Market & Economic Commentary

Used cars in the headlights

According to the latest Consumer Price Index (CPI) report from the Bureau of Labor Statistics, “the index for used cars and trucks rose 10.0 percent in April. This was the largest 1-month increase since the series began in 1953, and it accounted for over a third of the seasonally adjusted all items increase.”

Wow. That’s a big increase in used car prices, and a large influence for something that only makes up 2.7% of the overall CPI basket.

Based on questions we have heard from clients, we have compiled some data on used car sales and pricing as well as the financing of autos generally (new and used). This can help us gauge the relative impact of the auto market and whether the surge in used car prices is likely the result of demand or supply issues.

The chart below shows data since 2015 for various data series, as explained below.Source: Mill Street Research, Bloomberg

The top section shows the monthly retail sales data for used cars (in dollars, not number of cars), along with the 12-month average (since sales have seasonal fluctuations). Below that is the year-on-year growth rate of used car sales. In the third section is a widely followed index of used vehicle prices, the Manheim US Used Vehicle Index.


These three series tell us a few things. First, the surge in used car prices picked up by the CPI is also clearly evident in the Manheim index. That index (based on used car sales at wholesale auctions) is up 37% from its January 2020 (pre-COVID) level, and up 20% in the year-to-date alone (four months through April). Second, demand for used cars in the retail sales data has been fairly strong post-COVID, but not at the extreme level reflected in the pricing data. This combination of solid but unspectacular sales volume and skyrocketing prices suggests that limited supply is likely the key issue right now. Most likely supply has been limited by both COVID and reduced supply of new cars from auto makers, so that people who really want a car are paying up for them, and others are either priced out of the market or waiting for supply to improve.

The bottom three sections of the chart relate to auto financing, including both new and used cars. While overall bank lending growth has been weak recently, demand for auto loans has continued to rise, as shown in the fourth section. The fifth section shows auto loans at banks as a proportion of all consumer loans, and we see the big jump that started last year and has continued, now sitting at multi-year highs. So banks are clearly increasing their exposure to auto loans in both absolute and relative terms.

Is this a risk to banks if people do not pay their auto loans? While any loan can be a risk in theory, so far, the answer seems to be no. The bottom section shows the S&P/Experian auto loan default rate, which has tumbled to multi-year lows recently, currently running at roughly half the normal rate in the 2015-2019 period. The combination of stimulus and tighter lending standards by banks have thus far meant that borrowers are making their payments more reliably than before.

The bottom line is that demand for vehicles remains elevated as city-dwellers move to the suburbs and require cars, people have been uncomfortable using public transportation, and stimulus has made it easier for some buyers to afford a car. Supply has been constrained by people not selling their cars as usual during COVID, and the inability of auto makers to produce enough new cars to meet demand due to supply chain and labor issues. Banks have found an area of consumer lending that is growing (unlike, say, credit card debt) and are lending more, but so far the credit risk looks modest. Given the ongoing supply limitations in the auto industry, it seems likely these conditions could persist a while longer before demand and supply return to equilibrium.

Cyclical sectors still dominating globally on earnings trends

As we discussed in our last commentary, analysts continue to raise earnings estimates broadly as companies keep beating consensus expectations. Expectations of additional fiscal spending and ongoing easy monetary policy along with progress toward re-opening of the economy are key macro drivers, while certain sectors such as Energy and Financials which had been areas of weakness in pre-COVID and immediate post-COVID times are now contributing more positively to the earnings outlook. This is true in the US and also globally.

Within the strong headline aggregate earnings figures, the strength in earnings reports has been led by cyclical sectors, and those areas have maintained their strong dominance over defensive sectors in our global estimate revisions breadth measures. This is feeding through to the relative performance of cyclical sectors versus defensive sectors globally.

The top section of the chart below shows our aggregated sector earnings estimate revisions indicators for global cyclical sectors (Energy, Materials, Industrials, Consumer Discretionary) relative to global defensive sectors (Utilities, Real Estate, Consumer Staples, Health Care). We build these indicators in two steps. First, we use earnings estimate revisions breadth readings (i.e., the proportion of analysts raising versus lowering estimates for earnings over the next 12 months for each stock) to build equal-weighted aggregate revisions readings for each of the 11 GICS sectors using our broad (6000-stock) global stock database. Then, we create an average of the sector-level readings for the four cyclical sectors and the four defensive sectors, and calculate the rolling difference between those two averages. That difference is what is plotted in the blue line in the chart, such that readings above zero indicate cyclical sectors having stronger estimate revisions than defensive sectors, and readings below zero pointing to defensive sectors having stronger revisions globally.

