Sectors

The world economy is likely already in recession, defined as world growth dipping to sub-2 percent. So far, the world recession has been China-led, but in the coming months it will change to being developed economy-led. Hence, while metals and industrial commodities may get some brief respite, high yield credit and stocks will underperform government bonds. New tactical recommendations are to overweight French luxury goods versus US tech, and to overweight USD/COP.

In June, the rally gained momentum and broadened due to positive economic data, particularly in the housing market. We expect cheaper cyclical sectors and styles to mark a change in leadership as the rally broadens, helped on by excess cash on the sidelines. We upgrade Banks to equal-weight, and Homebuilders to overweight. The rally may continue but a soft landing continues to be elusive - disappointment may be in store.

Recession is on track to start around year-end. Stocks usually peak shortly before recession begins. So, position defensively but be prepared for a few more months of the rally.

We build a four-stage business cycle framework based on economic growth and capacity utilization, and then analyze historical returns for most major asset allocation decisions for each stage. Given that we are in the early recession stage (negative growth coupled and an overheated economy), our framework recommends a defensive positioning across all asset classes.

In this Strategy Outlook, we present the major investment themes and views we see playing out for the rest of 2023 and beyond.

Highlights Recent US housing market data has signaled a potential turnaround in housing construction and new home sales. Permanent site residential structures investment may begin to contribute positively to US real GDP growth if the recent pickup in housing starts is sustained. Housing construction is set to rise because of an extremely low level of existing homes available for sale. Existing home inventories and sales are low because US households do not want to give up low-rate mortgages. This underscores that US monetary policy is currently restrictive, not easy. To the extent that the recent pickup in housing starts, new home sales, and house prices reflects a negative supply shock in the US housing market, it could also cause rental market conditions to tighten again. Tight, supply-driven, housing market conditions could point to even more monetary policy tightening than we currently expect if sustained over the coming several months. This, for now, is a risk rather than our base case scenario. We do not yet believe that bond yields will hit a new cycle high, and would continue to recommend a long duration position within a fixed-income portfolio. The recent housing market data is far more consistent with the “no landing” economic scenario that we described in last month's report then the “soft landing” scenario that stocks are betting on. In our view, either the “hard landing” or “no landing” scenarios are negative for risky asset prices, implying that recent housing market developments are not bullish for stocks. US real residential investment has negatively contributed to economic growth for eight straight quarters. Initially, this contraction in activity reflected a normal unwinding of pandemic-related excesses and was not directly connected to the Fed’s monetary policy stance. Over the past year, however, the sharp rise in mortgage rates prolonged the weakness in housing activity, which has often been a harbinger of a recession. Chart II-1A Surprising Surge In Housing Construction A Surprising Surge In Housing Construction A Surprising Surge In Housing Construction Very recently, some signs of a potential turnaround in the US housing market have emerged. Chart II-1 shows that homebuilder confidence has rebounded in the first half of the year and, in May, new private housing starts surged significantly. In this report, we revisit the outlook for US housing as a follow-up to last year’s report on the topic.1 While permanent site residential structures investment may begin to contribute positively to US real GDP growth, we would caution against the view that the recent pickup in housing starts, if sustained, would raise the odds of a meaningful cyclical upswing in economic activity. What is occurring in the US housing market is best described as a negative supply shock, rather than a demand-driven boom that is surmounting high mortgage rates. We interpret that as highly inconsistent with the “soft landing” economic scenario that stocks are betting on, which underscores that conservative portfolio positioning is warranted. A Recent Surprise In The US Housing Data Chart II-2The Housing Market Index Has Lagged New Home Sales Over The Past Two Years. The Latter Only Recently Took Off. The Housing Market Index Has Lagged New Home Sales Over The Past Two Years. The Latter Only Recently Took Off. The Housing Market Index Has Lagged New Home Sales Over The Past Two Years. The Latter Only Recently Took Off. Until this month, there were signs of only a very tepid recovery in US housing-related activity. Chart II-2 highlights that, while the NAHB’s housing market index has been rising for six months, the index has lagged new single-family home sales since the start of the pandemic. As such, the recent pickup in the index to levels that were still below the average of the past decade looked like it was simply catching up to an uptrend in new home sales that themselves appeared modest until the release of the May data. However, the May housing market data caught the attention of many investors, especially those who argue that a US recession is not likely to occur over the coming year. Housing starts and new home sales rose 21% and 12%, respectively, in May, and the 20-city S&P Case/Shiller home price index rose at an 11% annualized rate in April (Chart II-3). To add to the confusion for