Leithner Letter No. 56
26 August, 2004

Prior to the past several decades, earnings forecasts were not nearly so important a factor in assessing the value of corporations. In fact, I do not recall price-to-earnings ratios as a prominent statistic in the 1950s. Instead, investors tended to value stocks on the basis of their dividend yields. ... [Subsequently], the sharp fall in dividend payout ratios and yields has dramatically shifted the focus of stock price evaluation toward earnings.

Alan Greenspan
Speech at New York University
(26 March 2002)

If ‘using trailing earnings is like driving a car looking out the rear window’ [as stated in the 5 August issue], then using projected earnings is like riding in a car with blacked-out windows and steering on the basis of what someone imagines is up ahead. Either way, you’re likely to crash. Maybe people who think investing is analogous to driving should be taking the bus.

David G. Danielson
Barron’s (12 August 2002)

People get fooled by Fed-Model thinking into paying high price-earnings ratios for stocks. ... A major underpinning of bullish pundits’ defence of this high valuation is the Fed Model I discredit. Sadly, the Fed Model perhaps offers a contemporaneous explanation of why P/Es are high, but no true solace for long-term investors.

Cliff Asness
“Fight the Fed Model” (2002)

Of Myths, Models and Intelligent Investing

“The Australian stock market has posted its best financial-year performance in seven years.” With these celebratory words The Australian Financial Review commenced its Market Wrap for the 2003-2004 financial year (1 July 2004). But it hastened to add “leading funds managers caution that the best gains have now been made and investors will require sharp stock-picking skills over the next twelve months.” Many of the people whom it quoted agreed that good results could and likely would be achieved. But they also believed that the upbeat forecasts of corporate profits, acceleration of the pace of economic activity in the U.S., booming Chinese demand for Australian commodities and historically low rates of interest that helped to generate such good results in 2003-2004 would not exert themselves so prominently in 2004-2005. Further, particular risks (i.e., higher interest rates and oil prices, a slowdown of residential and other construction, geopolitical tensions in Iraq and elsewhere) might become more prominent. Consequently, “the main message from market watchers is that there is still money to be made in 2004-2005, but it will be harder to find and there are more traps.”

Unsurprisingly, given that no one possesses a clear crystal ball, the views of the analysts, brokers, funds managers and journalists cited in the 2003-2004 Market Wrap are quite diverse. One journalist is quite cheerful. “The big picture is bright. The U.S. economy is growing at a four per cent plus clip, the global economy is improving, surging employment growth will underpin consumer spending whilst business investment is picking up, albeit slowly.” A funds manager is more cautiously optimistic. “We are through the sweet spot for equity returns. That’s not to say markets can’t go higher, but with monetary conditions tightening, valuations less attractive and profit growth slowing, investors should prepare themselves for lower returns this year.” But one strategist is unequivocally gloomy. “This is as good as it gets. The reporting season we’re about to start will be the last hurrah. The only issue for next year is how bad it gets. We’re on the cusp of a big turn. The miracle economy is going to become a very different place.”

Yet on the whole and, as usual, optimists are presently much more numerous than pessimists. According to The Weekend AFR (3-4 July), “market commentators remain upbeat on the outlook for Australian equities and reckon the share market’s advance has further to run.” One pundit enthused “the Australian economy has super-strong momentum. Not only is the job market firing on all cylinders and consumers spending freely ... the global economic skies are exceedingly bright. We expect the Aussie share market to scale new heights over coming months, underpinned by solid domestic and global economic growth and favourable valuations.” Another said “the near-term earnings outlook for the equity market remains attractive, supported by strong earnings momentum, attractive yields and undemanding valuations.” One of The AFR’s most senior columnists summarised the situation (5 July). “Wall Street and local investment managers seem convinced that stock markets are fairly priced and that the next six months or more should see further increases.”

Part of the trouble with newspapers (see Letter 55) is that they seldom justify the assertions they print with clear premises, assessments of their strengths and weaknesses and explicit patterns of reasoning. Another part is that they (surely unknowingly) often print cant, rubbish, misrepresentations, myths and outright lies. Given the present profusion of warm words and paucity of cold reasoning in Australia’s financial press, one is tempted to infer that the prices of stocks and bonds rise when economic conditions are perceived to be (or becoming) strong and are expected to remain robust. The quicker the pace of economic growth, in other words, the more favourable is the investment climate and the bigger the gains that investors can expect. To most people this generalisation is obvious – so much so that they regard it as an unwritten law. A constant, frenetic and expensive search for new and ever more arcane insights into economic and financial market conditions – and their course into the “foreseeable future” – encourages them to believe that a good economy means good investing.

A False Basis for Today’s Optimism?