Source: Mill Street Research, Factset

We see that cyclical sector revisions have been above those of defensives since the middle of last year, and the spread has continued to widen. The latest readings are at historic highs in favor of cyclical sectors (since at least 2003) and so far are showing no signs of a turn.

The bottom section of the chart shows the relative performance of cyclical sectors versus defensives. It is based on a similar aggregate index of stock returns in each global GICS sector, then aggregating the returns for cyclical sectors and defensive sectors and plotting the relative performance. 

The rising red line in recent months indicates that cyclical sectors are outperforming defensive sectors on a global basis (using equal-weighted returns). And again, there is no sign yet of a turn in the relative performance trend. Notably, the upturn that started last year came after a period of several years of underperformance by cyclical sectors, so there is reason to think that the performance trend could continue a while longer.

Overall, while the recent revisions tilt toward cyclicals is extreme by historical standards, so is the amount of stimulus and the economic rebound from last year’s COVID-driven recession. Thus it could continue a while longer, and shows no signs of turning just yet.

Analysts still can’t keep up with surging earnings

Earnings reports for Q1 are coming in very hot once again, even after several consecutive quarters of beating consensus expectations. Analysts seem to still be struggling to keep up with the strength in US earnings, and continue to raise their earnings estimates.

To be fair, analysts have never had to deal with the level of volatility and uncertainty in the macroeconomy that has been seen in the last year or so. The shock of COVID-19 and associated shutdowns in economic activity, followed by unprecedented levels of fiscal and monetary stimulus, and the record-breaking speed of vaccine development are all extraordinary events that most analysts following individual companies are not traditionally prepared to incorporate into their earnings forecasts. The limitations on travel and uncertainty among company executives themselves are also likely hampering analysts in producing their earnings forecasts.

The latest data from Factset show that about half of S&P 500 companies have now reported Q1 earnings.  The bottom line is that earnings reports have been extremely positive so far, beating estimates broadly and by wide margins once again.

  • 47% of companies have reported (238 of 505)
  • 86.5% have reported positive earnings surprises – one of the highest readings on record and in line with the big beats in Q2 and Q3 of last year
  • 80.3% have reported positive sales surprises
  • Aggregate year-on-year quarterly EPS growth for Q1 is running at 43% right now (blending actual reports with estimates for those not reported yet)
  • Financials has been the biggest contributor so far, followed by Technology

The aggregate S&P 500 estimates for calendar 2021 are also rising. Full-year EPS growth for 2021 is now expected to be +31%, up from a consensus of +18% at the start of this year.

Turning to actual earnings levels, current S&P 500 consensus EPS is $179.45 for 2021 as reported by Factset, and $203.03 for 2022. For context, the final 2020 EPS figure was $137.99, and 2019 was $160.41, so 2021 earnings will be substantially above pre-COVID 2019 levels.

And mid- and small-caps are also beating estimates broadly so far. For the S&P 1000 index (which is the combination of S&P 400 MidCap and S&P 600 SmallCap indices), with 34% of the companies having reported so far, 87% have reported positive surprises, also led by Financials.

When we look at our own aggregated indicators of earnings estimate revisions, as shown in the chart below, which measure the breadth (red line) and magnitude of changes (blue bars) to estimates over the next 12 months, we see continued net positive revisions trends in our broad (2300-stock) US universe. And we again see the impact of the quarterly earnings season showing up in the data as companies beat estimates and analysts have to raise their future forecasts further.

Source: Mill Street Research, Factset

With such strong earnings momentum, it is less surprising that stock prices have been rising steadily with few meaningful pullbacks recently. And with the Fed yesterday reiterating their stance that they will keep their current aggressive monetary support in place for some time despite the latest signs of economic recovery and inflation pressures, asset prices should continue to have tailwinds until all the good news becomes fully priced in.

Tech fundamentals still favor Hardware over Software

While the Technology sector has been less dominant in terms of returns this year than it was last year, it remains the largest sector in the US market by value and the focus of much investor attention.

Our view within the Technology sector for some time now has been to favor hardware-related industries over software-related or services areas, and the latest update of both bottom-up and top-down indicators continues to support this view.

