sub-sectors such as machinery, mining and chemicals have also begun to underperform (Chart I-6). Chart I-5Global Cyclicals Have ##br##Begun Underperforming... Chart I-6...Including Machinery ##br##And Chemical Stocks Among both global and U.S. traditional cyclicals, only the technology sector is outperforming the benchmark. However, we do not think tech should be treated as a cyclical sector, at least for now. In brief, the underperformance of global cyclical equity sectors and sub-sectors following last month's equity market correction corroborate that China's capital spending is beginning to slump. Notably, this is occurring as our global growth sentiment proxy rests on par with its previous apexes (Chart I-7). Previous tops in this proxy for global growth sentiment have historically coincided with tops in EM EPS net revisions, as shown in this chart. Chart I-7Global Growth Sentiment: As Good As It Gets All told, we may be finally entering a meaningful slowdown in China that will dampen commodities prices and EM corporate earnings. The latter are still very strong but EPS net revisions have rolled over and turned negative again (Chart I-8). Chart I-8EM EPS Net Revisions Have Plummeted EM share prices typically lead EPS by about nine months. In 2016, EM stocks bottomed in January-February, yet EPS did not begin to post gains until December 2016. Even if EM corporate profits are to contract in the fourth quarter of this year, EM share prices, being forward looking, will likely begin to wobble soon. Poor EM Equity Breadth There is also evidence of poor breadth in the EM equity universe, especially compared to the U.S. equity market. First, the rally in the EM equally-weighted index - where all individual stocks have equal weights - has substantially lagged the market cap-weighted index since mid 2017. This suggests that only a few large-cap companies have contributed a non-trivial share of capital gains. Second, the EM equal-weighted stock index's and EM small-caps' relative share prices versus their respective U.S. counterparts have fallen rather decisively in the past six weeks (Chart I-9, top and middle panels). While the relative performance of market cap-weighted indexes has not declined that much, it has still rolled over (Chart I-9, bottom panel). We compare EM equity performance with that of the U.S. because DM ex-U.S. share prices themselves have been rather sluggish. In fact, DM ex-U.S. share prices have barely rebounded since the February correction. Third, EM technology stocks have begun underperforming their global peers (Chart I-10). This is a departure from the dynamics that prevailed last year, when a substantial share of EM outperformance versus DM equities was attributed to EM tech outperformance versus their DM counterparts and tech's large weight in the EM benchmark. Chart I-9EM Versus U.S. Equities: Relative ##br##Performance Is Reversing Chart I-10EM Tech Has Started ##br##Underperforming DM Tech Finally, the relative advance-decline line between EM versus U.S. bourses has been deteriorating (Chart I-11). This reveals that EM equity breadth - the advance-decline line - is substantially worse relative to the U.S. Chart I-11EM Versus U.S.: Relative Equity Breadth Is Very Poor Bottom Line: Breadth of EM equity performance versus DM/U.S. has worsened considerably. This bodes ill for the sustainability of EM outperformance versus DM/U.S. We continue to recommend an underweight EM versus DM position within global equity portfolios. Three Pillars Of EM Stocks EM equity performance is by and large driven by three sectors: technology, banks (financials) and commodities. Table I-1 illustrates that technology, financials and commodities (energy and materials) account for 66% of the EM MSCI market cap and 75% of MSCI EM total (non-diluted) corporate earnings. Therefore, getting the outlook of these sectors right is crucial to the EM equity call. Table I-1EM Equity Sectors: Earnings & Market Cap Weights Technology Four companies - Alibaba, Tencent, Samsung and TSMC - account for 17% of EM and 58% of EM technology market cap, respectively. This sector can be segregated into hardware tech (Samsung and TSMC) and "new concept" stocks (Alibaba and Tencent). We do not doubt that new technologies will transform many industries, and there will be successful companies that profit enormously from this process. Nevertheless, from a top-down perspective, we can offer little insight on whether EM's "new concept" stocks such as Alibaba and Tencent are cheap or expensive, nor whether their business models are proficient. Further, these and other global internet/social media companies' revenues are not driven by business cycle dynamics, making top-down analysis less imperative in forecasting their performance. We can offer some insight for technology hardware companies such as Samsung and TSMC. Chart I-12 demonstrates that semiconductor shipment-to-inventory ratios have rolled over decisively in both Korea and Taiwan. In addition, semiconductor prices have softened of late (Chart I-13) Together, this raises a red flag for technology hardware stocks in Asia. Chart I-12Asia's Semiconductor Industry Chart I-13Semiconductor Prices: A Soft Spot? Finally, Chart I-14 compares the current run-up in U.S. FANG stocks (Facebook, Amazon, Netflix and Google) with the Nasdaq mania in the 1990s. An equal-weighted average stock price index of FANG has risen by 10-fold in the past four and a half years. Chart I-14U.S. FANG Stocks Now ##br##And 1990s Nasdaq Mania A similar 10-fold increase was also registered by the Nasdaq top 100 stocks in the 1990s over eight years (Chart I-14). While this is certainly not a scientific approach, the comparison helps put the rally in "hot" technology stocks into proper historical perspective. The main take away here is that even by bubble standards, the recent acceleration in "new concept" stocks has been too fast. That said, it is impossible to forecast how long any mania will persist. This has been and remains a major risk to our investment strategy of being negative on EM stocks. In sum, there is little visibility in EM "new concept" tech stocks. Yet Asia's manufacturing cycle is rolling over, entailing downside risks to tech hardware businesses. Putting all this together, we conclude that it is unlikely that EM tech stocks will be able to drive the EM rally and outperformance in 2018 as they did in 2017. Banks We discussed the outlook for EM bank stocks in our February 14 report,2 and will not delve into additional details here. In brief, several countries' banks have boosted their 2017 profits by reducing their NPL provisions. This has artificially boosted profits and spurred investors to bid up bank equity prices. We believe banks in a number of EM countries are meaningfully under-provisioned and will have to augment their NPL provisions. The latter will hurt their profits and constitutes a major risk for EM bank share prices. Energy And Materials The outlook for absolute performance of these sectors is contingent on commodities prices. Industrial metals prices are at risk of slower capex in China. The mainland accounts for 50% of global demand for all industrial metals. Oil prices are at risk from traders' record-high net long positions in oil futures, according to CFTC data (Chart I-15, top panel). Traders' net long positions in copper are also elevated, according to the data from the same source (Chart I-15, bottom panel). Hence, it may require only some U.S. dollar strength and negative news out of China for these commodities prices to relapse. Chart I-15Traders' Net Long Positions In ##br##Oil And Copper Are Very Elevated How do we incorporate the improved balance sheets of materials and energy companies into our analysis? If and as commodities prices slide, share prices of commodities producers will deflate in absolute terms. However, this does not necessarily mean they will underperform the overall equity benchmark. Relative performance dynamics also depend on the performance of other sectors. Commodities companies could outperform the overall equity benchmark amid deflating commodities prices if other equity sectors drop more. In brief, the improved balance sheets of commodities producers may be reflected in terms of their relative resilience amid falling commodities prices but will still not preclude their share prices from declining in absolute terms. Bottom Line: If EM bank stocks and commodities prices relapse as we expect, the overall EM equity index will likely experience a meaningful selloff and underperform the DM/U.S. benchmarks. Exchange Rate Pegs Versus U.S. Dollar With the U.S. dollar depreciating in the past 12 months, pressure on exchange rate regimes that peg their currencies to the dollar has subsided. These include but are not limited to Hong Kong, Saudi Arabia and the United Arab Emirates (UAE). As a result, these countries' interest rate differentials versus the U.S. have plunged (Chart I-16). In short, domestic interest rates in these markets have risen much less than U.S. short rates. This has kept domestic liquidity conditions easier than they otherwise would have been. However, maneuvering room for these central banks is narrowing. In Hong Kong, the exchange rate is approaching the lower bound of its narrow band (Chart I-17). As it touches 7.85, the Hong Kong Monetary Authority (HKMA) will have no choice but to tighten liquidity and push up interest rates. Chart I-16Markets With U.S. Dollar Peg: ##br##Policymakers' Maneuvering Window Is Closing Chart I-17Hong Kong: Interest ##br##Rates Are Heading Higher In Saudi Arabia and the UAE, the monetary authorities have used the calm in their foreign exchange markets over the past year to not match the rise in U.S. short rates (Chart I-18A and Chart I-18B). However, with their interest rate differentials over U.S. now at zero, these central banks will have no choice but to follow U.S. rates to preserve their currency pegs.3 Chart I-18ASaudi Arabian Interest Rates Will Rise Chart I-18BThe UAE Interest Rates Will Rise If U.S. interest rates were to move above local rates in Saudi Arabia and the UAE, those countries' currencies will come under considerable depreciation pressure because capital will move from local currencies into U.S. dollars. Hence, if U.S. short rates move higher, which is very likely, local rates in these and other Gulf countries will have to rise if their exchange rate pegs are to be preserved. Neither the Hong Kong dollar nor Gulf currencies are at risk of devaluation. The monetary authorities there have enough foreign currency reserves to defend their respective pegs. Nevertheless, the outcome will be domestic liquidity tightening in the Gulf's and Hong Kong's banking system. In addition, potentially lower oil prices will weigh on Gulf bourses and China's slowdown will hurt growth and equity sentiment in Hong Kong. All in all, equity markets in Gulf countries and Hong Kong have probably seen their best in terms of absolute performance. Potential negative external shocks and higher interest rates due to Fed tightening have darkened the outlook for these bourses. Bottom Line: Local liquidity in Gulf markets and Hong Kong is set to tighten. Share prices in these markets have probably topped out. However, given these equity markets have massively underperformed the EM equity benchmark, they are unlikely to underperform when the overall EM index falls. Hence, we do not recommend underweighting these bourses within an EM equity portfolio. For asset allocators, a neutral or overweight allocation to these bourses is warranted. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 Please see Emerging Markets Strategy Special Report "China's "De-Capacity" Reforms: Where Steel & Coal Prices Are Headed," dated November 22, 2017; the link is available on page 16. 2 Please see Emerging Markets Strategy Special Report "EM Bank Stocks Hold The Key," dated February 14, 2018; the link is available on page 16. 3 Please see BCA's Frontier Markets Strategy Special Report "United Arab Emirates: Domestic Tailwinds, External Headwinds," dated March 12, 2018. The link is available on fms.bcaresearch.com. Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Special Report

Feature The existence of 'mini-cycles' in economic and financial variables is an empirical fact. We encourage readers to plot for themselves the change in global bank credit flows, the global bond yield, global inflation, and metal price inflation. The very clear and regular mini-cycles should shout out at you (Chart I-2, Chart I-3, Chart I-4, Chart-5). Feature ChartThe Cobweb Theory Explains The Regular Mini-Cycles In Economic And Financial Variables Chart I-2Mini-Cycles In Global Credit Flows Chart I-3Mini-Cycles In The Global Bond Yield Chart I-4Mini-Cycles In Global Inflation Chart I-5Mini-Cycles In Metal Price Inflation Identifying these mini-cycles is very useful because it helps us to predict the future. Just as we know when the tide will go out and come back in, we can predict the mini-cycle's downswings and upswings. And if most market participants are unaware of the next turn in the mini-cycle, it will not be discounted in today's price - providing a compelling investment opportunity. The obvious question is: if the existence of mini-cycles is an empirical fact, what is its theoretical foundation? Dusting Down The Cobweb Theory A likely answer comes from an economic model called the Cobweb Theory, first proposed in the 1930s by several economists, among them Althus Hanau and Nicholas Kaldor. The Cobweb Theory is so called because when its predicted pattern of price and output mini-cycles is traced out on a standard price/quantity diagram, it resembles a cobweb (Chart I-6, Chart 7, Chart I-8). Chart I-6Cobweb Theory Case 1: ##br## Regular Mini-Cycles Chart I-7Cobweb Theory Case 2: ##br##Divergent Mini-Cycles Chart I-8Cobweb Theory Case 3: ##br##Convergent Mini-Cycles The Cobweb Theory is based on a simple premise: lagging supply. The demand for an item depends on its price in the current period, but the supply of the item depends on its price in the previous period. Or equivalently, the price in the current period influences the supply in the next period. In the 1930s, economists used the theory to explain the mini-cycles in agricultural output and prices. Most crops can be sown and reaped only once a year. Therefore, an unanticipated increase in demand will cause a sharp rise in price - because there can be no immediate increase in supply. This high price may lure farmers to increase their output more than is justified by future demand. So when this supply eventually comes on the market, it will cause a sharp fall in price. In turn this will result in a decrease in output for the next period to a greater extent than is justified. And so on. More generally, the Cobweb Theory applies in any market where supply lags demand. Under this simple premise, the market price will produce a two-period oscillation with the actual price being alternately above and below the equilibrium price. When the price is above equilibrium, it falls in the next period as supply adjusts upwards; and when the price is below equilibrium, it rises in the next period as supply adjusts downwards. But supply tends to over-adjust, causing both the quantity and price to overshoot and undershoot equilibrium repeatedly - effectively creating a mini-cycle (see Box I-1). Box I-1The Cobweb Theory Of Cycles The Cobweb Theory Of Credit Demand And Supply We now come to a key point: credit demand and supply often meet the conditions of the Cobweb Theory. Chart I-9 illustrates that the credit demand cycle is perfectly coincident with the bond yield cycle. Whereas Chart I-10 and Chart I-11 demonstrate that the credit supply cycle can often lag the credit demand cycle - and therefore the bond yield cycle - by several months. One obvious explanation is that unless you have an (unexpended) existing credit line to draw upon, there will be a lag between applying for credit and receiving it. Chart I-9The Credit Demand Cycle Is Coincident ##br##With The Bond Yield Cycle... Chart I-10...But The Credit Supply Cycle Lags ##br##The Credit Demand Cycle... Chart I-11...And The Bond ##br##Yield Cycle With credit demand and supply meeting the conditions of the Cobweb Theory, both the quantity and the price of credit (the bond yield) should exhibit mini-cycles. And as the charts in this report attest, they do. What about the mini-cycles in commodity inflation and broader CPI inflation? Given that these closely track the credit impulse mini-cycle (Feature Chart), we can deduce that they must be mostly a reflection of the mini-cycle in global demand growth. Still, could the commodity inflation mini-cycle also be impacted by the supply-side, as postulated for agricultural prices in the original Cobweb Theory? Interestingly, a recent paper, "The cobweb theorem and delays in adjusting supply in metals" markets,1 does "link the dynamics of raw material markets and commodity price fluctuations to a delayed adjustment of supply." However, the