Alas, there are ample grounds to question and perhaps reject this generalisation. Setting aside criticisms about how they conceptualise and measure economic growth and the results of investment operations, and recognising their incessant bias towards the aggregate and against the individual (Letter 53), a growing literature finds little evidence that supports it and much that contradicts it.

Elroy Dimson, Paul Marsh and Mike Staunton, the authors of The Triumph of the Optimists: 101 Years of Global Investment Returns (Princeton University Press, 2002, ISBN: 0691091943), for example, studied 16 countries’ economies and stock markets during the past century. In each country, they found that real stock returns are either unrelated to or inversely related with growth of GDP. Statistically, the correlation is -0.27 for 1900-2000 and -0.03 for 1951-2000. This result is based upon much longer periods than earlier research, but it is not new. Jeremy Siegal (Stocks for the Long Run: The Definitive Guide to Financial Market Returns and Long-Term Investment Strategies, McGraw-Hill Trade, 2003, ISBN: 007137048X) found that between 1970 and 1997 the average correlation between stock returns and GDP growth was -0.32 in 17 developed countries and -0.03 in 18 emerging markets. Jay Ritter (Economic Growth and Equity Returns), using the data assembled by Dimson et al., obtained a correlation of -0.42 for these 16 countries during the twentieth century. The more robust a country’s overall rate of economic growth, in other words, the lower its stock market’s subsequent rate of return; and rates of growth explain between 1% and 18% of the variation in stock returns in these countries during these 100 or so years (see also Alan Wood’s column in The Weekend Australian 17-18 July 2004).

Ritter adds a caveat. In the short run (which seems to be one year or less) unexpected changes in the rate of growth influence the prices of stocks. Prices decline when market participants suddenly anticipate a recession (or discover that a recession has begun) and they increase when investors abruptly sense a recovery (or realise that a recession has ended). Ritter believes that these cyclical or business cycle effects “should rationally have an effect on equity valuations, but the effects should be largely transitory.” The effects associated with recessions “are partly due to higher risk aversion at the bottom of a recession, but also due partly to an irrational overreaction.” In a passage that warms the heart of the value investor, Ritter says this “irrationality” generates short-term volatility “and mean reversion over multi-year horizons.” According to Ritter, then, “whether the Chinese economy grows by 7% per year or by 3% per year for the foreseeable future [there’s that bloody phrase again!] is largely irrelevant for the future returns on Chinese stocks. There is also an asymmetry – if a country has negative growth, this is probably bad for stocks. But for positive rates of long-term growth, whether the growth rate is 3% or 7% shouldn’t matter.”

Why does economic growth (as it is conventionally defined and measured) not seem to benefit investors (considered as an homogenous group)? Indeed, why does robust growth apparently depress their returns? Siegel hypothesises that market participants tend to anticipate higher rates of growth by paying higher prices for each dollar of anticipated income, i.e., by increasing stocks’ P/E and price-to-dividend multiples. To do so is to lower realised returns because more money must be invested today in order to secure a given stream of earnings and dividends tomorrow. Hence “Dow 10,000 Means It’s Time to Prepare for the Hangover” (The Wall Street Journal 23 March 1999) and “Forget the Party Hats: Why Dow 10,000 Should Be No Cause For Celebration” (The Wall Street Journal 10 December 2003).

Ritter provides another clue. In every country certain industries have grown over time and others have declined. In the U.S., for example, the twentieth century’s “growth” industries included motor cars, air transport, pharmaceuticals, computer hardware and software and biotechnology. Industries in decline included railways, steel and tobacco. The airline industry has not made many people rich; nor (during the past 30-40 years) have auto companies; nor have biotech firms. Cigarette companies, on the other hand, have generated significant wealth for their shareholders (subject to the favourable settlement of lawsuits). So too have pharmaceuticals firms. “Growth industries,” in other words, typically contain both successful and unsuccessful businesses; ditto declining industries. According to Ritter (note that this point’s conclusion is reasonable but its premise is not), companies earn profits only if entry into the industry is restricted and consumers and employees (including executives) do not extract rents; and that these conditions are not related to industries’ growth or decline. In summary, and in Ritter’s words, “technological change does not increase profits unless firms have lasting monopolies, a condition that rarely occurs; [and] countries with high growth potential do not offer good equity investment opportunities unless valuations are low.”

If higher economic growth does not better investment returns, how do we make sense of the torrent of generally upbeat words from Australian analysts, brokers, funds managers and journalists? No doubt each analyst, etc. has his own reasons to justify his views; but it is possible that they might abide premises and reasoning along lines such as the following (and as a thought experiment it is interesting to assume that they do and to see where the experiment takes us):

  1. On 1 July 2004, 10-year Commonwealth Bonds yielded approximately 5.9%. On that date, in other words, these bonds sold at a price equivalent to 16.9 times the coupons they are expected to generate during the next year (i.e., 1 ÷ 0.059 = 16.9).