The table below shows some of the key bottom-up fundamental metrics we track for all sectors and industries, and focuses on the six US Technology sector industries (using the GICS industry classifications) within our broad 2300-stock US universe.

The revisions breadth measures shown in the table are based on the proportion of analysts covering each stock who have raised their earnings estimates over the last three months net of those who have lowered earnings estimates. That is, the number of upward revisions to estimates minus the number of downward revisions relative to the total number of analysts covering a stock. We use a weighted average of estimates for each company’s current fiscal year (2021 for most companies right now) and next year (2022) to calculate a consistent 12-month forward earnings figure.

Source: Mill Street Research, Factset

The first column shows the average breadth reading for all stocks in each industry on an equal-weighted (“EW”) basis, while the second column shows the same data on a cap-weighted (“CW”) basis. Naturally, the cap-weighted figures give much more weight to the largest stocks in each industry, while the equal-weighted figures will be relatively more tilted toward smaller names. All stocks in our US universe require a minimum market capitalization of $200 million and at least three analysts reporting estimates, among other criteria, so there are no micro-caps or stocks with very few analysts included.

A few key points jump out from the table:

  • Revisions remain net positive in most areas of Technology, in line with the broadly positive revisions activity in US stocks overall.
  • The average revisions breadth readings on a cap-weighted basis are higher in every industry than the equal-weighted readings, and often much higher. The sector-wide average of about 51% far exceeds the equal-weighted average of about 19%. This means the largest Tech stocks have much more positive analyst activity than the average stock in the sector, i.e., a big-cap bias.
  • The strongest industries are those which focus on hardware right now, while software and services industries are substantially weaker.

We see that on an equal-weighted basis, the Technology Hardware, Storage & Peripherals industry has the highest proportion of positive revisions, followed closely by Electronic Equipment, Instruments & Components and then Semiconductors & Semiconductor Equipment. All three of these hardware-related industries also have the highest revisions breadth on a cap-weighted basis.

At the bottom of the list we see the large Software industry, which has marginally negative net revisions breadth on an equal-weighted basis. This means the average Software stock has a roughly equal number of analysts raising versus lowering estimates. However, on a cap-weighted basis, the revisions figure is strongly positive. This means that a few mega-cap names in the industry have strong revisions, while the majority of mid- and smaller-cap names have relatively weak revisions.

A similar but less dramatic picture is seen in IT Services, where revisions are much lower than in the hardware related industries on both equal-weighted and cap-weighted metrics. The cap-weighted figure is still higher than the equal-weighted, as it is for all Tech industries, but by a much smaller margin than in Software.

The overall picture from the bottom-up, stock level analyst revisions data is that hardware makers continue to surprise analysts positively across large and smaller companies alike, while providers of software and tech-related services are much more mixed and dominated by the largest names in those areas.

The second chart below provides some macro, top-down context for the bottom-up fundamentals we see. The top section of the chart plots the year-on-year percent changes in aggregate spending on investment in computer hardware (blue line) and in software (red line) over the last 30 years. The data come from the quarterly GDP data produced by the US Bureau of Economic Analysis. As of the end of 2020, the growth rate of investment in computer hardware had jumped to 16%, the highest reading since the previous recessionary rebound in 2009-10. Growth in software investment spending, by contrast, had only grown 5%, which is near the lower end of its recent historical range.

Source: Mill Street Research, Bureau of Economic Analysis

The bottom section of the chart plots the ratio of software to hardware spending over time. We see that the line has tended to rise, indicating higher growth in software spending than hardware spending on average over the long run. But the latest readings show a drop, reflecting the jump back toward hardware spending that occurred last year and likely is still ongoing. In the longer-term, it seems likely that software spending will again resume its higher growth rate versus hardware, but the preference for hardware is still very visible in current analyst earnings forecast behavior and could last a while longer as the world continues to adapt to the post-COVID landscape.

Labor market improving but still shows plenty of slack

In the longer-run, a key measure of inflation pressure is the amount of labor market slack (unemployed or underemployed people), which heavily influences the ability of workers to demand higher wages.

