Commodities & Energy Sector

Highlights Crude-oil fundamentals stand out among commodities because of the active efforts by critical producers to rein in supply since the end of last year. This can be seen in even-higher compliance with the production accord - a supply shock in many ways - negotiated by the Kingdom of Saudi Arabia (KSA) and Russia: Last month, Reuters estimated 94% compliance on the 1.2mm b/d in cuts pledged by OPEC states. We expect compliance to remain high, which will strengthen the divergence between oil prices and the USD, as markets look toward the upcoming summer driving season in the Northern Hemisphere. Active supply management and robust demand growth wrought by lower prices could continue to overwhelm a strong USD's influence on oil prices, if this Agreement becomes a durable modus operandi for KSA and Russia going forward. We give a high probability to this outcome, even as the Fed leans into its interest-rate normalization. Energy: Overweight. This past Thursday, we closed our long WTI Dec/17 vs. short Dec/18 backwardation spread at +$0.96/bbl (Dec/17 over); it was initiated February 9 at -$0.11/bbl (Dec/17 under), resulting in a 972.7% gain. We also closed our Dec/19 short WTI vs. long Brent spread, elected February 6 at +$0.07/bbl (WTI over) at -$1.17/bbl (WTI under), for a gain of 1,771.4%. Base Metals: Neutral. Any demand uptick for base metals' coming from U.S. fiscal stimulus will not hit markets until 2H18 at the earliest. We remain neutral. Precious Metals: Neutral. Based on last week's analysis, we are tactically long a Jun/17 gold put spread (long the $1200/oz put vs. short the $1150/oz puts) and call spread (long the $1275/oz call vs. short the $1325/oz calls) at a net debit of $21/oz. Ags/Softs: Underweight. The USDA expects continued demand from China to keep soybeans relatively well bid versus corn and wheat in the 2017/18 crop year. Total planted area for these crops is expected to be the lowest since 2011, keeping ending stocks flat to lower. Feature Prior to the end of the 1990s, crude-oil prices were, to use one of the most popular catch-phrases in finance, mean-reverting: The price of crude oil imported to the U.S. averaged just over $19/bbl from Mar/83, when WTI futures began trading, to 1999 (Chart of the Week). This meant WTI traded at ~ $20/bbl on average over that period. Prices were volatile, but pretty much returned to $20ish/bbl, which allowed traders to take a view on how soon prices would revert to their mean. Whenever prices were too far removed from that level, markets expected producers - OPEC mostly - to adjust output to meet current and expected demand conditions. Since roughly 2000 - maybe a little earlier - oil prices have followed a random walk.1 During this time, oil prices have been negatively correlated with the broad trade-weighted index (TWI) for the USD. One striking characteristic of oil prices and the USD TWI during this time is both followed random walks, which "like the walk of a drunken sailor, wanders indefinitely far, listing with the wind," to borrow Paul Samuelson's well-turned metaphor (Chart 2).2 Chart of the WeekOil's Past As Prelude: ##br##A Return To Mean Reversion? Chart 2Oil Prices And The USD Followed ##br##A Common Long-term Trend Until 1Q16 We believe this was caused by OPEC's decision to become a price-taker at the end of the 1990s - shortly after Dec/98 or thereabouts - after years of unsuccessfully trying to manage oil prices via production adjustments. After the price of oil imports in the U.S. dropped below $10/bbl (nominal), it appears the Cartel took the decision to respond to prices set by market forces (supply, demand, inventories and exchange rates), and to abandon its price-management efforts. The long-term correlation between oil and the USD was due to the fact that while oil prices and the USD followed random walks, they followed a common long-term trend as they wandered indefinitely about. This held up to the end of 1Q16, when a massive sell-off in risky-asset markets globally took oil prices below $30/bbl (Chart 3).3 This came on the heels of a price collapse brought about by OPEC's Nov/14 decision to launch a market-share war. By no means did this high correlation mean oil and the USD were always moving in lock step. The collapse in oil prices at the end of the last century led to a production-cutting agreement among OPEC states, Norway and Mexico, which lifted U.S. import prices from less than $10/bbl at the end of 1998 to $30/bbl by Nov/00. Likewise, export disruptions in Venezuela in 2002 - 2003 and, to a lesser extent, hurricane losses in the U.S. Gulf in 2005 sharply curtailed supply and lifted oil prices above what could have been expected given the USD's level at the time, as the Chart of the Week shows.4 End Of Oil's Random Walk? The price collapse of 1Q16 marked the bottom of the price move begun a few months prior to the Nov/14 market-share war declaration. The subsequent divergence between oil prices and the USD since then has been remarkable (Chart 4). The market-share strategy, which essentially allowed Cartel members to produce full-out and grab as much market share as possible, was engineered by KSA, and, we believe, initially was directed at undermining Iran's efforts to restore oil production lost to nuclear-related sanctions. From time to time, it also appeared OPEC was trying to retard the continued growth of shale-oil production in the U.S., which, by 2014, was increasing at an annual rate of more than 1mm b/d, enough to replace the entire output of Libya. Chart 3Close-up Of USD vs. ##br##Brent Divergence Chart 4The Divergence Between ##br##Oil Prices And The USD Is Remarkable This strategy was a complete failure. The price collapse that ensued brought KSA and Russia - both highly dependent on oil revenues - to the brink of financial ruin, compelling them to find a way to work together.5 After several false starts in 2016, they succeeded late in the year with a negotiated production cut. OPEC pledged to reduce output by as much as 1.2mm b/d, and non-OPEC producers agreed to cut output by close to 600k b/d, half of which is expected to come from Russia. Recent tallies by Reuters indicate 94% of the cuts from OPEC states that signed on to the deal have actually been realized.6 Should KSA and Russia find a way to coordinate their and their allies' production in a way that maintains the backwardation we expect later this year - the result of production cuts (Chart 5), and robust demand growth (Chart 6) - we could see oil prices become mean-reverting once again. Chart 5If KSA And Russia Can ##br##Coordinate Production ... Chart 6... And Demand Continues To Grow, ##br##The Oil-Price Backwardation Could Persist This likely requires the forward curves for WTI and Brent to remain backwardated, so as to moderate the growth in shale production, and for prices to remain between $55/bbl and $65/bbl, so as not to set off another shale boom. Gulf sources have indicated KSA prefers prices this year of ~ $60/bbl, which, we believe would allow it to keep some control over the rate at which shale production revives.7 Chart 7Supply Destruction And Robust Growth ##br##Rallied Oil Despite A Strong USD Investment Implications We are not calling for a return to mean-reversion in oil prices just yet. We are, however, highlighting the possibility for such a sea-change in the market if all the supply-side pieces fall into place - i.e., KSA, Russia and their respective allies find a way to work together to moderate U.S. shale-oil production. That said, we will be watching closely to see whether the KSA - Russia Agreement becomes a durable modus operandi in the oil market, particularly as regards the management of inventories and production in the market generally. If these states are able to keep prices ~ $60/bbl, and gain some control over the forward curve's slope - i.e., literally manage their production for backwardation - then there is a chance oil prices could once again become mean-reverting. In a mean-reverting world with backwardated oil prices, commodity-index exposure is favored, since investors would, once again, earn positive roll yields as the indices are rebalanced monthly in the underlying futures markets. Bottom Line: The persistent negative correlation between oil prices and the USD broke down following the global asset sell-off in 1Q16. Since then, the combination of supply destruction and robust demand growth has allowed oil prices to rally despite a strong USD (Chart 7). If KSA and Russia can continue to cooperate in their production-management deal - i.e., find a way to manage production so that prices remain closer to $60/bbl than not - and Brent and WTI forward curves backwardate, markets could once again become mean-reverting. In such a world, commodity-index exposures are favored - particularly those heavy on crude-oil and refined-products price exposure - for their positive roll yield. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com 1 Technically, oil prices have been I(1) variables (integrated of order 1) since about 2000, meaning they are mean-reverting in first differences (e.g., today's price minus yesterday's price). Please see Geman, Helyette (2007), "Mean Reversion Versus Random Walk in Oil and Natural Gas Prices," pp. 219 - 228, in Advances in Mathematical Finance. Haidar, Imad and Rodney C. Wolff (2011) obtained similar results, reporting crude prices were mean-reverting from Jan/86 - Jan/98, then random-walking since then; please see pp. 3 - 4 of "Forecasting Crude Oil Price (revisited)," presented at the 30th USAEE/IAEE North American Conference in Washington, D.C., during October 2011. Our own research corroborates these results - we find WTI and Brent were mean-reverting from Mar/83, when WTI futures started trading, to Mar/98; and were random-walking I(1) variables after that. 2 Please see Samuelson, Paul A. (1965), "Proof That Properly Anticipated Prices Fluctuate Randomly," in Industrial Management Review, 6:2. 3 This is to say, these variables were cointegrated, and could be expressed in a linear combination using an error-correction model. 4 Our colleague, Mathieu Savary, who runs BCA Research's Foreign Exchange Strategy, addressed these oil-USD divergences in "Party Like It's 1999," published November 25, 2016. It is available at fes.bcareseach.com. 5 We discuss this at length in the feature article of Commodity & Energy Strategy published September 8, 2016, entitled "Ignore The KSA - Russia Production Pact, Focus Instead On Their Need For Cash." Both states were burning through cash reserves, and were trying tap foreign markets for additional funds by selling interests in their most valuable holdings - via the IPO of, and via the sale of just under 20% of Rosneft held by the Russian government. Russia placed its Rosneft shares late last year with Glencore and Qatar's sovereign wealth fund, while KSA is expected to IPO Aramco in late 2018. 6 Please see "OPEC compliance with oil curbs rises to 94 percent in February: Reuters survey," published by the news service online February 28, 2017. 7 Please see "Exclusive: Saudi Arabia wants oil prices to rise to around $60 in 2017 - sources," published by Reuters online February 28, 2016. Investment Views and Themes Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed In 2017 Summary of Trades Closed in 2016

Highlights The supply of U.S. dollar outside America has been curtailed, yet there is large pent-up demand for dollars. This warrants another upleg in the greenback. The Trump administration's desire to shrink America's current account deficit will be very deflationary for the rest of the world, and mildly inflationary for the U.S. Such policies, if adopted, will exaggerate the paucity of U.S. dollars beyond America's borders and lead to notable dollar appreciation. The RMB is at risk because Chinese banks have created too many yuan, and deposit rates in real terms have turned negative as inflation has risen. Our negative view on EM has been and continues to be driven by our outlook on EM/China domestic demand, commodities prices and the U.S. dollar - not growth in advanced economies. Feature In recent weeks we met with clients in Asia and Australia. This week's report addresses some of the more common questions that we were asked to address. Question: You have written about "global U.S. dollar liquidity shortages." Why have these "global dollar shortages" occurred given the Fed expanded its balance sheet enormously from 2008 until 2014? How does one measure "global dollar shortages," and what does it mean for financial markets? By "global U.S. dollar shortages," we refer to deficiency in U.S. dollars outside the U.S., where U.S. dollar supply growth has fallen short of growth in demand for the greenback. We have the following pertinent observations on this issue: U.S. dollar shortages in the global banking system (eurodollar market) can be represented by U.S. banks' and other financial firms' claims on foreigners. This measure has been shrinking since early 2015 (Chart I-1). This corroborates the fact that U.S. banks, prime money market funds and other financial institutions have been unable/unwilling to supply dollars to the eurodollar market. This is consistent with rising LIBOR rates, which still continue to climb. U.S. non-financial entities' foreign assets have also fallen in the past year and a half but they are much smaller than banks and other financial institutions claims. As to U.S. banks' and other financial firms' claims on EM, they have also been shrinking since early 2015 (Chart I-2). Chart I-1Weak Supply Of U.S. Dollars To Rest ##br##Of World By U.S. Financial Institutions Chart I-2Shrinking Supply Of U.S. Dollars ##br##To EM By U.S. Financial Institutions Another way that the U.S. emits dollars to the rest of the world is by running a current account deficit. The U.S. current account deficit as a share of global GDP is now much smaller now than it was before the Great Recession (Chart I-3). This also means a smaller U.S. dollar supply relative