  2. Based upon its projected earnings for 2004-2005, on 1 July the All Ordinaries Index yielded approximately 6.5%. To make sense of this, regard the index as if it were a company (“All Ordinaries Ltd”). Analysts expect, in effect, that during this financial year All Ords Ltd will earn approximately $232 per share. Extending the analogy, on 1 July the price of its shares was roughly $3,550. Accordingly, the earnings yield of this proxy “company” is $232 ÷ $3,550 = 6.5%.

  3. Multiply this “12-month forward earnings estimate” for the All Ords by the “forward coupon multiple” of the 10-year Commonwealth Bond, i.e., 16.9 x 232 = 3,921. This product was the “fair value” of the index on 1 July. Above 3,921 it would be “overvalued” relative to the 10-year bond; and below this level it would be “undervalued” relative to the bond.

  4. Subtract the actual level of the All Ords on 1 July from the “fair value.” We therefore have 3,921 - 3,550 = 371.

  5. Divide that number by the “fair-value” number, i.e., 371 ÷ 3,921 = 9.5%, and conclude that on 1 July the All Ordinaries Index was “undervalued” by 9.5% relative to 10-year Commonwealth bonds.

These premises and reasoning, to those who accept them, subsume the apparently widespread view that during 2004-2005 Australian investors can expect positive (and, compared to years such as 2001 and 2002, reasonably good) results, but that they will not be as good as those achieved in 2003-2004. More generally, one way to evaluate stocks is to compare the yields of stocks and bonds. If the yield of stocks exceeds the yield of 10-year government paper, then stocks are generally cheap relative to these “risk free” bonds (or, equivalently, bonds are dear relative to stocks). Conversely, if stocks’ yield is lower than bonds’, then stocks are generally expensive relative to bonds.

It is interesting to note that Benjamin Graham said roughly similar things (except for the last sentence of point #3) in The Intelligent Investor: A Book of Practical Counsel (1949, rev. ed. 2003, HarperBusiness, ISBN: 060555661). “Our basic recommendation is that the stock portfolio, when acquired, should have an overall earnings/price ratio – the [inverse] of the P/E ratio – at least as high as the current high-grade bond rate. This would mean a P/E ratio of no higher than 13.3 against an AA bond yield of 7.5%.” Graham elaborated this general point with a more stringent rule of thumb. In order to provide a “margin of safety” a stock portfolio’s overall P/E ratio should be approximately 20% below the maximum. Given the yields of quality corporate bonds presently prevailing in Australia (roughly 7.0%), for example, it follows that Grahamite investors should select portfolios selling at no more than 11.4 (i.e., 14.2 · 0.8) times their earnings.

Value Investors, Meet the Fed Model

But why should contemporary mainstream analysts, brokers, funds managers and journalists accept these premises? Whence comes these premises’ pedigree? Arguably, it derives from none other than the U.S. Federal Reserve. On 5 December 1996 its Chairman, Dr Alan Greenspan, famously expressed his concern about “irrational exuberance” in financial markets. It is reasonable to suppose that his disquiet stemmed at least partly from the findings of research conducted by his staff. One critical result, buried within the Fed’s Monetary Policy Report to the Congress, accompanied Dr Greenspan’s Humphrey-Hawkins testimony on 22 July 1997.

The Chairman’s biannual testimony is widely – indeed, obsessively – dissected, analysed and discussed. Conversely, relatively few seem to study or even notice the report that accompanies his testimony. The gist of what would become known as the Fed Model was summarised in one of the 1997 Report’s paragraphs and charts. The chart depicted a strong correlation between the 10-year Treasury bond yield and the S&P 500’s current earnings yield, i.e., the ratio of 12-month “forward consensus expected operating earnings” to the level of the S&P 500. The Fed noted “the ratio of prices in the S&P 500 to consensus estimates of earnings on the coming twelve months has risen further from levels that were already unusually high. Changes in this ratio have often been inversely related to changes in long-run Treasury yields.”

Pondering this passage, Dr Edward Yardeni, a market strategist at Prudential Securities, formalised it into a stock valuation model (see also his series of reports on the subject of Stock Valuation Models). Yardeni dubbed this model the “Fed’s Stock Valuation Model (FSVM)” or Fed Model for short, but was careful to disclaim that no one at the Federal Reserve had publicly endorsed it. To avoid any misunderstanding on this point, in subsequent reports Yardeni renamed it Stock Valuation Model #1 (SVM-1).

