Neither efficient nor animally spirited, but eventually adjusting: the stock market according to L.A. Hahn.

Bragues, George

1. INTRODUCTION

Until recently, the Efficient Markets Hypothesis (EMH) demonstrated an impressive resiliency in the face of discordant events. It emerged from the 1987 stock market crash only slightly bruised, though the Dow Jones Industrial Average fell a record 22.6 percent in a single day when the only news that might have possibly accounted for such a cataclysm was a disagreement among industrialized nations about currency and interest rate levels. Somehow, though it was left tottering, it managed to survive the denouement of the late-1990's dot-com bubble. During this bubble, the NASDAQ Composite Index nearly quadrupled in an eighteen month period. Internet companies, such as eToys and TheGlobe.com, were accorded multi-billion valuations despite generating limited revenues and no profits. With the financial crisis of 2007-2009, however, it seems that the EMH has finally succumbed. Here was a situation, after all, in which a multitude of sophisticated analysts and investors, operating in the world's leading financial institutions, grossly overvalued the mortgage backed securities at the heart of the crisis and systematically undervalued the risk in their portfolios, all the while relying on models quantitatively constructed on the assumptions of the EMH (Dowd and Hutchison, 2010).

Just as the crisis has revived Keynesian ideas in macroeconomics, so it has led observers to revisit the sections of The General Theory of Employment, Interest, and Money that address the behavior of securities markets. There, Keynes (1964) compares the stock market to a giant casino, describing investors as fundamentally driven by "animal spirits," making decisions to buy and sell based on the estimated direction of crowd psychology as opposed to the intrinsic value of financial assets. Today, this Keynesian perspective is being taken up by Behavioral Finance (BF), a school of thought that began percolating at the margins of financial economics in the mid-1980's and which has since dethroned the EMH of its monopoly status in the discipline to become a formidable alternative. Based on the work of Amos Tversky and Daniel Kahneman (1979, 1982), BF applies the findings of psychology to explain price formation on securities markets. Opposing the EMH presupposition of human beings as utility maximizing calculators, BF sees the mind as inextricably swayed by emotions, feelings, social influences, cognitive biases, and heuristics (Baker and Nofsinger, 2010; Ackert and Deaves, 2010; Schleifer, 2004).

No doubt, BF offers a useful corrective to the EMH. Yet common sense, in tandem with a bit of elementary logic, suggests that it cannot fully account for market phenomena. If everyday observation amply confirms that we are not cool logicians, it also reveals examples in which people manage to overcome their biases and control their passions. An investor is often enticed by greed to buy a penny stock touted on an Internet newsgroup only to be brought back to reason by the realization that the deal is too good to be true. What is more, markets surely do regain their senses after bouts of extreme pessimism and optimism. And while prices may not always be exactly right, it would be hard to deny that, on occasion at least, certain securities, if not stocks in general, are correctly valued. Yet if the human mind is the servant of sub-rational forces, as BF seems to claim, these moments of rationality are a puzzle. (1) If it is such a challenge to impartially reason, how then do we manage to get things right from time to time?

This suggests we ought to explore the possibility of a middle way between the EMH and BF. In search of this, we return to the writings of Lucien Albert (L.A.) Hahn, a self-styled common-sense economist who first gained notice with his Economic Theory of Bank Credit (1921). In a career that traversed the worlds of academia and banking, Hahn attained his greatest level of fame with The Economics of Illusion (1949), a critique of Keynes, before laying out his own economic theory in Common Sense Economics (1956). Since then, Hahn's work has been almost forgotten; only two articles on Hahn come to sight in the scholarly literature over the last twenty years (Selgin and Boudreaux, 1990; Leeson, 1997).

Hahn's views on the stock market are set forth in the final part of Common Sense Economics. Embarking from an Austrian understanding of the business cycle and uncertainty, fleshed out with insights from psychology, Hahn argues that stock prices result from a combination of objective and subjective factors. On his account, the influence of mass opinion and mental inertia over most people's psyches generates sustained divergences from intrinsic values. Sooner or later, Hahn observes, these distortions are corrected by the pull of the objective facts in a process led by a few alert, independently minded investors. In Hahn's analysis--which this paper finds has stood the test of time--financial markets are neither perfectly efficient, nor animally spirited, but eventually adjusting.

2. CURRENT THEORIES OF PRICE FORMATION IN FINANCIAL MARKETS

Anyone putting himself forward as a common sense economist must have their views evaluated against not just the reigning academic theories, but also the way in which the relevant practitioners understand their own activity. Among stock market professionals, the myriad of investing and trading strategies followed all belong to either one, or a mix, of two approaches: fundamental analysis (FA) and technical analysis (TA). The first is the more predominant and respected of the two, though the second has gained adherents and stature over the last two decades with the development of faster computers and increased access to short-term trading opportunities. While most investors rely exclusively on either one of the two methodologies, more than a few combine them, typically by using FA to identify which stocks to purchase or sell and then turning to TA in deciding when precisely to execute a transaction.

