Sentiment Disconnect

Every few months, we’ve revisited sentiment metrics to see how broader sentiment trends, generally contrarian in nature, line up with the prevailing market direction.  One of the more interesting aspects of the global equity market rally off last summer’s lows has been the consistently low readings registered in various investor sentiment gauges.  Investors love to rely on a bevy of simple clichés to explain market action; one of the most used is the notion that bull markets climb a “wall of worry.”  It surely seems like “wall of worry” behavior is a prominent part of the market advance over the past few months.  

One of the more prominent investor sentiment indicators in the US is the American Association of Individual Investors Bull/Bear sentiment indicator which, according to the AAII website, “measures the percentage of individual investors who are bullish, bearish, and neutral on the stock market for the next six months.”  With data going back to 1987, the numbers provide an interesting history of investor sentiment through several major market crises, such as the ’87 crash, the August 1990 Kuwait invasion, the Asian crisis in 1997/1998, the massive bear market from 2000 to 2002, the financial crisis and recession from 2007 to early 2009, and the sovereign debt crises over the past few years.  In a fascinating development, at almost the exact point the S&P was reaching new all-time nominal price highs (and current cycle highs) last week, the AAII Bull/Bear numbers showed a bullish reading of only 19.3, the lowest reading since the week the S&P 500 bottomed near 660 in March 2009 (the reading then was 18.9) and the 26^th lowest reading in the database of 1,342 weekly readings.  On the flip side, the bearish reading of 54.5 was the 31^st most bearish reading in the data set.  And, taking the differential between the two, subtracting bear from bull, the -35 reading was the 13^th most negative weekly reading over the past quarter century, only exceeded by readings during the reaction to the 1990 Iraq invasion, the aforementioned week of 3/5/2009 when the equity markets hit their absolute low during the financial crisis, other readings generated during the 2008/2009 meltdown, and one week from summer 2010 when the European sovereign debt crisis reached a frenzy.  Below is a list of the most bearish numbers:

Source: AAII, Bloomberg, IronHorse Capital 

As seen above, extremely low readings have generally been precursors to intermediate to longer-term equity market rallies.  Markets rallied smartly after the 1990 swoon, as well as following the 2010 summer declines (though we got a nasty repeat of market disruption the following summer and fall).  Like any other extensive dataset, exceptions exist.  In the above, the readings from early 2008 perhaps provided false hope to contrarians that the worst was over in the months surrounding the Bear Stearns collapse; the bulk of the equity market losses in the crisis occurred later in the year.  Still, and most importantly, we’re currently observing a level of disgust with equity markets that’s more associated with severe equity market disruption, not with new market highs.  This is a situation that belies the notion that sentiment in markets is far too frothy.  

Other longstanding indicators are also showing a healthy dose of skepticism.  The CBOE’s composite put/call ratio, which “tracks the ratio of total equity and index put/call volume traded on the Chicago Board of Exchange,” remains elevated.  The 10-week moving average is currently 0.73 standard deviations above normal, not an alarming result, but notable considering the current positioning of the market (barely off the highs).  The ISE Sentiment All Equities Index shows, “the number of calls traded for every 100 puts.”  Lower numbers show lower sentiment in equity markets, again contrarian.  The 10-day moving average here is currently 1.21 standard deviations below the mean.  They’ve remained in this posture consistently since last summer.  The chart for the ISE Index follows:

Source: Bloomberg, IronHorse Capital

We don’t want to draw any definitive conclusions on the future path of equities based on a few pieces of sentiment data.  In general, the statistical connections between sentiment data and future equity market performance have been somewhat inconclusive.  Even so, we’re always interested when we see readings in data sets that are at or near extremes.  Today, we’re seeing data on sentiment that is very rarely if ever observed at the peaks of long-term bull markets, and is more often observed with significant equity market lows.  Perhaps there are factors distorting some of the data, such as the fact that investors are bombarded by market news through numerous sources, and much of that news is negative in nature in light of the constant crisis posture in Europe and elsewhere.  There’s no way to come to solid conclusions without having complete access to the underlying survey data from AAII. On the surface, though, we think the data provides an interesting insight into individual investors, namely that the individual investor remains decently underweight equities.  In the past, significant upward moves have been sparked when individual investors on the sidelines en masse decide they can’t take the pain of being out of the market anymore and throw their collective hats in the ring.  This type of activity could provide an underlying bid to markets going forward, keep corrections relatively contained, and keep markets grinding higher.  We’ll become very worried when sentiment indicators like those listed above become uniformly and excessively bullish, the flip side of the current marketplace condition.