Yardeni also intended that this nomenclature suggest that there are plenty of alternative stock valuation models. But the name “Fed Model” has stuck. Whether by that name or SVM-1, it has become a prominent means of assessing the value of markets as a whole; and major publications such as Barron’s have based assessments of overall stock prices upon it. Perhaps most notably, its cover on 24 September 2001 trumpeted that the stock market, which it reckoned was undervalued by 17%, was “the biggest bargain in years.” The bullish article that accompanied the cover, entitled “Buyers’ Market,” was premised largely and explicitly upon SVM-1.

That cover story and article seemed to prompt major investment institutions to launch a three-month buying spree. The so-called Fed Model reading (which quantifies in percentage terms the extent to which stocks are “undervalued” or “overvalued” relative to government bonds), together with Barron’s general bullishness, apparently converted many of the gloomy investors who had jettisoned stocks immediately after the attacks on 11 September 2001. Shortly thereafter the S&P 500 and other major indices rose strongly – before faltering and then falling to new lows in 2002 and 2003. Other Wall Street analysts, such as Ed Hyman at ISI, Doug Cliggot at JP Morgan and Byron Wien at Morgan Stanley, have devised variations of the Fed Model and used them to justify their bullishness during some or most of these lean years.

Since 2002 relatively little analysis and commentary about the Fed Model has appeared in the American financial press. Further, it is only infrequently – or, at any rate, inexplicitly – applied to Australian conditions (it has also been exported to Britain, Canada, France, Germany and Japan). Yet it is presently emitting bullish signals in both countries. Accordingly, now is perhaps a good time to evaluate the Fed Model’s relevance to and suitability for value investors.

FSVM’s American version formalises the idea that the expected returns (defined implicitly to include unrealised capital gains) of the S&P 500 and a 10-year U.S. Treasury note should not just gravitate towards one another – they should be equal. This is because stocks and bonds are competing assets: if one is dear relative to the other, then owners of the relatively expensive class of asset will sell some of their holdings (placing downward pressure on its price) and buy some of the cheaper class (placing upward pressure upon its price). The greater the overvaluation of one relative to the other (i.e., the greater the disparity of their yields), the greater the subsequent shift from lower yields to higher yields; and this arbitrage continues until the yields of the two classes of asset correspond (i.e., an equilibrium is obtained). In these respects FSVM fits comfortably within mainstream neoclassical economic theory and Modern Portfolio Theory. For that reason alone, value investors should handle it with care and regard it sceptically (see also Letter 53).

Does the Fed Model Work?

According to Ned Davis Research, which has applied it to a series of American data that begins in 1980, the Fed Model is a reasonably accurate predictor of the S&P 500. Specifically, when the model shows that stocks are more than 5% undervalued the index increases by an average of 31.7% in the following year; and when it indicates that stocks are more than 15% overvalued the market falls by an average of 8.7% during the next twelve months. Stuart Chaussee (Revisiting the Fed Model) examined weekly data beginning in 1979. He sought to ascertain the Fed Model’s ability to predict the general level of stock prices in subsequent 13-, 26- and 52-week periods. He found that when stocks were undervalued by 15% or more the subsequent 52-week return of the S&P 500 was about 16%. When stocks were within a 5% range on either side of undervaluation and overvaluation, the subsequent 52-week return of the S&P 500 was about 12%. When stocks were overvalued by 25% or more the subsequent 52-week return was -6%.

Perhaps unsurprisingly, its author, Ed Yardeni, agrees that the Fed Model “retrodicts” the past reasonably accurately. Had it existed between 1979 and 1982, for example, it would have correctly noted that stocks were extremely cheap relative to bonds. “Back testing” the model would also have indicated – correctly – that before the Crash of October 1987 stocks were very dear relative to bonds. After the crash, when the model suggested that stocks were underpriced relative to bonds, the prices of stocks obligingly rose; and in the early 1990s, when the model showed that stocks were modestly overvalued, prices increased at a lacklustre pace.

Yardeni also notes that during the mid-1990s, when the model would have indicated that stocks were moderately undervalued, a great bull market gathered pace on stock markets. Ironically, late in 1996 (when Dr Greenspan warned of “irrational exuberance”) the model signalled that the S&P 500 was fairly valued. During the summers of 1997 and 1998, a sharp drop of stock prices corrected the overvaluation of stocks relative to bonds. A period of undervaluation beginning in September 1998 was quickly and dramatically reversed. By the beginning of 2000 equities had soared to a record (70% overvalued) reading. The crash of technology stocks between 2000 and 2002 brought the market closer to fair value. Concerns about the quality of corporate earnings and the economic outlook subsequently depressed stock prices. By late 2002, SVM-1 was undervalued by a record 49%. Since then the market has rallied – but not, according to the model, to the point of overvaluation.