to the size of the world economy. On the demand side, the widening in cross currency basis swaps indicates structural demand for U.S. dollar funding among euro area and Japanese investors (Chart I-4). Chart I-3The U.S. Emits Less ##br##Dollars To World Via Trade Chart I-4Pent-Up Demand For Dollars From Japanese ##br##And European Fixed-Income Investors These investors have been opting for exposure to dollar assets due to the higher yield on U.S. dollar fixed-income instruments - but they have been reluctant to take on exchange rate risk. In brief, they have avoided getting long exposure to the U.S. dollar. The reluctance to accept the exchange rate risk by European and Japanese investors means they are not bullish on the dollar. This goes against the widespread opinion among investors that the overwhelming majority of global investors are bullish on the U.S. currency. By hedging the exchange rate risk - in this case the risk of potential greenback depreciation - these investors are giving up a considerable portion of higher yield that they obtain in U.S. fixed-income market. In fact, if these basis swaps continue to widen or remain wide it might make sense for European and Japanese fixed-income investors to buy U.S. fixed-income securities and not hedge the currency risk. If and when these investors stop hedging their exchange rate risk, the U.S. dollar will appreciate versus the euro and the yen. Provided European and Japanese fixed-income investors are sizable players in global fixed income and hence currency markets, they have the potential to make a difference in exchange rate markets. In short, there is potential pent-up demand for U.S. dollars from these European and Japanese institutions. Such a widening in basis swaps is also consistent with the above observations that U.S. banks have been reluctant to take the other side of this trade - i.e., offer U.S. dollars to European and Japanese investors - even though it is a very profitable opportunity. Finally, the drop in EM central banks' foreign exchange reserves reflects demand for U.S. dollars in their economies, primarily in China (Chart I-5). The Chinese central bank has sold U.S. securities to meet mushrooming demand for U.S. dollars from Chinese households and companies. This entails there has been and remains considerable pent-up demand for dollars by mainland companies and households. With respect to the supply of currency, it is important to note that it is up to commercial banks - not the central bank - to create money. Central banks provide liquidity for commercial banks, but it is the latter that creates money.1 In a nutshell, by undertaking QE, the Fed provided reserves for U.S. commercial banks (Chart I-6), yet the latter have been reluctant to create too much money. Banks create money by originating loans and other types of claims. Chart I-5China: Selling U.S. Securities To ##br##Meet Domestic Demand For Dollars Chart I-6The Fed's Balance ##br##Sheet In Perspective U.S. banks have been very conservative in money creation especially outside America. In the U.S., banks shrunk their balance sheets and loans in the 2009-2011 period. That is why the Fed's QE programs have not led to inflation. Notably, U.S. banks' total assets - including bank loans - and broad money (M2) growth have lately rolled over (Chart I-7). This worsens the lingering dollar scarcity outside the U.S., which should in turn prop up the value of the dollar. The reasons why U.S. banks and financial institutions have been conservative is due to their own deleveraging objectives and because of regulatory changes in the financial industry. In regard to interest rates, U.S. nominal and real (inflation-adjusted) interest rates are very low yet they are high relative to European and Japanese real rates (Chart I-8). Given a relatively tight labor market, odds are that U.S. interest rate expectations will rise further in both absolute and relative terms. This will cause the dollar to appreciate. Chart I-7U.S. Banks Control ##br##The Supply Of U.S. Dollars Chart I-8U.S. And German ##br##Inflation-Adjusted Interest Rates Bottom Line: The pace of supply of dollars beyond the U.S. is falling short of growth in demand for this currency. Typically, this warrants greenback appreciation. Question: What about the U.S. administration's preference for a weaker dollar to improve America's trade position? Won't the greenback depreciate as the Trump administration expresses its desire for a weaker currency? Certainly U.S. officials can verbally influence the exchange rate and drive markets for a (short) period of time. Yet fundamentals and flows will re-assert themselves and the greenback will ultimately appreciate even if its rally is delayed by policymakers. The new U.S. administration intends to run mercantilist policies to create jobs in America and doing so will shrink the current account deficit. Nevertheless, a narrowing U.S. current account deficit ultimately entails diminishing flows of U.S. dollars to the rest of the world, which is bullish for the greenback. In brief, the U.S. administration can delay the dollar rally, but it will not be able to prevent it if and when it shrinks the U.S. current account deficit. This will be enormously deflationary for the rest of the world and ultimately for the global economy. The supply of dollars outside U.S. borders will become even more dearth. As their exports tumble, manufacturing-heavy Asian and European economies will have to run even more stimulative policies - reduce their real interest rates further - to offset such a deflationary shock to their economies. In the case where the Trump administration successfully manages to weaken the U.S. dollar, the ensuing boost to U.S. manufacturing and employment will be mildly inflationary given the already relatively tight labor market. Thereby, trade protectionism or policy-driven currency depreciation, if these occur, will lift U.S. inflation and U.S. interest rates will go up. Rising U.S. interest rates and lower interest rates throughout the rest of the world will propel the dollar's value higher. On the whole, in the case of U.S. trade restrictions, the exchange rates have to adjust to mitigate deflation in the rest of world and cap inflation in America. This ultimately entails a stronger U.S. dollar and weaker currencies abroad. A final note on exchange rates valuation. Based on unit labor costs, the U.S. dollar is not yet expensive (Chart I-9A). The same measure for other currencies is also shown in Chart I-9A and Chart I-9B. Chart I-9AReal Effective Exchange ##br##Rates Based On Unit Labor Costs Chart I-9BReal Effective Exchange ##br##Rates Based On Unit Labor Costs Financial markets tend to overshoot and undershoot before a major trend reversal. We believe the U.S. dollar is in a genuine bull market and will likely become more expensive before topping out. Bottom Line: The U.S.'s desire to shrink its current account deficit is very deflationary for the rest of the world. Such policies, if adopted in the U.S., will exaggerate the scarcity of U.S. dollars beyond America's borders and lead to notable dollar appreciation. Question: The RMB/USD exchange rate has been stable lately. Does this mean the authorities have reasserted their control over the exchange rate and will not allow it to depreciate? The authorities in China have partial and temporary control over the exchange rate. Ultimately, it will be Chinese households and companies that drive the exchange rate, barring full-out government controls over all export/import transactions, money transfers as well as financial and capital account flows. If mainland households and companies opt to convert a small portion of their liquid savings (deposits at banks) into foreign currency, there is little the authorities can do to defend the RMB, barring a complete closing of balance-of-payments transactions to companies and households. The primary risk to the yuan exchange rate is not currency valuation but an overflow of yuan in the system - i.e., excess supply of RMBs is the main factor that will cause currency depreciation. Unlike U.S. banks, Chinese banks have created too many yuan. Broad money (M2) in China has risen from RMB 48 trillion as of December 2008 to RMB 158 trillion currently - i.e., it has surged by 3-fold. M2 has risen from 150% to 210% of GDP in the past eight years (Chart I-10). In the meantime, the ratio of foreign exchange reserves to M2 has dropped to 14% (Chart I-11). Chart I-10Chinese Banks Have ##br##Created Too Many Yuan Chart I-11China: Foreign Reserves Are ##br##Small Relative To Money Supply The latter ratio implies that if Chinese companies and households decide to convert 14% of their deposits at banks into foreign currencies and the People's Bank of China (PBoC) sells its international reserves to offset it, the latter will simply evaporate. We are not suggesting this will actually happen. The point to emphasize is that mainland banks have created so much money that even the country's US$ 3 trillion foreign exchange reserves are not sufficient to back those deposits up. Chinese households and companies may already be sensing there is too much in the way of RMBs floating around, and intuitively may not trust the currency. They have paid astronomical multiples for real assets like property in China, and have recently been willing to shift assets into foreign currencies/assets. Importantly, the one-year deposit rate at banks is 1.5% in nominal terms but in real terms it has now become negative as inflation has picked up. Chart I-12 (top panel) demonstrates that the deposit rate deflated by core inflation is negative for the first time in the past 10 years. The bottom panel of Chart I-12 shows that the deposit rate deflated by headline CPI inflation is also negative. Interestingly, any time the real deposit rate turned negative in the past, the central bank hiked interest rates. It is impossible to know whether the latest pick up in China's inflation represents a temporary spike or is the beginning of a major and lasting uptrend (Chart I-13). We are surprised by how fast and sharply inflation has risen lately, given the growth improvement has so far been modest. Chart I-12China: Real Deposit ##br##Rates Have Turned Negative Chart I-13China: Inflation ##br##Is Rising, For Now The trillion- dollar question is what is the true output gap in China and, correspondingly, whether the latest rise in inflation is genuine and lasting or simply a statistical aberration. No one including Chinese policymakers knows the answers to these very essential questions. What type of adjustment China embarks on depends on monetary policy and banks in China. As and if Chinese banks slow down money creation, economic growth will tumble and deflationary tendencies will resurface. This scenario is good for creditors - households and companies with large amounts of deposits - because deposit rates in real terms will rise again. Yet this is a bad outcome for indebted companies, capital spending and employment. If mainland banks continue to create money at a double-digit pace as they have been doing, inflation will likely become persistent and durable. These dynamics are positive for debtors as real borrowing costs will drop further/stay negative, and growth will hold up. However, in such a case, negative real rates will buttress capital outflows and pressure the value of the RMB. By and large, the Chinese authorities are facing a profound choice: Policymakers can choose to help debtors (indebted companies) by accommodating continuous money supply expansion by banks, i.e., opt for negative real interest rates. The outcome will be much stronger downward pressure on the RMB. The latter will depreciate at a double-digit pace annually in the next several years. They can opt to force the banking system to slow down the pace of money/credit creation. This will lead to some sort of debt deflation. Money growth and inflation will drop and the currency will not be at a risk of major depreciation. Yet, economic growth/profits/employment will tumble. A third choice for the authorities is to resort to full-out government controls over all trade, transfers as well as financial and capital account transactions - i.e., take the country back to socialism. Only in such a case can the authorities control the exchange rate and interest rates simultaneously - i.e., they can inflate the credit bubble away while preventing households from converting their liquid savings into foreign currency. In brief, this entails financial repression, and it will erode the real value of Chinese deposits. It is not clear to us whether this is a politically more viable option than allowing some bankruptcies/layoffs and debt deflation. Besides, this will devastate China's vibrant private sector as businessmen and high-income employees become reluctant to invest and expand as they observe the real value of their savings/wealth decline. Chart I-14U.S. Dollar And Commodities ##br##Prices Unusual Decoupling As if there were not enough domestic challenges, Chinese policymakers are also facing a hawkish Trump administration on the issue of trade and the exchange rate. Putting it all together, we conclude it will be extremely difficult for the Chinese authorities to navigate through these challenges. One area where we disagree with many investors is that the Chinese authorities have a viable plan and strategy. Given the above constraints, there are no easy choices and it is hard to know which route the Chinese government will take. The latest bout of stability in the RMB has been due to a notable shutdown in outflows. Yet this is a temporary solution. The inability to convert liquid savings into foreign currency will only make households and companies more set on converting their yuan. Odds are that capital outflows will skyrocket on any relaxation of recent harsh restrictions. Bottom Line: In any country, the monetary authorities cannot simultaneously control the price of money (interest rates), the quantity of money, and thereby the exchange