It’s a dreary time to be an investor. Stock market returns are well below the range investors were seeing a decade ago. Barely a third of professional portfolio managers beat the stock market in a given year, and those who outperform the market over time are even fewer. Either way, they charge fees. And there’s nowhere to go because bond and money market returns are even lower. Faced with these challenges, what can investors do?
“In a tough environment, one of the things you may have to adjust is your expectations,” said Bruce Greenwald, the Robert Heilbrunn Professor of Finance and Asset Management and director of the Heilbrunn Center for Graham and Dodd Investing, at the 2005 Alumni Reunion in April. “There are stocks out there that won’t return the 10 to 11 percent you’re used to, but they will do better than the 6 to 7 percent you’re looking at.”
Finding those stocks requires a good valuation process, Greenwald said. Value investing, developed by Columbia Business School Professor Benjamin Graham — and exploited most famously by Warren Buffett, MS ’51 — is that process. Value investors compare a company’s fundamental value to its stock price and buy only stocks that are undervalued by the market. But while this approach seems so intuitive as to be obvious, it is by no means the only way to invest.
Value investing is based on an assumption of market efficiency, an understanding of human behavior and a reliance on statistical evidence. Value investors calculate a company’s asset value and earnings-power value, assess the company’s management and competitive position in its industry and then factor in growth. They do not rely on the discounted cash flow (DCF)/capital asset pricing model (CAPM) approach that business schools teach in introductory corporate finance courses.
“There is an absolutely fundamental sense in which markets are efficient,” Greenwald said. “It is this: When someone buys a stock, another active investor is selling the stock… The inevitable result is that one of them is wrong.”
Value investors concentrate their search for investments so as to improve their chances of being on the right side of that trade. They look for obscure stocks — small caps, spin-offs and boring stocks with low analyst coverage; those that are distressed, bankrupt or otherwise disappointing; and those that have low growth, low price-to-earnings and low market-to-book ratios because of company or industry problems. They understand that most investors — corporate and individual — avoid these stocks, and that this produces bargains.
“You want to be oriented toward races where you have a chance to be the fastest runner,” Greenwald said. “Even if you’re the fastest guy in the world, you raise your probability of winning races by competing against slow runners, not other fast runners.”
Statistics also help. About 30 studies done in Europe and the United States from the 1930s to the present show that portfolios with low market-to-book and low price-to-earnings stocks outperform the market by 3 to 5 percent, while portfolios with low-growth stocks outperform it by 1 percent and those with small-cap stocks outperform it by 2 to 3 percent.
“It’s very hard to do better than these simple statistical approaches,” Greenwald said. “But if you’re going to do better, you better have a really good technology for valuing stocks.”
The DCF/CAPM methodology that business schools teach “is a theoretically elegant formulation,” Greenwald said. But in practice, the margin of error makes it worthless for investing. These models depend not only on near-term cash flows, which can be projected reliably, but also on long-term cash flows and terminal values, which cannot. Terminal values rely on highly subjective assumptions of costs of capital and growth rates. Any error, however slight, in these variables can dramatically throw off valuations.
“A valuation approach is like a machine,” Greenwald said. “It’s like a meat grinder. You put in your assumptions, and it generates a valuation or a range of valuations. What you would like to put into that meat grinder are assumptions you can actually make.”
Furthermore, DCF models ignore balance sheets, throwing away some of the most tangible, reliable and therefore valuable information available. In contrast, the value investing approach organizes information by the reliability of assumptions, starting with the balance sheet. Within the balance sheet, reliability decreases from cash to accounts receivable, inventory, property, plant and equipment to intangibles like customer acquisition, branding and advertising costs.
Value investors use estimates of sustainable, steady-state earnings levels, based on current earnings and cash flows, to value a company’s earnings power. This analysis yields three possible scenarios: a company’s earnings-power value is lower than its asset value, its earnings-power and asset values are about equal or its earnings-power value is greater than its asset value.
In the first case, which usually stems from poor management, the company’s earnings are too low to justify its assets. If the company’s management is energetic about growth but actually incurs losses through mismanagement, investors should stay away. However, if the firm’s stock price is lower than its earnings-power value, the stock could be a bargain.
The second case is the most common and reassuring. Such stocks offer zero net growth, but they could be worth buying if the price is below the calculated value and industry forecasts are positive.
In the third case, investors should buy a stock only if the company has sustainable earnings as well as a long-term strategic and competitive advantage. Coca-Cola — one of Buffett’s most lucrative investments in the late 1980s — is an example of this kind of company, and Greenwald predicted that it would outperform the S&P 500 over the long term. But in general, he said, these kinds of companies, particularly the well-known ones, tend to be overpriced, so he recommends looking for stocks no one else has spotted.
The value investing approach — looking first at asset value, then earnings-power value, then competitive advantage and managerial ability and then growth — is in every way more accurate than the DCF method, and value investors tend to do much better than the market as a whole, Greenwald said. Portfolio managers who add value to the statistical model have industry specialization as well as a good technique.
“My advice to you would be to find people who do statistically cheap stocks and do it with low management fees,” Greenwald said. “If you want to go beyond just the statistical measures, my advice to you is just listen to what they say and see how sensible they are. If they say they’re going to do it all over the world in every industry and they’re going to use the standard technology, save your money.”