According to its critics, however, the Fed Model’s ability to predict is far less impressive than it appears at first glance. Perhaps most notably, Clifford Asness (Fight the Fed Model: The Relationship Between Stock Market Yields, Bond Market Yields, and Future Returns) concluded that the model has little or no ability to predict stock returns. The unadorned price-earnings ratio is a far better predictor of “performance” – the higher (lower) the ratio, the lower (higher) the subsequent returns. According to Asness, “the crucible for testing a valuation indicator is how well it forecasts long-term returns, and the Fed Model fails this test, while the Traditional Model [which uses Price/Earnings ratios without regard to the level of interest rates to evaluate the overall market] has strong forecasting power. Long-term expected real stock returns are low when starting P/Es are high and vice versa, regardless of starting nominal interest rates.” Jay Ritter agrees: “if past stock returns are irrelevant for predicting future stock returns, and future economic growth rates are also irrelevant, what does matter? The answer is simple: earnings yields [i.e., the inverse of the P/E ratio] ... P/E ratios revert towards the mean through price changes rather than earnings changes.”

Asness also examines the Fed Model’s ability to explain how investors set stock market P/Es. “That is, does the market contemporaneously set P/Es higher when interest rates are lower? Note the subtle but important difference between testing whether the Fed Model makes economic sense and accurately forecasts future long-term returns versus testing whether it explains how investors set P/Es. Asness confirms that investors have indeed historically bid P/E ratios to high (low) levels when nominal interest rates have been low (high). He also notes “the stock market’s P/E (based on trend earnings) is still very high versus history.” To Asness, then, the Fed Model induces investors to buy stocks at high P/E ratios – and thereby generates poor returns for those investors. “The evidence strongly suggests that the Fed model is fallacious as a tool for long-term investors. ... The bottom line is that for forecasting long-term stock returns, the Fed model is an empirical failure.”

Kicking the Tyres and Looking Under the Bonnet

Like any model, the Fed Model is underpinned by certain assumptions. Like all good assumptions, they render tractable an otherwise unmanageably complex reality. All assumptions have their weaknesses. Unfortunately, those of the Fed Model may be particularly troublesome in circumstances such as the present. For the sake of brevity, let us leave aside important issues such as the supposed suitability of mathematical models to the study of human action and the allegedly inverse relation between earnings and interest rates. Instead, note that FSVM relies upon “forward consensus expected operating earnings.” This formidable mouthful contains several possible sins. “Forward earnings” do not refer to actual (i.e., “trailing”) earnings during the most recent period; rather, they refer to the average (“consensus”) of analysts’ expectations about what these earnings will be during the next (“forward”) period.

How well can analysts predict companies’ earnings? Alas, not very well. David Dreman, Manitoban, Forbes columnist, Chairman of Dreman Value Management and author of Contrarian Investment Strategies: The Next Generation (Simon & Schuster, 1998, ISBN: 0684813505), reviewed almost 100,000 earnings forecasts made between 1973 and 1996. He found that the average forecast errs by no less than 30-60%. The odds are 1 in 130 that an analyst can predict a company’s earnings to within 5% of its true value four times in succession. Dreman concluded that companies’ earnings are “utterly unpredictable” and – what an understatement! – that analysts’ predictions of earnings are “not [of] much value.” William Sherden, author of (The Fortune Sellers: The Big Business of Buying and Selling Predictions, John Wiley & Sons, 1999, ISBN: 0471358444), concluded that despite their advanced methods, (often) incomprehensible jargon and (almost invariably) lofty salaries, economic and financial market “experts” are no better at predicting the future than astrologers, gizzard-squeezers and fortune-tellers.

In the long run, corporate profits cannot grow more quickly than overall economic activity. Over the decades, in other words, corporate profits have comprised a varying but roughly constant slice of the growing economic pie. (In Anglo-American countries, the earnings of listed companies have historically grown at an average of roughly 3-5% per year; and the economy, as it is conventionally measured, has grown at approximately 2-4% per year. The takeover of smaller, unlisted and less profitable firms by larger, listed and more profitable predators seems to explain part of the discrepancy. The “rebalancing” of stock market indices, whereby up-and-coming firms are added and down-and-out ones are ejected, may also account for part of it.)