rate. This will prove to be true in China too. We continue betting on further RMB depreciation. Question: Why do you not think this commodities rally has further to go, given supply has been curtailed and demand is picking up as global growth improves? The strength in commodities prices in recent months when the U.S. dollar has been firm is a major departure from historical correlations (Chart I-14). Remarkably, oil forward prices have recently dropped and global energy share prices have relapsed in absolute terms, even though the spot price has held up (Chart I-15). This foretells that the marketplace does not believe in the sustainability of the current spot price level of crude. As to industrial metals, our hunch is that Chinese demand will weaken again as the nation's credit and fiscal impulse relapses (Chart I-16). Besides, the recent resilience in copper has been due to supply disruptions that may be temporary. Chart I-15Has Sell Off In Oil Market Begun? Chart I-16China's Growth To Peak Later This Year Notably, hopes that U.S. infrastructure spending - even if such spending turns out to be considerable - will boost demand for industrial metals are misplaced, because the U.S. is a small consumer of metals. China consumes six to seven times more copper, nickel, zinc, aluminum, tin and lead than the U.S. Hence, we view industrial metals as a pure play on China's capital spending. Bottom Line: We expect a combination of a stronger dollar, weaker Chinese growth and elevated oil inventories to produce a major reversal in industrial metals and oil prices. Chart I-17EM Stocks And U.S. ##br##TIPS Yields: Negative Correlation Question: Is your negative stance on EM contingent on weakness in DM growth? No, our negative stance on EM is not contingent on a relapse in DM growth. Some combination of the following key factors will trigger and drive weakness in EM risk assets: Higher U.S. real rates or a stronger U.S. dollar. Chart I-17 demonstrates the strong negative correlation between higher U.S. TIPS yields and EM share prices in the recent years. Lower commodities prices. Renewed weakness in China's economy. Our negative view on EM has and continues to be driven by our views on EM/China domestic demand/credit cycles, commodities and the U.S. dollar. Investment Conclusions Chart I-18EM/China Plays Are At Critical Juncture Exchange rates have been critical to financial market dynamics in recent years. This is unlikely to change. Odds favor another upleg in the U.S. dollar and a weaker RMB. As such, the outlook for EM risk assets is poor. EM currencies will be driven by a stronger dollar, a weaker RMB and lower commodities prices. EM share prices as well as global mining, and machinery stocks are at a critical juncture (Chart I-18). China-plays may soon start reacting to the PBoC's recent modest tightening as well as regulatory credit curtailment and begin to sell off in anticipation of weaker growth later this year. Global equity portfolios should continue underweighting EM stocks. Similarly, global credit (corporate bonds) portfolios should underweight EM sovereign and corporate credit. Finally, the outlook for weaker currencies does not bode well for EM local currency bonds. However, for fixed income investors we have several swap rate trades, relative value recommendations and yield curve positions that are published regularly in our Open Position Table on page 16. Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 Please refer to Trilogy of Special Reports on money/loan creation, savings and investment, titled "Misconceptions About China's Credit Excesses," dated October 26, 2016, and "China's Money Creation Redux And The RMB," dated November 23, 2016, links available on page 17. Equity Recommendations Fixed-Income, Credit And Currency Recommendations

How Long Is The Sweet Spot? Table 1Recommended Allocation The sweet spot on a baseball bat, scientists find,1 is the small area about two inches (5 cm) long, some six inches from the tip. The sweet spot for global risk assets may not be much bigger. The 22% rise in global equities since February last year has been driven by a "goldilocks" combination of recovering economic activity, quiescent inflation, and still-accommodative monetary policy. But, after such a strong rally, markets must walk a fine line - no slowdown in growth and no surprising tightening of monetary conditions - for prices to rise further. Our analysis suggests that they can, but the risk of a correction is rising. A lot of the better news of the past year has already been priced in. The price-to-sales ratio for U.S. stocks is close to an all-time high, and even the plain-vanilla 12-month forward PE ratio has reached 17.5x, the highest since 2002 (Chart 1). Volatility has fallen to a low level, with the VIX not rising above 12 over the past month, and the S&P500 index going 98 days without a one-day decline of 1% or more, the longest such period since 1995 (Chart 2). To a degree, this is justified by the recent strong pick-up in global growth. Sentiment indicators have accelerated since the election of President Trump, and even hard data is now showing the first signs of recovery (Chart 3) with, for example, U.S. retail sales rising 5.6% year-on-year in January, and core durable goods orders starting to follow the rise in companies' capex intentions (Chart 4). Similar positive economic surprises are visible in Europe, Japan, China and elsewhere. The problem is that further upside surprises are likely to be limited. Regional Fed NowCast surveys for Q1 real GDP growth are already at 2.5-3.1%. Consensus forecasts for S&P500 earnings growth in 2017 look about right at 10.5% but, with a stronger dollar and rising wages, are unlikely to be beaten. Chart 1Historically High Valuations Chart 2Time For A Pull-Back? Chart 3Hard Data Starting To Recover Too Chart 4Orders To Follow Capex Intentions Headline inflation has picked up (to 2.5% in the U.S. and 1.9% in the Eurozone), mainly because of higher oil prices, but core inflation remains sufficiently under control that central banks don't need to slam on the brakes. The rise in unit labor costs in the U.S. suggests that core PCE inflation will gradually move up to 2% during the year (Chart 5). The latest FOMC minutes revealed that members want a further rate hike "fairly soon", and BCA expects the Fed to raise three times this year (to which the futures market ascribes only a 36% probability). But Fed policy remains very accommodative (Chart 6), the European Central Bank is unlikely to end its asset purchases soon on account of political and banking system concerns, and the Bank of Japan remains committed to its 0% yield target for 10-year government bonds until inflation is well above 2%. Absent a powerful fiscal stimulus in the U.S. or a move by the "hard money" advocates in the Trump administration to change the Fed's modus operandi, we think its unlikely that a tightening of monetary policy will drag down asset prices. Chart 5Labor Costs Putting Pressure On Prices Chart 6Fed Policy Still Accomodative Risks certainly abound. The Trump administration could start a trade war with China. Its proposals for corporate and personal tax cuts could disappoint both in terms of their details and the timing of Congress's passing them. European politics remain a concern, with the probability of Marine Le Pen becoming French President increasing recently (though it remains small). But risk markets tend to rise on a wall of worry. Investor sentiment is not particularly bullish at the moment, with the bull/bear ratio among individual investors barely above 1 (Chart 7) and flows into equity funds in recent months not reversing the outflows of last year (Chart 8). Chart 7Retail Investors Not So Bullish Chart 8Equity Flows Are Still Tepid After a year of a strong cyclical risk-on rally, progress from now on will get tougher. A short-term change of direction is quite possible (and has already happened in some assets, with the yen moving back to 112 and the 10-year Treasury yield to 2.3%). But we expect economic growth to remain robust this year - with U.S. real GDP growth likely to come in close to 3% on the back of surprises in capex - which will push the 10-year Treasury yield above 3% by year-end. In this environment, we continue to favor equities over bonds, and maintain our pro-risk tilt in equity sectors, credit and alternative assets. Equities: U.S. equities have outperformed Eurozone ones by 5% year-to-date, mainly because of worries about Europe's political risk and the fragility of its banking sector. Though we think the political risks are overstated (except perhaps in Italy), we continue to prefer the U.S. in common currency terms because of our expectations of further dollar appreciation and because the lower volatility of the U.S. helps reduce the beta of our recommended portfolio. Emerging markets have outperformed global equities by 3% YTD, mainly on the back of stronger commodities prices. But we remain underweight EM because of the risks from a stronger dollar and rising global rates, concerns about protectionism and debt refinancing, and because of the likelihood that China's rebound will run out of steam over the next 12 months (Chart 9). Fixed Income: Rates have pulled back recently: long-term institutional investors have begun to find attraction in the long end of the U.S. Treasury yield curve at 2-3%, though speculative investors remain short (Chart 10). With the Fed likely to raise rates three times this year, inflation expectations to pick up further, and nominal GDP growth in the U.S. to reach 4.5-5%, we expect the U.S. 10-year yield to rise above 3%. We therefore remain underweight duration and prefer inflation-linked over nominal bonds. In the improving economic environment, we continue to like credit, but find valuations more attractive for investment-grade bonds than for high-yield. Currencies: Dollar appreciation has been on hold since January but we think the long-term trend remains in place because of the probable direction of relative interest rates. Neither Japan nor the Eurozone is likely to move towards monetary tightening over the next 12 months. Even if the Trump administration were to want a weaker dollar, a few tweets would not be enough to offset monetary fundamentals. And, while it is true that sentiment towards the dollar is already bullish, this has historically not precluded further appreciation, for example in the late 1990s (Chart 11). Chart 9EM Equities Correlated With China PMIs Chart 10Divergent Views On U.S. Bonds Chart 11Optimism Need Not Stop USD's Rise Commodities: The oil price remains close to its equilibrium level at around $55 a barrel, with the OPEC agreement largely holding but being offset by a production increase from the U.S. shale drillers, whose rig count has doubled since last May. We are neutral on industrial commodities: Chinese demand resulting from last year's reflationary policy is likely to be offset by the stronger dollar. Gold remains a useful portfolio hedge in a world of elevated geopolitical worries and inflation tail-risk, but is also negatively correlated with the U.S. dollar. Garry Evans, Senior Vice President Global Asset Allocation garry@bcaresearch.com 1 Please see, for example, "The Sweetspot of a Hollow Baseball or Softball Bat", by Daniel A. Russell, Pennsylvania State University, available at www.acs.psu.edu/drussell/bats/sweetspot.html Recommended Asset Allocation Model Portfolio (USD Terms)

The Tactical Asset Allocation model can provide investment recommendations which diverge from those outlined in our regular weekly publications. The model has a much shorter investment horizon - namely, one month - and thus attempts to capture very tactical opportunities. Meanwhile, our regular recommendations have a longer expected life, anywhere from 3-months to a year (or longer). This difference explains why the recommendations between the two publications can deviate from each other from time to time. Highlights In February, the model underperformed global equities and the S&P 500 in USD and local-currency terms. For March, the model slightly increased its allocation to stocks and cut its weighting in bonds (Chart 1). Within the equity portfolio, the allocation to Europe was increased. The model boosted its weightings to French and Australian bonds at the expense of Canadian and Swedish paper. The risk index for stocks, as well as the one for bonds, deteriorated in February. Feature Performance In February, the recommended balanced portfolio gained 2.1% in local-currency terms, and 0.2% in U.S. dollar terms (Chart 2). This compares with a gain of 3% for the global equity benchmark and a 3.3% gain for the S&P 500. Given that the underlying model is structured in local-currency terms, we generally recommend that investors hedge their positions, though we provide suggestions on currency risk exposure from time to time. The high allocation to bonds continued to hold back the model's performance. Chart 1Model Weights Chart 2Portfolio Total Returns Weights The model increased its allocation to stocks from 53% to 57%, and cut its bond weighting from 47% to 43% (Table 1). Table 1Model Weights (As Of February 23, 2017) The model increased its equity allocation to Dutch and Swedish equities by 4 points each, Germany and New Zealand by 2 points each, and France and Emerging Asia by 1 point each. Weightings were cut in Italy by 4 points, Latin America by 3 points, Spain by 2 points, and Switzerland by 1 point. In the fixed-income space, the allocation to Australia was boosted by 8 points, France by 6 points, and Germany by 4 points. The model cut its exposure to Swedish bonds by 9 points, Canadian bonds by 6 points, U.S. and U.K. bonds by 3 points each, and Kiwi bonds by 1 point. Currency Allocation Local currency-based indicators drive the construction of our model. As such, the performance of the model's portfolio should be compared with the local-currency global equity benchmark. The decision to hedge currency exposure should be made at the client's