Stock prices are 14 times as volatile as businesses’ profitability and financial health. Value investors try to profit from this, buying the stocks of out-of-favor companies with excessively cheap valuations. When they find an undervalued stock, they hold it for years while waiting for the market’s unwarranted pessimism to subside and for the valuation to revert to a more reasonable level. Most of their investments are long-term, so value investors prefer stocks with stable dividends. This lets them earn income while waiting for valuations to revert.
Columbia Business School professors Benjamin Graham and David Dodd advocated for value investing in their 1934 book Security Analysis. This book is likely the most influential investing book ever written, still celebrated by money managers today. In it, Graham and Dodd create a fictional character, “Mr. Market," to represent how most investors behave.
Mr. Market struggles to control his emotions and frequently misprices stocks. His excessive euphoria and fear can cause valuations to be absurdly expensive or cheap. Graham recommends having a portfolio of at least 10, but less than 30, drastically undervalued stocks and checking up on them annually. If any of them rise above their intrinsic value or stop delivering consistent earnings, investors should sell them and replace them with other undervalued securities if available. This contrarian strategy requires that investors bet against the crowd, buying into pessimism and selling into optimism. This opposes human nature and requires emotional discipline.
Graham felt that most investors probably won’t earn higher raw returns than the overall stock market, but that they can reduce their risk of loss by buying underappreciated stocks. He looked for them by evaluating each stock’s margin of safety, the amount that its market valuation falls below its intrinsic value. He approximated a stock’s margin of safety using valuation ratios. These compare a company’s share price to one of its fundamental metrics, such as earnings, sales, cash flow, annual dividends, or book value, the value of a company’s physical assets minus its liabilities. Graham focused on the price-to-earnings (PE) ratio and the price-to-book (PB) ratio. He searched for statistical bargains, seeking companies with historically low PE and PB ratios to help ensure he bought stocks with a large margin of safety.
To calculate a company’s PE, you can divide its market valuation by its average annual earnings, or you can divide its share price by its average annual earnings per share. Both methods give you the same value; the second method is just a per-share version of the former and is faster to calculate. If a company’s PE is 11, the company’s market valuation is 11 times the size of its annual earnings, and its share price is 11 times the size of its average annual earnings per share. This means the company will take 11 years to generate a sum of profits equal to its valuation, assuming its average profitability continues.
For value investors, investing in stocks with exciting growth stories is a fool’s errand. Graham stressed that investors should avoid glamor stocks, companies with high valuations due to unreasonably optimistic earnings expectations. Too often, investors overpay for promises of future growth. Analysts’ growth forecasts have little predictability, if any. Graham also felt that the principles of successful long-term investment will remain constant; investors should be skeptical when commentators argue we are in a new era where the classic principles no longer apply.
Value investors often fall into two camps. One focuses on relative value, favoring stocks with low valuations compared to the rest of the market even if those companies’ valuations are not historically low. The other looks for absolute value, buying a stock if its valuation is low compared to its historical range. As a statistical bargain hunter, Graham falls in the second group. But the first camp can do extremely well too. From 1980-2005, a strategy that annually buys the 20% of global stocks with the lowest Schiller PEs and holds them for a year outperforms the global market by about 10% annually on average. A similar strategy that also shorts the 20% of the global market with the highest Schiller PEs outperforms by 25% on average.
Strategy Summary
Graham’s original recommendations depend on investors’ risk tolerance. He suggested conservative investors buy stocks with:
● A relatively large underlying business, with annual sales > $340 million for industrial companies or > $170 million for public utilities (1934 numbers shown in 2019 dollars)
● PE ratio below 15, based on the average earnings over the past three years
● PB ratio below 1.5, based on the last reported book value
○ Stocks with a PE under 15 can have a PB above 1.5 as long as PE x PB < 22.5
● Reasonable debt dependence; current assets should be at least two times current liabilities, and long-term debt should be less than the company’s net current assets
● Positive earnings for each of the last five years; the longer the better.
● At least twenty years of uninterrupted dividend payments
● Earnings per share growing at least 33% in the past 10 years
Graham relaxed these criteria for aggressive investors, suggesting stocks with:
● Reasonable debt dependence; current assets should at least 1.25 times current liabilities, and long-term debt should be less than 1.1 times the company’s net current assets.
● Some history of dividend payments
● PB ratio below 1.2
Graham allows aggressive investors to look at companies of all sizes and allow a stock to be lacking in some criteria if it exceeds other criteria.
He recommends that investors calculate the PE ratio by comparing the company’s share price to its long-term average annual earnings per share, preferably the average over the past 7-10 years. Companies’ success is highly seasonal, producing large earnings fluctuations that can mislead investors. The long-term average is less volatile and more indicative of a firm’s earnings power. Research shows that the 10-year average works the best, and profitability significantly decreases the more that investors decrease that time horizon. When calculating that average, investors should adjust the earnings for inflation. This produces the cyclically-adjusted price-earnings (CAPE) ratio. Researchers John Campbell and Robert Schiller released an influential study demonstrating the benefits of using the 10-year CAPE, so many investors refer to the ratio as the Schiller PE.