Yet analysts unabashedly and consistently predict that corporate earnings will grow at an impossibly rapid rate. During the ten years from 1985 to 1995, for example, American analysts estimated that the earnings of companies comprising the S&P 500 would grow between 10.8% and 12.1% per year. Expectations rose steadily, to a rate of 14.9% per year, by the end of 1998. They then peaked at 18.7% in August 2000. Analysts, urged or at least unhindered by their institutional and retail clients, ignored the ineluctable limit upon the growth of profits that is imposed by the overall size of the economy and pace of activity. They forgot or disregarded their expectations’ absurd implication – namely that if profits really did grow at their expected rate then before many years companies’ earnings would exceed their revenues.

When The Great Bubble burst in 2000, analysts began to restrain their wildly optimistic projections. By the end of 2002, consensus expectations about long-term earnings growth for the S&P 500 had fallen from the all-time peak to a rate of “only” 12.8% per year. The reversal in the tech sector of the S&P 500 was even more dramatic: growth expectations fell from the 2000 peak rate of 28.7% to 16% at the end of 2002. American businesses, it is important to recognise, are not significantly more profitable today than they were in 2001; yet analysts’ “reassessments” of their prospects remain absurdly confident. And so it goes in Australia: in early July 2004, the analysts of a major international brokerage house were “optimistic about the coming reporting season, expecting an average 17.7% rise in earnings compared with the previous period.

Clearly, then, analysts’ collective (“consensus”) forecast of long-term earnings growth is severely and seemingly permanently biased in an overly sanguine direction. If so, then so too is the Fed Model’s assessment of the market’s “fair value.” In Grahamite terms, the Fed Model encourages investors to disavow the “margin of safety” and buy stocks at excessively high P/E ratios.

Why are analysts’ estimates of earnings so skewed towards over optimism? One reason is that caution generates less attention – and business – than either unqualified buoyancy or unmitigated gloom. Another is that analysts tend to extrapolate the potential earnings growth of “their” companies in “their” industries. Few wish to cover laggard companies and industries, and so attention is concentrated upon “winners.” A common problem, perhaps a bit less prevalent today than it was a few years ago, is that a “winner” tends to be defined as a company with whom the analyst’s firm has or seeks business. If so, then the long-term outlook with respect to these companies is likely to be overly rosy, and overconfident outlooks with respect to individual companies beget unrealistically bullish “consensus” outlooks.

Another reason is that analysts, particularly in the U.S., do not utilise “reported” earnings (which conform to Generally Accepted Accounting Principles). Instead, they use “operating” earnings (which, among other things, exclude most write-downs and many non-cash charges). GAAP has its defenders and critics and its strengths and weaknesses. Whatever one thinks of its rules, they are relatively firm. In sharp contrast, there are no accounting standards with which one can ascertain operating earnings. Which expenses, charges, etc., are included in or excluded from operating earnings? The answer varies from company to company and situation to situation. As a result, if executives ascertain that a particular asset should be written down and its effect excluded from operating earnings, and if “non-recurring” write-downs occur regularly, then so shall it be.

Operating earnings began to win analysts’ affections during the late 1980s and their use grew like topsy during The Great Bubble. Before then, “earnings” meant reported earnings; and (surprise, surprise) for a given company in a given reporting period, GAAP earnings are usually lower – and often much lower – than operating earnings. More generally, and as a consequence of things such as the increased use of options over stock as a form of compensation, a frenzied pace of acquisitions (and its associated torrent of “non-recurring charges”) and the increased use of various esoteric rationalisations to pad income statements, during the 1990s the quality of earnings, particularly in the U.S., declined. The questionable – from the point of view of many a Grahamite value investor – validity of operating earnings provides another reason to regard the Fed Model with misgivings. Its reliance upon (probably inflated) operating earnings reinforces the temptation to discount the “margin of safety” and buy stocks at excessively high P/E ratios.

Another criticism concerns the equity risk premium, i.e., the additional return (in the form of higher yield) which one requires from stocks in order to offset the risk that inheres in their ownership relative to “risk free” (in the sense that the chance of default is miniscule) government bonds. According to FSVM, stocks are fairly valued when their earnings yield equals that of the 10-year Treasury. The owner of a Commonwealth Government bond knows exactly what (nominal) coupon he will receive and when he will receive it; and the likelihood that the bond will default is effectively zero. The owner of an Australian (or American or British, etc.) company’s stock, or of a portfolio of stocks, in contrast, knows neither of these things. Traditionally, therefore, investors demanded compensation for these risks; and this compensation took the form of lower stock prices (i.e., higher yields). Successful investors of yesteryear, such as Graham would therefore have baulked at the idea that the earnings yield of stocks should match that of government paper. To accept the Fed’s Model’s premise about the equity risk premium is to be predisposed, other tings equal, to pay higher prices for stocks.

What Can Go Wrong?