discretion, though from time to time, we do provide our recommendations. The most recent bout of dollar depreciation was halted in February. Our Dollar Capitulation Index is below neutral levels. However, it is not extended, meaning that it does not preclude renewed dollar weakness in the near term. That said, assuming no major negative economic surprises, a relatively more hawkish Fed versus its peers should provide support for the dollar (Chart 3). Chart 3U.S. Trade-Weighted Dollar* And Capitulation Capital Market Indicators The risk index for commodities was little changed in February. The model continues to avoid this asset class (Chart 4). The risk index for global equities rose to its highest level since early 2010, mostly on the back of deteriorating value. Despite this, the model slightly increased its allocation to equities (Chart 5). Chart 4Commodity Index And Risk Chart 5Global Stock Market And Risk The rally in U.S. stocks - driven by optimism about the economic outlook - pushed the value component of the risk index into expensive territory. The model kept a small allocation in U.S. equities. A change in the perception about the ability of the new U.S. administration to boost growth remains a risk for this market (Chart 6). The risk index for euro area equities continues to deteriorate. However, it remains lower than its U.S. counterpart. The continued flow of solid economic data and a weaker currency should bode well for euro area stocks, although political uncertainty is a potential headwind (Chart 7). Chart 6U.S. Stock Market And Risk Chart 7Euro Area Stock Market And Risk All three components of the risk index for Dutch equities are close to neutral levels. As a result, despite the recent deterioration in the overall risk index, it remains one of the lowest among the markets the model covers (Chart 8). The risk index for Swedish stocks worsened. However, the model increased its allocation to this bourse. Swedish equities would be a beneficiary of the continued risk-on environment (Chart 9). Chart 8Netherlands Stock Market And Risk Chart 9Swedish Stock Market And Risk The momentum indicator for global bonds is less stretched in February. Meanwhile, despite its latest decline, the cyclical indicator continues to signal that the positive global economic backdrop is firmly bond-bearish. Taken all together, the risk index for bonds deteriorated in February, although it still remains in the low-risk zone (Chart 10). U.S. Treasury yields moved sideways in February as investors await more guidance from the Fed on the timing of the next hike. A bond-negative cyclical indicator coupled with the unwinding of oversold conditions - as per the momentum measure - led to a deterioration in the risk index for U.S. Treasurys. The latter is almost back to neutral levels. The model trimmed the allocation to this asset class (Chart 11). Chart 10Global Bond Yields And Risk Chart 11U.S. Bond Yields And Risk The momentum indicator remains the main driver of the risk index for Canadian bonds. As a result, the less extreme momentum reading translated into an increase in the risk index for this asset class. (Chart 12). The risk index for Australian bonds moved lower in February, reflecting improvements in all three of its components. The model included the relatively high-yielding Aussie bonds in the portfolio. (Chart 13). Chart 12Canadian Bond Yields And Risk Chart 13Australian Bond Yields And Risk The cyclical indicator for euro area bonds is near expensive levels, and the momentum indicator shows heavily oversold conditions. These two measures are offsetting the cyclical one that is sending a bond-bearish message. While the overall risk index for euro area bonds is in the low-risk zone, the country allocation is concentrated in French paper (Chart 14). The risk level for French bonds is seen as low thanks to oversold momentum. French presidential elections are probably the most important political event in Europe this year. Whether the models' heavy allocation to this asset pans out hinges to a certain extent on the reduction of investor anxiety about this political risk (Chart 15). Chart 14Euro Area Bond Yields And Risk Chart 15French Bond Yields And Risk The 13-week momentum measure for the dollar broke below the zero line, and is currently sitting on its upward-sloping trendline, drawn from the 2010 lows, that has been broken only once before. Meanwhile, the 40-week rate of change measure is still suggesting that the dollar bull market has more legs on a cyclical horizon. Monetary divergences should lend support to the dollar over the cyclical horizon, although the new administration's attempts to talk down the dollar as well as heightened policy uncertainty could translate into more volatility (Chart 16). The weakening trend in the yen hit a snag two months ago, as the 13-week momentum measure reached the lows that previously foreshadowed a consolidation phase after sharp depreciations. This short-term rate-of-change measure has bounced smartly this year reaching a critical level. Meanwhile, the 40-week rate-of-change measure is not warning of a major change in the underlying trend which remains dictated by BoJ's dovish bias (Chart 17). EUR/USD has been gravitating towards 1.05 over the course of February. The short-term rate-of-change measure seems to be holding at the neutral level, while the 40-week rate-of-change measure is in negative territory, but hardly stretched. Political uncertainty has the potential to drive the euro in near term, but the longer-term outlook is mostly a function of the monetary policy divergence between the ECB and the Fed (Chart 18). Chart 16U.S. Trade-Weighted Dollar* Chart 17Yen Chart 18Euro Miroslav Aradski, Senior Analyst miroslava@bcaresearch.com

Highlights Gold volatility is trending lower, suggesting unresolved economic and political issues are diminishing, and investors' confidence in the global economy is improving. This is a false positive. Uncertainty is elevated. "Known unknowns" loom large: U.S. and Chinese fiscal policy, which drive USD dynamics and commodity supply and demand, are unresolved; The outcome of French and Italian elections could shock the euro zone; The reaction functions of systemically important central banks as they navigate these risks remain opaque. Given gold's exquisite sensitivity to political and policy nuances globally, our attention naturally turns to it when we look for ways to position in the face of this political and policy-related uncertainty. Our analysis suggests the low volatility in gold markets is the result of traders and investors being driven to the sidelines, where they await clarity re politics and policy. This is keeping trading volumes low: No one wants to be long or short lacking critical information necessary to take a view on the evolution of asset-price paths. Lower trading volumes, therefore, reflect a paucity of information in the price-discovery process, which, all else equal, will tend to keep commodity prices range bound until new information arrives to propel them in one direction or the other. With fewer prints going up across markets, realized and implied volatilities remain low ... for the moment. Energy: Overweight. We are taking profits basis today's close on our WTI Dec/17 vs. Dec/18 backwardation spread initiated February 9 at -$0.11/bbl. We also will be taking profits on our Dec/19 WTI vs. Brent spread, elected February 6 at +$0.07/bbl, after WTI traded premium to Brent in anticipation a U.S. border-adjusted tax would be enacted. Base Metals: Neutral. Copper remains well bid amid transitory supply outages. Workers resumed their strike at BHP's Escondida mine in Chile, while Anglo American temporarily suspended work at its El Soldado mine in a regulatory dispute, according to Metal Report. Freeport-McMoRan declared force majeure on Grasberg deliveries. Precious Metals: Neutral. We are withdrawing our gold buy-stop, and are recommending long gold options spreads to position for higher volatility (see below). Ags/Softs: Underweight. Corn and wheat came under selling pressure over the past week, but still are holding trend-line support from end-2016. We continue to monitor these markets for signs of a short-term rally. We remain strategically bearish, however. Feature While we believe the Efficient Market Hypothesis (EMH) holds most of the time - at least in semi-strong form (i.e., all public information is fully reflected in prices) - traders and investors now find themselves in something of a quandary.1 Much of the information needed to assess future paths for asset prices and form expectations for returns has yet to be revealed. In other words, there are large parts of markets' information sets made up of "known unknowns," which, once resolved, will be of enormous consequence to the paths taken by different asset prices. This is particularly true for gold. Our analysis suggests this lack of information is keeping trading volumes in gold markets low. As a result, the price-discovery process is stymied, which, all else equal, tends to keep prices range bound until new information arrives to propel them in one direction or the other. With fewer prints going up across markets, realized and implied volatilities remain low. Investors accustomed to viewing low volatility as an indication unresolved economic and political issues are diminishing therefore have to adapt to a new reality, one in which low volatility actually is the product of heightened uncertainty (Chart of the Week). Granted, financial stress is low. This contributes to lower volatility, particularly in gold, which is highly sensitive to U.S. real rates and USD levels. However, we find low trading volumes in gold markets also are responsible for the lower-trending realized and implied volatility prevailing in in gold markets (Chart 2).2 Chart of the WeekVolatility Is Low, Despite Uncertainty Being High Chart 2Realized And Implied Gold Vols Are Trending Lower This suggests there is an opportunity to position ahead of the resolution of these "known unknowns" in the gold market, given the low volatility levels we see. This is driven largely by our view that there are numerous risks in near- and longer-term price distributions, which imply much fatter tails than markets are pricing in at the moment.3 Indeed, the CBOE Gold VIX is running at ~ 13.5% presently vs. a post-Global Financial Crisis (GFC) average of 18.8% p.a.4 First, The Fat Left Tail There are a number of risks pumping up the left tails of many commodity price distributions - e.g., how long China will continue to tighten fiscal policy (Chart 3), and the effect this will have on the prices of base metals and bulk commodities like iron ore and steel. And, of course, markets will continue to hang on every utterance of Federal Reserve officials, attempting to discount rate-hike probabilities and their implications for the USD and real rates, the critical drivers of gold prices (Chart 4). Chart 3China Fiscal Stimulus Grinds To A Halt Near term, these risks will continue to loom large, but they are dwarfed by a possible border-adjusted tax (BAT) being imposed in the U.S. In our estimation, this is the largest left-tail risk we've identified for commodity markets over the near term. It is being championed by Republican leaders in the U.S. House of Representatives - led by Speaker Paul Ryan and Ways and Means Chairman Kevin Brady.5 A BAT would raise the price of commodities subject to the tax in the U.S. Domestically, producers of commodities subject to the tax would benefit from this increase in prices, since it would boost their revenues and incentivize increased domestic production. This would be used to displace imports and take market share from exporters to the U.S. Once the domestic market has been saturated with the higher domestic output, U.S. producers would turn to export markets to sell their increased output. A BAT would shrink the U.S. trade deficit, which would, all else equal, raise the trade-weighted value of the USD. Our expectation is there is a 50:50 chance a BAT is enacted, but that it will exclude oil and apparel. We expect the USD would appreciate ~ 10% on the back of this scheme, on top of the 5% increase in the value of the dollar we already were expecting from the Fed's continued push to normalize monetary policy. On the back of this 15% appreciation in the USD over the next year or so, commodity prices ex U.S. would increase in local-currency terms, which would crimp demand in EM and DM economies. On the supply side, the cost of producing commodities ex U.S. would fall in local-currency terms, which would increase supply at the margin. Net, net: A BAT would cause global commodity demand to fall and supply to increase, which would, all else equal, send a deflationary impulse back to the U.S., and DM and EM economies. Fat Right Tails Permanent and transitory commodity supply losses constitute large right-tail risks for investors, in our estimation, as does stronger-than-expected demand. Chief among these are ongoing losses in copper markets in the near term, which we believe to be transitory. The massive $1 trillion+ capex cuts registered in the oil markets in the wake of the price collapse induced by OPEC's market share war leave us with low confidence our oil-price expectation of $55/bbl will prevail beyond 2018. Near term, however, the timing and type of infrastructure projects that will be funded under the Trump administration's forthcoming fiscal roadmap, and whether Congress will be supportive represent the largest right-tail risk for gold markets. Highly expansive fiscal stimulus could spur inflation in the U.S., given this stimulus will be hitting an economy that already is at or near full employment. Given the synchronized global economic recovery currently underway, we believe an inflationary impulse could percolate into near-term inflation realizations, and into inflation expectations longer term. Chart 4Markets Will Continue To Hang On Every Fed Utterance This elevated inflation risk will be bullish for gold, as we