supply lags mentioned in the paper are too long to explain the half-cycle lengths typically observed in the commodity inflation mini-cycle. This would confirm that this mini-cycle is mostly a demand-side phenomenon. But the paper does also point out that speculation on futures markets may lead to higher volatility. This implies that while the phases of the mini-cycles should stay closely aligned, the amplitudes of the commodity inflation and credit impulse mini-cycles can deviate. Which is precisely what we observe in the data (Chart I-12). Chart I-12The Various Mini-Cycles Have Similar Periods But Different Amplitudes What Is The Current Message? Chart I-13The Bond Yield Cycle Explains##br## The Sector Selection Cycle To sum up, global credit flows, the global bond yield, global inflation, and metal price inflation exhibit clear and regular mini-cycles with a consistent half-cycle length averaging around 8 months, but not necessarily a consistent amplitude. We propose that all of these mini-cycles will continue indefinitely, and that they are manifestations of the lagging supply of credit and the Cobweb Theory. In the context of these clear and regular mini-cycles, the current mini-upswing in activity which started last May is getting long in the tooth, and we would expect it to end in early 2018. Having said that, given that the recent upswing in the global bond yield is quite modest, the next mini-downswing in the global credit impulse, and thereby activity, should be quite shallow. Nevertheless, in terms of investment implications, any mini-upswing in price since last May that has displayed an outsize amplitude would be more vulnerable to a setback. Industrial metal prices might be in this vulnerable category. Furthermore, the mini-cycle framework has been an important driver of cyclical versus defensive sector performance over the past few years (Chart I-13), and likely will continue to be an important driver. On a 6-9 month horizon, the current message would be to pare back exposure to cyclical sectors and to tilt towards defensive-biased equity markets such as Switzerland and Denmark. Dhaval Joshi, Senior Vice President Chief European Investment Strategist dhaval@bcaresearch.com 1 System Dynamics Review, April 2017: 'The cobweb theorem and delays in adjusting supply in metals' markets' by Glöser-Chahoud, Hartwig, Wheat and Faulstich. Fractal Trading Model* This week's trade is to expect a countertrend reversal in S&P500 versus Eurostoxx50 performance. Set a profit target of 2.0% with a symmetrical stop-loss. In other trades, long IBEX35 / short Eurostoxx50 closed in profit while short WTI crude closed at its stop-loss. For any investment, excessive trend following and groupthink can reach a natural point of instability, at which point the established trend is highly likely to break down with or without an external catalyst. An early warning sign is the investment's fractal dimension approaching its natural lower bound. Encouragingly, this trigger has consistently identified countertrend moves of various magnitudes across all asset classes. Chart 14 The post-June 9, 2016 fractal trading model rules are: When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. Use the position size multiple to control risk. The position size will be smaller for more risky positions. * For more details please see the European Investment Strategy Special Report "Fractals, Liquidity & A Trading Model," dated December 11, 2014, available at eis.bcaresearch.com Fractal Trading Model Recommendations Equities Bond & Interest Rates Currency & Other Positions Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields Interest Rate Chart II-5Indicators To Watch##br## - Interest Rate Expectations Chart II-6Indicators To Watch ##br##- Interest Rate Expectations Chart II-7Indicators To Watch ##br##- Interest Rate Expectations Chart II-8Indicators To Watch ##br##- Interest Rate Expectations

Special Report

Highlights The financial system / banks cannot and do not lend out or intermediate national or households "savings". In any economy, new money/new purchasing power is originated by commercial banks "out of thin air". The term "savings" in macroeconomics denotes an increase in the economy's capital stock, not deposits at the banks. The Chinese banking system has enormous amount of deposits because banks have created them "out of nothing" not because households save a lot. Hence, the narrative that justifies China's money, credit and property market excesses by high national and household "savings" is incorrect. The maneuvering room for China is diminishing as inflationary pressures are rising, productivity is slowing and speculative leverage is high. Feature The debate on China's macro outlook continues to linger both within and outside BCA. The focal point of the debate centers on the role of national "savings" in China in spurring credit origination and debt formation. Many of my colleagues at BCA and the majority of commentators outside BCA argue that China's high "savings" rate, or so-called "excess savings", has been an important contributor to its exponential credit and money growth. Contrary to this narrative, we within BCA's Emerging Markets Strategy team maintain that the dramatic surge in credit and money in China has been the result of speculative behavior by banks and debtors. As such, the boom in money and credit growth has produced large imbalances and excesses, if not outright bubbles (Chart I-1). Chart I-1An Unprecedented Credit ##br##And Money Boom In China Every financial bubble in history has had its justifications. Last decade, the common narrative about U.S. real estate was that nationwide, U.S. house prices had historically never deflated in nominal terms. In the late 1990s, the tech bubble was vindicated by the "new productivity" era. In the meantime, in the 1980s in Japan and the mid-1990s in Hong Kong, sky high property prices were rationalized by limited amounts of land, given that these are islands. Despite these validations, all of these bubbles ultimately burst. We feel that vindicating China's enormous credit, money and property market excesses - which are all interrelated - by the nation's high "savings" is another attempt to endorse overextended and unsustainable macro imbalances. This report is a continuation of our series discussing these issues in great depth.1 The objective of this piece is to illuminate on the confusion between national "savings" and credit / deposits / money. Intuitively, many investors and commentators use the term "savings" to refer to bank deposits. Yet, in macroeconomics, national and household "savings" are not about deposits or money in the banking system at all. The term "savings" in macroeconomics denotes an increase in the economy's capital stock. Therefore, the financial system in general, and banks in particular, cannot and do not lend out or intermediate national or households "savings." The Chinese banking system has enormous amount of deposits because banks have created them "out of nothing" not because households save a lot. In an economy where banks exist, "savings" and financing are very different things. Commercial banks (hereafter referred to as banks) provide financing by expanding their balance sheets - creating deposits "out of thin air" as and when they originate loans. We previously elaborated on this money creation process,2 but given its importance to the topic of this report, we revisit it here. Banks Create New Purchasing Power "Out Of Thin Air" When a bank originates a loan, it simultaneously creates a deposit, or new money. Importantly, this does not represent a transfer of an existing deposit to the new borrower. This is a new deposit - new purchasing power - that did not previously exist (Figure 1). Figure I-1Credit / Money Creation Process The borrower can immediately use this new deposit to purchase goods and services or buy assets. At the same time, all owners of existing deposits at the bank still have their deposits too, and can use them as, when, and how they prefer. Thereby, the bank has created new purchasing power "out of nothing" when it originated a loan. Traditional macroeconomic theory presumes that for a person or company to invest in productive capacity, another person/unit must save. This assumption is true for a barter economy with no banks and money - where some entities produce but do not consume, so that others can acquire their output (goods) and in turn use them as investment. Nevertheless, in an economy with banks, one does not need to save in the form of a deposit in a bank in order for the latter to lend money to another entity. When a bank grants a loan or acquires an asset, it simultaneously creates new deposit/money - which is de facto new purchasing power originated by the bank "out of thin air." We use the terms deposit and money interchangeably because broad money supply is computed as the sum of all deposits in the commercial banks. Let's consider an example of how a bank loan leads to new income creation. A company borrows from a bank to build a bridge, it then pays its suppliers and contractors for their work. As a result, the suppliers and contractors, and consequently their employees and shareholders, earn income. Without this loan, the bridge would not have been built, and the suppliers, their employees and business owners would not have received income. In short, the loan comes first, then the investment - and only after the investment is carried out do employees and business owners earn income. Thereafter, they can consume, acquire assets and save in forms of bank deposits. Critically, this income is realized because the bank originated a loan / new purchasing power "out of nothing." Chart I-2 illustrates that the Chinese banking system has created RMB 140 trillion of broad money/deposits since January 2009. This is equivalent to US$21 trillion at today's exchange rate. This is twice as much as aggregate broad money - equivalent to $10.5 trillion - generated by commercial and central banks in the U.S., the euro area and Japan combined since early 2009 - even amid their respective QE programs. Chart I-2Helicopter Money In China The unprecedented new purchasing power of Chinese companies and households has been primarily due to this enormous balance sheet expansion by mainland commercial banks (Chart I-3). Chart I-3China: Commercial Banks ##br##Assets And Money Multiplier Bank Versus Financial Intermediaries Banks perform a unique function in the economy and financial system. There are considerable differences between a bank lending money or buying assets and a non-bank doing the same. This is unfortunately not reflected in mainstream economic theory and macro models. Unlike banks, non-banks - such as pension funds, insurance companies, households, businesses and all other non-bank entities - do not create new money/new purchasing power when they grant a loan or acquire an asset. The act of lending by non-banks simply constitutes a transfer of an existing deposit from a creditor to a borrower. Banks are not intermediaries of deposits into loans as the Loanable Funds Theory (LFT) alleges. They create deposits themselves by making loans and acquiring assets. The LFT, nonetheless, applies to non-bank lenders - the latter are indeed financial intermediaries, i.e., they channel existing deposits into loans or other assets. The institutional and legal differences that make commercial banks unique and allow them to create money are discussed in detail in "How Do Banks Create Money, and Why Can Other Firms Not Do the Same?," Werner (2014b).3 The theory of fractional banking is not applicable to modern banking as well.4 It is the theory of money creation by banks that we subscribe to and present here that accurately describes the process of money creation. Bottom Line: Banks differ vastly from non-bank financial institutions, and are unique in their ability to create money/new purchasing power by originating loans or acquiring assets. Money Versus Credit Remarkably, there is also an important analytical distinction between credit/leverage and money. New money matters when one is attempting to gauge the (nominal) growth outlook because it represents new purchasing power. New money can only be originated by banks, including the central bank. Central banks can create broad money in circulation (i.e. beyond central bank reserves) when they buy financial assets from or lend to non-bank entities. Doing so creates a deposit in the commercial banking system. By contrast, the degree of credit/leverage is critical when evaluating the risk of financial distress in both the economy and the financial system. Credit can be extended not only by banks but also by non-banks. Hence, lending or buying corporate bonds by non-banks creates leverage/credit but not new money. The banking system is the only one capable of originating new money, and in turn, new purchasing power. In China, the outstanding stock of total non-financial debt (private plus public) is close to the amount of money supply (Chart I-4). Even though non-bank credit growth has risen in importance since 2010, it seems that without banks' money creation, non-bank credit would not have expanded. Chart I-4China: Money Versus Credit/Debt On another note, household propensity to save alters the velocity of money, not the amount of money in the banking system. A decision by a household to spend more rather than save does not change the amount of deposits in the banking system. As an example, a person who gets paid $1000 might spend $800 of her income and decide to save the remaining $200. The amount of deposits in the banking system does not change; $800 will be transferred to another bank account as she pays for her purchases, while the remaining $200 stays in her existing bank account. Hence, there is no change in the amount of deposits and money supply in the banking system in this scenario. On the whole, the amount of deposits, and hence, broad money supply, in any banking system is equal to the cumulative net money creation by banks and the central bank over the course of their history. This has nothing to do with household and national "savings." The Chinese banking system has enormous amount of deposits because banks have created them "out of nothing" not because households save a lot. Interestingly, changes in household propensity to save are reflected not in money supply but in the velocity of money. When households or companies decide to spend their deposits, the velocity of money rises. Conversely, when households and companies decide to save (retain) their deposits, the velocity of money drops. Bottom Line: Money is distinct from credit and leverage. Changes in the propensity to save alter the velocity of money, but not the amount of deposits/money supply in the banking system. True Meaning Of "Savings" In Macroeconomics What is the true meaning of "savings"5 in macroeconomics, given the amount of deposits in the banking system has no bearing on "savings?" The confusion between national "savings" and deposit/money creation is dealt with nicely by Fabian Lindner. Having modelled it, Lindner6 argues: "... the aggregate economy's saving is equal to the newly produced tangible assets and inventories. That total saving is equal to just the increase in tangible assets ... (because) all changes in net financial assets in the economy add up to zero... Thus, for every economic agent increasing her net financial assets, there is a corresponding decrease in net financial assets of all other economic agents in the economy. Put in more general terms: An economic agent can only save financially if other agents dis-save financially by the same amount... That is why in the entire economy (that is the world economy or a closed economy) only the increase in tangible assets, thus investment, is saving (emphasis is added). Thus, saving and investment are equivalent in the aggregate... The equivalence of investment and saving however does not mean - as claimed by LFT - that household saving (or the sum of household and government saving) is equal to total saving and thus to investment. No matter how high one group's financial saving is, the financial dis-saving of the rest of the economy has to be just as high. The only thing remaining is the creation of tangible assets." (Lindner 2015) In another paper,7 Lindner asserts: "Investment is the production of any non-financial asset in an economy and thus is always directly and unambiguously savings: it increases the economy's net worth... The economy as a whole cannot change its net financial wealth since it always equals zero. The aggregate economy can only save in the form of non-financial assets...The only way an economy can save is by increasing its non-financial wealth, i.e., its physical capital stock." (Lindner 2012) Bottom Line: For a country to raise its domestic "savings" rate, it needs to build its capital stock by using domestically produced investment goods and raw materials. Thereby, domestic "savings" have nothing to do with the absolute level