It is instructive to look at Australian conditions on 1 July and see – if only roughly – whether investors who responded to the Fed Model’s cues might subsequently regret their decision. Given that the ten-year Commonwealth Bond yielded 5.9%, the Fed Model’s “fair value” P/E for the All Ords on that date was 16.9 times its earnings. Applying this P/E to expected operating earnings of $232 per share yields an All Ords with a “fair value” of 3,921 or roughly 9.5% above its level on 1 July.

Let us assume, however, that earnings in 2004-2005 will be $220 (i.e., 5% less than the initial estimate) and grow at 5% per year over the next five years (that is, more or less at their historical average). Let us also assume that during the year the yield on 10-year Commonwealth paper increases to 6.5%. Under these assumptions, which are cautious but not unrealistic, the All Ordinaries Index would be “fairly valued” at 3,385, i.e., a level that is 4.6% less than that on 1 July. Plus 9.5% or minus 4.6%? Clearly, to change one’s view about the future is to change the Fed Model’s implications, and significantly different implications can follow from different-but-still-plausible outlooks.

From this example emerges another imperfection – like the foregoing, it is not a deficiency of the Fed Model per se but of some of the people who might use it to inform their investment decisions. Like much of contemporary finance, the model does not necessarily assume but certainly gives the most pleasing predictions during interest rate environments like the one that has prevailed during the past 10-20 years. During these years in Anglo-American countries, rates have tended to fall, for the most part gently and steadily, to their present historic lows. Moreover, rates have been so low for so long that many people apparently believe they have a “right” to large amounts of artificially cheap money – and their politicians vociferously agree. But today’s hyper-low rates hardly guarantee similar rates in the future. Quite the contrary: Austrian School economists demonstrate that artificially low rates set in train a range of deleterious consequences, and that most of these consequences eventually generate higher rates (see Letter 27, Letter 30, Letter 41 and Letter 51). For the last ten years in Australia, a bet on low and falling rates has been a very good bet; but the laws of human action tell us that this wager is hardly foolproof.

It follows that, whether they are doing so consciously or not, many and perhaps most of today’s market participants are relying upon the Fed Model to generate favourable cues at what might be an inopportune moment – that is, when the model’s variables could disappoint investors. More specifically, if you accept the possibility that the consensus forward operating earnings estimates for the All ordinaries, S&P 500, etc., remain too high (i.e., that during the next several years the rates of growth of corporate earnings in Anglo-American countries will approximate rather than exceed their long-term averages); and if you believe that bond yields in these countries, “thanks” to their central banks, are artificially low, then you might think long and hard before you rely upon the Fed Model to justify your bullishness about stocks.

Some Conclusions for Value Investors

“He who lets the world, or his own portion of it, choose his plan of life for him has no need of any other faculty than the ape-like one of imitation” said John Stuart Mill in On Liberty (1859). Conversely, “he who chooses his plan for himself employs all his faculties. He must use observation to see, reasoning and judgment to foresee, activity to gather materials for decision, discrimination to decide, and when he has decided, firmness and self-control to hold to his deliberate decision.” These are wise words for value investors as well as classical liberals (see also The Austrian School and Classical Liberalism by Ralph Raico).

Leithner & Co. evaluates individual companies, securities, etc., and seeks to locate quality at a price significantly lower than a cautious assessment of value. These assessments depend neither upon the overall level of stock or bond markets nor upon forecasts about their level any particular point in the future. Indeed, the Company disclaims any ability to foretell general business conditions, the overall level and direction of financial markets or the market prices of securities. It has no crystal ball and ignores (apart from the occasional scorn) the many who claim, whether implicitly or overtly, that they do. (Stop and ask yourself: exactly how far into the future can you foresee?) On the basis of these individual assessments, during the past several years Leithner & Co. has generally found it difficult to locate businesses with requisite good quality and low price. Hence the case for caution remains strong.

What, then, of things such as the Fed Model? The trouble is partly with the model (and its assumptions and data), and largely with the people who calibrate it such that it sings to them the cheerful tune they want to hear. Even if the model’s assumptions were sensible, its data were valid and reliable and it presciently signalled that stocks were cheap relative to 10-year government bonds, it would not follow that stocks are cheap on an absolute basis. To say, for example, that stocks are undervalued relative to bonds can also be to say that stocks are not as crazily overvalued as bonds – but that both are nevertheless prohibitively dear. Analysts often obscure or forget this critical point (see also Rebecca Byrne, Adding Up the Fed Model’s Flawed Math).