showed in recent research.6 Indeed, we noted, "All else equal, with the U.S. labor market at or close to full employment, and the Trump administration signaling its desire for stimulative fiscal policy, we would be inclined to look for inflation hedges within commodities that are highly sensitive to rising inflation." Topping that list is gold, in our estimation. Taking A View On Volatility Chart 5Gold Provides A Good Hedge For Equity Volatility Volatility is trading-market shorthand for the annualized standard deviation of expected returns for an underlying asset. It is a parameter used to price options. Options markets are unique in that they allow investors to take a view on the dispersion of the expected returns of the asset against which the option is written.7 Volatility is a calculated value, whereas the other components of an option's price - i.e., the underlying asset's price, the strike price, time to expiration, and interest rates - all are known inputs. Volatility, like the price of the underlying asset, therefore is "discovered" when a trade occurs. After an option trades and its premium becomes known, an option-pricing model - e.g., the Black-Scholes-Merton model - can be run backwards, so to speak, to see what level of volatility solves the pricing model for the value that cleared the market. This is known as the option's implied volatility, because it is the expected standard deviation of returns implied by the price at which the option clears the market. One reason investors and traders buy and sell options is to express a view on implied volatility. Option buyers who think the market is underestimating the likelihood of sharply higher or sharply lower returns can express this view by buying out-of-the-money options - calls or call spreads on the upside, puts or put spreads on the downside. This can arise for any number of reasons, but they all boil down to one essential point: Option buyers think there is a higher probability returns will be higher or lower during the life of an option than what is being priced in the options market presently.8 Option sellers, on the other hand, are expressing the opposite view. We believe the fat-tail risks we've discussed in this article are not being fully reflected in the options markets most sensitive to this information, among them the gold market. Our own assessment of these risks implies much fatter tails than we currently observe in the out-of-the-money gold options, as noted above. For this reason, we are recommending investors consider buying put spreads and call spreads against June-delivery gold. We will look to get long Jun/17 $1,200/oz puts vs. selling $1,150/oz puts, and getting long $1,275/oz calls vs. selling $1,325/oz calls, basis tonight's closing levels for the underlying contract. This is a low-risk strategic recommendation, with the put and call spreads roughly equidistant from where the Jun/17 gold contract is trading. The motivation for this recommendation is simple: We believe volatility is low, given the "known unknowns" and their associated fat tails, which are not being accounted for in options prices. This makes these options cheap. Gold Can Hedge Equity Risk As Well Our analysis reveals gold provides a good edge against rising equity volatility, as measured by the CBOE's equity volatility index (CBOE VIX).9 From 1995 to the present, gold's monthly percentage returns outperformed those of the S&P 500 61% of the time when the VIX was increasing, and 36% of the time when the VIX was decreasing (Chart 5). Over the entire sample, gold outperformed the S&P 500 in average by 2.25% in periods of increasing equity volatility as measured by the VIX. However, if we focus only on sub-sample periods where the VIX was increasing but from an already-elevated level (20% or above), gold returns outperformed S&P 500 returns by 4.57% on average. Given our assessment that current volatility is abnormally low, particularly for gold, we believe the gold options exposure recommended here will provide investors protection against increasing equity volatility, as well. Moreover, if market sentiment changes and volatility begins to increase significantly, our analysis provides evidence that gold's volatility-risk-mitigation properties increase even more when the VIX is already at a high level. Bottom Line: Markets lack sufficient information to fully price the risks in potential fat-tail events on the down- and up-side of commodity price distributions. We believe gold options - particularly put and call spreads - offer a low-risk way to position for the eventual resolution of this uncertainty. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com Hugo Bélanger, Research Assistant hugob@bcaresearch.com 1 For an excellent discussion of the EMH, please see Timmermann, Allan, and Clive W.J. Granger (2004), "Efficient market hypothesis and forecasting," in the International Journal of Forecasting, Vol. 20, pp. 15 - 27. 2 When we regress CBOE gold volatility on first-nearby gold futures volume using daily data over January 2016 to February 2017 using an error-correction model, we find trading volume explains ~ one-third of the CBOE implied gold volatility's level. 3 Many of these risks are geopolitical in nature, which our colleague Marko Papic considers at length in "A Fat-Tails World," published February 22, 2016, in BCA Research's Geopolitical Strategy Weekly Report, available at gps.bcaresearch.com. 4 We mark the post-GFC period as Jan/10 to present. 5 A BAT essentially would tax imports coming in to the U.S. and subsidize exports, using proceeds to reduce corporate taxes. We are not ready to pronounce the BAT dead, as some pundits already have. We think the market's 20% probability that such a tax becomes law is too low: We give it a 50:50 chance of passage, albeit in a watered down form likely calling for a 10% tax on imports, which likely will not include oil or apparel. Base metals and agricultural imports likely would be taxed under this scheme. We analyzed the commodity impacts of this proposed scheme in "Taking a BAT To Commodities," which was published in the January 26, 2017, issue of BCA Research's Commodity & Energy Strategy Weekly Report, available at ces.bcaresearch.com. 6 Please see issue of BCA Research's Commodity & Energy Strategy Weekly Report, "Gold Will Perform...," dated February 2, 2016, available at ces.bcaresearch.com. 7 Call options give the buyer the right to go long an underlying asset at the price at which an option contract is struck - i.e., the option's strike price. Puts give option buyers the right to go short the underlying asset at the price at which the contract is struck. While an option buyer is not required to ever exercise an option, option sellers must take the other side of the deal if the buyer chooses to exercise. Option buyers pay a premium for the put or call they purchase. 8 This probability also can be expressed in terms or price levels, which allows investors to take an explicit view on the likelihood of a particular price being realized during the life of the option being purchased. Please see Ryan, Bob and Tancred Lidderdale (2009), "Energy Price Volatility and Forecast Uncertainty," published by the U.S. Energy Information Administration, for a discussion of options markets and implied volatility. "Appendix II: Derivation of the Cumulative Normal Density for Futures Prices" beginning on p. 22 shows how to transform the returns distribution into a price distribution. It is available at https://www.eia.gov/outlooks/steo/special/pdf/2009_sp_05.pdf. 9 Our results are similar to those reported in "Gold is still a good hedge when volatility rises," by Russ Koesterich, CFA, published by Blackrock on its Blackrock Blog September 9, 2016. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed In 2017 Summary of Trades Closed in

Highlights Price inflation is paradoxically deflationary for European consumers, because there is no feedthrough from price inflation to wage inflation. Whenever price inflation has risen towards the ECB's highly-misguided 2% target, euro area real wages have gone into recession. The same is true in the U.K. Do not expect a structural sell-off in high-quality bonds. Go overweight the broad-based Eurostoxx600 versus the bank-heavy Eurostoxx50. Stay overweight the international dollar-earning FTSE100 versus the more domestic pound-earning FTSE250. Feature We have a love-hate relationship with inflation. Love, if the inflation refers to our wages. Hate, if the inflation refers to our weekly grocery bill. Put another way, inflation is good for our purchasing power when wages are going up faster than prices; it is bad when prices are going up faster than wages. Unfortunately, recent inflation has been unequivocally bad for European purchasing power. Through the past 7 years, euro area nominal wages have been growing at a remarkably steady 1-2% clip. Whereas price inflation has swung between -0.5% and 3% (Chart I-2). Therefore, whenever price inflation has stayed close to 0% (the true definition of price stability), real wages have grown very healthily. But whenever inflation has risen towards the ECB's highly-misguided 2% target, euro area real wages have gone into recession (Chart of the Week). Chart I-1The Inflation Paradox: When Price Inflation Rises to 2%, Real Wages Go Into Recession Chart I-2Nominal Wages Have Been Growing At A Remarkably Steady 1-2% The same is true in the U.K. There has been no feedthrough from price inflation to wage inflation (Chart I-3 and Chart I-4). If anything, an inverse relationship has existed. Hence, whenever inflation has declined, it has boosted real wages. And whenever inflation has risen, it has choked real wages (Chart I-5 and Chart I-6). Chart I-3Very Little Connection... Chart I-4...Between Price Inflation And Wage Inflation Chart I-5When Price Inflation Has Declined,##br## It Has Boosted Real Wages Chart I-6When Price Inflation Has Increased,##br## It Has Choked Real Wages Households Dislike 2% Price Inflation An argument we frequently hear is that highly indebted economies need higher inflation to 'inflate away their high debts'. But this logic only works if inflation is boosting the incomes of those burdened with the high debt, such as households. The problem, as we have just seen, is that there has been very little connection between the price inflation that central banks are targeting and the wage inflation that eases households' debt burdens. To its credit, the Bank of England recognises this paradox. "Continued moderation in pay growth and higher import prices following sterling's depreciation are likely to mean materially weaker household real income growth over the coming few years" 1 Inflation is ultimately a transfer of resources from those paying the higher prices to those receiving them. In a closed economy, the winners and losers might balance out. However, Europe is a large net importer of food and energy, whose demand is inelastic and whose prices are denominated in dollars. Therefore, currency weakness transfers resources from domestic consumers to foreign producers. As the BoE goes on to say: "Over the next few years, a consequence of weaker sterling is that the higher imported costs resulting from it will boost consumer prices... and the hitherto resilient rates of household spending growth will slow as real income gains weaken." Exactly the same dynamic applies to the euro area as a consequence of the weaker euro. The difference is that sterling's Brexit-induced slump was out of the BoE's control, whereas the euro's weakness is a direct consequence of the ECB's extreme and experimental monetary easing. The ECB is keen to tell us about the benefits of its extreme monetary easing; it is less keen to tell us about the costs. However, we believe that the benefits have diminished while the costs are rapidly rising. And absent a major shock, the ECB should end its risky experiment. What's Up With Wage Growth? The intriguing question is: why has there been little connection between price inflation and wage inflation? The BoE observes that pay growth has remained persistently subdued by historical standards - strikingly so in light of the decline in the rate of unemployment to below 5%. This outcome is likely to reflect a substantial decline in the 'equilibrium unemployment rate', the point at which wage pressures start to bubble up. The explanation comes from the type of jobs created in recent years. ECB research points out that the dynamics of wages not only reflect changes in wages at the individual level, but are also influenced by changes in the composition of employment. "The structure of recent employment creation may have contributed to low wage growth in the euro area. Since the second quarter of 2013, employment creation in the euro area has been stronger in sectors associated with relatively lower wage levels and wage growth rates. This employment composition effect puts a drag on average wage growth." 