The standard reported unemployment rate data (i.e., the U-3 measure in the US) is useful in measuring labor market slack, but has limitations due (in part) to its definitions of “unemployed” and the “labor force” in the calculation: “unemployed” people as a percentage of the “labor force”. That is, to be counted as “unemployed”, a working-age person must be considered actively looking for a job (when asked if they have sought employment in the last four weeks in the monthly household surveys done by the Bureau of Labor Statistics, BLS). If they are not currently seeking employment for any reason, they are “not in the labor force” and thus do not count in the standard unemployment rate, even if they consider themselves unemployed.

Another unemployment rate put out by the BLS, known as the U-6 measure, attempts to address this issue to some degree. The U-6 rate is much broader and includes anyone who has been looking for work within the previous 12 months but has been unable to find a job and has not looked for work in the past four weeks. It also includes anyone who has gone back to school, become disabled, and people who are underemployed or working part-time hours and want to work more. The first chart below plots the traditional U-3 rate and the broader U-6 rate since 1994. We can see that they typically follow a very similar pattern, but by construction the U-6 rate is always higher, and likely a more accurate description of slack in the labor market. What is more notable right now is that even after a dramatic recovery in employment from spring of last year (supported by historically large fiscal and monetary stimulus), the current U-6 unemployment rate of 10.7% is still higher than the worst reading of the 2001-02 recession, and far from the December 2019 low of 6.8%. By comparison, the current U-3 rate of 6.0% is likely understated (too optimistic) relative to the pre-COVID (Feb 2020) low of 3.5% due to measurement issues.

However, these metrics still rely on classifying people into various employment categories with somewhat arbitrary conditions.  An even broader and simpler measure of unemployment can help address these issues, and may be a better overall metric of labor market slack: the employment/population ratio (often abbreviated as “EPOP”) among “prime age” people aged 25-54 is simply the proportion of all persons in that age range who are employed, regardless of whether they are actively seeking a job.

The second chart below shows the total EPOP ratio over the last 50 years in the top section, and the bottom section shows the breakdown between male and female EPOP readings. Even after a sharp rebound from last year’s extreme lows and a very strong employment report last week, the overall reading of 76.8% is still far from the recent “full employment” levels in the 80-82% range in recent decades. Note that the true level of “full employment” is unknown and possibly higher than previous peaks in this ratio. This pattern can be seen in the data for both men and women, with men having a downward drift in EPOP over time while women joined the workforce up until about 2000 and have then remained in a range about 10 percentage points below that of men.

Because we are looking at people in the 25-54 age range, when adults are typically working, retirement and schooling should not have much impact on the ratios (as may be the case in the whole adult population). With a population of about 126 million in this age range in the US, the 4% difference in EPOP from current (76.8%) to previous peak levels (~80-81%) suggests five million or more people who are still unemployed potential prime-age workers. The average monthly gain in employment in this age group in 2016-2019 was about 86K/month, so even at an average of 100K jobs per month on a sustained basis it would take roughly four years to return to the previous peak. Further stimulus and a decisive suppression of COVID could accelerate that pace, but there is likely to be slack in the overall labor market for some time. This, in turn, suggests that labor cost-driven inflation is unlikely to accelerate on a sustained basis, beyond the short-term impacts from supply chain disruptions and re-opening.

Inflation is steady on the surface, volatile underneath

Inflation expectations have been a topic of growing interest thanks to the extraordinary fiscal and monetary support in place for much of the last year, most recently the huge $1.9 trillion American Rescue Plan that is currently sending checks out to millions of Americans.

All of this new spending by the federal government, along with the economic recovery permitted by the rollout of COVID-19 vaccines, is provoking more discussion about whether demand for goods and services will outpace the economy’s ability to produce them and push prices higher.

While we have previously discussed why we do not think inflation will be a major problem in the next few years, investors in the Treasury market appear to have raised their forecasts for future inflation, at least in the near-term. This is reflected in the “breakeven rate”, which is the difference between the yield on nominal Treasury bonds and that of the corresponding TIPS (Treasury Inflation Protected Security) bond. TIPS provide a “real” interest rate (currently negative) plus whatever the rate of inflation is over the bond’s duration. The difference between nominal and TIPS yields indicates the level of inflation that would produce the same return to holding either type of bond after accounting for inflation.

The chart below shows the breakeven rates for five-year and 10-year maturities, and we see that the five-year inflation expectations in particular have risen sharply recently, now back to levels around 2.5% seen after the last two recessions. The 10-year breakeven rate has risen less, as investors appear to expect a temporary jump in inflation that will subsequently ease. Both figures are roughly in line with the Fed’s stated inflation target, and thus are unlikely to provoke a tightening response at this point.