or changes in amount of deposits/money in the banking system. China's Great Wall Of "Savings" China has been investing tremendous amounts for many years, and its capital stock has been mushrooming (Chart I-5, top panel). Yet, the incremental capital-to-output ratio (ICOR) has surged and, its inverse, the output-to-capital ratio has plunged since 2010 (Chart I-5, middle and bottom panels). These developments signify deteriorating efficiency in the Chinese economy and worsening capital allocation. They also entail that companies might have difficulties servicing their debt. When its export machine faltered in 2008 due to the Global Financial Crisis, China offset it by boosting its domestic investments. These investments - incremental additions to the nation's capital stock - defined by macroeconomics as domestic "savings"- offset the decline in external "savings." As such, the composition of national "savings" has changed dramatically since 2008: the share of external "savings" (net exports) have declined while the share of domestic "savings" has risen (Chart I-6). Chart I-5China: Capital Stocks Has Surged Chart I-6China: Domestic And External 'Savings' In China, the augmentation of its capital stock and, hence, its domestic "savings," have been largely financed by loans from Chinese banks. This may sound like nonsense, but only because we are using the term "savings" in a way used in macroeconomics. Yet, new purchasing power originated by the banking system is not in and of itself a sufficient condition to generate domestic "savings." The sufficient condition for having high domestic "savings" is the ability to produce domestic capital goods and raw materials that go into investment. If a country does not build its capacity to produce capital goods and raw materials, it would need to rely on imports - in other words it has to acquire foreign "savings" to invest. Encouraging domestic "savings" entails enhancing capacity to produce goods that are used in capital spending like raw materials, chemicals, steel, cement, machinery, and various equipment and instruments. This is what China has done exceptionally well over the past 20 years. The following points illustrate how China achieved very high "savings" and investment rates (Chart I-7): China devalued its currency in January 1994 by 32% and relied on a cheap currency to produce large trade surpluses (Chart I-8). It used the foreign currency proceeds to purchase foreign technologies and equipment to boost its capital stock. Chart I-7Savings And Investment Ratios Chart I-8China: The 1994 Currency ##br##Devaluation Started New Era It also attracted FDI to build its productive capacity both for consumer goods as well as capital goods. FDI inflows surged since China's acceptance into the WTO in 2001. Since 2009, however, China has been relying on new purchasing power created by banks to expand its industrial capacity to produce commodities, raw materials, industrial equipment and machinery. Meanwhile, mainland banks have been originating new loans, and hence deposits/money - new purchasing power - to finance real estate development and infrastructure construction, utilizing these domestically produced raw materials and machinery. This has allowed China to sustain high levels of domestic "savings." On the whole, China indeed has had "excess savings" as its economy has been suffering from excess industrial capacity. Initially, China invested to create such excess capacity. Then, its banking system originated enormous amount of money/new purchasing power to support and keep zombie companies alive in these industries with excess capacity. The banking system is still involved in this function up until today. While this is a reasonable economic policy in the short run, it is not a good growth strategy in the long term. The problem is that easy money and credit support inefficient enterprises and encourage unproductive investment. As a result, productivity growth will slow and potential growth will decelerate considerably. Bottom Line: The countries that produce a lot of goods and services for domestic investment are said to have high domestic "savings." By definition, the more excess industrial capacity a country has, the more "excess savings" that economy will carry. Yet, uncontrolled money/credit origination to support zombie enterprises in over-capacity sectors entails inefficient allocation of capital that necessarily slows productivity growth and hence economic growth potential in the long term. Limits On Money Creation A natural question that arise from all this is what are the limits on money creation? We list some of major ones here, but these issues have been addressed in our previous three reports,8 and we will address them again in forthcoming reports. Inflation and/or deprecation pressures on the currency that could lead to monetary tightening; Bank regulation and various regulatory ratios; Shareholders of banks - who are highly leveraged to non-performing assets/loans - might order reduced lending; Removing the implicit government "put" that encourage irresponsible borrowing and lending. Inflationary pressures are presently rising and more entrenched in China now than at any time in the past decade or so (Chart I-9). In the context of negative real interest rates (Chart I-10) and barring major growth slowdown, the authorities are unlikely to stimulate anytime soon. Chart I-9Beware Of Rising Inflation In China... Chart I-10...Making Interest Rates Negative Negative real local interest rates undermine Chinese households' willingness to hold the currency. China's foreign exchange reserves at $3 trillion, while high, are equal only to 10% broad money (M3) and 14% of official M2. This signifies how much money the banking system has created. At the moment, mainland banking regulations are being tightened. This as well as liquidity tightening by the People's Bank of China and the government's anti-corruption crackdown that is moving into the financial industry will further dampen money creation and leverage expansion. This triple tightening amid lingering money and credit excesses constitutes the main rationale behind our negative stance on China's growth and China-related plays in global financial markets. Policy tightening is especially dangerous amid the existing credit, money and property market imbalances and excesses. Downgrade Chinese Stocks From Overweight To Neutral The Chinese MSCI Investable equity index - which unlike H-shares includes mega-cap tech companies - has rallied massively and outperformed the EM benchmark (Chart I-11). Chart I-11Downgrade Chinese Investable Stocks ##br##From Overweight To Neutral Relative performance is overbought, and we recommend dedicated EM equity portfolios downgrade their allocation from overweight to neutral. Our overweight position was initiated on November 26, 2014, and has generated an 18.5% gain. The freed-up capital should be allocated proportionally to our remaining overweights, which are Taiwan, Thailand, Korean tech stocks, Russia and central Europe. We are contemplating upgrading Chile. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com Andrija Vesic, Research Assistant andrijav@bcaresearch.com 1 Please refer to the Emerging Markets Strategy Special Reports from October 26, 2016, November 23, 2016 and January 18, 2017; available on ems.bcaresearch.com 2 Please refer to the Emerging Markets Strategy Special Report titled "Misconceptions About China's Credit Excesses," dated October 16, 2016, available on available on ems.bcaresearch.com 3 Werner, R. (2014b), "How Do Banks Create Money, and Why Can Other Firms Not Do the Same?", International Review of Financial Analysis, 36, 71-77. 4 Werner, R. (2014a), "Can banks individually create money out of nothing? -- The theories and the empirical evidence", International Review of Financial Analysis, 36, 1-19. 5 We use "savings" in parenthesis because as this term does not really mean households' and companies' and governments' financial assets or deposits at the banks. "Savings" signifies the amount of goods and services produced but not consumed by an economy. 6 Lindner, F. (2015), "Did Scarce Global Savings Finance the US Real Estate Bubble? The Global Saving Glut thesis from a stock flow Consistent Perspective", Macroeconomic Policy Institute, Working Paper 155, July 2015. 7 Lindner, F. (2012), "Savings does not finance Investment: Accounting as an indispensable guide to economic theory", Macroeconomic Policy Institute, Working Paper 100, October 2012. 8 Please refer to the Emerging Markets Strategy Special Reports from October 26, 2016, November 23, 2016 and January 18, 2017; available on ems.bcaresearch.com Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights China stands out as the most likely candidate to send negative shock waves through EM and commodities in 2018. Granted the ongoing policy tightening in China will likely dampen money growth further, the only way mainland nominal GDP growth can hold up is if the velocity of money rises meaningfully, offsetting the drop in money growth. Assigning equal probabilities to various scenarios of velocity of money, the outcome is as follows: one-third probability of robust nominal growth (continuation of the rally in China-related plays) and two-third odds of a non-trivial slowdown in nominal growth with negative ramifications for China-related plays. Hence, we reiterate our negative stance on EM risk assets Feature The key question for emerging markets (EM) in 2018 is whether a slowdown in Chinese money growth will translate into a meaningful growth deceleration in this economy, and in turn produce a reversal in EM risk assets. This week we address the above question in detail elaborating on what could make China's business cycle defy the slowdown in its monetary aggregates and how investors should approach such uncertainty. Before this, we review the status of financial markets going into 2018. Priced To Perfection Or A New Paradigm? Several financial markets are at extremes. Our chart on the history of financial market manias reveals that some parts of technology/new concept stocks may be entering uncharted territory (Chart I-1). Tencent's share price, for instance, has surged 11-fold since January 2010. Chart I-1History Of Financial Markets Manias: They Lasted A Decade This is roughly on par with the prior manias' average 10-year gains. As this chart indicates, the manias of previous decades run wild until the turn of the decade. It is impossible to know whether technology/new concept stocks will peak in 2018 or run for another two years. Regardless whether or not the mania in tech/new concept stocks endures up until 2020, some sort of mean reversion in their share prices is likely next year. This has relevance to EM because the magnitude of the EM equity rally in 2017 has been enormously boosted by four large tech/concept stocks in Asia. Our measure of the cyclically-adjusted P/E (CAPE) ratio for the U.S. market suggests that equity valuations are reaching their 2000 overvaluation levels (Chart I-2, top panel). The difference between our measure and Shiller's measure of CAPE is that Shiller's CAPE is derived by dividing share prices by the 10-year moving average of EPS in real terms (deflated by consumer price inflation). Our measure is calculated by dividing equity prices by the time trend in real EPS (Chart I-2, bottom panel). Our CAPE measure assumes that in the long run, U.S. EPS in real terms will revert to its time trend. Meanwhile, the Shiller CAPE is based on the assumption that real EPS will revert to its 10-year mean. Hence, the assumptions behind our CAPE model are quite reasonable if not preferable to those of Shiller's P/E. Remarkably, the U.S. (Wilshire 5000) market cap-to-GDP ratio is close to its 2000 peak (Chart I-3). With respect to EM equity valuations, the non-financial P/E ratio is at its highest level in the past 15 years (Chart I-4). EM banks have low multiples and seem "cheap" because many of them have not provisioned for NPLs. Hence, their profits and book values are artificially inflated. In short, excluding financials, EM stocks are not cheap at all, neither in absolute terms nor relative to DM bourses. Chart I-2A Perspective On U.S. Equity Valuation Chart I-3The U.S. Market Cap-To-GDP ##br##Ratio Is Close To 2000 Peak Chart I-4EM Non-Financial Equities Are Not Cheap Such elevated DM & EM stock market valuations might be justified by currently low global long-term bond yields. Yet, if and when long-term bond yields rise, multiples will likely shrink. The latter will overpower the profit growth impact on share prices, as multiples are disproportionately and negatively linked to interest rates - especially when interest rates are low - but are proportionately and positively linked to EPS.1 As a result, a small rise in long-term bond yields will lead to a meaningful P/E de-rating. Despite very high equity valuations, U.S. advisors and traders are extremely bullish on American stocks. Their sentiment measures are at all time and 11-year highs, respectively. So are copper traders on red metal prices (Chart I-5). The mirror image of the strong and steady rally in global stocks is record-low implied volatility. The aggregate financial markets' implied volatility index is at a multi-year low (Chart I-6). Finally, yields on junk (high-yield) EM corporate and sovereign bonds are at all-time lows (Chart I-7). They are priced for perfection. Chart I-5Bullish Sentiment On Copper Is Very Elevated Chart I-6Aggregate Global Financial Markets ##br##Implied VOL Is At Record Low Chart I-7EM Junk Bond Yields Are At Record Low Are we in a new paradigm, or are we witnessing financial market extremes that are unsustainable? In regard to the timing, can these dynamics last throughout 2018 or at least the first half of next year, or will they reverse in the coming months? We have less conviction on the durability of the U.S. equity rally, but our bet is that EM risk assets will roll over in absolute terms and begin underperforming their DM peers very soon. What could cause such a reversal in EM risk assets? China stands out as the most likely candidate to send negative shock waves through emerging markets and commodities. China: "Financial Stability" Priority Entails Tighter Policy The Chinese authorities are facing unprecedented challenges: The outstanding value of broad money in China (measured in U.S. dollars) is now larger than the combined U.S. and euro area broad money supply (Chart I-8, top panel). Chart I-8Beware Of Money Excesses In China As a share of its own GDP, broad money in China is much higher compared to any other nation in history (Chart I-8, bottom panel). In brief, there is too much money in China and most of it - $21 trillion out of $29 trillion - has been created by the banking system since early 2009. We maintain that the enormous overhang of money and credit in China represents major excess/imbalances and has nothing to do with the nation's high savings rate.2 Rather, it is an outcome of animal spirits running wild among bankers and borrowers over the past nine years. Easy money often flows into real estate and China has not been an exception. Needless to say, property prices are hyped and expensive relative to household income. Policy tightening amid lingering excesses and imbalances makes us negative on China's growth outlook. In a nutshell, we place more weight on tightening when there are excesses in the system, and downplay the importance of tightening in a healthy system without excesses. Importantly, excessive money creation seems to finally be pushing inflation higher. Consumer price services and core consumer price inflation rates are on a rising trajectory (Chart I-9, top and middle panels). As a result, banks' deposit rates in real terms (deflated by core CPI) have plunged into negative territory for the first time in the past 12 years (Chart I-9, bottom panel). Remarkably, the People's Bank of China's existing $3 trillion of international reserves is sufficient to "back up" only 13% and 11% of official M2 and our measure of M3, respectively (Chart I-10). If Chinese households and companies decide to convert 10-15% of their deposits into foreign currency and the PBoC takes the other side of the trade, its reserves will be exhausted. Chart I-9China: Inflation Is Rising And ##br##Real Deposit Rate Is Negative Chart I-10China: Low Coverage Of ##br##Money Supply By FX Reserves Therefore, reining money and credit expansion is of paramount importance to China's long-term financial and economic stability. "Financial stability" has become the key policy priority. "Financial stability" is policymakers' code word for containing and curbing financial imbalances and bubbles. Having experienced the equity bubble bust in 2015, policymakers are determined to preclude another bubble formation and its subsequent bust. Consequently, the ongoing tightening campaign will not be reversed in the near term unless damage to the economy becomes substantial and visible. By the time the authorities and investors are able to identify such damage in the real economy, China-related plays in financial markets will be down substantially. Chart I-11China: Corporate Bond Yields And Yield Curve Faced with significant excesses in money, leverage and property markets, the Chinese authorities have been tightening - and have reinforced their policy stance following the Party's Congress in October. There is triple tightening currently ongoing in China: 1. Liquidity tightening: Money market rates have climbed, and onshore corporate bond yields are rising (Chart I-11, top panel). Remarkably, the yield curve is flat, pointing to weaker growth ahead (Chart I-11, bottom panel). 