October 17, 2005
Return on Equity
One of the important facts of investing is that year-to-year stock market movements have very, very little to do with the year-to-year growth of corporate earnings. In fact, the correlation is just about zero. Stocks aren't thermometers. While it's true that individual stocks can often be affected substantially by earnings surprises that “reset” the expectations of investors in an abrupt way, it's equally true that over the long-term, fundamentals like earnings, dividends, book values, and revenues for the market as a whole are closely tied to long-term GDP growth, with growth rates that are surprisingly stable when measured from peak-to-peak across economic cycles.
But if short-term earnings results can affect short-term returns for individual stocks, and individual stocks make up the market, doesn't it follow that short-term earnings prospects for the S&P 500 are crucial? The answer is, not really. See, investors are generally good at distinguishing earnings developments that are isolated to a given company and have some risk of permanence, versus earnings developments that are tied to general economic fluctuations. That's why, for example, blowout earnings in an economic expansion don't get reflected in the prices of auto makers. Instead, their P/E ratios drop to the single digits. It's also why depressed earnings for the market as a whole are generally accompanied by very high market P/E ratios during a recession. Investors generally realize that economic fluctuations cause transitory earnings fluctuations, and they price stocks accordingly.
In short, stocks are a claim on a very, very long-term stream of cash flows. Short-term fluctuations in earnings don't have nearly as much impact as TV would have you believe. In fact, if you do the present value calculations, wiping out 20% of the S&P 500's earnings, dividends and cash flows for 5 consecutive years would only be worth a 3% decline in the index.
Short-term prospects for S&P 500 earnings are not very useful for making investment decisions (and year-ahead analyst estimates are worse than a Ouija board), but we can make good use of earnings by understanding how they relate to long-term returns. First, it's important to filter out short-term fluctuations that wreak havoc on year-to-year growth rates, which is why I measure earnings growth from peak-to-peak across economic cycles and vastly prefer using price/peak earnings measures to raw P/E calculations. Second, it's important to recognize factors that might make those peak earnings more or less reliable than other fundamentals like revenues, book values, or dividends. Everything is part of a big picture.
Some of my views on these fundamentals are contained in recent weekly market comments:
September 12, 2005 - The S&P 500 as a Stream of Payments ,
This week, let's take a short look at the relationship between earnings and book values – something known as “return on equity.”
Return on equity is defined as the ratio of earnings to book value. For the Dow Industrials, for example, the book value is 3359.70, earnings are 585.52, and the index is 10,287.54. Those put the price/book ratio of the Dow at 3.06, and ROE at 17.4%.
To most investors, those figures probably mean very little, and there's nothing surprising about that. Book values and ROE don't get much attention, particularly because the late 90's tech bubble encouraged investors to imagine boundless earnings power based on intangible ideas, information, and a “dot com” at the end of the company name. Book value is even further out of favor than dividends. But the fact remains that for the vast majority of U.S. companies, it takes money to make money, and book value is the value of identified tangible and intangible assets belonging to shareholders after all debt is subtracted. In a competitive economy, it's difficult to maintain the returns on invested equity capital at permanently high levels. Like profit margins, return on equity has historically been a strongly mean-reverting proposition.
In contrast, P/E ratios are carefully followed. The P/E on the Dow Industrials is currently 17.57 on the basis of 12-month trailing net earnings (you'll see lots of other figures depending on whether earnings are “forward” earnings for 2006 or 2007, or “operating” or as Google just decided to use, “pro-forma.” Hmmm).
A 17.57 P/E doesn't seem so bad compared with a historical average of about 15, except that you have to recognize that the historical average is calculated including periods of extremely high P/E ratios that occurred during recessions, when earnings were depressed. When earnings have been at a record, as they are now, the historical average P/E works out closer to 12. Moreover, as noted below, earnings are currently unusually elevated compared to revenues and book values. Elevated earnings relative to sales means that profit margins are currently unusually high. Elevated earnings relative to book values means that ROE is also unusually high.
As Adam Barth noted last July in Barron's magazine, over 20 year periods, the overall relationship between earnings and book value "barley varies in the slightest, and has remained basically immune to inflation, wars, massive changes in the tax code or any other external factor.” The historical average ROE for the Dow Industrials? 11%.
Now, it's an identity that Price = ROE x P/E x Book Value (prove this to yourself). So if we took that 11% as law, and took the historical average P/E of 15 as justified as well, the resulting fair value for the Dow would be .11 x 15 x 3359.70 = 5544. But that's just plain depressing.