Breakevens 5yr 10yr_032521

Source: Mill Street Research, Bloomberg

Reported inflation, as measured by the Personal Consumption Expenditure (PCE) Price Index (the Fed’s preferred measure), has remained low thus far. Year-on-year comparisons will soon be skewed by the price drops around this time last year when the economy buckled under the weight of COVID and the associated limits on activity. Overall, however, there are thus far few signs of significant overall inflation in the reported data.

When we look under the surface of the inflation data, however, we find that there are a lot of significant changes going on among the various components, which are thus far mostly offsetting each other. This is not unusual around recessions, and certainly consistent with the extreme and unusual kind of shock that COVID-19 has caused in the economy.

The Federal Reserve Bank of San Francisco publishes metrics that capture the dispersion and amount of underlying change in the inflation rates of the various categories of goods and services measured in the PCE data. The charts below show the increased levels of “rotation” under the surface of the muted headline PCE inflation measures.

The first chart below shows the last 15 years of data on the PCE headline (total, dark blue line) and core (excluding food and energy, red line) inflation rates along with measures of the dispersion among the 100+ granular subcomponents of the PCE inflation data. It shows the inflation rate for the 90th and 10th percentile subcomponents of the PCE price index and below that the spread between the 90th and 10th percentiles as a measure of the cross-sectional dispersion or variation of inflation rates among products at any given time. We can see that dispersion in inflation rates among the categories of PCE components has widened significantly recently to 6-7%, up from readings closer to 4% for much of the last several years.

Also notable is the fact that even the subcomponents with the highest inflation rates (the 90th percentile, purple line) have price gains of only about 4%, marking a fairly modest upper bound on inflation pressures so far.

PCE Inflation and DispersionSource: Mill Street Research, Federal Reserve Bank of San Francisco

With some prices falling substantially, the current inflation spread (90th vs 10th percentile) is among the widest in the last decade, and it quantifies and corroborates the intuition that the supply and demand impacts of the COVID-driven recession and recovery have led to more divergent outcomes for different industries (some hurt badly, some benefiting strongly). The inflation spread widened even more dramatically around 2008, driven by enormous volatility at that time in energy prices in particular, and there was less direct fiscal stimulus support in that recession than the current one.  This analysis helps make the point that the gap between winners and losers seen in the stock market has a fundamental driver visible in the inflation data.

The final chart below plots another way of looking at the underlying divergences beneath the surface of the headline inflation data. The same San Francisco Fed data set also calculates the percentage of PCE price index subcomponents that have had large changes in their 12-month inflation rates relative to their previous five-year averages (“large” meaning two standard deviations from the five year average). Industries that have seen their inflation rates lurch upward would be captured in the blue line, showing the percentage of all PCE components with large upward movements in inflation, while those facing big downward changes would be captured in the red line.

Proportion of Big Changes in PricesSource: Mill Street Research, Federal Reserve Bank of San Francisco

The grey line below plots the sum of the two series to provide a measure of how many industries have seen major shifts in their respective inflation trends. Here we see that 25-30% of industries have recently had major shifts (up or down) in pricing power relative to their longer-term trends, much higher than the usual 5-15% readings for most of the last decade.

We also see that current readings are at similar levels to those seen around the 2008 recession, though with a slightly different pattern: in 2008, there was initially a spike in price increases, then a spike in price decreases, while this year has seen concurrent but less extreme spikes in both series as demand and supply had to simultaneously adapt to COVID’s impact on activity.

The implication is that some industries that may have had weak pricing power pre-COVID now have stronger price trends, while others that had higher industry-level inflation are now under more pricing pressure. While there are always a few such industries at any given time, there are many more right now.

These data help provide some further fundamental corroboration for the rotation in stock market industry trends and wide gaps between winners and losers, even in an environment of continued low headline inflation rates.

S&P 500 earnings have fully recovered, but with wide differences among sectors

The S&P 500 has reported another strong earnings season for Q4, with 79% of companies beating consensus earnings estimates for the quarter. This would be the third highest such reading in Factset’s data since 2008. The beat rate for top-line sales was similarly high at 77%. Aggregate income for the index is about 4% above year-ago levels, indicating that net income on a quarterly basis has fully recovered pre-COVID levels based on Factset’s data.

However, when we drill down into the earnings trends in the S&P 500 and its component sectors, we see the wide differentials in earnings patterns among sectors.