2. Regulatory tightening: The China Banking Regulatory Commission (CBRC) is forcing banks to bring off-balance-sheet assets onto their balance sheets, and is reining banks' involvement in shadow banking activities. In addition, financial regulators are trying to remove the government's implicit "put" from the financial system, and thereby curb speculative and irresponsible investment behavior. Finally, many local governments are tightening investors' participation in the real estate market. 3. Anti-corruption campaign is embracing the financial institutions: The powerful anti-corruption commission is planning to dispatch groups of inspectors to examine financial institutions' activities. This could dampen animal spirits among bankers and shadow banking organizations. The Outlook: The "Knowns"... In China, broad money growth has already slumped to an all-time low (Chart I-12). The money as well as the credit plus fiscal spending impulses both point to a considerable slowdown in the mainland's industrial cycle and overall economic activity (Chart I-13). Chart I-12China: Broad Money ##br##Growth Is At All-Time Low Chart I-13China: Money And Credit & ##br##Fiscal Impulses Are Negative The slowdown is not limited to money growth; there are a few real business cycle indicators that are already weakening. For example, the growth rate of property floor space sold and started has slumped to zero (Chart I-14). Electricity output and aggregate freight volume growth have both decisively rolled over (Chart I-15). Chart I-14China: Property Starts Are Set To Contract Again Chart I-15China: A Few Signs Of Slowdown That said, based on the past correlation between money and credit impulses on the one hand and the business cycle on the other, China's economy should have slowed much more, and its negative impact on the rest of the world should have already been felt (Chart I-13, on page 9). This has been the key pillar of our view on EM, but it has not yet transpired. Is it possible that the relationship between money/credit impulses and the business cycle has broken down? If so, why? And how should investors handle such uncertainty? Bottom Line: China's ongoing policy tightening will ensure that money and credit impulses remain negative for some time. Can the country's industrial sectors de-couple from its past tight correlation with money and credit? ...And The "Unknowns" By definition, the only way to sustain nominal economic growth in the face of a decelerating money supply is if the velocity of money increases. This is true for any economy. Nominal GDP = Money Supply x Velocity of Money Provided China's policy tightening will likely further dampen money growth, the only way nominal GDP growth can hold up is if the velocity of money rises meaningfully, offsetting the drop in money growth. This is the main risk to our view and strategy. Chart I-16 portrays all three variables. Chart I-16China: Money, Nominal GDP ##br##And Velocity Of Money Even though the velocity of money has fallen structurally over the past nine years (Chart I-16, bottom panel), it has risen marginally in 2017, allowing the mainland's nominal economic growth to hold up despite a considerable relapse in money supply growth. Notably, this has been the reason why our view has not worked this year. What is the velocity of money, and how can we forecast its fluctuations and, importantly, the magnitude of its variations? The velocity of money is one of the least understood concepts in economic theory. The velocity of money is anything but stable. In our opinion, the velocity of money reflects animal spirits of households and businesses as well as government spending decisions. Forecasting animal spirits and the magnitude of their variations is not very a reliable exercise. In a nutshell, the banking system (commercial banks and the central bank) creates money via expanding its balance sheet - making loans to or acquiring assets from non-banks. However, commercial banks have little direct influence on the velocity of money. The latter is shaped by non-banks' decisions to spend or not (i.e., save). Significantly, non-banks' spending and saving decisions do not alter the amount of money in the system. Yet they directly impact the velocity of money. The banking system creates money, and non-banks churn money (make it circulate). At any level of money supply, a rising number of transactions will boost nominal output, and vice versa. Further, there is a great deal of complexity in the interaction between money supply and its velocity. Both are sometimes independent, i.e. they do not influence one another, but in some other cases one affects the other. For example, with the ongoing triple tightening in China and less money being originated by the banking system, will households and businesses increase or decrease their spending? Our bias is that they will not increase spending. This is especially true for the corporate sector, which has record-high leverage and where access to funding has been tightening. It is also possible that rising velocity will lead to more money creation as more spending leads to higher loan demand and banks accommodate it - i.e., originating more loans/money. These examples corroborate that money supply and the velocity of money are not always independent of each other. On the whole, it is almost impossible to reliably forecast the magnitude of changes in velocity of money. In the same vein, it is difficult to forecast animal spirit dynamics in any economy. Chart I-17U.S.: The Rise In Velocity Of Money ##br##Overwhelmed Slowdown In Money One recent example where nominal GDP has decoupled from broad money growth is the U.S. Chart I-17 demonstrates that in the past 12 months, U.S. nominal GDP growth has firmed up even though broad money (M2) growth has slumped. This decoupling can only be explained by a spike in the velocity of M2. In other words, soaring confidence and animal spirits among U.S. households and businesses have boosted their willingness to spend, even as the banking system has created less money and credit growth has slowed considerably over the past 12 months. Going back to China, how should investors consider such uncertainty in changes in the velocity of money? Investing is about the future, which is inherently uncertain. Hence, an investment process is about assigning probabilities to various scenarios. Provided the velocity of money is impossible to forecast, we assign equal probabilities to each of the following scenarios for China in 2018 (Figure I-1): One-third odds that the velocity of money rises more than the decline in broad money growth, producing robust nominal GDP growth; One-third probability that the velocity of money stays broadly flat - the outcome being meaningful deceleration in nominal GDP growth; A one-third chance that the velocity of money declines - the result being a severe growth slump. Figure I-1How Investors Can Consider Uncertainty Related To Velocity Of Money In short, a positive outcome on China-related plays has a one-third probability of playing out, while a negative outcome carries a two-thirds chance. This is why we continue to maintain our negative view on EM and commodities. Commodities Our view on commodities and commodity plays is by and large shaped by our view on China's capital spending. Given the credit plus fiscal spending impulse is already very weak, the path of least resistance for capital expenditures is down. Besides, the government is clamping down on local governments' off-balance-sheet borrowing and spending (via Local Government Financing Vehicles). A deceleration in capital expenditures in general and construction (both infrastructure and property development) in particular is bearish for industrial metals (Chart I-18). Money and credit impulses herald a major downturn in Chinese imports values and volumes (Chart I-19). Chart I-18Industrial Metals / Copper Are At Risk Chart I-19China Will Be A Drag On Its Suppliers As to China's commodities output reductions, last week we published a Special Report3 on China's "de-capacity" reforms in steel and coal. The report concludes the following: The path of least resistance for steel, coal and iron ore prices is down over the next 12-24 months. China's "de-capacity" reforms in steel and coal will continue into 2018 and 2019, but the scale and pace of "de-capacity" will diminish. Importantly, the mainland's steel and coal output will likely rise going forward as new capacity using more efficient and ecologically friendly technologies come on stream. The capacity swap policy introduced by the authorities has been allowing steel and coal producers to add new capacity in order to replace almost entirely obsolete capacity. The combination of demand slowdown and modest production recovery will weigh on non-oil raw materials. As for oil, the picture is much more complicated. Oil prices have been climbing in reaction to declining OECD inventories as well as on expectations of an extension to oil output cuts into 2018. One essential piece of missing information in the bullish oil narrative is China's oil inventories. In recent years, China has been importing more crude oil than its consumption trend justifies. Specifically, the sum of its net imports and domestic output of crude oil has exceeded the amount of refined processed oil. This difference between the sum of net imports and production of crude oil and processed crude oil constitutes our proxy for the net change of crude oil inventories. Chart I-20 shows that our proxy for mainland crude oil inventories has risen sharply in recent years. This includes both the nation's strategic oil reserves as well as commercial inventories. There is no reliable data on the former. Therefore, it is impossible to estimate the country's commercial crude oil inventories. Chart I-20China: Beware Of High Chinese Oil Inventories Nevertheless, whether crude oil inventories have risen due to a build-up of strategic petroleum reserves or commercial reserves, the fact remains that crude oil inventories in China have surged and appear to be reaching the size of OECD total crude and liquid inventories (Chart I-20). In short, China has been a stabilizing force for the oil market over the past three years by buying more than it consumes. Without such excess purchases from China, oil prices would likely have been much weaker. Going forward, the pace of Chinese purchases of crude oil will likely slow due to several factors: (a) China prefers buying commodities on dips, especially when it is for strategic inventory building. With crude oil prices having rallied to around $60, the authorities might reduce their purchases temporarily, creating an air pocket for prices, and then accelerate their purchases at lower prices; (b) Commercial purchases of oil will likely decelerate due to tighter money/credit, possibly high inventories and a general slowdown in industrial demand for fuel. Bottom Line: Raw materials and oil prices4 are at risk from China and overly bullish investor sentiment. Beyond Commodities The slowdown in China will impact not only commodities but also non-commodity shipments to the mainland (Chart I-21). In fact, 47% of the nation's imports are commodities and raw materials and 45% are industrial/capital goods - i.e., China's imports are heavily exposed to investment expenditures, not consumer spending. This is why money/credit impulses correlate so well with this country's imports. Consistently, China's broad money (M3) impulse leads EM corporate profit growth by 12 months - and currently heralds a major EPS downtrend (Chart I-22). In addition, aggregate EM narrow money (M1) growth also points to a material slump in EM EPS (Chart I-23). Chart I-21China Is A Risk To ##br##Non-Commodity Economies Too Chart I-22Downside Risk To EM EPS The only EM countries that are not materially exposed to China and commodities are Turkey and India. The former is a basket case on its own. Indian stocks are expensive and will have a difficult time rallying in absolute terms when the EM equity benchmark relapses. As for Korea and Taiwan, their largest export destination is not advanced economies but China. China accounts for 25% of Korea's exports and 28% of Taiwan's. This compares to a combined 22% of total Korean exports and 20% of total Taiwanese exports going to the U.S. and EU combined Can robust growth in the U.S. and EU derail the growth slowdown in China when capital spending slows? This is very unlikely, in our view. Chart I-24 portends that China's shipments to the U.S. and EU account for only 6.6% of Chinese GDP, while capital spending and credit origination constitute 45% and 25% of GDP, respectively. Chart I-23EM M1 And EM EPS Chart I-24What Drives Chinese Growth? A final word on tech stocks. EM's four large-cap tech stocks (Tencent, Ali-Baba, Samsung and TSMC) have gone exponential and are extremely overbought. At this juncture, any strong opinion on tech stocks is not warranted because they can sell off or continue advancing for no fundamental reason. We have been recommending an overweight position in tech stocks, and continue recommending overweighting them, especially Korean and Taiwanese semiconductor companies. As for Tencent and Alibaba, these are concept stocks, and as a top-down house we have little expertise to judge whether or not they are expensive. These are bottom-up calls. Investment Strategy EM Stocks: Asset allocators should continue to underweight EM versus DM, and absolute-return investors should stay put. Our overweights are Taiwan, China, Korean tech stocks, Thailand, Russia and central Europe. Our underweights are Turkey, South Africa, Brazil, Peru and Malaysia. Chart I-25EM Currencies: A Canary In ##br##Coal Mine For EM Credit? Stay short a basket of the following EM currencies: ZAR, TRY, BRL, IDR and MYR. We are also shorting the COP and CLP. Unlike in 2014-2015, EM currencies will depreciate not only versus the U.S. dollar but also the euro. For traders who prefer a market neutral currency portfolio, our recommended longs (or our currency overweights) are TWD, THB, SGD, ARS, RUB, PLN and CZK. INR and CNH will also outperform other EM currencies. Continue underweighting EM sovereign and corporate credit relative to U.S. investment grade bonds. The mix of weaker EM/China growth, lower commodities prices and EM currency depreciation bode ill for already very tight EM credit spreads (Chart I-25). Within the sovereign credit space, our underweights are Brazil, Venezuela, South Africa and Malaysia and our overweights are Russia, Argentina and low beta defensive credits. The main risk to EM local currency bonds is EM currency depreciation. With foreign ownership of EM domestic bonds at all-time highs, exchange rate depreciation could trigger non-trivial selling pressure. Among local currency bond markets, the most vulnerable are Turkey, South Africa, Indonesia and Malaysia. The least vulnerable are Korea, Russia, China, India, Argentina and Central Europe. Other high-conviction market-neutral recommendations: Long U.S. banks / short EM banks. Long U.S. homebuilders / short Chinese property developers. Long the Russian ruble / short oil. Long the Chilean peso / short copper. Long Big Five state-owned Chinese banks / short small- and medium-sized banks. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 For example, given that interest rates are in the denominator of the Gordon Growth model, a one percentage point change in interest rates from a low level can have a significant impact on the fair value P/E ratio. 2 Please refer to the Emerging Markets Strategy Special Reports from October 26, 2016, November 23, 2016 and January 18, 2017; available on ems.bcaresearch.com 3 Please refer to the Emerging Markets Strategy Special Report titled "China's 'De-Capacity' Reforms: Where Steel & Coal Prices Are Headed," dated November 22, 2017, link available on page 22. 4 This is the Emerging Markets Strategy team's view and is different from BCA's house view on commodities. Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights The current mini-upswing in the global mini-cycle started in May and is likely to end around January. On a 6-month horizon, lean against the rally in industrial metals. Equity investors should underweight Basic Resources, and especially Industrial Metals and Mining. The contrasting economic fortunes of Spain and Italy may switch. The peak bank credit impulse for Spain is almost certainly behind it, while for Italy it likely lies ahead. On this hope, we will dip our toes into a small pair-trade: long Italian BTPs versus French OATs. Feature Key to the medium-term behaviour of markets is the existence of what we call 'mini-cycles' in global activity. The evolution of these perpetual mini-cycles explains much of what has happened, what is happening, and what will happen, to financial markets both in Europe and more broadly. Chart of the WeekExpect A Trend-Reversal In The Metals Market Mini-cycles are not a hypothesis. They are an indisputable empirical fact. Just look at the global bond yield (Chart I-2), metal price inflation (Chart I-3), global inflation (Chart I-4), and the bank credit impulse (Chart I-5 and Chart I-6). The regular mini-cycles shout out at you! Furthermore, given that these clearly observed mini-cycles show the same half-cycle length of about 8 months, Investment Reductionism strongly suggests that there is a common over-arching driver. Chart I-2The Global Bond Yield Exhibits Mini-Cycles Chart I-3Metal Price Inflation Exhibits Mini-Cycles Chart I-4Inflation Exhibits Mini-Cycles Chart I-5The Global Credit Impulse Exhibits Mini-Cycles Chart I-6Individual Credit Impulses Exhibit Mini-Cycles Explaining Mini-Cycles Previously,1 we explained that the distinct mini-cycles are interconnected parts of the same never-ending feedback loop. A lower bond yield accelerates bank credit flows... which boosts economic growth... which pushes up commodity inflation and overall inflation... causing the bond market to raise the bond yield, at which point the cycle reverses. And then the alternate cycles repeat ad perpetuam (see Box I-1). Box I-1The Mathematics Of Mini-Cycles One common question we get is: why focus on bank credit analysis and not on bond-intermediated credit analysis too? The simple answer is that bank credit expands the broad money supply whereas bond-intermediated credit usually does not. When a bank issues a new loan, fractional reserve banking allows it to create money 'out of thin air'. In contrast, when a company or government issues a new bond, no new money is created, unless the primary issue is financed by the central bank - which is generally forbidden. Usually, when a bond is issued, existing money just moves from one account - that of the bond buyer - to another account - that of the bond issuer. This means that bond-intermediated credit cannot increase demand by creating new money, but only by increasing the velocity of existing money. Whereas bank credit can increase demand by increasing both the amount of money and its velocity. Therefore, changes in bank credit are the much bigger driver of the mini-cycle in economic activity. If a bank issues 100 euros of credit today, then we know that this new money will be spent in the coming days and weeks - because nobody borrows money just to sit on it. If, in the previous period, the bank had issued 90 euros which was spent, it means that economic activity in the coming period will grow by 10 euros. But if the bank had previously issued 110 euros, it means that economic activity in the coming period will contract by 10 euros. In this way, the cycles in credit and activity are interconnected. Mini-upswings in the credit impulse mini-cycle tend to signal mini-upswings in commodity inflation (Chart I-7), overall inflation and bond yields. So if we can identify turning points in the credit impulse then we can correctly position the cyclical stance of our investment strategy. Chart I-7The Same Mini-Cycle: The Global Credit Impulse And Metal Price Inflation The problem is that the bank credit data is slow to come out. For example, although we are in the middle of November, the last bank credit data for the euro area refers to September. This means that if the mini-cycle is turning now, we might not find out until January. Nevertheless, we can still use the mini-cycle framework. We know that the current mini-upswing started in May and that mini-upswings have an average length of 8 months. Hence, we can infer that the mini-upswing is likely to end around January. That said, upswing lengths do have some degree of variation: the current upswing might be longer or shorter than the average. How to avoid being too early or too late? Combining Mini-Cycles With Fractal Analysis To optimise our proprietary mini-cycle framework, we propose combining it with our proprietary fractal analysis framework. As regular readers know, fractal analysis measures whether herding in a specific investment has become excessive, signalling the end of its price trend. The combined mini-cycle and fractal framework works best if we use a 130-day herding indicator (fractal dimension), as it broadly aligns with the mini half-cycle length. Excessive herding signals that an investment's trend is approaching exhaustion because the liquidity that has fuelled the trend is about to evaporate. Liquidity is plentiful when the market is split between different herds - say, short-term momentum traders and long-term value investors. This is because the herds disagree with each other. If the price fluctuates up, the momentum trader wants to buy while the value investor wants to sell; and vice-versa. So the herds trade with each other with plentiful liquidity. But liquidity starts to evaporate when too many value investors join the momentum herd. Instead of dispassionately investing on the basis of value, value investors get sucked into chasing a price trend, and their buy orders add fuel to the trend. The tipping point comes when all the value investors have joined the momentum herd. If a value investor then suddenly reverts to type and puts in a sell order, he will find that there are no buyers left. Liquidity has evaporated, and finding new liquidity might require a substantial reversal in the price to attract a buy order from an ultra-long-term deep value investor. Earlier this year, our combined frameworks signalled that the aggressive rise in bond yields was likely to reverse (Chart I-8). Therefore, on February 2 we correctly advised: "Lean against the rise in bond yields and bank equities." Chart I-8Excessive Herding In Bonds Always Signals A Trend Reversal Today, we see the same dynamic in parts of the commodity rally - and specifically the move in the LME Index (Chart of the Week). Hence, on a 6-month horizon, lean against the rally in industrial metals. Equity investors should underweight Basic Resources, and especially Industrial Metals and Mining. Could Italy Be A Good Surprise? Returning to the concept of the bank credit cycle, the evolution of longer-term impulses also explains the contrasting recent fortunes of Spain and Italy. In 2013, Spain recapitalized its banking system and ring-fenced bad assets within a 'bad bank'. In effect, it finally did what other economies - most notably the U.S., U.K. and Ireland - had done several years earlier in response to their own housing-related banking crises. As Spanish banks' aggressive deleveraging ended, the bank credit impulse rebounded very sharply and has remained positive for several years. This undoubtedly explains why Spanish real GDP has grown by 13% since mid-2013 (Chart I-9). In contrast, Italy's banking system remained dysfunctional - which meant that its own credit impulse stayed much more muted and barely positive over the past four years (Chart I-10). But now, the Italian banking system is slowly recuperating. Italian banks' equity capital is rising, their solvency is improving, and the share of non-performing loans has fallen sharply this year. Chart I-9Spain's Peak Credit Impulse##br## Is Probably Behind It Chart I-10Italy's Peak Credit Impulse##br## Is Likely Ahead Of It So the contrasting economic fortunes of Spain and Italy may switch. The peak bank credit impulse for Spain is almost certainly behind it, while for Italy it likely lies ahead. On this hope, we will dip our toes into a small pair-trade: long Italian BTPs versus French OATs. Dhaval Joshi, Senior Vice President Chief European Investment Strategist dhaval@bcaresearch.com 1 Please see the European Investment Strategy Weekly Report 'Credit Slumps While Animal Spirits Soar. Why?' March 30, 2017 available at eis.bcaresearch.com Fractal Trading Model* There are no new trades this week, leaving us with six open positions. For any investment, excessive trend following and groupthink can reach a natural point of instability, at which point the established trend is highly likely to break down with or without an external catalyst. An early warning sign is the investment's fractal dimension approaching its natural lower bound. Encouragingly, this trigger has consistently identified countertrend moves of various magnitudes across all asset classes. Chart I-11 The post-June 9, 2016 fractal trading model rules are: When the fractal dimension approaches the lower limit after an investment has been in an established trend it is a potential trigger for a liquidity-triggered trend reversal. Therefore, open a countertrend position. The profit target is a one-third reversal of the preceding 13-week move. Apply a symmetrical stop-loss. Close the position at the profit target or stop-loss. Otherwise close the position after 13 weeks. Use the position size multiple to control risk. The position size will be smaller for more risky positions. * For more details please see the European Investment Strategy Special Report "Fractals, Liquidity & A Trading Model," dated December 11, 2014, available at eis.bcaresearch.com Fractal Trading Model Recommendations Equities Bond & Interest Rates Currency & Other Positions Closed Fractal Trades Trades Closed Trades Asset Performance Currency & Bond Equity Sector Country Equity Indicators Bond Yields Chart II-1Indicators To Watch - Bond Yields Chart II-2Indicators To Watch - Bond Yields Chart II-3Indicators To Watch - Bond Yields Chart II-4Indicators To Watch - Bond Yields Interest Rate Chart II-5Indicators To Watch -##br## Interest Rate Expectations Chart II-6Indicators To Watch - ##br##Interest Rate Expectations Chart II-7Indicators To Watch -##br## Interest Rate Expectations Chart II-8Indicators To Watch -##br## Interest Rate Expectations

Highlights The private sectors in Brazil, Russia and India have indeed experienced some deleveraging. Yet in China, deleveraging has not even begun. In fact, the money and credit excesses have become ever larger in the past two years. China's broad money (M3) is as large as the entire outstanding stock of broad money in the U.S. and euro area banking systems combined. In China, the triple tightening - higher corporate bond yields and money market rates, ongoing tightening by banking regulators and the anti-corruption campaign - will lead to lessened money and credit origination. That in turn will weigh on mainland capital spending and growth in general. Chart I-1Some Deleveraging In Brazil, Russia, And India Feature A judgment on the sustainability of the rally in EM/China-related risk assets, from a big picture perspective, should include whether deleveraging in these economies is in late stages - i.e., whether credit and debt excesses accumulated following the 2008 global financial crisis have been unwound, at least partially. The objective of this week's note is to provide an update on the status of deleveraging within EM/China. Herein, deleveraging is defined as a falling debt-to-GDP ratio. The private sectors within Brazil, Russia and India have indeed experienced some deleveraging, with their private sector debt-to-GDP ratio either falling or moving sideways (Chart I-1). However, in China, deleveraging has not yet even begun (Chart I-2). Excluding Korea, Taiwan, and the BRIC economies, the rest of EM has not seen much deleveraging either (Chart I-3) - we exclude Korea and Taiwan because their equity markets are contingent on global demand rather than domestic dynamics. Note that this debt-to-GDP aggregate is weighted by each country's respective market cap in the MSCI EM stock index. The latest stabilization in this ratio might be due to these countries' currency appreciation, which has reduced their foreign currency debt burden relative to GDP. While deleveraging in many individual developing economies will not affect the rest of the world, deleveraging in China will have an impact on global trade in general and EM economies in particular. This remains one of the most important reasons why we believe the current recovery in EM growth will not be sustained. Chart I-2Deleveraging Has Not Started Yet In China... Chart I-3...Nor In The Rest Of EM Some investors and commentators have remarked that in the U.S., the euro area and Japan, there was no deleveraging following their respective credit bubbles and crises. As such, they argue that there is no compelling reason to expect deleveraging in EM/China. The point about the lack of deleveraging in Japan, the U.S. and Europe following their credit bubbles is only true when one includes public debt (Chart I-4). Yet, their private sectors did deleverage as can be seen in Chart I-5. Chart I-4DM: Deleveraging Concealed By ##br##Acceleration In Public Credit Chart I-5Private Sector Deleveraged ##br##Meaningfully In DM In the U.S. and euro area, deleveraging lasted an average of about seven years. As to Japan - which had a larger credit bubble but never experienced an acute credit crisis - private sector deleveraging endured over more than 21 years (Chart I-5, bottom panel). Did deleveraging in these DM economies involve outright nominal contraction in private credit and bank loans, or only decline in private debt-to-GDP ratio? Both bank loans and private credit nominal growth contracted, as demonstrated in Chart I-6. In short, despite massive policy support - i.e. monetary and fiscal easing and various bank recapitalization programs - private credit growth shrunk in nominal terms in the U.S. and euro area for a couple of years, and for many more years in Japan. China An update on China's debt burden is in order: Despite the vast local government financing vehicle (LGFV) debt swap of about RMB13 trillion conducted over the past two years the corporate debt-to-GDP ratio has not dropped (Chart I-7, top panel). Chart I-6DM: Bank Loans & Private Sector Credit ##br##Contracted In Nominal Terms Chart I-7China's Breakdown ##br##Of Debt By Sector The corporate debt-to-GDP ratio has stopped rising because LGFV debt - which belonged to SOEs and was classified as corporate debt - has been converted into provincial government debt. Since the onset of the Chinese equity market crash in the summer of 2015, our measure of broad money (M3) has expanded by RMB38 trillion ($6 trillion). Similarly, total social financing excluding equity issuance and including local government debt issuance - our so-called TSF+ measure - has surged by RMB49 trillion ($7.4 trillion). In terms of annual growth rates, M3 and TSF+ are still expanding at 10% and 14%, respectively. Chart I-8China's Money Impulse Points ##br##To Growth Deceleration We do not expect China's credit growth to contract in nominal terms, but we do expect credit/money growth to slow further. If and when this occurs, the money and credit impulses - the second derivatives - will become negative. The growth rates of GDP, industrial production, capital spending, profits and imports are impacted by the second derivatives of money and credit, which have been declining. In fact, the M3 impulse is already negative, which is consistent with deceleration in China's business cycle (Chart I-8). Some commentators and strategists have argued that debt should be compared with debtors' assets not GDP. This is a very weak argument. The sustainability of debt is contingent on borrowers' ability to service it. In turn, the ability to service debt is determined by debtors' cash flow generation which can be measured / approximated by nominal GDP. This is why the debt-to-nominal GDP ratio is the best metric for debt sustainability on a macro scale. It factually measures debt relative to corporate nominal revenues and household income. What about assets? Just because a company has assets does not mean it can service its debt. Note that in China, debt sustainability concerns are primarily around companies not households or government. First, if a company's assets do not generate sufficient cash flow to service debt, the value of these assets will be low. Second, asset valuations in EM state-controlled companies in general and among Chinese SOEs in particular, where most of the debt is concentrated, cannot be taken at face value. When evaluating the creditworthiness of a debtor, should investors rely on the accounting value of buildings that a debtor owns, or on the cash flow that these assets generate? We believe the latter is a much more prudent approach to investment analysis than the former. Third, if assets indeed need to be liquidated to service debt across many debtors, the situation is already very dire. Finally, we acknowledge that the Chinese government has a lot of fiscal room to bail out corporate debtors. When the