To some extent, it's proper to criticize that simple an analysis. First, U.S. corporations are a whole lot more leveraged now than they were in the past. Greater debt means that a given amount of book value can control a larger amount of total assets, and it's really total assets that companies use to generate revenue and earnings. In fact, we can break down ROE itself and write it this way:
Earnings/Book = Earnings/Revenues x Revenues/Assets x Assets/Book
So ROE is effectively driven by profit margins, “asset turnover” or the amount of revenues that can be generated by a given level of assets, and leverage – the number of dollars in assets a company can control with a dollar of shareholder equity.
Today's high ROE essentially reflects 1) unusually high (and mean reverting) profit margins, 2) various efficiencies that have allowed companies to generate more revenues for a dollar of assets, and finally, 3) unusually high corporate leverage.
Even though there have been a lot of competitive pressures from China, India and elsewhere, U.S. companies have been able to achieve very high profit margins because of low inflation (so they have not been, until recently, forced to pass along higher prices to consumers), and because the share of corporate revenues going to labor compensation is at a historic low. Those aren't things that investors should count on indefinitely. Moreover, to the extent that China revalues the yuan in the coming years, energy demand from China may put further upward pressure on oil prices. A stronger yuan means that a yuan buys more dollars (and more oil as well, as oil is internationally priced in dollars). Meanwhile, I've long argued that the U.S. is likely to experience a “deleveraging cycle” in the coming years, which will suppress another driver of ROE.
In any event, one needn't assume that ROE will revert to its historical norm of 11-12% to understand that current ROE levels are unlikely to be sustained indefinitely. That's another way of saying that P/E ratios, currently benefiting from high profit margins and other factors, may start looking more consistent with other fundamentals like price/book, price/revenue, and price/dividend ratios: really, really high.
As of last week, the Market Climate for stocks continued to be characterized by unusually unfavorable valuations and unfavorable market action. Note that this is a discrete change from recent weeks. Until and unless market action improves enough, we have to infer that investors have moved to a more skittish view of risk. At present, we have very low risk premiums, increasing risk aversion, and rising interest rates. This is a really bad combination. That's not to forecast a crash (though current conditions match only the 4% of market history that contains most significant crashes on record) or even the necessity of a market decline here. But very emphatically, this is not an environment in which we are willing to take a significant exposure to market risk.
This is a very flexible view, and will change quickly if the quality of market action improves enough to infer that investors are still willing to extend their risk taking. But here and now, it is much more preferable (especially from the standpoint of historical experience) to be patient for that evidence to arrive than to assume that it necessarily will. Importantly, we'll miss out on some amount of market gains if current conditions turn out to be, for example, an intermediate low in this specific instance. My hope is that shareholders are familiar enough with the full-cycle, risk-managed investment approach of the Funds to recognize that I don't attempt to closely track market fluctuations in environments that have historically delivered a poor return/risk tradeoff on average.
At present, so many internals are poor and the market is so oversold that we have to allow for a typical “fast, furious and prone to failure” rally to clear that oversold condition. A dropoff of trading volume on such a rally would be a particularly unfavorable development, as would a substantial widening of credit spreads. For now, we're well hedged, but open to a more constructive investment stance if the evidence allows.
In bonds, the Market Climate remains characterized by modestly unfavorable valuations and unfavorable market action. Yields are becoming somewhat more attractive, but not to a sufficient extent to take on substantial duration risk. A widening of credit spreads would improve the case for taking on at least some amount of interest rate exposure in Treasuries. Though I don't trade on opinions about specific future market direction, my impression is that a reasonable yield curve expectation, maybe a year from now, would be a flat curve slightly above 5% at most maturities. In any event, here and now the evidence continues to suggest a muted exposure to interest rate fluctuations, so the duration of the Strategic Total Return Fund continues to be about 2 years, with slightly less than 20% of assets in precious metals shares, and about 5% in foreign currency denominated notes (primarily the Japanese yen).
|Home | The Funds | Open an Account | Account Access | Research & Insight | Site Map|
For more information about investing in the Hussman Funds, please call us at
513-587-3440 outside the United States
Site and site contents © copyright 2003 Hussman Funds. Brief quotations including attribution and a direct link to this site (www.hussmanfunds.com) are authorized. All other rights reserved and actively enforced. Extensive or unattributed reproduction of text or research findings are violations of copyright law.
Site design by 1WebsiteDesigners.