The charts below use quarterly index operating earnings per share as reported by S&P (S&P’s data differ slightly from Factset and other sources). The grey shaded areas are the current bottom-up consensus forecasts for the four quarters of this year.

S&P 500 Quarterly Sector EPS p1Source: Mill Street Research, S&P Dow Jones Indices

The top section of the first chart above shows us that the overall S&P 500 index is expected to see its quarterly EPS rise steadily over the next few quarters and see quarterly EPS exceed the previous peak of Q2 2019 by Q2 of this year, before reaching a new high of $47.06 by Q4 (which would be up 37% from Q4 2020).  The aggregate index figures, however, hide the variation among the 11 sectors that comprise it.

We see that Consumer Discretionary is expected to see earnings recover but not decisively exceed recent highs. Consumer Staples shows a choppy upward pattern, with Q4 expected to be slightly below Q3 of this year.

The Energy sector is expected to rebound from heavily negative earnings during 2020 to moderately positive EPS this year, but still well below the levels seen in 2018 and early 2019 (though if crude continues to rise, that could change).

Financials are expected to see earnings generally plateau around current levels, with no improvement in EPS from Q4 2020 levels, and still near levels seen in 2018-2019. Low net interest margins on lending have been a heavy headwind, partly offset by profits from stock and bond market activity and mortgage-related fees.

Health Care is expected to see a jump in earnings in Q1 of this year and then hold steady at new highs. As we will see, it is thus one of the few sectors where the consensus sees new highs in earnings occurring this year.

Turning to the second chart below with the remaining sectors, see that Industrials (as a classic cyclical sector) had a big drop in earnings in 2020 (with the Transportation component weighing heavily) but is expected to fully recover by the end of this year. Naturally, Industrials should be among the biggest beneficiaries of both fiscal stimulus and re-opening from COVID limitations.

S&P 500 Quarterly Sector EPS p2

Source: Mill Street Research, S&P Dow Jones Indices

The Technology sector, as one might guess, has been least affected by COVID and almost certainly got a boost in the second half of 2020 due to a surge in spending on IT as workers and companies scrambled to work remotely wherever possible. Consensus calls for earnings to continue growing and reach new further new highs this year.

The Materials sector is seeing a big rebound in earnings from a decline that began well before COVID hit when commodity prices and global growth were already relatively weak. Analysts expect the sector’s earnings to hit a near-term peak in Q2 of this year (matching the 2018 peak) before easing by year-end.

Communication Services has been under earnings pressure for some time, as a mix of some big Tech-related firms along with Telecom, Entertainment and Media firms. Earnings are expected to rise steadily from the current (Q4 2020) lows through the end of this year.

Earnings for the Utilities sector look to remain in a range, with no discernible growth trend visible. Demand for electricity, natural gas, and other utilities has seen little overall growth for years, and regulations in many areas keep revenues and profits relatively stable.

Finally, the Real Estate sector (principally REITs), has seen earnings hit hard by COVID, and the consensus is for only very modest recovery this year as demand for commercial real estate is likely to remain depressed.

Overall, it appears that only the Growth-oriented sectors of Technology and Health Care will see earnings reach true new all-time highs this year. Other sectors may approach or just exceed 2019 levels, while a few have further to go. While fiscal stimulus and re-opening should provide a tailwind for earnings in many sectors, those winds are not blowing at the same speed for all.

Reviewing current stock vs bond sentiment

Despite what you might hear or read some places, investor surveys do not show an extreme level of optimism toward US stocks. Bullishness on stocks has in fact declined somewhat recently and is not far from long-term average readings.

Sentiment toward bonds, by contrast, has moved quite sharply and is now approaching extreme bearishness by the standards of recent years. This is not too surprising, given that long-term Treasury bond yields have recently risen to their highest levels since COVID hit early last year. The result has been that investors in long-term (20+ year) Treasury bonds have lost about 13% since the end of November and about 18% since the end of July.

One of our key measures of survey sentiment for equities is shown in the top section of the chart below. It is an average of several sources, each of which monitors the proportion of market commentators who are bullish or bearish on equities. They include Investors Intelligence, Market Vane, and Consensus Inc.Stock vs Bond Sentiment

Source: Mill Street Research, Investors Intelligence, Market Vane, Consensus Inc.