authorities do so and overall corporate debt declines, we will seriously contemplate changing our view and investment strategy. So far, corporate indebtedness has not declined. For all of the above reasons, the debt-to-nominal GDP ratio is a much more reasonable measure than the debt-to-assets ratio. To be clear, we are not suggesting that Chinese companies are heading into a massive default and liquidation cycle. Our key premise as it relates to China's debt burden is as follows: overleveraged companies that could potentially struggle to service their debt are unlikely to embark on major capital spending initiatives. And in fact, their creditors should not lend to these debtors. As a result, capital spending will slow, weighing on commodities and other related areas. Conclusions The credit and money excesses in China and EM have been increasingly getting larger. Not only does China have too much corporate debt, but its stock of outstanding broad money is very high compared to any other economy in the world (Chart I-9). Chart I-9China's 'Money Bubble' Money is created by banks "out of thin air" (subject to regulatory capital ratios and other constraints) when they lend or buy assets from non-bank entities. Banks' ability to originate money does not relate to or depend on consumers or national savings. We have explored these issues in detail in Trilogy of reports in the past.1 Chart I-10China: Beware Of Rising Inflation Chart I-9 illustrates that China's official broad money (M2) is equivalent to $25 trillion while our measure of broad money (M3) is equivalent to about $29 trillion. This compares with broad money of $14 trillion in each of the U.S. and the euro area. Hence, China's broad money (M3) is as large as the U.S. and euro area's aggregate broad money combined. Furthermore, as of January 1, 2009, China's M2 and M3 were only $7.3 trillion and $8 trillion, respectively. This entails that the Chinese banking system has increased the broad money supply by the equivalent of $18-21 trillion. The triple tightening - higher corporate bond yields and money market rates, ongoing tightening by banking regulators and the anti-corruption campaign that is moving into the financial system - will lead to lessened money and credit origination. This will weigh on capital spending and growth in general. The odds are that tightening will escalate. First, after the party Congress, President Xi has consolidated power and can now enact meaningful structural reforms. Second, as we documented several weeks ago, core consumer inflation is rising (Chart I-10). Producer prices inflation is holding up around 7%. This is not surprising, given the amount of money that has been created in the economy in the past two years. Even marginal policy tightening amid lingering credit excesses is very dangerous. Yet a moderate slowdown in credit growth can translate into a notable drop in the credit impulse, weighing on growth as a result. This is especially true for capital spending and construction and is one of the primary reasons why we maintain a negative stance on China-related and EM risk assets. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 Please refer to the Emerging Markets Strategy Special Reports from October 26, 2016, November 23, 2016 and January 18, 2017; available on ems.bcaresearch.com Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Special Report

Highlights Three factors point to stable or narrower USD cross-currency basis swap spreads: the improving health of global banks, the end of the adjustment to the regulatory change affecting prime-money market funds, and the relaxation to the Supplementary Leverage Ratio rules by the U.S. Treasury. Four factors point to wider basis swap spreads: BCA's forecast that U.S. loan growth will pick up, our view on U.S. inflation, the coming decline in the Federal Reserve's balance sheet, and the potential for U.S. repatriation. We expect USD basis swap spreads to widen again, which suggests increasing FX vol. This would hurt carry trades, EM currencies and dollar bloc currencies. Feature The rather arcane topic of cross-currency basis swap spreads has periodically surfaced in the news in the past few years. The widening in cross-currency basis swap spreads has been highlighted as one of the key factors explaining why covered interest rate parity relationships (the link between the price of FX forward, spot prices and interest rate differentials) have not held as closely after the Great Financial Crisis (GFC) as before. The widening of cross-currency basis swap spreads has also been highlighted as a factor behind the strength in the U.S. dollar in 2014 and 2015. Similarly, the recent narrowing in the cross-currency basis swap spread has been highlighted as a factor behind the weakness in the USD this year. This week we delve a little deeper into what cross-currency basis swap spread measures, and what some of its major determinants are. We ultimately expect the USD cross-currency basis swap spread to widen again, which should contribute to a stronger dollar and increased global FX volatility. What Is A Cross-Currency Basis Swap? To examine what drives cross-currency basis swap spreads, one first needs to understand what these instruments are. Let's begin with a regular FX swap. An FX swap in EUR/USD is a contract through which two counterparties agree to exchange EURs for USDs today, with a reversal of that exchange at the maturity of the contract - a reversal set at a predetermined exchange rate simply equal to the forward value of the EUR/USD. So, if counterparty A lends X million EURs to counterparty B, the former receives in U.S. dollars the equivalent of X million EURs times the prevalent EUR/USD spot rate from counterparty B today. The transaction does not end there. Simultaneously, the FX swap forces B to give back the X million EURs to counterparty A at maturity, while counterparty A gives back X million EUR times the EUR/USD forward rate in U.S. dollars to counterparty B. This forward rate is the rate prevalent when the contract was agreed upon. The transactions are illustrated in the top panel of Chart 1. Chart 1FX Swaps Vs. Cross Currency Basis Swaps The problem with regular FX swaps is that they offer little liquidity at extended maturities. If market players want to hedge long-term liabilities and assets, they tend to do so using a cross-currency basis swap, where much more liquidity is available at long maturities. A EUR/USD cross currency basis swap begins in the same way as a regular FX swap: counterparty A lends X million EURs to counterparty B, and the former receives in U.S. dollars the equivalent of X million EURs times the prevalent EUR/USD spot rate from counterparty B today. However, this is where the similarities end. A cross-currency basis swap has exchanges of cash flows through its term. Counterparty B, which provided USDs to counterparty A, receives 3-month USD Libor, while counterparty A, which provided EURs to counterparty B, received 3-month EUR Libor + a (alpha being the cross-currency basis swap spread). At the maturity of the contract, counterparty A and B both receive their regular intermediary cash flows, and also re-exchange their respective principal - but this time at the same spot rate as the one that existed at the entry of the contract (Chart 1, bottom panel). Chart 2A Bigger Funding Gap Equals##BR##A Wider Basis Swap Spread In both regular FX and cross-currency basis swaps, counterparties have removed their FX risks, except that in the latter, the interest differentials have been paid during the life of the contract instead of being factored through the forward premium/discount. This is fine and dandy, but it leaves a unexplained. The cross currency basis swap spread (a), is a direct function of the relative supply and demand for each currency. If investors demand a lot of EUR in the swap market relative to its supply, a will be positive. If they demand more USDs, a will be negative. A good example of this dynamic is the funding gap of banks. Let's take the Japanese example. Japanese banks have a surplus of domestic deposits (thanks to the massive savings of the Japanese corporate sector) relative to their yen lending. As a result, they have large dollar lending operations. To hedge their dollar assets, Japanese banks borrow USD in large quantities in the cross-currency swap market. This tends to result in a negative swap spread in the yen (Chart 2). This is particularly true if both the banking sector and the other actors in the economy (institutional investors and non-financial firms) also borrow dollars in the swap market to hedge dollar assets, which is the case in Japan (Chart 3). Chart 3Japanese Investors Are Accumulating Assets Abroad Additionally, if there are perceived solvency risks in the European banking sector, this should further weigh on the cross-currency basis swap spread, pushing it deeper into negative territory, as the viability of the main EUR counterparties becomes at risk. The same dance is true for any currency pair. The other factor that affects USD cross-currency basis swap spreads is the supply of U.S. dollars, especially the room on large banks' balance sheets to service these markets. The cross-currency basis swap spread could be close to zero if large arbitrageurs take offsetting positions to arbitrage the spread away, doing so until the spread disappears. However, with the imposition of Basel III and Dodd-Franks, banks have been constrained in their capacity to do this. Indeed, increased leverage ratio requirements (now banks need to post more capital behind repo transactions as well as collateralized lending and other derivatives) mean that arbitraging cross-currency basis swap spreads and deviations from covered interest rate parity has become much more expensive. Furthermore, the increase in Tier 1 capital ratios associated with these regulations has forced banks to de-lever; however, engaging in arbitrage activities still requires plenty of leverage (Chart 4). Chart 4The Structural Gap In The Basis Swap##BR##Spread Reflects Regulation Economic Factors Driving The Spread The factors that we look at essentially relate to the supply of USD available for lending in offshore markets, as well as determinants of relative counterparty risks between the U.S. and the rest of the world. Factors Arguing For Narrower Cross-Currency Basis Swap Spreads 1. Global Banks Health Chart 5Banks Perceived Health##BR##Determines Basis Swap Spreads The price-to-book ratio of global banks outside the U.S. has been largely correlated with USD cross-currency swap spreads. When global banks get de-rated, spreads widen, and it becomes more expensive to hedge USD positions in the swap market (Chart 5). This is because as investors perceive the solvency of global banks deteriorating, they impose a penalty as the Herstatt risk increases. Additionally, solvency problems can force banks to scramble to access USD funding, prompting deeper spreads. BCA is positive on global financials and sees continued improvement in European NPLs. This means that solvency risk concerns are likely to remain on the backburner for now, pointing to narrower basis swap spreads. 2. Supplementary Leverage Ratio Changes In June, the U.S. Treasury announced a relaxation of some of its rules on supplementary leverage ratios, lowering the amount of capital required to support activity in the repo market behind initial margins for centrally cleared derivatives, and behind holdings of Treasurys. This means that commercial banks in the U.S. can have bigger balance sheets and more room to engage in arbitrage activity, implying a greater supply of dollars in the USD cross-currency basis swap market. In response to last June's proposal, basis swap spreads narrowed by 11 basis points. BCA believes these changes will continue to support dollar liquidity, and will further help in narrowing cross-currency basis swap spreads. 3. Prime Money-Market Funds Debacle Is Over Chart 6More Expensive Bank Funding##BR##= Wider Basis Swap Spreads In October 2016, regulatory changes were implemented that allowed prime money market funds to have fluctuating net asset values. Obviously, this meant that prime money-market funds would be not-so-prime anymore. As a result, to remain the ultra-safe vehicles that they once were, prime money-market funds de-risked. As a result, they cut their exposure to risky activities in anticipation of these changes. In practice, a key source of short-term funding for banks evaporated from the market, putting upward pressure on bank financing costs. As the LIBOR-OIS spread increased, so did basis-swap spreads (Chart 6): as it became more expensive for banks to finance themselves, they had to curtail the supply of USDs provided to the swap market, an activity normally requiring intense demand on banks' balance sheets. This adjustment is now over, suggesting limited potential widening in USD basis swap spreads. Factors Arguing For Wider Cross-Currency Basis Swap Spreads 1. U.S. Loan Growth When U.S. banks increase their loan formation activity, USD cross-currency basis swap spreads widen (Chart 7). As banks increase their extension of credit through loans, they decrease the amount of securities they hold on their balance sheets (Chart 8). This means there is less supply of liquidity available for balance sheet activities, particularly providing dollar funding in the offshore market. In the Basel III / Dodd-Frank world, less-liquid bank balance sheets are synonymous with wider USD basis-swap spreads. As we argued last week, increasing U.S. capex, easing lending standards for firms and rising household income levels should result in increasing loan growth in the U.S. which will result in lower abundance of liquid assets and a widening basis swap spreads.1 Chart 7More Bank Loans Lead##BR##To Wider Swap Spreads Chart 8More Debt Equals Less##BR##Securities In Bank Credit 2. U.S. Inflation There is a fairly close relationship between U.S. inflation and the USD basis swap spread, where a higher core CPI tends to lead to a wider spread (Chart 9). The fall in U.S. inflation this year likely contributed to the narrowing in basis swap spreads. Our take on this is that as inflation falls, it gives an incentive for banks to hold low-yielding liquidity on their balance sheets as real returns on cash improve. This fuels a gigantic carry trade through the basis-swap market. We expect inflation to pick up meaningfully by mid-2018, which should widen cross-currency basis swap spreads.2 Chart 9When U.S. Inflation Increases, Swap Spreads Widen 3. Central Bank Balance Sheets When the Federal Reserve increases the size of its balance sheet relative to other balance sheets, this tends to lead to a narrowing of the USD basis swap spread as the global supply of dollars relative to other currencies increases. The opposite is also true. This relationship did not work after late 2016 (Chart 10). However, during that episode, as the change in prime money-market funds caused a dislocation in banks' funding, commercial banks exhibited cautious behavior and increased their reserves with the Fed. As Chart 11 illustrates, there is a tight relationship between the change in commercial banks' reserves held at the Fed and cross-currency basis swap spreads. Going forward, as the Fed lets it balance sheet run off, we expect to see a decrease in commercial banks' excess reserves. This could contribute to upward movement in the basis swap spread. Chart 10Smaller Fed Balance Sheet Leads##BR##To Wider Basis Swap Spreads Chart 11Fed Runoff Could Widen##BR##Basis Swap Spreads 4. U.S. Repatriations Chart 12U.s. Repatriations Support Wider##BR##Basis Swap Spreads The most revealing relationship unearthed in our study was that when U.S. entities repatriate funds at home, this tends to put strong widening pressure on the USD cross-currency basis swap spread (Chart 12). U.S. businesses hold large cash piles abroad - by some estimates more than US$2.5 trillion. However, most of these funds are held in highly liquid, high-quality U.S.