investors, housing permits rose only modestly in May, existing home sales declined, and the MBA Mortgage Application Purchase index remains very low (Chart II-4). Chart II-3A Big Jump in Starts, New Home Sales, And House Price Appreciation... A Big Jump in Starts, New Home Sales, And House Price Appreciation... A Big Jump in Starts, New Home Sales, And House Price Appreciation... Chart II-4...But A Tepid, If Any, Rise In Existing Home Sales, Permits, And Mortgage Applications ...But A Tepid, If Any, Rise In Existing Home Sales, Permits, And Mortgage Applications ...But A Tepid, If Any, Rise In Existing Home Sales, Permits, And Mortgage Applications This raises some important questions for investors. The first question is whether the recent pickup in housing construction and new home sales in May is a head fake, or the beginning of a major uptrend. The second question is whether there is any significance to the divergence between the new and existing segments of the housing market. And the third is should investors wonder whether a pickup in housing market activity signals that monetary policy is not as tight as some investors (and we) believe, potentially arguing that a US recession is further away than we currently expect. While it remains early days, and we are reluctant to draw firm conclusions in response to one month of data, for now our answers to these questions are as follows: There is mounting evidence that US housing construction may have bottomed, suggesting that the permanent site structures portion of real residential investment may begin to contribute positively to US economic growth. Rather than reflecting strong demand and a robust capacity for US households to tolerate high mortgage rates, this improvement in housing construction likely reflects the opposite: housing construction is set to rise because of an extremely low level of existing homes available for sale. Existing home inventories and sales are low because US households do not want to give up low-rate mortgages. This underscores that US monetary policy is currently restrictive, not easy. Weak existing home sales and extremely poor home affordability will cause the other components of real residential investment (brokerage commissions and home improvements) to be weak, and will likely act as a tax on US spending. If significant US house price appreciation / housing market tightness continues, it could threaten or blunt the downtrend in core PCE inflation that we have been forecasting. Were that to occur, it would further reduce the odds of a “soft landing” economic outcome and could cause a more substantial rise in long-maturity government bond yields than we currently expect. Housing Supply, House Prices, And New Home Construction The monthly pace of housing starts and new home sales are volatile series, and it is possible that the June housing market data will come in significantly worse than what occurred in May. It is also true that housing starts cannot truly decouple from building permits (Chart II-5). The muted pickup of the latter in May suggests that the sharp acceleration in starts and new home sales is possibly exaggerated. However, it is probably the case that US housing construction has bottomed, at least until the unemployment rate begins to rise and a recession takes hold. First, on the question of whether the tepid rise in building permits suggests that housing starts have overshot, Chart II-6 shows that our US state building permit diffusion index is in a solid uptrend. Unsmoothed, the series has risen to a value of 66%, suggesting that housing permits in the US are likely to trend higher. Chart II-5Housing Starts Cannot Sustainably Decouple From Permits... Housing Starts Cannot Sustainably Decouple From Permits... Housing Starts Cannot Sustainably Decouple From Permits... Chart II-6...But There Are Signs That Permits Are Set To Improve ...But There Are Signs That Permits Are Set To Improve ...But There Are Signs That Permits Are Set To Improve But, more broadly, the pickup in starts, even if the pace of advance observed in May is not sustained, reflects underlying structural rigidities and a legacy of underbuilding in the US housing market that are giving the false appearance of strong underlying housing demand. Both a decade of too-low residential housing construction following the global financial crisis, and the fact that 30-year fixed-rate mortgages in the US allow homeowners to lock in interest rates for the entire amortization term of the mortgage, have contributed to a housing supply crisis in the US. Chart II-7 shows the standardized trend in permanent site residential structures investment (both single- and multi-family) relative to both the US population and the number of households. As we noted in last year’s report, the chart makes it clear that the US vastly overbuilt homes from the late-1990s to 2007, but also vastly underbuilt since 2008. Relative to the number of households, real permanent site residential structures investment is still half of a standard deviation below its long-term average – even after the surge in construction that occurred in 2020. This has contributed to a very low level of single-family housing inventory (Chart II-8) and a record low homeowner vacancy rate (Chart II-9). Chart II-7Over A Decade Of Underbuilding Homes... Over A Decade Of Underbuilding Homes... Over A Decade Of Underbuilding Homes... Chart II-8...Has Led To A Negative Supply Shock In The US Housing Market ...Has Led To A Negative Supply Shock In The US Housing Market ...Has Led To A Negative Supply Shock In The US Housing Market On top of this, the structure of the US mortgage market is exacerbating the housing supply problem by reducing the incentive for homeowners to list existing houses for sale. In sharp contrast to most other (if not all) major developed countries, US homeowners are able to secure a constant fixed mortgage rate for the entire