Investor Psychology - Going Against the Herd
Value investing's superior long-term performance has endured the test of time. So long as markets continue to be driven by human behavior and biases, it will remain relevant.
We are by nature bad investors
Late 1990s, access to the internet expanded and computer technologies became increasingly intertwined in daily lives. Stock values soared. The value of the NASDAQ, representing many of the largest tech stocks, grew from around 1,000 in 1995 to more than 5,000 in 2000. New IPOs were fetching extravagant prices, without any revenues or cash flows. Companies such as eNotes, simplayer and Constellation 3D appreciated by more than 1000% with zero sales. Investors increasingly jumped on the bandwagon, fearing of losing out, despite concerns over stretched valuations. The Merrill Lynch analyst Henry Blodget or “King Henry” as the Wall Street Journal called him, encouraged investors to bid technology stocks up from already elevated levels. Publications like Fortune, Money and SmartMoney added fuel to fire by repeatedly making calls for the Millennium’s hottest stocks. From the peak of Mar 2000 to Oct 2002, the S&P 500 suffered a draw-down of 49%, the tech heavy NASDAQ lost 78%.
Between 2007 and 2009, U.S. equity mutual funds had seen more than $236 billion in net outflows, whereas U.S. bond funds had seen their assets swell by $504.6 billion. Late 2008, CNBC ran a series titled “The Death of Buy and Hold” and Barron’s Electronic Investor published an article in 2009 titled “Modern Portfolio Theory Ages Badly”. Both the average investor and institutions questioned the future out-performance of equities vs bonds as panic ensued. 8 years later, from the troughs of Mar 2009 to present day Mar 2018, the S&P 500 has returned over 22% per year, a cumulative gain of 392.8%. Investment grade corporate bonds with an average rating of A, returned 7.8% per year, a cumulative gain of 82.25%.
Oct 2016, the UK financial regulator launched an investigation into the workings of the asset management industry, to examine fund fees and profitability. The consultancy – Create Research conducted analysis which showed that 99% of actively managed U.S. equity funds sold in Europe had failed to beat the S&P 500 over the preceding 10 years, only two in every 100 global equity funds have outperformed the S&P Global 1200 since 2006. And almost 97% of emerging market funds have under-performed.
The average investor, news media, institutions, analysts, portfolio managers. None are spared. Humans are bad investors and it is our very nature that fails us.
Humans more prominently recall and emphasize recent events and observations than those in the near or distant past. During bull markets, we tend to forget about bear markets and believe that it will continue to go up as has been the case. We increase our equities allocation and forsake the principles of risk management and portfolio diversification.
Conversely, in a bear market, when our net worth tumbles, we sell our equity holdings at rock bottom because we feel the market’s decline will continue.
We delude ourselves
Surely, a quick glance at stock market histories reveals that following extended bull markets, will come bear markets and subsequently, the beginning of a new bull market. (Since 1930s for the S&P 500, on average bull markets last 97 months and bear markets last 18 months) Stocks that have extended valuations unmatched by fundamentals will eventually crater, and the unloved but solid companies will eventually rise.
Concerns mounted over valuations during the dot com craze, and assuredly our rational minds questioned the sky-high stock prices. Greed won. Instead of putting a check on our emotions, our rational minds “rationalized” our greed, discarding any semblance of doubt. Even portfolio managers who rely on empirically proven strategies doubt themselves and change courses as the market reacts differently.
Numbers don't lie
James O’Shaughnessy, Chairman & CIO of O’Shaughnessy Asset Management and author of the book “What Works on Wall Street”, tracked U.S. stock market returns from 1927 to 2009 and detailed the success of various strategies over multiple rolling time periods. The results show that over long-time horizons, there exists a set of simple strategies that have consistently outperformed the market or sector index. Both on an absolute and risk-adjusted basis.
In my description of these strategies, I have omitted those which distinguish between micro-cap, small-cap, large-cap and market leader stocks, include combinations of metrics, and which target a set number of investable companies, to provide a list of simple to apply strategies. I encourage reading the book’s fourth edition for the full analysis. It includes composite strategies that combine multiple value factors together with growth factors to achieve superior returns to the simple one-factor strategies shown below.
The All Stocks index represents all companies listed in the U.S. exchanges (including ADRs) with a minimum market capitalization of $200 million (1995 dollars). All single factor and composite strategies described in this article selects stocks from this index.
Single Factor Strategies tested (by 1st and 10th deciles)
Stocks were re-balanced on a yearly basis based on the strategies’ criteria.
James assumed no trades are made throughout the year, with one exception – if a stock goes bankrupt or is acquired before the formal re-balance. Then the proceeds are invested on a pro-rated basis across the remaining stocks in the portfolio. He also examined annual returns and removed stocks with extreme returns or data that were inconsistent with outside information.