-dollar assets offshore. These assets are perfect collaterals for various transactions in the interbank market. The funds held abroad by U.S. firms are a source of supply for U.S. dollars in the offshore markets. When U.S. entities bring assets back home, the widening in the basis swap spread essentially reflects a decline in the supply of USD in offshore markets, and vice versa when Americans export capital abroad. BCA's base case is that tax cuts are likely to hit the U.S. economy in 2018, even if the growing feud between Trump and the establishment Republican party members is a growing risk. BCA still views a tax repatriation as a higher-likelihood event, as it is the easiest way for the U.S. government to bring funds into its coffers. The 2004 tax repatriation under former President George W. Bush did result in substantial fund repatriation in the U.S. This time will not be different. We expect any such tax repatriation to cause a potentially large deficit of supply in the USD offshore markets, which could create a strong widening basis on the cross-currency basis swap spread in favor of the dollar. Bottom Line: Three factors argue for USD cross-currency basis swap spreads to stay at current levels, or even narrow further. These factors are the health of global banks, the easing in U.S. supplementary leverage ratios and the end of the adjustment of U.S. bank funding to new regulations affecting prime money-market funds. On the other hand four factors points to wider USD cross-currency basis swap spreads: BCA's positive outlook for U.S. credit growth; BCA's positive outlook on U.S. inflation; the run-off of the Fed's balance sheet; and the potential for U.S. entities repatriating funds from abroad. Potential Direction And Investment Implications We anticipate USD cross-currency basis swap spreads to widen over the coming 12 months. We think the easing in the Supplementary Leverage Ratios rules by the U.S. Treasury is the most important factor pointing to narrower USD cross-currency basis swap spreads. However, Basel III rules and most of Dodd-Frank are still in place, which suggest there remains large constraints on the balance-sheet activities of global banks, which will limit the potential for a narrowing of the USD basis swap spread as U.S. banks will remain constrained in their ability to supply U.S. dollars in the offshore market. Chart 13Wider Basis Swap Spreads Equals Higher Vol On the other hand many factors support wider USD cross-currency basis swap spreads, most important of which is the potential for more credit growth. This is in our view a very strong force as it requires banks to ration the use of their balance sheets, limiting their activity in the offshore market. Moreover, we do foresee a high probability of tax repatriation, which would put strong widening pressure on the swap spreads. In terms of implications, wider USD basis swap spreads tend to be associated with rising FX vols (Chart 13). As we highlighted in a Special Report last year, higher FX vols are poison for carry trades.3 As such, we think that widening swap spreads could spur a period of trouble for traditional carry currencies. This means EM and dollar-block currencies are likely to suffer in this environment. Additionally, in China, Xi Jinping is consolidating power and has taken control of the Politburo. This implies he now has more room to implement reforms. Removal of growth targets after 2020, removal of growth as a criterion for grading local officials, a focus on balanced growth, and a focus on combatting pollution all suggest that Chinese growth is unlikely to follow the same debt-fueled, capex-led model.4 This will weigh on Chinese imports of raw materials, and hurt export volumes and prices for many EM countries and commodities producers. This means these policies represent a headwind for many carry currencies. Moreover, historically, wider USD funding costs have been associated with a stronger dollar, as it makes it more expensive to hedge dollar assets. Thus, in an environment where U.S. interest rates are rising relative to the rest of the world - making U.S. assets attractive - wider basis swap spreads are an additional factor that could lift the dollar. Bottom Line: We anticipate the USD cross-currency basis swap spread to widen over the next 12 months. This will be associated with higher FX vols, which hurt carry trades, EM currencies and dollar-block currencies. Chinese reforms will reinforce these risks. Additionally, wider basis swap spreads will create support for the USD. Mathieu Savary, Vice President Foreign Exchange Strategy mathieu@bcaresearch.com 1 Please see Foreign Exchange Strategy Weekly Report, titled "All About Credit", dated October 20, 2017, available at fes.bcaresearch.com. 2 Please see Foreign Exchange Strategy Weekly Report, titled "Conflicting Forces For The Dollar", dated September 8, 2017, and "Is The Dollar Expensive?", dated October 13, 2017, available at fes.bcaresearch.com. 3 Please see Foreign Exchange Strategy Special Report, titled "Carry Trades: More Than Pennies And Steamrollers", dated May 6, 2016, available at fes.bcaresearch.com. 4 Please see Geopolitical Strategy Weekly Report, titled "Xi Jinping: Chairman Of Everything", dated October 25, 2017 and Special Report, titled "How To Read Xi Jinping's Party Congress Speech", dated October 18, 2017, available at gps.bcaresearch.com. Appendix Implications For The Global Fixed Income Investor Chart A1FX Basis Swaps Boosting##BR##Hedged European Yields The outlook for cross-currency basis swap spreads has important implications for global fixed income investors. Chiefly, a wider (more negative) basis swap spread makes it more profitable for U.S. investors to lend U.S. dollars. For example, the top panel of Chart A1 shows that if a U.S.-based investor swaps dollars for euros on a 3-month horizon, and then invests those euros in 10-year German bunds, they will earn a hedged yield of 2.5% (annualized). This compares to a current yield of 2.3% on the 10-year U.S. Treasury note. If the basis swap spread were zero, then the U.S. investor would face a hedged German 10-year yield of only 2.1%. Conversely, a deeply negative basis swap spread works against non-U.S. investors looking to gain exposure to the U.S. bond market. If a Eurozone-based investor swaps euros for dollars on a 3-month horizon and then invests those dollars in 10-year U.S. Treasuries, he will earn a hedged yield of 0.1% (annualized). This compares to a current yield of 0.4% on 10-year German bunds. If the basis swap spread were zero, then the European investor would face a more enticing hedged U.S. 10-year yield of 0.6%. The middle three panels of Chart A1 show the 10-year yields in other Eurozone bond markets from the perspective of a U.S.-based investor who has hedged his currency risk on a 3-month horizon, as per the strategy explained above. The bottom panel of Chart A1 shows that the deviation of the EUR/USD basis swap spread from zero currently adds 42 basis points to the hedged yields faced by a U.S. investor. Charts A2, A3, A4 and A5 present the same analysis for other major bond markets, again from the perspective of a U.S. based investor.5 Chart A2FX Basis Swaps Boosting Hedged Gilt Yields Chart A3FX Basis Swaps Boosting Hedged JGB Yields Chart A4FX Basis Swaps Boosting##BR##Hedged Canadian Yields Chart A5FX Basis Swaps Are NOT Boosting##BR##Hedged Australian Yields The Impact Of Hedging Costs On Returns Of course, the basis swap spread is only one input to hedging costs. Once again, using the example of a U.S.-based investor looking for exposure in European bond markets, we calculate the hedging cost as: (1 + Hedging Cost) = (1 + 3-month EUR LIBOR + basis swap spread) / (1 + 3-month USD LIBOR) Right now the hedging cost in the above example is below zero. This is why German bund yields actually appear more attractive to U.S. investors after taking hedging costs into account. But what's more interesting is that total returns in 7-10 year German bunds (hedged into USD) relative to total returns in 7-10 year U.S. Treasury notes track hedging costs very closely over time (Chart A6). Chart A6Hedging Costs Will Continue To Boost Hedged German Bond Returns As The Fed Hikes Rates This is highly logical. As hedging costs become more negative, it means that U.S.-based investors make more money swapping U.S. dollars for euros. Therefore, a strategy of swapping dollars for euros, and then placing the proceeds in 7-10 year German bunds should continue to be a profitable one for U.S. investors as long as hedging costs continue to decline. Fortunately for U.S. investors, hedging costs should become even more negative during the next 12 months. In our base case scenario, we assume that the Federal Reserve will lift rates by 100bps by the end of 2018. We also assume that the ECB will not lift rates during this timeframe. That divergence in policy rates on its own will drive hedging costs further into negative territory, and it will only be exacerbated if the cross-currency basis swap spread widens as we anticipate. We illustrate the impact of the cross-currency basis swap spread on hedging costs in the bottom panel of Chart A6. The panel shows where hedging costs will go between now and the end of 2018, assuming policy rates move as we described above, and that the basis swap spread either widens to -100 bps or tightens back to zero. It is evident that a sharp widening in basis swap spreads would be a boon for U.S. investors in foreign bond markets. Bottom Line: Deeply negative basis swap spreads make it more profitable to lend dollars on a short-term horizon. This presents an opportunity for U.S. investors to swap dollars for foreign currencies and invest in non-U.S. bond markets. Ryan Swift, Vice President U.S. Bond Strategy rswift@bcaresearch.com 5 While the basis swap spread between the U.S. and most countries is negative, it is actually positive between the U.S. and Australia. So in this case the basis swap spread makes Australian bonds look less attractive to U.S. investors. Conversely, the basis swap spread makes U.S. bonds look slightly more attractive to Australian investors.

Special Report

Highlights Three factors point to stable or narrower USD cross-currency basis swap spreads: the improving health of global banks, the end of the adjustment to the regulatory change affecting prime-money market funds, and the relaxation to the Supplementary Leverage Ratio rules by the U.S. Treasury. Four factors point to wider basis swap spreads: BCA's forecast that U.S. loan growth will pick up, our view on U.S. inflation, the coming decline in the Federal Reserve's balance sheet, and the potential for U.S. repatriation. We expect USD basis swap spreads to widen again, which suggests increasing FX vol. This would hurt carry trades, EM currencies and dollar bloc currencies. Feature Three factors point to stable or narrower USD cross-currency basis swap spreads: the improving health of global banks, the end of the adjustment to the regulatory change affecting prime-money market funds, and the relaxation to the Supplementary Leverage Ratio rules by the U.S. Treasury. Four factors point to wider basis swap spreads: BCA's forecast that U.S. loan growth will pick up, our view on U.S. inflation, the coming decline in the Federal Reserve's balance sheet, and the potential for U.S. repatriation. We expect USD basis swap spreads to widen again, which suggests increasing FX vol. This would hurt carry trades, EM currencies and dollar bloc currencies. The rather arcane topic of cross-currency basis swap spreads has periodically surfaced in the news in the past few years. The widening in cross-currency basis swap spreads has been highlighted as one of the key factors explaining why covered interest rate parity relationships (the link between the price of FX forward, spot prices and interest rate differentials) have not held as closely after the Great Financial Crisis (GFC) as before. The widening of cross-currency basis swap spreads has also been highlighted as a factor behind the strength in the U.S. dollar in 2014 and 2015. Similarly, the recent narrowing in the cross-currency basis swap spread has been highlighted as a factor behind the weakness in the USD this year. This week we delve a little deeper into what cross-currency basis swap spread measures, and what some of its major determinants are. We ultimately expect the USD cross-currency basis swap spread to widen again, which should contribute to a stronger dollar and increased global FX volatility. What Is A Cross-Currency Basis Swap? To examine what drives cross-currency basis swap spreads, one first needs to understand what these instruments are. Let's begin with a regular FX swap. An FX swap in EUR/USD is a contract through which two counterparties agree to exchange EURs for USDs today, with a reversal of that exchange at the maturity of the contract - a reversal set at a predetermined exchange rate simply equal to the forward value of the EUR/USD. So, if counterparty A lends X million EURs to counterparty B, the former receives in U.S. dollars the equivalent of X million EURs times the prevalent EUR/USD spot rate from counterparty B today. The transaction does not end there. Simultaneously, the FX swap forces B to give back the X million EURs to counterparty A at maturity, while counterparty A gives back X million EUR times the EUR/USD forward rate in U.S. dollars to counterparty B. This forward rate is the rate prevalent when the contract was agreed upon. The transactions are illustrated in the top panel of Chart 1. Chart 1FX Swaps Vs. Cross Currency Basis Swaps The problem with regular FX swaps is that they offer little liquidity at extended maturities. If market players want to hedge long-term liabilities and assets, they tend to do so using a cross-currency basis swap, where much more liquidity is available at long maturities. Chart 2A Bigger Funding Gap = ##br##A Wider Basis Swap Spread A EUR/USD cross currency basis swap begins in the same way as a regular FX swap: counterparty A lends X million EURs to counterparty B, and the former receives in U.S. dollars the equivalent of X million EURs times the prevalent EUR/USD spot rate from counterparty B today. However, this is where the similarities end. A cross-currency basis swap has exchanges of cash flows through its term. Counterparty B, which provided USDs to counterparty A, receives 3-month USD Libor, while counterparty A, which provided EURs to counterparty B, received 3-month EUR Libor + a (alpha being the cross-currency basis swap spread). At the maturity of the contract, counterparty A and B both receive their regular intermediary cash flows, and also re-exchange their respective principal - but this time at the same spot rate as the one that existed at the entry of the contract (Chart 1, bottom panel). In both regular FX and cross-currency basis swaps, counterparties have removed their FX risks, except that in the latter, the interest differentials have been paid during the life of the contract instead of being factored through the forward premium/discount. This is fine and dandy, but it leaves a unexplained. The cross currency basis swap spread (a), is a direct function of the relative supply and demand for each currency. If investors demand a lot of EUR in the swap market relative to its supply, a will be positive. If they demand more USDs, a will be negative. A good example of this dynamic is the funding gap of banks. Let's take the Japanese example. Japanese banks have a surplus of domestic deposits (thanks to the massive savings of the Japanese corporate sector) relative to their yen lending. As a result, they have large dollar lending operations. To hedge their dollar assets, Japanese banks borrow USD in large quantities in the cross-currency swap market. This tends to result in a negative swap spread in the yen (Chart 2). This is particularly true if both the banking sector and the other actors in the economy (institutional investors and non-financial firms) also borrow dollars in the swap market to hedge dollar assets, which is the case in Japan (Chart 3). Chart 3Japanese Investors Are Accumulating Assets Abroad Additionally, if there are perceived solvency risks in the European banking sector, this should further weigh on the cross-currency basis swap spread, pushing it deeper into negative territory, as the viability of the main EUR counterparties becomes at risk. The same dance is true for any currency pair. Chart 4The Structural Gap In The Basis Swap Spread##br## Reflects Regulation The other factor that affects USD cross-currency basis swap spreads is the supply of U.S. dollars, especially the room on large banks' balance sheets to service these markets. The cross-currency basis swap spread could be close to zero if large arbitrageurs take offsetting positions to arbitrage the spread away, doing so until the spread disappears. However, with the imposition of Basel III and Dodd-Franks, banks have been constrained in their capacity to do this. Indeed, increased leverage ratio requirements (now banks need to post more capital behind repo transactions as well as collateralized lending and other derivatives) mean that arbitraging cross-currency basis swap spreads and deviations from covered interest rate parity has become much more expensive. Furthermore, the increase in Tier 1 capital ratios associated with these regulations has forced banks to de-lever; however, engaging in arbitrage activities still requires plenty of leverage (Chart 4). Economic Factors Driving The Spread The factors that we look at essentially relate to the supply of USD available for lending in offshore markets, as well as determinants of relative counterparty risks between the U.S. and the rest of the world. Factors Arguing For Narrower Cross-Currency Basis Swap Spreads Global Banks Health The price-to-book ratio of global banks outside the U.S. has been largely correlated with USD cross-currency swap spreads. When global banks get de-rated, spreads widen, and it becomes more expensive to hedge USD positions