So too, it seems, do policymakers. “The 1990s” said Dr Greenspan on 17 June 1999, “have witnessed one of the great bull stock markets in American history. Whether that means an unstable bubble has developed in its wake is difficult to assess. A large number of analysts have judged the level of equity prices to be excessive, even taking into account the rise in ‘fair value’ resulting from the acceleration of productivity and the associated long-term corporate earnings outlook. But bubbles generally are perceptible only after the fact. To spot a bubble in advance requires a judgment that hundreds of thousands of informed investors have it all wrong. Betting against markets is usually precarious at best.”

This is an example par excellence of The Maestro’s many ambiguous insights – one that might have helped to inflate the very bubble he dreaded. (“I know you believe you understand what you think I said,” The Maestro once chided an inquisitive Congressman, “but I am not sure you realise that what you heard is not what I meant.”) In this 1999 utterance Dr Greenspan seemed to say that during the late 1990s the stock market might have been gripped by mania. But if (and as the academics reassure us) financial markets efficiently reflect the expectations of countless “informed investors,” then the crowd must right. After all, so many people – including formidably intelligent central bankers, finance academics, brokers and strategists – cannot be collectively wrong.

In sharp contrast stands Benjamin Graham. Armed not with statistical models and computers but with Emersonian self-reliance, he stoutly emphasised “you are right because your data and reasoning are right.” If your premises and sensible, data valid and reliable and reasoning valid, in other words, then it is completely irrelevant whether important people, vocal people or great numbers of people agree or disagree with you. The basis of sound investing is sound method – not “consensus.” Correspondence with current fashion and the prevailing wisdom, it is vital to recognise, are no substitutes for cautious premises, hard evidence and valid reasoning; and consensus is no protection against error and overconfidence.

The point is that during The Great Bubble the crowd was wrong. It was egregiously mistaken, and so too was Dr Greenspan’s judgment about the crowd and its expert shepherds. At the time, all but a few – whose thin ranks tended to be populated by value investors such as Warren Buffett – were clearly convinced that stocks were worth buying; and they did not realise otherwise until after the bubble had burst. Otherwise intelligent actors in allegedly “efficient” markets, in other words, can succumb to a millennial mentality and bubble-think when they uncritically extrapolate absurdly bullish assumptions.

To say that prices during The Great Bubble were (on the basis of traditional, Graham-and-Buffett premises and reasoning) ridiculously high was in effect to say that most people were wrong; and to say that today’s prices remain too high is to say that many people continue to err. If so, it is not because they are dumb: it is because they heed politicians and are predisposed to view the world through overly rosy spectacles. (Plus ça change: During the 1930s, Albert Jay Nock, author of Our Enemy the State (Hallberg, 2001, ISBN: 0873190513), derided the “regular pre-election effort to start a boom in the stock market.” He added “Americans have a strange notion that the ordinary laws of economics do not apply to them. So doubtless they will think they are prosperous if the boom starts, and that deficits and indebtedness are merely signs of how prosperous they are.” Do not gloat: Australians and Britons are no different.)

Given analysts’ overoptimism, humans’ (including value investors’) inability to foresee financial events and developments with any useful degree of accuracy means that they should discount or ignore the analysts, strategists and economists who (or models that) implicitly claim that they have a clear crystal ball. It hardly means, however, that value investors should ignore the future. Quite the contrary: they must plan not by making specific predictions about what will happen but by considering general scenarios – and concentrating upon pessimistic circumstances – of what might conceivably transpire. Taking these pessimistic scenarios into account – i.e., building a hefty margin of safety into their assessments – they then structure their investments in order to reduce the risk of permanent loss of capital. Cater to the downside and the upside will mind itself. In this context it is significant that, according to Prof. Terrance Odean of the University of California-Davis, “psychologists have found that people who are mildly depressed tend to have the most realistic outlook” (quoted in The Wall Street Journal 22 September 1998).

“In the old legend,” Graham wrote in The Intelligent Investor, “the wise men boiled down the history of mortal affairs into the single phrase ‘this too shall pass.’ Confronted with a like challenge to distil the secret of sound investment into three words, we venture the motto ‘margin of safety.’ This is the thread that runs through all the preceding discussion of investment policy.” It is a theme that finds disappointingly little resonance in the Fed Model and many of the people who use it – and in today’s Australian financial press and the many people who peruse it. The purchase of quality assets at excellent prices is the essence of Graham-and-Buffett value investing. The more the investor follows its principles, the more he can respond with equanimity to models’–and people’s – bullish or bearish noises. In Graham’s words, “basically price fluctuations have only one significant meaning for the true investor. They provide him with an opportunity to buy when prices fall sharply and to sell wisely when they advance a great deal. At other times he will do better if he forgets about the stock market and pays attention to his dividend returns and to the operating results of his companies.

Chris Leithner


Contact | Disclaimer | Subscription
Chris © 2014-2017 by Artist Web Design