In real time, he held a list of six “red flags” that provided alert to remove a stock from a portfolio and replace it with a new one that met the criteria of the strategy. Such include:
• If a company failed to verify its numbers as required by Sarbanes-Oxley, it was replaced.
• If a company was charged with fraud by the Federal government, it was replaced.
• If a company restated numbers such that they would not have qualified at the time of purchase, it was replaced.
• If a company received a takeover offer from a third party, it was replaced if its stock price moved to within 95% of the takeover offer price.
• If the company dropped by 50% from purchase price and was in the bottom 10% of all stocks for the previous 12 months of price performance, it was replaced.
• If a stock based on dividend strategies had cut its dividend by 50% or more, it was replaced.
Table 1: Top 5 strategies, Aug 31, 1965 - Dec 31, 2009. Strategies sorted by compound average annual return. Excludes composite strategies.
Table 2: Worst 5 strategies, Aug 31, 1965 - Dec 31, 2009. Strategies sorted by compound average annual return. Excludes composite strategies.
Table 3: Top 5 strategies Percentage of wins against the All Stocks benchmark for all rolling one-, three-, five-, seven-, and ten-year periods, Dec 31, 1967 - Dec 31, 2009
Table 4: Worst 5 strategies Percentage of wins against the All Stocks benchmark for all rolling one-, three-, five-, seven-, and ten-year periods, Dec 31, 1967 - Dec 31, 2009
Table 5: Summary results for various strategies applied to the various sectors (Bench-marked against the returns of all stocks in the sector) Dec 31, 1967 - Dec 31, 2009
Value trumps growth
Over lengthy periods of time, Value trumps Growth. All top 5 single factor strategies were value-oriented (tables 1 and 5). However, Value can get significantly out of sync during bull markets. Stocks that are very cheap from a valuation perspective, be it EV/EBITDA, Price/Earnings, Price/NCF, EV/FCF or indeed a combination of these factors, are so for a reason. Perhaps analysts or management teams share negative forward outlooks, or the company faces intense competition from competitors.
At first glance, it may appear that Value investing is inherently riskier, and hence should yield a higher rate of return. Such logic would after all be consistent with Risk-Return theories, which dictate that investors should expect higher returns from taking on higher levels of risk. However, Value strategies as depicted above, have lower standard deviations and betas than the index and Growth strategies.
It is our flawed nature as humans, that in turn renders the market inefficient. We over-exaggerate the negatives for troubled companies and over-estimate the future of top performing ones. The efficient market hypothesis, even in its semi-strong form definition, remains a dream.
Avoid companies with the following attributes
While selecting companies on aggregate using value drivers, it is useful to avoid certain attributes which have led to truly abysmal long-term returns.
• Excessive debt levels relative to cash flow
• Suspect accounting policies
• Very negative price momentum
• Very poor return on assets
• Very poor operating margins / net margins
• Worst Value factors (highest EV/Sales, EV/EBITDA, Price/Earnings, Price/FCF, EV/FCF, Price/Book, and lowest buyback and shareholder yields)
Companies in the highest Debt/CF decile gravely under-performed the All Stocks index and even U.S. T bills at 2.08% per year and Sharpe ratio of (0.11). Companies with the highest coverage ratio only outperformed the market slightly, with the 6th decile performing best. This suggests the market rewarded companies that are aggressive with debt, but severely punishes those that take on excessive amounts that cannot be covered by their cash flows.
Net operating assets are a company’s operating assets minus operating liabilities, derived by separating operating activity from financing activity. It represents the culmination of the discrepancies between cumulative net operating income and cumulative free cash flow. Elevated levels of net operating assets are indication of a lack of sustainability of recent earnings performance. In their paper “Do Investors Overvalue Firms with Bloated Balance Sheets?” David Hirshleifer, Kewei Hou, Siew Hong Teoh and Yinglei Zhang state, “When cumulative net operating income outstrips cumulative free cash flow, subsequent earnings growth is weak. If investors with limited attention focus on accounting profitability, and neglect information about cash profitability, then net operating assets, the culmination of the discrepancies between the two, measures the extent to which reporting outcomes provoke over-optimism. During the 1964-2002 sample period, net operating assets scaled by total assets is a strong negative predictor of long-run stock returns.” Companies in the top decile of NOA change % registered abysmal returns at 3.03% per year and Sharpe ratio of (0.08).
Price momentum, while inferior to Value factors in determining long run performance, is a better indicator than other growth factors such as earnings growth rates. It reflects where the market is actually allocating capital, and high momentum stocks may not be significantly more expensive valuation-wise, than low momentum stocks. Selecting stocks in the highest decile of price momentum returned 14.24% per year and Sharpe ratio of 0.37, beating the All Stocks index. However, it is when we select companies with the worst price momentum that truly yields terrible results. In the study, companies in the worst 6-month price momentum decile registered returns of 3.36% per year and Sharpe ratio of (0.06).