2 Automation and Artificial Intelligence (AI) are major drivers of this composition effect. Moreover, as we argued in The Superstar Economy: Part 2,3 the effect has much further to run. "Many of the jobs that AI will destroy - like credit scoring, language translation, or managing a stock portfolio - are regarded as skilled, have limited human competition and are well-paid. Conversely, many of the jobs that AI cannot (yet) destroy - like cleaning, gardening, or cooking - are relatively unskilled and are low-paid." With well-paid jobs being displaced by low-paid jobs, job creation itself might still seem very healthy and the unemployment rate might be falling to levels associated with 'full employment' - prompting some people to warn that wage inflation is about to take off. Except it won't, for two reasons: first, the AI-displaced formerly well-paid workers are downshifting to lower-paid work; second, the added supply of labour competing for the lower-paid work keeps a lid on the wages for that lower-paid work. In the U.S., the Federal Reserve Board of San Francisco points out that: "As long as employers can keep their wage bills low by replacing or expanding staff with lower-paid workers, labour cost pressures for higher price inflation could remain muted for some time." 4 A further point is that if employment creation is in jobs with lower wages, wage growth, and job security, then it will also constrain credit growth. Lacking income growth or security, households will be unwilling to borrow and banks will be unwilling to lend. Absent strong credit growth, we subscribe to a monetarist conclusion: a generalised and sustained inflation - a wage-price spiral - cannot take hold. Some Investment Considerations For the foreseeable future, there will be little feedthrough from price inflation to wage inflation. So whenever price inflation picks up - as is now happening in the U.K. and the euro area - it will choke real wages. Therefore paradoxically, price inflation will be deflationary for European consumers. This will prevent a structural sell-off in high-quality bonds. For a U.K. equity portfolio at this juncture, it means tilting towards international exposure. Stay overweight the international dollar-earning FTSE100 versus the more domestic pound-earning FTSE250 - especially given that sterling could come under renewed pressure after the U.K. formally files for its divorce from the EU (Chart I-7). For a broader European equity portfolio, prefer non-financials over financials. A very easy way to implement this is to go overweight the broad-based Eurostoxx600 versus the bank-heavy Eurostoxx50 (Chart I-8). Chart I-7Overweight The International Dollar-Earning ##br##FTSE100 Versus The FTSE250 Chart I-8Overweight The Broad Eurostoxx600##br## Versus The Bank-Heavy Eurostoxx50 Dhaval Joshi, Senior Vice President European Investment Strategy dhaval@bcaresearch.com 1 From the Bank of England Monetary Policy Summary and minutes of the Monetary Policy Committee meeting on February 1, 2017. 2 From the ECB Economic Bulletin, Issue 3 / 2016: Recent wage trends in the euro area. 3 Published on January 19, 2017 and available at eis.bcaresearch.com 4 From the FRBSF Economic Letter March 7, 2016: What's Up with Wage Growth? Fractal Trading Model* This week's recommendation is a commodity pair-trade: long tin / short copper. 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-9 * 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 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. 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 Global manufacturing inventories are low but this does not guarantee higher share prices for global cyclical stocks. If an increase in inventories is accompanied by strengthening final demand, it will be very bullish for the global business cycle. If final demand growth falters, global cyclical plays will relapse amid rising inventories. China's inventory depletion has been due to the large fiscal and credit impulse in the past 12 months - i.e., improving final demand has been instrumental to inventory shedding. Looking forward, the mainland's aggregate credit and fiscal impulse seems to have topped out raising the odds of a reversal in EM/China plays sooner than later. The risk/reward of EM/China plays remains unattractive. Feature Global Manufacturing Inventories Global manufacturing inventories have been depleted over the past 12 months, and inventory levels are generally low (Chart I-1 and Chart I-2). Chart I-1Global Manufacturing Inventories Are Low Chart I-2Global Manufacturing Inventories Are Low Could inventory re-stocking extend the current manufacturing cycle recovery worldwide? Will low inventories and re-stocking in China lengthen the nation's business cycle upswing? Chart I-3 demonstrates inventory cycles and manufacturing production within manufacturing-intensive economies. The correlation is not stable. Currently, this entails that low manufacturing inventories and a potential rise in inventories over the course of this year do not guarantee acceleration in industrial output growth. Having reviewed manufacturing inventory cycles and their correlation with share prices, we conclude that the key to share prices is final demand - not inventory swings. Manufacturing inventories have dropped in the past 12 months because final demand has been robust (Chart I-4). Historically, periods of re-stocking have often coincided with poor equity market performance. Indeed, Taiwanese, Korean, Japanese and German non-financial share prices have no stable correlation with their respective manufacturing inventory cycles (Chart I-5). In short, manufacturing inventories could rise in the months ahead, but this does not guarantee higher share prices in cyclical industries. Chart I-3Inventories And Production ##br##Are Not Always Correlated Chart I-4Robust Demand Has Led ##br##To Inventory Depletion Chart I-5Non-Financial Share Prices And##br## Inventories: Little Correlation By and large, the outlook for corporate profits is contingent on final demand rather than re-stocking. All of the above confirms that inventories are a residual of demand and supply. Stronger-than-expected demand is bullish for share prices, though it also often coincides with declining inventories. By contrast, rising inventories typically reflect demand falling behind output growth (one can define it as involuntary re-stocking) and these periods are not favorable for share price gains in cyclical industries. One caveat is that there could be a re-stocking cycle amid strengthening demand or, in other words, voluntary re-stocking. If this transpires in the coming months, it will be extremely bullish for share prices as it will supercharge output growth. While the latter scenario - inventory re-stocking amid strengthening final demand - could very well occur within the advanced economies this year, odds of such positive dynamics are low in EM/China. Bottom Line: Share prices in global cyclical sectors are driven by swings in final demand - not in inventories. Going forward, global manufacturing inventories will rise. If this rise is accompanied by strengthening demand, it will be very bullish for the global business cycle. Otherwise, global cyclical plays will relapse as inventories rise. What Drives China's Inventory Cycles Chart I-6 shows that China's manufacturing inventories typically deplete when the credit and fiscal impulse is rising, and vice versa. China's manufacturing inventories have been exhausted because demand has been strong in the past 12 months. In turn, demand strength has originated from the country's massive fiscal and credit stimulus push from the first half of 2016. Chart I-6China: Strong Policy Stimulus Led To Manufacturing Inventories Reduction That said, China's aggregate fiscal and credit impulse seems to have recently rolled over, pointing to a top in its manufacturing mini-cycle and commodities prices (Chart I-7). This signals a potential deceleration in final demand. On the whole, the ongoing modest tightening by the People's Bank of China and by the bank regulator (the China Banking Regulatory Commission) amid a lingering credit bubble is raising the odds of a moderate credit slowdown in the months ahead. Even modest credit growth deceleration will result in a negative credit impulse (Chart I-8, top panel). Meanwhile, the mainland's fiscal impulse has already dropped (Chart I-8, bottom panel). Chart I-7China: Aggregate Credit And Fiscal##br## Stimulus Has Topped Out Chart I-8China: A Breakdown Of Credit ##br##And Fiscal Impulses On the whole, these developments are leading us to maintain our negative bias toward EM risk assets and China plays. What has gone wrong in our view/analysis on China in the past 12 months is that the nation's credit growth has stayed much stronger than we expected. In our April 13, 2016 report,1 we did a scenario analysis and argued that China's large fiscal stimulus push would be offset by a negative credit impulse if credit growth slowed from 11.5% to below 10%. In reality, credit growth has been between 11.5-12.5%, producing a positive credit impulse. Barring tightening by the central bank or bank regulators, mainland banks can continue originating loans/money at a double-digit pace, as they have been doing for many years (Chart I-9). In general, commercial banks do not need savings to create money/loans and there are few limits on Chinese banks originating loans "out of thin air," as we argued in our Trilogy of Special Reports on money/loan creation, savings and investment.2 Chart I-9China's Credit/Money Growth##br## Remains Rampant Therefore, if credit growth does not slow, our negative view on China's growth will be off-the-mark again. The pressure point in such a case will be the exchange rate. Unlimited money creation/oversupply of local currency is bearish for the value of the RMB. The RMB will continue depreciating, but it is not certain if it will hurt EM risk assets. It is a major consensus view nowadays that the Chinese authorities will not allow growth to suffer ahead of the Party Congress in autumn of this year. Yet, the PBoC and bank regulators are modestly tightening to "normalize" credit growth. Some clients may wonder why we are placing so much emphasis on the rollover of credit and fiscal impulses now, while placing little emphasis on these same indicators in 2016 when they were recovering. The rationale is as follows: when there is a credit bubble - as there is in China now - we tend to downplay the importance of policy easing and put more significance on policy tightening. The opposite also holds true: when the credit/banking system is healthy, we tend to downplay the impact of moderate policy tightening and put greater emphasis on policy easing. In a credit bubble, it does not take much tightening to trigger a downtrend that unwinds excesses. Similarly, moderate tightening in a healthy credit system should not be feared. From a big picture perspective, we turned bearish on China's growth several years ago due to the formation of a credit bubble. The bubble has only gotten larger and an adjustment has not yet even started. This does not justify altering our fundamental assessment of China's growth outlook. It would have been ideal to turn positive tactically on EM/China plays a year ago. Unfortunately, we did not do that. Presently, chasing the market higher might not be the best investment idea. Based on all this and given: the sharp rally in EM/China plays and widespread investor complacency and consensus that "everything" will be fine before the end of this year; modest tightening in Chinese monetary policy amid lingering credit and asset (property and the corporate bond market) bubbles; our outlook for higher U.S. bond yields and a stronger U.S. dollar; the fact that financial markets are forward looking, and timing is impossible; We believe the risk/reward of EM/China plays remains unattractive. In regard to EM ex-China, as we documented in last week's report, domestic demand in the developing economies has not recovered at all, or is mixed at best. DM final demand strength and global manufacturing inventory rebuilding will certainly help Korea and Taiwan, but not other emerging economies. The most important variables for other EM economies including China are domestic demand and/or commodities prices. If commodities prices relapse along with China's credit and fiscal impulse (Chart I-7, bottom panel), EM financial markets will suffer regardless of the growth trends within advanced economies. In fact, strong U.S. growth could lead to higher U.S. interest rate expectations and prop up the U.S. dollar. This will also be a bad omen for EM and commodities. Bottom Line: China's inventory depletion has been due to the large fiscal and credit impulse in the past 12 months - i.e., improving final demand has been instrumental to inventory shedding. Looking forward, the mainland's aggregate credit and fiscal impulse seems to have topped out, raising the odds of a reversal in EM/China plays sooner than later. Industrial Metals Inventories And Prices There is no good data reflecting industrial metals inventories globally. London Metal Exchange and Shanghai Futures Exchange data are likely not indicative of global metals stockpiles. China accounts for close to 50% of global demand for industrial metals, and its demand is critical to prices. Given that the large spike in metals prices in the past several months has coincided with improving Chinese economic data, one would expect the mainland to be the driving force behind the rally. However, Chart I-10 demonstrates that China's imports of industrial metals actually contracted in 2016. This is puzzling, but we have to take it at face value. The top panel of Chart I-11 depicts that traders' net long positions in copper are at a six-year high. This might partially explain the rally in copper in the recent months. Chart I-10China's Import Of Base Metals##br## And Base Metals Prices Chart I-11Traders Are Long ##br##Copper And Oil Clearly, China has been depleting its stock of industrial metals, and is likely primed to increase its imports. Nevertheless, periods of metals re-stocking by the mainland have historically not entailed higher industrial metals prices (Chart I-10). On the contrary, rising Chinese imports of metals have actually coincided with falling prices. One can interpret this relationship as China buying industrial metals when prices are falling. This is consistent with China attempting to buy commodities on dips. As to metals inventories in China, the picture is as follows: Steel inventories have plummeted and are low (Chart I-12). One can safely argue that there will be an inventory re-stocking cycle in China. Nevertheless, it is highly uncertain if this will be bullish for steel prices and steel stocks. In fact, there has been a mild negative correlation between steel prices and inventories; historically, when inventories have risen, prices declined (Chart I-12, top panel). This confirms that inventory levels are a residual of demand and supply, and prices are often driven by final demand - not inventories. This is also