Small Cap Performance vs. Large Cap Performance

A few weeks ago, we looked at the performance history of Value vs. Growth and observed that the history was prone to longer streaks.  This week, we’ll look at the performance history between large-cap names and small-cap names using the Russell 1000 as a proxy for large-cap, and the Russell 2000 index as our small-cap index.  According to Russell Investments, the Russell 1000 Index , “…measures the performance of the large-cap segment of the U.S. equity universe…” and,  “…represents approximately 92% of the U.S. market.”  Alternately, the Russell 2000 Index, “…measures the performance of the small-cap segment of the U.S. equity universe.”  

As with the data series pertaining to growth and value, small-cap performance relative to large-cap performance has also been subject to longer trends.  Let’s begin by presenting the yearly performance data (simple, i.e. no dividends included) for both indices going back to 1979.  Below, we show the annual return for both indices, and the difference in performance for each year.  A negative number in the “Difference” column represents underperformance by the larger-cap Russell 1000 in that given year.  

Source: IronHorse Capital and Bloomberg

From 1979 through the end of last year (34 years of data), the small-cap Russell 2000 index outperformed its larger-cap counterpart by approximately 1% per year, 9.36% per annum vs. 8.36% per annum.  These performance numbers belie a wide range of performance outcomes when the numbers are examined by decade, or when broken down by the winning streaks for each series identified in the data set.  

Going by decades, you can see that larger-cap names outperformed significantly during the secular bull market years of the 1980s and 1990s, but have underperformed during the secular bear market we have experienced since 2000.  

Eyeballing the annual data series above, and moving beyond the confines of tidy decades, it appears that outperformance and underperformance regimes run for approximately 15 years or so, again roughly in-line with the broader secular bull/bear positioning in the market.  From 1999 through 2012, 14 years, small-cap names outperformed 10 times with a cumulative return over that time frame of 101.29% vs. 22.87% for the larger-cap Russell 1000.  On the flip side, from 1984 to 1998, 15 years, the Russell 1000 outperformed 9 times.  Cumulative performance: 611.3% for the Russell 1000, 275.8% for the Russell 2000 small-cap index.  Of note, prior to the 1984 turn towards a large-cap streak, the Russell 2000 small-cap index had outperformed for 5 straight years from the late 1970s through the early 1980s recession years, which happened to mark the end of the 1968 to 1982 secular bear market.  

It seems counterintuitive that small-caps would outperform large-caps during secular bear markets in light of the fact that small-caps would seem to benefit more from consistent, strong economic growth, usually a feature of secular bull periods, lower volatility, another characteristic of secular bull markets, and better access to debt and equity capital markets (theoretically better during secular bulls).  Various analysts ascribe performance differentials to everything from the direction of interest rates and inflation in bull and bear periods to growth in real GDP.  Looking back at various data sets, there doesn’t seem to be a consistent pattern to create a storyboard when it comes to macro data.  For instance, small-caps outperformed during the 1970s and early 1980s according to Ned Davis Research, a period defined by rising interest rates and inflation/stagflation (again, our available data set ends in 1979; we’ll have to take Ned Davis’ word).  Small-caps outperformed during the 80s and 90s as seen above, a period defined by strong economic growth, declining interest rates, and declining inflation.  This led some to conclude that higher interest rate environments turned out better for small-caps at the expense of large-caps.  However, the 2000s have been defined by a continue drop in interest rates, and even lower inflation metrics.  Nonetheless, small-caps reversed their underperformance and resumed a leadership position.  We’ll leave it up to academics and others to ascertain the exact reasons why small-caps have been outperforming during poor overall market periods.  Suffice to say, it’s a curious quirk in the data, but one that investors should pay attention to.  

Based on the fact that large cap names have been mired in a long period of underperformance (nearly 15 years) that matches the length of past streaks, it seems small-caps may be pushing the limits with the current winning streak.  Valuation may confirm this as well.  At the last major performance turn, small-caps were consistently overvalued versus large-caps on an EV/EBITDA basis.  At the end of 1998, for instance, the Russell 2000 was trading at 10x EV/EBITDA vs. 13x for the Russell 1000.  That situation is now reversed.  The Russell 2000 is now trading at approximately 12x vs. 9.5x for the Russell 1000.  

Taking all into consideration, we believe large-cap will outperform small-cap in coming years, and based upon the data on growth and value we outlined in an earlier post, believe value will take the performance baton back from growth.  As with our last post, we’ll leave you with a chart that gives a visual representation of small and large-cap out/under performance through the years.

Source: IronHorse Capital and Bloomberg