Companies in the highest decile for returns on assets (ROA) do not significantly outperform the market. It is the middle decile (decile 6) that performs the best, registering returns of 12.67% per year vs the highest decile’s 11.81% per year from 1963 to 2009. Overall, returns do not vary significantly in the 2nd to 10th deciles. However, when we consider the worst decile, the returns are horrific at 4.57% per year, with Sharpe ratio of (0.02) and largest draw-down of (91.36%).
Operating and net margins, like return on assets (ROA), has not shown to be a reliable predictor of out-performance. Companies in the top decile for operating margins, have in fact slightly under-performed the All Stocks index. But we should avoid the companies in the lowest decile for operating margins as if it were the plague. This group collectively yielded a poultry 4.58%, with Sharpe ratio of (0.02), but more appallingly had the largest draw-down of any single factor strategy at (93.34%). Companies with the lowest net margins tell a similar story, yielding returns of 5.02% per year, with Sharpe ratio of (0.00) and the second largest draw-down of any single factor strategy at (93.05%).
Companies with the worst value factors (e.g. highest EV/Sales, EV/EBITDA, Price/Earnings, Price/FCF, EV/FCF, Price/Book, and lowest buyback and shareholder yields), severely under-performed the All Stocks index with returns ranging from 4.73% to 8.13% per year.
Value can be improved by incorporating growth elements
James O’Shaughnessy tested a hypothetical combination of value factors and a single growth factor (25 and 50 stocks with the best 6-month price appreciation). The resulting “Trending Value Portfolio” crushed any single factor value portfolio on an absolute and risk-adjusted basis. It achieved a Sharpe ratio of 0.93 and had never had a five-year period in which it lost money. The 25-stock version beat the All Stocks index in 100% of all rolling five and ten-year periods and in 99% of all rolling three-year periods. The 50-stock version was not far behind.
For the combined group of value factors, James assigned a percentile ranking (from 1 to 100) for each stock in the All Stocks universe. If the stock has an EV/EBITDA ratio that is in the lowest 1% of the universe, it receives a rank of 100; if a stock has an EV/EBITDA ratio in the highest 1% for the universe, it receives the rank of one. And follows a similar convention for the other factors, except for shareholder yield, where the inverse is true. If a value is missing, it is assigned a neutral rank of 50. The ranks are added up and the stocks are assigned to deciles. Those with the highest scores are assigned to decile 1, while those with the lowest scores are assigned to decile 10.
The Value factors include:
• Shareholder yield
The Trending Value Portfolio takes the top value composite decile ranking of all stocks and selects the 25 to 50 stocks with the best 6-month price appreciation.
Table 6: Top 6 strategies for monthly data series (including Trending Value Portfolio), Aug 31, 1965 to Dec 31, 2009. Strategies sorted by compound average annual return. Excludes other composite strategies
Table 7: Top 6 strategies (including Trending Value Portfolio) Percentage of wins against the All Stocks benchmark for all rolling one-, three-, five-, seven-, and ten-year periods, Dec 31, 1967 - Dec 31, 2009
Keep it simple
For the retail investor, without paid software and financial database access, one can still take advantage of James’s research by including the above-mentioned metrics in your stock picking criteria, as well as sector-specific best strategies shown in table 5. The Value and Growth indicators can often be found on free sources such as Yahoo Finance, Finviz, or brokerage platforms and paid for services such as TD Ameritrade and Capital Cube. Let the statistical odds work in your favor, not against you.
To leverage statistical advantages, it is imperative we build sufficient number of occurrences. Investing in only a handful of stocks using these strategies could end badly if one or two companies suffers a catastrophic decline. The larger our number of occurrences (within our strategies’ criteria), the better our chances to achieve out-performance relative to the benchmark.
Further, we should maintain diversification across asset classes and sectors. Due to non-perfect correlations between sectors and risk assets, and negative correlations between risk and safe-haven assets, we can achieve superior risk-adjusted returns by holding well diversified portfolios. For retail investors with limited funds, using ETFs or futures to gain exposure to government bonds, corporate bonds, commodities and overseas equities may be a promising idea.
Think long term
We are our worst enemies in investing. During periods of under-performance, we will doubt the validity of our strategies, and worst still, change it or revert to what everyone else is doing. We will convince ourselves that our original strategy is no longer valid. Markets have changed, the political landscape is shifting, quantitative funds are creating more efficient markets or artificial intelligence is taking over finance. That may all be true, but one thing is for certain: Times will change but we will not. So long as markets are determined by human participants influenced by their natural biases, value-oriented strategies will have their relevance. Thus, we should select a suitable strategy and stick with it. Consistency and risk management are critical to our long-term investing success.