corroborated by the bottom panel of Chart I-12, which illustrates that share prices of global steel companies are sometimes negatively correlated with China's steel inventories. Stock prices occasionally sell off when inventories rise, and rally when inventories are shrinking. In contrast to steel and steel products, iron ore inventories have risen, and it seems the re-stocking cycle is well advanced (Chart I-13). Chart I-12China: Steel Inventories And Prices Chart I-13China: Iron Ore Inventories And Prices Yet, again there is no strong correlation between inventories and prices of iron ore (Chart I-13). In our discussions with clients, investors often attribute the rally in industrial metals in general and steel prices in particular over the past 12 months to supply cutbacks in China. While supply reductions have helped in the case of certain metals, it is also evident that the rally in industrial commodities has been driven by rising demand globally and in China. First, China's aggregate credit and fiscal impulse was positive until very recently, implying strengthening demand and thereby higher metals prices. Second, if there were only production cutbacks in steel and other commodities and not demand recovery, the mainland's manufacturing PMI would not have risen (Chart I-14). Finally, steel production has risen both in China and the rest of the world (Chart I-15). Hence, world steel supplies have expanded in the past 12 months. Given this has coincided with rising steel prices, it confirms there has been notable improvement in demand for steel. Chart I-14China: Steel Prices Are Up ##br##Because Of Strong Demand Chart I-15Chinese And Global ##br##Steel Production We are not experts in the ebbs and flows of commodities supplies, but it seems the Chinese government's mandated steel capacity cutbacks have not prevented rising steel output in China. In the meantime, rising prices amid rising production and falling inventories are indicative of robust final demand for many metals. Bottom Line: Industrial metals prices have risen because demand in the real economy and among financial investors has been strong. That said, a rollover in China's fiscal and credit impulse and a strong U.S. dollar will likely create headwinds for industrial metals prices over the course of this year. A Word About Oil Inventories OECD oil product inventories have continued to rise, despite supply cuts (Chart I-16, top panel). At the same time, our proxy for change in China's oil inventories has been very elevated for a while, depicting strategic and/or commercial inventory building on the mainland (Chart I-16, bottom panel). It is true that supply curtailments have been instrumental to the rally in oil prices, but the continued inventory buildup also indicates that supply is still outpacing demand. Besides, traders' net long positions in crude have spiked close to their 2014 highs (Chart I-11, bottom panel). This corroborates that demand for crude, like for copper, has partially been financial rather than from final consumers. Finally, U.S. rig counts have recovered somewhat, which may be indicative of a continued rise in America's oil output (Chart I-17). Chart I-16Oil Inventories Keep On Rising Chart I-17U.S. Rig Counts And Oil Production Bottom Line: While we do not have expertise to follow or forecast oil supply dynamics, we are biased in believing that the risk-reward for oil prices is unattractive because of a strong U.S. dollar and potentially weak EM/China asset prices, which could trigger a reduction in net long positions in crude. Investment Conclusions Complacency reigns in the global financial markets. EM equity volatility has fallen close to its cycle lows, the U.S. VIX is depressed, U.S. equity investor sentiment is very elevated and EM corporate credit spreads have plummeted to a ten-year low (Chart I-18). While the timing of a reversal is impossible, the risk-reward profile of EM financial markets is greatly unattractive. The U.S. trade-weighted dollar has consolidated recently, and might be primed for another upleg. As the U.S. dollar resumes its uptrend, EM risk assets will likely sell off. Finally, EM share prices have failed to outperform the developed bourses much, despite the rally in commodities and amelioration in Chinese growth (Chart I-19). Chart I-18Complacency Reigns Chart I-19EM Equities Have Not Yet Outperformed Remarkably, analysts' net earnings revisions for EM stocks have so far failed to turn positive (Chart I-20). Either analysts' EPS expectations were originally still too high, or companies are failing to deliver profits. Whatever the reason, the implication is that the consensus is more bullish on EM than is suggested by the underlying fundamentals. Within an EM equity portfolio, our overweights remain Taiwan, Korea, India, China, Thailand, Russia and central Europe. Our underweights are Malaysia, Indonesia, Turkey, Brazil and Peru. We are neutral on other bourses. Finally, the EM equity benchmark is at a critical technical resistance level (Chart I-21) but odds do not favor a sustainable breakout. Chart I-20EM EPS Net Revisions Are Still Negative Chart I-21EM Stocks: A Breakout Attempt Arthur Budaghyan, Senior Vice President Emerging Markets Strategy arthurb@bcaresearch.com 1 Please refer to the Emerging Markets Strategy Special Report titled, "Revisiting China's Fiscal And Credit Impulses", dated April 13, 2016, available at ems.bcaresearch.com 2 Trilogy of Special Reports on money/loan creation, savings and investment, titled, "Misconceptions About China's Credit Excesses" dated October 26, 2016, "China's Money Creation Redux And The RMB", dated November 23, 2016 and "Do Credit Bubbles Originate From High National Savings?", dated January 18, 2017, available at ems.bcaresearch.com Equity Recommendations Fixed-Income, Credit And Currency Recommendations

Highlights We expect the high level of compliance with the OPEC - non-OPEC production agreement engineered by the Kingdom of Saudi Arabia (KSA) and Russia will endure, leading to significant reductions in global oil inventories this year and next. All else equal, this should backwardate WTI and Brent forward curves later this year. However, recent developments in the North American pipeline market - i.e., U.S. President Donald Trump's orders to revive development of the Keystone XL (KXL) and completion of the Dakota Access (DAPL) pipelines - could send as much as 1mm barrels/day (bbl/d) of crude south from Canada and the Bakken, which would boost inventories at Cushing and other Midwest storage facilities later in this decade. Depending on when these pipelines are completed - likely by 2020 in the case of KXL - the WTI forward curve could return to a sustained contango.1 The expanded flows of heavy crude via KXL, and light-tight oil south via the DAPL could undo a subtle benefit arising from the backwardation induced by the KSA - Russia production pact, which we uncovered in our modeling. Energy: Overweight. At Tuesday's close, our short Dec/19 WTI vs. long Dec/19 Brent spread elected last week at $.07/bbl (WTI over) was up 700%. Our long Dec/17 WTI vs. short Dec/18 WTI front-to-back spread, entered into at -$0.11/bbl on Feb 9/17, was up 263%. Base Metals: Neutral. BHP declared force majeure at its Escondida mine, which accounts for ~ 5% of global supply, after workers voted to strike. Union leaders agreed to another round of government-mediated talks with BHP management. Precious Metals: Neutral. Fed Chair Yellen's Senate Banking Committee testimony was more hawkish than expected, which rallied the USD and muted gold's overnight strength. We continue to look to get long gold at $1,180/oz. Ags/Softs: Underweight. The USDA revised grain and soybean supply/demand estimates last week, showing markets tightening slightly, with ending stocks for the 2016/17 crop year expected to be a touch lower. We remain bearish. Feature Chart of the WeekStorage Drawdowns Should Accelerate ##br##As U.S. Oil Imports Slow Regular readers of BCA's Commodity & Energy Strategy service will not be surprised by the very high compliance levels seen in the wake of the OPEC - non-OPEC production Agreement engineered by KSA and Russia late last year.2 Because the stakes are so high for KSA and Russia - and their respective oil-producing allies - we expect compliance to remain high into June, resulting in a drawdown of global oil storage, the stated goal of the deal. We believe the pact will result in both WTI and Brent forward curves returning to backwardation, as global storage levels fall some 300mm bbl (Chart of the Week). We are positioned for this outcome by being long Dec/17 WTI vs. short Dec/18 WTI. We are expecting to see the last of the Persian Gulf export surge to the U.S. this month, as the 45- to 50-day sailing time from the Gulf to the U.S. implies the last of these vessels will be arriving this week or next. This backwardation will, in all likelihood, restrain the rate at which U.S. shale-oil producers return rigs to the market next year. Chart 2Curve Shape Can Affect Rig Counts WTI Term Structure And Rig Counts: It's Complicated Recent modeling we've completed suggests curve shape can affect rig counts in the U.S. light-tight oil fields. When we regress U.S. rig count on the WTI forward curve, we find rig counts can be expected to increase when the forwards are in contango, and to decline when the market is backwardated. A flat forward curve can be expected to keep rig counts fairly constant (Chart 2).3 Obviously, the starting point for these outcomes is critical. We simulated rig counts by assuming Monday's closing prices for March through June WTI futures, then assumed different levels for July WTI futures as a starting point for estimating rig counts to end-2018. We used $50, $55 and $60/bbl in July as our starting point. All else equal, with the July/17 WTI at ~ $55/bbl and the forward curve backwardated by 10% 18 months out, we would expect to see average rig counts fall by 4.38 rigs/month in 2018, given the three-to-four month lag between rigs actually being deployed and the price signal being sent by the futures market. A contango term structure produces the opposite result. With the July/17 WTI at ~ $55/bbl and the forward curve in a contango of 10% 18 months out, we would expect to see rig counts increase by 4.57 rigs/month in 2018. There obviously is a price threshold from which the forward curve originates in this analysis, which we believe to be between $50 and $55/bbl. Below this level, we would expect shale producers to retreat back to their core production areas, and await a price signal to increase their rig counts. Above $60/bbl, backwardation and contango matters for rig counts over the next 2 to 2.5 years. A backwardated forward curve will, all else equal, incentivize a slightly lower level of rigs being deployed than a contango. For example, a 10% contango with a $60/bbl starting point results in 5.24 rigs/month being deployed, while 10% backwardation would lead us to expect 5.02 rigs/month being deployed. Sustaining Backwardation Will Be Difficult A sustained backwardation will be threatened later in this decade by the expansion of the North American pipeline grid, following U.S. President Trump's orders to revive the Keystone XL (KXL) pipeline's development and the completion the Dakota Access Pipeline (DAPL). The KXL and DAPL buildouts, if approved, will expand U.S. midcontinent crude deliveries by 1mm bbl/d, according to Genscape's tally.4 The KXL volumes would add close to 600k bbl/d to Canadian exports, and would flow directly into Cushing, OK. Another 400k bbl/d of light-tight oil from the Bakken LTO fields will flow to the midcontinent refining market via the DAPL. "Increased flows into Cushing due to the addition of Keystone XL could lead to a bottleneck of inventories at the hub, which would put downward pressure on crude prices," Genscape notes. Work on the KXL could start this year, and be completed before 2020. The DAPL is ~ 95% complete, and should be done in 6 months or less. Genscape believes the DAPL could be built and line fill could be in place in less than three months. Indeed, "drilling under Lake Oahe in southern North Dakota for Energy Transfer Partner's Bakken-to-Patoka, IL, Dakota Access (DAPL) crude pipeline began immediately upon receiving an easement from the U.S. Army Corps of Engineers on February 8, according to a company spokesman. It is expected to take 83 days for construction and linefill... ." We will monitor these pipeline buildouts closely, given the profound implications they have for U.S. midcontinent and Gulf Coast refiners, who could once again find themselves benefiting from a widening of the Brent vs. WTI differential, and Canadian E&Ps, who can be expected to increase production into this KXL buildout. The key market to watch as these pipelines are under construction will be the WCS vs. WTI spreads (Chart 3). As pipeline capacity opens up, exports of heavy crude from Canada will increase and the WCS - WTI differential will narrow, which will benefit Canadian E&Ps (Chart 4). A return of contango following the opening of these pipelines would benefit U.S. refiners, who can be expected to increase exports. Chart 3Expanding the N. American Pipeline Network##br## Will Widen WTI Differentials Chart 4Crude Differentials Will##br## Adjust To Pipeline Buildouts Bottom Line: The backwardation of the WTI and Brent forwards should accelerate as the last of the surge in exports from the Persian Gulf arrives in the U.S. President Trump's decision to expedite KXL and the completion of the DAPL in 6 months or less will have a profound impact on crude movements and storage levels in the U.S. later in the decade. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com 1 President Trump's decision to revive KXL was endorsed by House and Senate leaders in the U.S. last month, which greatly raises the odds it will go ahead. In addition, the DAPL received an easement from the U.S. Army Corps of Engineers to complete construction. 2 Please see issue of BCA Research's Commodity & Energy Strategy Weekly Report "Raising The Odds Of A KSA-Russia Oil-Production Cut," dated November 3, 2016, available at ces.bcaresearch.com. 3 Our previous modeling indicates Granger causality goes from WTI prices to rig counts - i.e., E&P companies drilling decisions are driven by price levels and curve shape. We believe this relationship arises from the hedging behavior of shale-oil producers, many of whom hedge their forward revenues in the futures markets over a two-year interval. 4 Please see "Keystone XL, Dakota Access Could Cause Bottlenecks at U.S. Mid-Continent Storage Hubs, Shift Crude Prices," published on Genscape's blog February 14, 2017. Genscape is a near-real-time pipeline, storage and shipping monitoring service. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed In 2017 Summary of Trades Closed in