in the swap market (Chart 5). This is because as investors perceive the solvency of global banks deteriorating, they impose a penalty as the Herstatt risk increases. Additionally, solvency problems can force banks to scramble to access USD funding, prompting deeper spreads. Chart 5Banks Perceived Health Determines ##br##Basis Swap Spreads BCA is positive on global financials and sees continued improvement in European NPLs. This means that solvency risk concerns are likely to remain on the backburner for now, pointing to narrower basis swap spreads. Supplementary Leverage Ratio Changes In June, the U.S. Treasury announced a relaxation of some of its rules on supplementary leverage ratios, lowering the amount of capital required to support activity in the repo market behind initial margins for centrally cleared derivatives, and behind holdings of Treasurys. This means that commercial banks in the U.S. can have bigger balance sheets and more room to engage in arbitrage activity, implying a greater supply of dollars in the USD cross-currency basis swap market. In response to last June's proposal, basis swap spreads narrowed by 11 basis points. BCA believes these changes will continue to support dollar liquidity, and will further help in narrowing cross-currency basis swap spreads. Prime Money-Market Funds Debacle Is Over Chart 6More Expensive Bank Funding Equals ##br##Wider Basis Swap Spreads In October 2016, regulatory changes were implemented that allowed prime money market funds to have fluctuating net asset values. Obviously, this meant that prime money-market funds would be not-so-prime anymore. As a result, to remain the ultra-safe vehicles that they once were, prime money-market funds de-risked. As a result, they cut their exposure to risky activities in anticipation of these changes. In practice, a key source of short-term funding for banks evaporated from the market, putting upward pressure on bank financing costs. As the LIBOR-OIS spread increased, so did basis-swap spreads (Chart 6): as it became more expensive for banks to finance themselves, they had to curtail the supply of USDs provided to the swap market, an activity normally requiring intense demand on banks' balance sheets. This adjustment is now over, suggesting limited potential widening in USD basis swap spreads. Factors Arguing For Wider Cross-Currency Basis Swap Spreads 1. U.S. Loan Growth When U.S. banks increase their loan formation activity, USD cross-currency basis swap spreads widen (Chart 7). As banks increase their extension of credit through loans, they decrease the amount of securities they hold on their balance sheets (Chart 8). This means there is less supply of liquidity available for balance sheet activities, particularly providing dollar funding in the offshore market. In the Basel III / Dodd-Frank world, less-liquid bank balance sheets are synonymous with wider USD basis-swap spreads. As we argued last week, increasing U.S. capex, easing lending standards for firms and rising household income levels should result in increasing loan growth in the U.S. which will result in lower abundance of liquid assets and a widening basis swap spreads.1 Chart 7More Bank Loans Lead To Wider Swap Spreads Chart 8More Debt Equals Less Securities In Bank Credit 2. U.S. Inflation There is a fairly close relationship between U.S. inflation and the USD basis swap spread, where a higher core CPI tends to lead to a wider spread (Chart 9). The fall in U.S. inflation this year likely contributed to the narrowing in basis swap spreads. Our take on this is that as inflation falls, it gives an incentive for banks to hold low-yielding liquidity on their balance sheets as real returns on cash improve. This fuels a gigantic carry trade through the basis-swap market. We expect inflation to pick up meaningfully by mid-2018, which should widen cross-currency basis swap spreads.2 3. Central Bank Balance Sheets When the Federal Reserve increases the size of its balance sheet relative to other balance sheets, this tends to lead to a narrowing of the USD basis swap spread as the global supply of dollars relative to other currencies increases. The opposite is also true. This relationship did not work after late 2016 (Chart 10). However, during that episode, as the change in prime money-market funds caused a dislocation in banks' funding, commercial banks exhibited cautious behavior and increased their reserves with the Fed. As Chart 11 illustrates, there is a tight relationship between the change in commercial banks' reserves held at the Fed and cross-currency basis swap spreads. Going forward, as the Fed lets it balance sheet run off, we expect to see a decrease in commercial banks' excess reserves. This could contribute to upward movement in the basis swap spread. Chart 9When U.S. Inflation Increases, ##br##Swap Spreads Widen Chart 10Smaller Fed Balance Sheet Leads To##br## Wider Basis Swap Spreads Chart 11Fed Runoff Could##br## Widen Basis Swap Spreads 4. U.S. Repatriations The most revealing relationship unearthed in our study was that when U.S. entities repatriate funds at home, this tends to put strong widening pressure on the USD cross-currency basis swap spread (Chart 12). U.S. businesses hold large cash piles abroad - by some estimates more than US$2.5 trillion. However, most of these funds are held in highly liquid, high-quality U.S.-dollar assets offshore. These assets are perfect collaterals for various transactions in the interbank market. The funds held abroad by U.S. firms are a source of supply for U.S. dollars in the offshore markets. When U.S. entities bring assets back home, the widening in the basis swap spread essentially reflects a decline in the supply of USD in offshore markets, and vice versa when Americans export capital abroad. Chart 12U.s. Repatriations Support Wider Basis Swap Spreads BCA's base case is that tax cuts are likely to hit the U.S. economy in 2018, even if the growing feud between Trump and the establishment Republican party members is a growing risk. BCA still views a tax repatriation as a higher-likelihood event, as it is the easiest way for the U.S. government to bring funds into its coffers. The 2004 tax repatriation under former President George W. Bush did result in substantial fund repatriation in the U.S. This time will not be different. We expect any such tax repatriation to cause a potentially large deficit of supply in the USD offshore markets, which could create a strong widening basis on the cross-currency basis swap spread in favor of the dollar. Bottom Line: Three factors argue for USD cross-currency basis swap spreads to stay at current levels, or even narrow further. These factors are the health of global banks, the easing in U.S. supplementary leverage ratios and the end of the adjustment of U.S. bank funding to new regulations affecting prime money-market funds. On the other hand four factors points to wider USD cross-currency basis swap spreads: BCA's positive outlook for U.S. credit growth; BCA's positive outlook on U.S. inflation; the run-off of the Fed's balance sheet; and the potential for U.S. entities repatriating funds from abroad. Potential Direction And Investment Implications We anticipate USD cross-currency basis swap spreads to widen over the coming 12 months. We think the easing in the Supplementary Leverage Ratios rules by the U.S. Treasury is the most important factor pointing to narrower USD cross-currency basis swap spreads. However, Basel III rules and most of Dodd-Frank are still in place, which suggest there remains large constraints on the balance-sheet activities of global banks, which will limit the potential for a narrowing of the USD basis swap spread as U.S. banks will remain constrained in their ability to supply U.S. dollars in the offshore market. On the other hand many factors support wider USD cross-currency basis swap spreads, most important of which is the potential for more credit growth. This is in our view a very strong force as it requires banks to ration the use of their balance sheets, limiting their activity in the offshore market. Moreover, we do foresee a high probability of tax repatriation, which would put strong widening pressure on the swap spreads. In terms of implications, wider USD basis swap spreads tend to be associated with rising FX vols (Chart 13). As we highlighted in a Special Report last year, higher FX vols are poison for carry trades.3 As such, we think that widening swap spreads could spur a period of trouble for traditional carry currencies. This means EM and dollar-block currencies are likely to suffer in this environment. Chart 13Wider Basis Swap Spreads Equals Higher Vol Additionally, in China, Xi Jinping is consolidating power and has taken control of the Politburo. This implies he now has more room to implement reforms. Removal of growth targets after 2020, removal of growth as a criterion for grading local officials, a focus on balanced growth, and a focus on combatting pollution all suggest that Chinese growth is unlikely to follow the same debt-fueled, capex-led model.4 This will weigh on Chinese imports of raw materials, and hurt export volumes and prices for many EM countries and commodities producers. This means these policies represent a headwind for many carry currencies. Moreover, historically, wider USD funding costs have been associated with a stronger dollar, as it makes it more expensive to hedge dollar assets. Thus, in an environment where U.S. interest rates are rising relative to the rest of the world - making U.S. assets attractive - wider basis swap spreads are an additional factor that could lift the dollar. Bottom Line: We anticipate the USD cross-currency basis swap spread to widen over the next 12 months. This will be associated with higher FX vols, which hurt carry trades, EM currencies and dollar-block currencies. Chinese reforms will reinforce these risks. Additionally, wider basis swap spreads will create support for the USD. Mathieu Savary, Vice President Foreign Exchange Strategy mathieu@bcaresearch.com 1 Please see Foreign Exchange Strategy Weekly Report, titled "All About Credit", dated October 20, 2017, available at fes.bcaresearch.com. 2 Please see Foreign Exchange Strategy Weekly Report, titled "Conflicting Forces For The Dollar", dated September 8, 2017, and "Is The Dollar Expensive?", dated October 13, 2017. 3 Please see Foreign Exchange Strategy Special Report, titled "Carry Trades: More Than Pennies And Steamrollers", dated May 6, 2016. 4 Please see Geopolitical Strategy Weekly Report, titled "Xi Jinping: Chairman Of Everything", dated October 25, 2017 and Special Report, titled "How To Read Xi Jinping's Party Congress Speech", dated October 18, 2017.

Please note that in addition to today's abbreviated Weekly Bulletin, we are also publishing a Special Report on Argentina. Feature Regarding recent financial market dynamics, it appears that the high-yielding EM currencies are breaking down as U.S. bond yields march higher. Several EM exchange rates have formed a tapering wedge pattern, as shown in Chart I-1. Such patterns eventually lead a major break out or break down. Our bias remains that we are witnessing a major breakdown in several EM high-yielding currencies. If this transpires, it would be a precursor for a wider selloff in EM risk assets. Below we discuss interesting dynamics that have emerged in India's onshore fixed-income market lately, and their implications for the nation's equity market. India Several signals tentatively indicate that the price of liquidity has risen at the margin in India. Onshore BBB corporate bond yields have increased and their respective credit spreads have widened (Chart I-2). In addition, the yield curve has steepened modestly. Chart I-1A Tapering Wedge: ##br##A Breakout Or Breakdown? Chart I-2India: Onshore BBB Corporate Bond ##br##Yields And Spreads Have Spiked Rising corporate bond yields and widening corporate credit spreads have been negative for share prices in the past (Chart I-3). Similarly, steepening yield curves have been associated with a pullback in equity prices in recent years (Chart I-4). Note that yields, spreads and the yield curve are shown inverted on Charts I-3 and I-4. Chart I-3India: Corporate Bond Yields ##br##And Spreads Versus Stocks Chart I-4India: Yield Curve ##br##And Share Prices Why has the market price of liquidity risen in India? In our opinion, it has to do with both the domestic and external environments. On the domestic side, the fiscal deficit has widened, implying that borrowing requirements by central and state governments have risen (Chart I-5). Increased demand for credit from the government would not have been a problem had the commercial banks accommodated for it by creating enough new money. Yet, broad money supply growth remains depressed (Chart I-6). Chart I-5India: Ballooning Fiscal Deficits ##br##And Weak Money Creation Chart I-6Indian Money Growth: ##br##New Record Low As a result, the diminished amount of new money relative to demand for money, among other reasons, pushed marginal borrowing costs higher. Chart I-7 shows our proxy for new money available to the private sector has dipped into negative territory. On the external side, the recent rise in U.S. bond yields and the rebound in the U.S. dollar against several EM currencies might have also contributed to higher borrowing costs in India. We expect this U.S. dollar rebound versus EM currencies to persist and U.S. Treasury yields to continue drifting higher. Hence, the global backdrop heralds marginally higher bond yields in India. Although the onshore corporate bond market - and its BBB segment - is not very large, investors should heed to its signals because it reflects the cost of borrowing for the marginal corporate borrower. Besides, its signals have worked quite well in the past as shown in previous Chart I-3 on page 2. Some commentators might argue that the mild rise in government bond yields has been driven by a rise in inflation and growth expectations. We will not disagree with that, but both economic growth and inflation variables are still muted. Chart I-8 shows economic activity is lukewarm at best. Chart I-7India: Proxy For New Money ##br##Available To Private Sector Chart I-8India's Growth Is ##br##Lukewarm At Best On the inflation outlook, the picture is mixed as well. Consumer price inflation, especially core measures, might have bottomed (Chart I-9). Critically, the government approved a draft bill in July that allows the central government to set minimum wages across all sectors and states. The central government is currently reviewing the formula used to set minimum wage and the new formula might lead to significant increases in minimum wages. These policy changes come on top of the pay raises that public sector workers saw earlier this year. Importantly, if consumer demand accelerates while capital spending remains in the doldrums, inflationary pressures will mount. Chart I-10 shows that since 2012 consumer spending has outpaced investment by a large margin. Chart I-9India: Consumer Inflation ##br##Might Be Bottoming Chart I-10India: Consumer Spending ##br##Has Outpaced Investment Provided India has been, and remains, an underinvested economy, if this gap persists, it will produce either inflation or a widening current account deficit. Rising consumption without an equal increase in the supply of goods and services will either lead to higher prices or mushrooming consumer goods imports. Both scenarios bode ill for the macro dynamics, the currency, and ultimately equity multiples. As to financial markets, the Indian bourse is one of the most expensive in the EM space, so it is not very surprising that share prices could react negatively to marginally higher interest rates. For dedicated EM equity investors, we downgraded India from overweight to neutral on August 23, and this stance remains intact. While near-term underperformance cannot be ruled out, the medium-term outlook for relative performance warrants a neutral stance. Bottom Line: There are signals that liquidity is tightening on the margin in India's fixed-income markets due to domestic and external reasons. This will likely hurt share prices. Dedicated EM equity investors should keep a neutral allocation on India's bourse. Mexico: Close Currency, Rates, And Credit Overweights NAFTA risks to Mexico are escalating again. According to our Geopolitical Strategy team, there is non-trivial probability that the NAFTA negotiations will become negative for Mexican financial markets. The recent relapse in Mexico's financial markets will likely endure. We are closing the following positions: long MXN / short BRL; long MXN / short ZAR; receive Mexican 2-year / pay 2-year swap rates as well as overweight positions in Mexican sovereign credit versus Colombia and Indonesia. Dedicated equity investors should stay neutral on this bourse. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com Ayman Kawtharani, Associate Editor ayman@bcaresearch.com Equity Recommendations Fixed-Income, Credit And Currency Recommendations