amortization period of a mortgage, effectively locking in a particular mortgage rate when purchasing a home. In most other countries, fixed-rate mortgages exist, but the term of the mortgage contract itself (which is separate from the amortization term of the mortgage) is typically limited to a few years (usually not more than five). That means that homeowners in developed markets outside of the US mechanically “reset” to prevailing interest rates at the end of their mortgage contract term. In the US, as in other countries, 30-year fixed-rate mortgages are not portable, meaning that the mortgage contract ends and the interest rate is reset if homeowners choose to sell their home. It is this latter fact that is seemingly restricting existing home supply, which helps explain why existing home sales are not rising alongside new home sales and why mortgage applications remain low (Chart II-10). This also helps explain why existing house prices are starting to rise, as higher prices are needed in order to induce homeowners to give up their lower interest rates and reset to prevailing market rates. Chart II-9The Lowest Homeowner Vacancy Rate On Record The Lowest Homeowner Vacancy Rate On Record The Lowest Homeowner Vacancy Rate On Record Chart II-10US Homeowners Are Hoarding Existing Low-Rate Mortgages US Homeowners Are Hoarding Existing Low-Rate Mortgages US Homeowners Are Hoarding Existing Low-Rate Mortgages Housing And The Neutral Rate Of Interest One possible interpretation of the May surge in housing starts is that it reflects a much higher neutral rate of interest than many investors would expect. In other words, some might argue that the sharp pickup in construction highlights that US monetary policy is not very tight or not tight at all. We disagree with this interpretation, and offer the following points in response: 1. To the extent that our analysis of the cause of rising house prices and starts is accurate, what is occurring in the US housing market is best described as a negative supply shock, rather than a demand-driven boom despite high mortgage rates. We agree that a housing supply shock is less negative for the US economy than a typical shock involving food & energy prices, given that households own homes and see their net worth increase in response to higher house prices. But the key point for investors is that the surge in starts and prices is not likely signaling anything significant about the neutral rate of interest unless it is demand-driven rather than supply-driven. 2. By any available measure, US housing affordability is extremely poor (Chart II-11). In addition, mortgage rates have also risen to above-average levels versus the prevailing rate of nominal growth. Mortgage rates have been higher relative to underlying growth in the past, but they are still in line with where they were prior to the 2008/2009 US recession (Chart II-12). Chart II-11US Housing Affordability Is Terrible US Housing Affordability Is Terrible US Housing Affordability Is Terrible Chart II-12US Mortgage Rates Are Not In Easy Territory US Mortgage Rates Are Not In Easy Territory US Mortgage Rates Are Not In Easy Territory     Charts II-11 and II-12 do not point to the recent surge in housing starts as having been driven by a strong underlying demand impulse that is being stimulated by still too-easy monetary policy. It is true that if excess savings are underpinning US aggregate demand then it is theoretically possible that US homeowners are comparatively immune to extremely poor housing affordability. But, as we showed in Section I of our report, excess savings are dwindling rapidly and it is not clear why they would be suddenly acting to stimulate housing demand after having fallen significantly over the past 18 months. 3. Consumer surveys specifically do not support the idea that housing construction and house prices are rising because of stimulative interest rates. Chart II-13 illustrates that US consumers still rank the current environment as the worst time to buy a home since the Volcker era. Importantly, while consumers reported this time last year that this was true mostly due to high prices, today “interest rates are high” as a reason cited for why it is a bad time to own a home has risen significantly, to levels that are close the highest recorded since the early 1980s. 4. It is true that homeowners who have locked-in mortgage rates are not directly affected by a sharp increase in market rates, and that is reflected in the fact that the effective mortgage interest rate paid has not yet increased significantly (Chart II-14, panel 1). However, that is true in every business cycle, meaning that it does not tell investors very much about the neutral rate of interest. Evidence of a historically low feedthrough from higher quoted mortgage rates to the effective rate paid would potentially be evidence of a higher neutral rate, but panels 2 and 3 of Chart II-14 do not suggest that this is the case. Panel 2 shows the rolling “beta” of annual changes in the effective rate relative to those of market rates, and panel 3 shows the residual of the relationship. The effective mortgage rate beta is on par with its historical median (lower beta = less feedthrough), and the residual of the relationship is only modestly negative. Chart II-13US Monetary Policy Is Tight, Not Stimulative US Monetary Policy Is Tight, Not Stimulative US Monetary Policy Is Tight, Not Stimulative Chart II-14That Relationship Between Effective And Actual Mortgage Rates Is No Different Today Than In The Past. That Relationship Between Effective And Actual Mortgage Rates Is No Different Today Than In The Past. That Relationship Between Effective And Actual Mortgage Rates Is No Different Today Than In The Past.   