Highlights Weekly swings in U.S. inventories notwithstanding, we believe global storage is on track to draw ~ 10% by early- to mid-3Q17, which will have achieved the goal of the OPEC - Russia production Agreement negotiated late last year. This will not require an extension of the pact beyond June, based on our modeling. Unexpectedly high compliance by OPEC producers to agreed cuts is being offset somewhat by increased production in those states exempted from the deal. Strong oil consumption on the back of a synchronized global uptick in GDP growth, which started to emerge late last year, provides the impetus for sustained storage draws. Markets are overestimating offshore production's resilience, particularly in the U.S. Gulf, where we see material declines beginning to set in next year. Backwardation likely persists in 2018, absent a U.S. policy-induced USD rally that crimps EM demand and spurs production ex U.S. Energy: Overweight. The return of contango in the WTI forward curve gives us the opportunity to reset our strategic front-to-back position (long Dec/17 vs. short Dec/18) at tonight's close. Our balances assessment supports our view backwardation will return in the deferred part of the curve. Our Dec/19 short WTI vs. long Brent spread buy stop was elected at $0.07/bbl. Base Metals: Neutral. We remain neutral base metals, but are keeping a close watch on copper. Unions working at BHP's Escondida mine, the world's largest, are set to strike today. Negotiations resumed this week, following BHP's request for government mediation. Precious Metals: Neutral. We continue to look to get long gold at $1,180/oz. Ags/Softs: Underweight. Grain fundamentals remain unsupportive for a rally. We remain underweight. Feature Chart of the WeekGlobal Oil Storage On Track For 10% Drop Global oil storage levels remain on track to hit the ~ 10% draw targeted in last year's OPEC - Russia production Agreement by early- to mid-3Q17, weekly gyrations in U.S. inventories notwithstanding. This means an extension of the agreement beyond its June expiry will not be required. Early reports suggest compliance with the deal is unexpectedly high by OPEC states that agreed to cut production by up to 1.2mm b/d - exceeding 80% by various accounts. However, OPEC states not required to cut - Libya, Nigeria, and Iran - have increased production and partially offset those declines, which took total reductions in OPEC output to ~ 840k b/d, based on a Bloomberg tally last week.1 This brought total Cartel compliance to ~ 60% of the agreed cuts, which, as we noted in our 2017 Commodity Outlook in December, would be sufficient to achieve the Agreement's goal of pulling inventories in the OECD down by ~ 10% by 3Q17.2 Non-OPEC producers also appear to be complying with the Agreement. Notable among them is Russia, which is ahead of its commitment with cuts of close to 120k b/d in January, due partly to extreme cold in Siberian fields. We expect cuts in Russia to average 200k b/d in 1Q17, going to 300k b/d in 2Q17. These cuts will allow demand to outstrip supply in 1H17 and into year-end. By early- to mid-3Q17, draws to OECD storage of 300mm bbl can be expected, without extending the OPEC - Russia production agreement (Chart of the Week). We expect to see these cuts show up in OECD inventory data this month and next and continue into the end of 2017. For non-OECD states, the draws will show up in JODI data beginning in March.3 The physical deficits - i.e., supply less than demand - will force storage to draw, backwardating the WTI forward curve later this year (Chart 2).4 If markets are not surprised by a policy-induced rally in the USD on the back of a U.S. border-adjustment tax (BAT), or a too-aggressive tightening by the Fed as it seeks to normalize monetary policy, we expect the drawdown in inventories to continue keeping markets backwardated. Even with production returning to pre-Agreement levels in 2H17 in states with the capacity to expand and reliably sustain production - Gulf Arab producers, Russia and U.S. shales - we expect storage to continue to draw through the year and into 2018 (Chart 3). Chart 2We Continue To Expect Backwardation Chart 3Storage Drawdown On Track In 4Q16 the impact of the higher Kuwaiti and UAE output is apparent, along with higher Russian production. This put more crude on the market, which found its way into storage late in 4Q16 and early 1Q17, reversing the trend in draws seen earlier in 2H16. This put the market back in a temporary surplus condition, with the result being more storage will have to be worked off in 1H17 than our earlier estimates indicated. But these draws will occur, following the implementation of the production accord. Extending The KSA - Russia Deal Beyond June Is Unnecessary In our estimates, OPEC crude production increases by ~ 850k b/d in 2H17 versus 1H17 levels. Despite this recovery, the storage drawdown continues. Our modeling assumes Gulf OPEC will account for slightly more than +1mm b/d growth, and non-Gulf OPEC will see production continue to fall by 170k b/d. Russia's total liquids production goes from 10.95mm b/d in 1H17 to 11.34mm b/d in 2H17. We estimate U.S. shale production grows at an average rate of ~ 300k b/d in 2H17, while total U.S. liquids production increases 720k b/d over the same interval. Setting aside the possibility of a policy-induced rally in the USD on the back of too-aggressive Fed tightening or a border-adjusted tax becoming the law of the land, both of which would depress demand and raise production ex U.S., we expect the crude-oil market to remain backwardated next year. The globally synchronized upturn in GDP will keep demand robust, with growth coming in close to even with this year's rate of ~ 1.50mm b/d. We have global liquids production and OPEC crude output growing less than 1.0% next year. We believe the market is overestimating the resilience of offshore production next year, particularly in the U.S. Gulf, based on the stout performance put in last year and expected for this year. Our colleague Matt Conlan notes in BCA's Energy Sector Strategy, U.S. production growth since October has almost exclusively been from the Gulf of Mexico's new projects. Output in the Gulf continues to increase due to the lagged effect of final investment decisions made during 2012 - 2014, when WTI prices were consistently trading above $100/bbl. GOM production will peak in 2017 then decline in 2018 due to lack of new investments made since 2014. Indeed, as "increasing decline rates overwhelm a shrinking inventory of new projects, GOM production should peak sometime in 2017 and then start decreasing. The EIA's estimate for another 200,000 b/d increase in GOM production in 2017 seems overly-optimistic."5 Once this becomes apparent to the market, we believe backwardation will reassert itself and persist into 2018. The backwardation of the forward curve structure will affect U.S. shale production economics in 2018. However, our base case is for U.S. shale-oil production in the "Big Four" basins - Permian, Eagle Ford, Bakken and Niobrara - to grow 700k b/d next year, given the current structure of the WTI forwards, which were taken higher along with the WTI price rally at the front of the curve. This triggered the revival of rig counts; however, we want to point out that different curve shapes at different price levels produce different expected rig-count responses.6 Chart 4Barring a Policy Shock Demand Will Remain Robust Global Demand Firing On All Cylinders Robust demand growth - ~ +1.50mm b/d in 2017 and 2018 in our modeling - provides the impetus for the continued draws in storage this year and next (Chart 4). We revised our demand estimates for 2015 - 16 in line with the IEA's just-revised assessment of global consumption published in its January 2017 Oil Market Report.7 The IEA brought 2016 oil demand growth up to 1.50mm b/d, in line with our earlier estimates, but significantly revised 2015 demand growth upward to 2.0mm b/d. The Agency expects higher prices to crimp demand this year, taking it to 1.30mm b/d; our estimate, however, is higher, largely on the back of the first global synchronized growth we've seen since the Global Financial Crisis, which will be supported by accommodative monetary conditions worldwide, all else equal.8 Investment Implications Our analysis suggests there will be no need to extend the OPEC - Russia production accord into 2H17. In addition, it reinforces our view markets will backwardate later this year and stay backwardated in 2018, provided we do not see a BAT-induced rally in the USD, or an overly aggressive Fed normalization trajectory. As we noted in previous research, a BAT would lift the value of the USD, which would lower demand ex U.S. and raise supply at the margin.9 We make the odds of a BAT becoming the law of the land in the U.S. this year 50:50, so this is a non-trivial risk. This would be unambiguously bearish for oil prices. While we do not expect oil to be included among the imported commodities subject to a BAT, we do, nonetheless, expect the imposition of a BAT to lift the USD by 10%. This, coupled with the 5% increase in the greenback we'd already penciled in due to the Fed's monetary-policy normalization, will lift the USD 15% if it goes through. Should this occur, we would be preparing for prices to again fall below $50/bbl and push back to the $40/bbl area, which would cause supply and capex to once again contract significantly. That said, we are reinstating our long front-to-back WTI recommendation (long Dec/17 WTI vs. short Dec/18 WTI), given our updated balances assessment. Our expectation for inventories to continue to draw after the OPEC - Russia production-cutting agreement expires in June supports this recommendation. In addition, if we do see a BAT in the U.S., we believe markets will take the deferred WTI curve significantly lower in expectation of reduced demand and higher marginal supplies that almost surely will ensue in 2018. While the Dec/17 contract also will trade lower, more damage to prices will occur in 2018 contracts. Robert P. Ryan, Senior Vice President Commodity & Energy Strategy rryan@bcaresearch.com 1 Please see "OPEC Cuts Oil Output, But More Work Needed to Fulfill Deal," published by Bloomberg February 2, 2017. Iraq stands out among OPEC producers agreeing to cut, but apparently not following through as diligently as the rest of the Gulf Arab states; we are assuming production of 4.5mm b/d for 1H17, going to 4.6mm b/d in 2H17 for Iraq. 2 Please see BCA Research's Commodity & Energy Strategy "2017 Commodity Outlook: Energy," dated December 8, 2016, available at ces.bcaresearch.com. 3 JODI refers to the Joint Organisations Data Initiative, a supranational producer-consumer oil-market data provider headquartered in Riyadh, Saudi Arabia. 4 "Backwardation" describes a forward price curve in which the price for a commodity for prompt delivery (e.g., tomorrow) exceeds the price of a commodity delivered in the future (e.g., next year). It is the opposite of a contango curve structure. 5 Please see issue of BCA Research's Energy Sector Strategy "Gulf Of Mexico Oil Production Likely To Peak In 2017," dated January 11, 2017, available nrg.bcaresearch.com. 6 In next week's report, we will present scenario analysis of shale-oil production as a function of WTI forward curve shape - i.e., the implications of backwardation for shale rig counts. This will update our assessments of price sensitivities to interest rates and USD movements. 7 Please see the IEA's Oil Market Report of 19 January 2017. 8 We discuss this in last week's Commodity & Energy Strategy feature article entitled "Gold Will Perform...," dated February 2, 2017, available at ces.bcaresearch.com. 9 Please see BCA Research's Commodity & Energy Strategy "Taking A BAT To Commodities," dated January 26, 2017, available at ces.bcaresearch.com. Investment Views and Themes Recommendations Strategic Recommendations Tactical Trades Commodity Prices and Plays Reference Table Trades Closed In 2017 Summary of Trades Closed in 2016

Gold mining shares look increasingly attractive, at least as a portfolio hedge. We took profits on our overweight position in the middle of last summer, just prior to the share price crunch, because tactical sentiment and positioning had gotten too stretched. Overzealousness no longer exists, and a revival in market volatility and intensification in policy uncertainty have created an attractive re-entry point in gold shares. The last playable rally began when the yield curve started to flatten, signaling doubts about the longevity of the business cycle. More recently, the yield curve stopped steepening when the Fed raised interest rates last month and may be signaling the start of another move up in gold shares. Importantly, sentiment toward the yellow metal is no longer overheated, as evidenced by both surveys and investor behavior. Flows into gold ETFs have been trending lower, reversing last summer's buying frenzy. Speculative positions have also been unwound. Against a positive structural backdrop for gold, we recommend reestablishing overweight positions in gold mining stocks as a portfolio hedge. Please see yesterday's Weekly Report for more details.