The average proportion of bulls on stocks right now is about 60%, down from a recent peak in January of 67%.  The average proportion over the last five years has been just over 57%, so the current reading is only slightly above average. The peak in the last five years (and in fact since at least 2003) was at 76% back in January 2018 (when the corporate tax cuts were going into effect), while the lowest point in the last five years was the brief reading around 33% in late March of last year amid the worst of the COVID-related panic. Readings below 50% on this metric are relatively rare outside of panic periods.

These survey readings stand in contrast to other sentiment-related indicators like put/call volume ratios, certain valuation metrics, or anecdotal “indicators” like the popularity of SPACs or the extreme behavior of “meme stocks” with heavy retail trading activity. Other market-based measures of risk appetite (e.g. relative performance of riskier stocks) suggest investors are still broadly in a “risk on” mood.

Sentiment readings that are not extreme in either direction tend to give less information about future returns. But we can certainly see that stocks have often done well when sentiment has been near current levels, as sentiment has been around these levels since August . Extreme bearishness tends to be the strongest signal for stocks, but occurs relatively rarely given the long-term uptrend in US stock prices.

The negative sentiment toward bonds suggests that the recent rise in bond yields may be getting somewhat stretched. If investors continue to assume that the Fed is not changing its current policy guidance and will keep short-term rates near zero for at least the next year or two, then long-term yields may not have too much further to rise.

Taking the next step, if equity investors are growing more concerned recently about risks that higher bond yields may cause regarding the level of stock prices (i.e., a higher discount rate reduces the present value of future earnings), then sentiment toward bonds is worth watching as part of the broader picture facing equities.

Bond market making some noise

The bond market has clearly awoken from what appeared to be a low-volatility Fed-induced slumber for much of last year. Longer-term bond yields in the US and elsewhere have jumped to their highest levels since just before the COVID crisis hit markets early last year (blue line in first chart below). Even after this rise, though, the 10-year Treasury yield remains below even the low points of previous cycles.

Short-term rates (0 to 2 year), meanwhile, have remained anchored near zero due to the Fed’s continued indications that any policy rate hikes are still at least two years away.

US Treasury Yields

There are two main components to Treasury yields (which do not have default risk): real yields (yield after subtracting expected inflation) and inflation expectations. The first component to move was inflation expectations (first chart below), and more recently real yields (second chart below) have also risen notably. However, real yields remain solidly negative and much lower than they were before COVID-19 hit last year.

Breakevens 5yr 10yr

Treasury 10-yr Real Yield

What have been the drivers of the recent move higher in long-term bond yields?

Some of the most prominent reasons include:

  • Additional fiscal stimulus is expected soon, which will potentially boost inflation and growth, and also require the issuance of a large amount of new Treasury bonds.
  • The economy’s recovery from COVID-related weakness is continuing thanks to ongoing vaccinations, which will combine with stimulus to push growth much higher than average this year.
  • Sharply higher prices for oil and other key commodities (copper, lumber, etc.) are raising concerns about inflation already in the pipeline, with further demand potentially driving prices even higher.
  • Selling or hedging activity by investors who trade leveraged positions in bonds, and/or own mortgage-backed securities that now appear riskier. Rising mortgage rates typically reduce pre-payments (due to less housing turnover) and refinancing of home mortgages, which increases the duration, and thus risk, of mortgage-backed bonds.

The jump in market-implied inflation expectations has been quite sharp, though something similar happened after the 2008 crisis period as well. The current inflation expectations reflected in 5-year and 10-year Treasuries (difference between nominal Treasury yields and inflation-protected TIPS yields) are now slightly above the Fed’s 2% inflation target. Indeed the 5-year implied inflation rate (breakeven rate) is at its highest level since 2013, and the 10-year rate touched its highest since 2014. Investors now appear to expect a rebound in inflation, which is likely not considered a problem by the Fed (given their stated inflation goals) as long as expectations do not become excessive. However, the Fed’s own buying of TIPS in the past year may itself have skewed the interpretation of “market expectations”: the Fed now owns well over 20% of the entire TIPS market, potentially influencing their prices (yields) more than usual.

For stock prices, a moderate amount of inflation is not necessarily a bad thing for corporate earnings, as long as it does not provoke Fed tightening. Since equities (via corporate profits) are partly hedged against inflation, real yields are often viewed as the more important rate. Thus the recent rebound in real bond yields has raised concerns about equities losing the tailwind they have had from falling real yields for more than two years. It has also provoked a shift in investor interest from Growth stocks that are attractive when economic growth and inflation are low toward Value (and other riskier) stocks that tend to do better when economic activity and inflation are accelerating.