Additionally, while locked-in homeowners are not directly affected by higher mortgage rates, the spending/savings behavior of those who plan on moving at some point in the foreseeable future is affected via expectations of eventually higher mortgage payments. Rising House Prices And US Monetary Policy As we presented in Section I of our report, a normalization in housing services inflation is one important reason to believe that US core PCE inflation will continue to decelerate. Based on a normalized rate of housing services inflation, core PCE inflation will decline to 3.5% even without any change in the current rate of core goods or core non-housing services inflation. Chart II-15Housing Inflation Set To Normalize Early Next Year, If The Housing Supply Shock Is Not Severe Housing Inflation Set To Normalize Early Next Year, If The Housing Supply Shock Is Not Severe Housing Inflation Set To Normalize Early Next Year, If The Housing Supply Shock Is Not Severe However, a pickup in house prices, if sustained, implies that housing services inflation may not soon fall back to its pre-pandemic level. Chart II-15 shows an attempt to gauge this risk analytically, by presenting a model for housing services inflation that was originally developed by our US Bond Strategy service to forecast the shelter component of the US Consumer Price Index. The model incorporates house prices, the unemployment rate, and two measures of apartment/rental market tightness. The good news from the model is that there is a long lag in the relationship between house prices and housing services inflation, suggesting that any pickup in house prices now will only impact inflation in over a year’s time. In addition, based on the current three-month annualized rate of change, the recent acceleration in house prices would only add between 40-50 basis points to housing services inflation. The bad news is that the model’s predicted level of housing services inflation, while meaningfully lower than its current pace of advance, is currently above its pre-pandemic level. The model suggests that housing services inflation will normalize early next year, but most of the modeled decline comes from the lagged effect of previously slowing house prices. Despite the existence of a long house price / housing inflation lag historically, a reacceleration in house prices could, theoretically, feed back into housing inflation sooner than historical lags would suggest. In addition, to the extent that the recent pickup in housing starts, new home sales, and house prices reflects underlying supply problems in the US housing market, it could also cause rental market conditions to tighten again. The bottom line for investors is that tight, supply-driven housing market conditions could point to even more monetary policy tightening than we currently expect if sustained over the coming several months. Investment Conclusions There are two key conclusions for investors that stem from our analysis of the US housing market. The first is that permanent site residential structures investment may begin to contribute positively to US real GDP growth if the recent pickup in housing starts is sustained, although we would expect the other elements of residential investment (brokers' commissions and home improvement) to remain weak. While a pickup in housing construction would normally raise the odds of a sustained cyclical upswing in economic activity, we would caution against that conclusion in the current environment. Our read of the housing market data points to new home construction and sales picking up because of a negative supply shock impacting US households, which may boost real residential investment but will likely act as a drag on US consumer spending. Ultimately, we believe the latter is more important in driving the US labour market, thus increasing the odds of a recession over the coming year. Chart II-16A Housing Inflation Shock Could Cause A New High In Yields, Which Would Cause Long Duration Positions To Lose Money A Housing Inflation Shock Could Cause A New High In Yields, Which Would Cause Long Duration Positions To Lose Money A Housing Inflation Shock Could Cause A New High In Yields, Which Would Cause Long Duration Positions To Lose Money The second conclusion is consistent with the first, in that a persistent reacceleration in house prices and/or renewed apartment / rental market tightness could prevent core PCE inflation from falling to the levels that we expect over the coming year. That could cause the Fed to raise rates even more than it recently projected, which could push long maturity bond yields above back above their October high and would cause investors to lose money on long-duration positions (Chart II-16). Were that scenario to develop, we would likely recommend reducing fixed-income portfolio duration to neutral (from long). This, for now, is a risk rather than our base case scenario. We do not yet believe that bond yields will hit a new cyclical high and would continue to recommend long duration positions on the expectation of pending economic weakness stemming from the Fed’s tight monetary policy stance. More generally, from an asset allocation perspective, the recent housing market data is far more consistent with the “no landing” economic scenario that we described in last month's report than the “soft landing” scenario that stocks are betting on. In our view, either the “hard landing” or “no landing” scenarios are negative for risky asset prices, implying that recent housing market developments are not bullish for stocks. As noted, the main investment implication of a major and sustained pickup in US housing construction and house prices would relate to one’s fixed-income portfolio. While we are watching housing market developments closely, we have not yet concluded that a change in duration stance is warranted. Stay tuned! Jonathan LaBerge, CFA Vice President The Bank Credit Analyst Footnotes 1  Please see The Bank Credit Analyst "Is The US Housing Market Signaling An Imminent Recession?" dated May 26, 2022, available at bca.bcaresearch.com