Though companies can use a variety of accounting tricks to exaggerate their earnings, book values are less easily manipulated and provide a more transparent view of financial health. However, Graham’s strict PB ratio criteria makes investors focus on companies that depend heavily on physical assets. This can make investors miss out on innovative companies with powerful intangibles like patents and software, which are important assets in the post-1934 world, because their small book values make their PB ratios appear too large.
Supporting Evidence
Does value investing work? Most evidence tests the relative value approach and demonstrates that it significantly outperforms the general market. Famed investor James Montier offers extensive proof in his book Behavioural Investing: A Practitioner's Guide to Applying Behavioural Finance. His team tested this strategy globally using the stocks tracked by the MSCI World Index. They found that from 1975 to 2004, a strategy that annually buys the 20% of global stocks with the lowest Schiller PEs and holds them for a year provides an average annual return of around 23%, annually outperforming the global index by 9.7%. The strategy was equal-weighted, making an equal initial investment in each stock. If instead the strategy weighted each stock position by the stock’s earnings growth, it would earn an average annual return of around 20%. His team also found that the majority of these global value stocks outperformed. Clearly, value investing works on a global scale.
In the U.S., the earnings growth-weighted value investing strategy outperforms the U.S. market by a more modest 2% annually. Why? Montier’s team found that the U.S. is plagued with value traps; the majority of U.S. value stocks (57%) underperform the U.S. market. This is why they recommend only buying U.S. value stocks with a high Piotroski score, a metric that measures firms' financial health. But despite being more modest, this 2% annual outperformance compounds, making a big difference over longer time horizons.
More importantly, value stocks in the U.S. have drastically superior risk-adjusted returns than the market. You can measure risk-adjusted performance by calculating the ratio between annual returns and volatility as measured by the standard deviation of returns. Wharton professor Dr. Jeremy Siegel provides this data in his book Stocks for the Long Run. He analyzes the stocks in the S&P 500 index from 1957 to 2012. For each year, he organized those stocks into five quintiles of cheapness based on their one-year trailing PE and noted each quintile’s future 1-year return and standard deviation.
On average, the cheapest 20% of the U.S. market had a risk-adjusted performance that’s 31.8% larger than the overall market's. That number drops slightly to 28% for the next-cheapest quintile, and 11% for the next. Had Siegel organized these stocks based on their Schiller PEs instead of their one-year trailing PEs, it’s likely that the raw returns and/or risk-adjusted returns could be even better than he reported.
Siegel also organized stocks into cheapness quintiles based on their dividend yields and saw similar but less impressive results. The two cheapest quintiles showed risk-adjusted outperformance compared to the general market, but each dividend yield quintile underperformed its corresponding PE quintile. The second-cheapest dividend yield quintile had the highest risk-adjusted performance by far. Overall, these results suggest you’re probably better off using PE ratios to implement these types of formulaic, relative value investing strategies.
Does value work at the sector level? Yes, but less than half as well. Montier and his team analyzed sector-level PB ratios and future returns in the U.S. since 1926 and found that the cheapest sector earned an average 4.4% annually compared to 10% for the cheapest quintile of stocks regardless of sector. Sectors almost always have both cheap and expensive stocks at any given time. If you avoid a supposedly overvalued sector you might miss some exceptional bargain stocks hiding inside. Buying a supposedly cheap sector saves time, but the sector's overvalued stocks water down your performance.
Overall, the evidence confirms Benjamin Graham’s suspicion; value investing has significantly higher risk-adjusted performance than the general market mainly because it reduces risk. This is consistent with another one of Siegel’s findings: value stocks usually fall much less than the general market during bear markets. So Graham’s predictions panned out, even more than 80 years after he shared them. It seems he was right to believe that the principles of sound investing will never change. Even the most modern trading tools require people to decide when to use them, and those people will continue to succumb to behavioral biases and follow the crowds. Stock prices will probably continue to deviate drastically from businesses’ values but eventually revert to the mean.
Modern Modifications
Most value investors have updated Graham’s original formula. Graham’s book came out in an era of unusually low stock prices, soon after the Great Depression and the 1929 Stock Crash. Many investors feel his thresholds for cheapness are too restrictive for modern times. Graham himself relaxed his PE criteria when he published the second edition of his book, raising the upper limit from 15 to 20.
Warren Buffett, Graham’s most famous pupil, added criteria to require companies have a certain level of quality. Buffett sought companies with skilled management and a strong brand, franchise, or market niche. Most of his investments have had simple business models and generated substantial cash flow. He also considers macroeconomic factors to see if any long-term trends support a company’s success. This diverges from Graham, who advised against evaluating macroeconomic factors or a company’s management quality. Graham felt a company’s earnings record already conveyed that information.