With short-term yields still pinned near zero by the Fed, the yield curve has steepened sharply recently. This is notionally good for Value-oriented banks and other “spread lenders” (who borrow at short-term rates and lend at long-term rates), though many other factors can affect the profitability of lending, and higher rates can dampen loan demand.

And one of the biggest impacts of higher long-term yields is in fact on mortgage rates and thus the housing market. Housing has been an extremely hot area recently, as demand for single-family homes is very high and supply has been unusually low. The average 30-year mortgage rate recently hit an all-time low of about 2.8% but has now jumped back up to 3.14% in just the last two weeks (and will likely rise somewhat further near-term). While this is still very low by historical standards and may not hamper housing too much, at the margin it may cool demand, and any further increases would put additional pressure on housing demand.

Overall, a moderate rise in long-term yields with well-anchored short-term rates is not necessarily a major cause for concern for stocks, or the economy, in general. It does however mean potentially more rotation among sectors and styles within equities. It will also bring up concerns about whether or when the Fed will decide to alter its current policy trajectory in response, as too much of a rise in yields could cause economic dislocations that the Fed would prefer to avoid.

Recent rally in “junk stocks” is not unusual

Financial headlines have been captivated recently by explosive behavior in certain “meme stocks” that have been the subject of intense speculation by online retail traders as well as some hedge funds. This has been accompanied by a general trend of outperformance by smaller, money-losing, heavily-shorted, and volatile stocks (sometimes referred to as “junk stocks”, similar to risky high-yield “junk bonds”).

Other signs of “froth” include aggressive use of SPACs (Special Purpose Acquisition Company, or “blank check” company that raises money to acquire private companies), historically high trading volumes and activity in short-term call options, and growing margin debt.

This has raised broader questions about why “junk stocks” seem to be rallying much more than “quality stocks” and whether this is historically unusual.

The short answer is: no, this is not historically unusual under these circumstances. The specifics vary, but similar patterns have been seen in the past when markets are recovering from a sharp decline and policy support is very aggressive.

The first chart below plots the recent (past year) returns of the Dow Jones Thematic Market-Neutral Style indices. These are hypothetical long-short indices (i.e., assuming equal dollars invested in offsetting long and short portfolios) based on widely-used factors, using the 1000 largest US stocks, and constructed sector-neutral.

The key points are:

  • Quality and Anti-Beta (low beta) are highly correlated, since quality stocks (defined by Dow Jones as those with high Return on Equity and low Debt/Equity ratios) also tend to be lower beta. Size (small-caps) is often negatively correlated with Quality and Anti-Beta (since small-caps are generally lower quality and higher beta). Risk is the key theme connecting these factors.
  • November 6th (where shaded area begins) is when Pfizer announced its first COVID-19 vaccine results.  This marked the point at which the recent themes really began: outperformance by low-quality, high-beta, small-cap stocks. This initially hurt the price momentum factor since those had not been the leadership areas previously.
  • The Value factor had a bounce in November, but since then has shown no net performance. Thus it is not Value that has been rewarded, but risk, in the period since November 6th.
  • It is also not coincidental that early November was when the US election occurred, and the results (fully decided in January) increased the perceived odds of additional aggressive fiscal stimulus. Such stimulus tends to benefit smaller, weaker (riskier) companies that had been hit hardest by COVID-19.
  • Thus riskier companies have had recent tailwinds from both COVID-19 developments and greater fiscal support.

Dow Jones Thematic Style Indices All

The long-term chart below shows the Quality and Anti-Beta factors since the data begin in 2001. We can see the correlation is clear over the longer-term, including the most recent few months.

Dow Jones Thematic Style Indices Quality AntiBeta

The key point here is that in each of the previous periods of post-recessionary aggressive stimulus (2002-03, 2009-13), higher risk (lower quality) stocks were rewarded as investors sought the biggest “bang for the buck” from the stimulus and recovery. Weaker, riskier stocks tend to get the most benefit from policy support, while stable, higher quality companies do not need it and get relatively less benefit. Thus the current conditions are not unusual, and fully consistent with a risk-on environment, in line with our other indicators that remain bullish for equities on a tactical basis.