Joel Greenblatt, a well-known finance professor and hedge fund manager, also expanded Graham’s criteria to focus on high-quality companies. He considers companies’ return on capital (ROC), which measures how efficiently companies use their physical assets to generate earnings. Greenblatt feels the most attractive stocks have a low valuation and a high ROC. When valuing stocks, he looks at a ratio comparing a company’s earnings before interest and tax (EBIT) to its market valuation, also known as enterprise value (EV). Low valuation stocks have a high EBIT/EV. Research shows that from 1993-2005, using the EBIT/EV worked significantly better than PE for finding undervalued stocks while adding minimal additional risk. Introducing his ROC requirement lowered profitability but also lowered downside risk.
Many value investors now consider another valuation ratio called the price-to-earnings growth (PEG) ratio, which is calculated by dividing a company’s current PE ratio by its expected or past earnings growth rate. When managing the Magellan Fund, Peter Lynch favored companies with low PEGs because their stock prices didn’t yet reflect the company’s growth prospects. Unfortunately, the expected growth rate comes from analysts who have dismal records of anticipating the future. If you use the past earnings growth rate, you risk the chance that the company has already seen its best days and will have disappointing future earnings growth. Supposedly, stocks with a PEG ratio under 1 are undervalued. Take this metric with a grain of salt.
Selecting Valuation Ratios
So the value premium is real. But which valuation metric should you use to profit from it? Kwag and Lee (2006) compared the risk-adjusted performance of several valuation ratios in the U.S. from 1954 to 2002, including PE, PB, dividend yield, and the price-to-cash flow (PC) ratio. Overall, stocks in the cheapest quintiles based on PE and PC ratios showed the best risk-adjusted performance according to all three risk-adjusted metrics they looked at: Sharpe, Treynor, and information ratios.
They also compared the valuation ratios’ performance in economic expansions and contractions as defined by the National Bureau of Economic Research (NBER). During contractions, the dividend yield performed the best, having a higher Sharpe and Treynor ratio than all other valuation metrics. During expansions, the PC ratio performed the best for all three performance metrics, followed by the PE. Sadly there’s no easy way to identify whether the economy is in a contraction or an expansion; the NBER defines those periods considerably after the fact. But you can use the Leading Economic Index (LEI), which tracks a robust group of statistics that have the best records of turning before the economy does. Every month, the Conference Board releases the LEI’s value. During economic expansions, the LEI is usually above 1 and is uptrending.
Don’t forget about the PB ratio just yet. Nissim (2013) suggested and confirmed that the PB ratio outperforms other valuation ratios when valuing insurance companies. This is likely because they don’t depend heavily on intangible assets, so their book values are a fairly accurate representation of those companies’ assets and overall intrinsic values. Additionally, Sehgal and Pandey (2010) found that the PB ratio outperforms other valuation ratios when valuing companies in several Asian countries including India, China, and South Korea, while the PE ratio works best in Brazil and South Africa. In the U.S., Siegel’s research found that the PB ratio generally underperformed the dividend yield, PC, and PE ratios from 1987 to 2012.
Disadvantages and Challenges
Sometimes, stocks are cheap for a reason; the market correctly predicts that these companies will soon be unprofitable and file for bankruptcy. These “value traps” look cheap but just keep getting cheaper. Investors can avoid most value traps by requiring stocks have a high Piotroski Score, a metric that measures a firm’s financial health. The Piotroski Score considers a firm’s profitability, efficiency, and debt use to produce a score from 0-9; the higher the better. The score is based on the following nine factors, each of which earn a score of either one or zero:
● Positive net income
● Positive return on assets in the current year
● Positive operating cash flow in the current year
● Cash flow from operations exceeds net income
● Lower ratio of long term debt in the current year compared to last year’s
● Higher current ratio this year compared to last year’s
● No new shares issued in the last year
● Gross margin in the current year exceeds last year’s
● Asset turnover ratio higher than last year’s
The evidence that this score works is compelling. Piotroski’s research showed that from 1976 to 1996, investors would’ve earned an average 23% annual return by buying stocks with high scores and shorting those with low scores. Their portfolio would experience only two down years.
Though value portfolios are less volatile, even the most successful value investors’ raw returns underperformed the market 30-40% of the time on average, according to research from Tweedy Browne (1998). Few investors are patient enough to wait out those temporary dry spells. Plus, bargain stocks often have the worst narratives and growth forecasts, so buying them defies the consensus and triggers social pain.
This strategy requires ample patience, performing drastically better the longer you hold onto value stocks. Montier and his team found that most successful value funds hold stocks for an average of 5 years, a minimum of 3, and a maximum of 17. But short-term market volatility hurts, often deterring investors from following their strategy even if they know it has a statistical edge in the long run.