FA insists that the best way to make investments decisions is to analyze the financial data pertaining to a security. It states, too, the necessity of accounting for industry and macroeconomic conditions that impinge on a security's value, the marketability of the firm's goods and services, as well as the quality of management. Thus, practitioners of FA pore over a firm's balance sheet, its cash flow statement, reported earnings and profit margins. They will evaluate how a company's products stack up against the competition and whether its strategy is adequately framed to boost profitability. They will gauge the firm's profit potential and risk exposures at different phases of the business cycle and ascertain whether it is part of an industry that is in a speculative, growth, maturity, or decline phase. All this is done with a view to determining the security's intrinsic value. Some try to arrive at this number by projecting future cash flows and then discounting these to their present value. Most FA practitioners, though, apply a valuation metric of some kind, most commonly the price to earnings (P/E) ratio, against a group of comparable securities. (2)

Clearly, FA assumes that financial markets are only sporadically efficient. According to the bible of FA, Graham and Dodd's Security Analysis, "market prices, like a stopped clock, are a correct representation of value twice in an investor's day" (Cottle, Murray, and Block, 1988, p. 26). However much information is now readily available to investors, FA proponents say, the fact remains that it is interpreted differently and that most observers fail to capture the more subtle and revealing bits of data within an evolving composite view. Not to mention that markets are subject to the oscillations of fashion, which bring different industry sectors in and out of favor alongside the waves of fear and greed that engulf the generality of stocks.

Diametrically opposed to FA, TA ignores all the financial, industry, and economic data (Edwards and Magee, 2001). Instead, TA focuses on the historical movement of prices and transactional volumes. This price and volume data is depicted on charts, which practitioners of TA examine for the presence of trends. Their modus operandi is to ride a trend until it shows signs of changing, at which point they reverse their market positions to exploit the new price move. To gauge these trends, TA refers to a set of patterns, such as the head and shoulders and pennant formations, in judging whether prices are consolidating within the prevailing trend or are at a critical turning point. Trend lines are drawn connecting significant high and low points, moving averages calculated, and indicators (i.e., stochastics, relative strength index, on balance volume) derived through a mathematical transformation of price and volume data into directional barometers. When asked to explain why charts are more instructive than financial statements, TA's supporters aver that all fundamental information relevant to a security--whether it be financial, economic, strategic, or political--is already reflected in the price. Charts, too, are said to disclose the historical reality that markets often trend. Or, to put it in statistical terms, financial asset returns exhibit autocorrelation. Finally, TA claims that the psychological laws governing human nature mean that chart patterns inevitably repeat themselves (Murphy, 1986, pp. 2-4).

While the seeds of the EMH can be found as early as Louis Bechelier (2006) and Alfred Cowles (1933, 1944), the theory came into prominence with Eugene Fama (1970) and Paul Samuelson (1965) in the 1960's and 1970's. From the fact that investors compete to find the best investment prospects, the EMH deduces that this search must equilibrate with the elimination of all misvaluations. For if any such exist, investors will immediately exploit the arbitrage opportunity thereby presented until their actions move prices until market value and intrinsic value are equal. So long as information is readily accessible and the barriers to trading are minimal, this no-arbitrage equilibrium will be reached quickly in response to any changes in market conditions. Rather than being a goal to which market forces are drawn and only rarely achieved, as the Austrian economics tradition holds, equilibrium is seen by the EMH in neo-classical terms as the ordinary state of affairs (Boettke, 2010). As such, EMH asserts that stock prices reflect all available information. In the weak version of the EMH, this claim is restricted to information about historical prices, thereby rejecting TA. As for FA, the EMH questions this in its semi-strong version, according to which all publicly information about a company's financial and economic situation is also encompassed in the stock price. There is also a strong version of the theory that states that all relevant information whatsoever, including that possessed by insiders, is assimilated into prices, though hardly any EMH advocate subscribes to it. So other than conceding an edge to insiders, the EMH claims that investors earn returns, not by buying undervalued securities and selling (or also shorting) overvalued ones, but by assuming non-diversifiable risk in their portfolios (Malkiel, pp. 180-215).

Despite its portrayal of investors as emotional and biased, BF admits this description does not apply to every single investor. A few are rational. Having made this exception, BF gives itself a means of solving the aforementioned puzzle of how to explain the periodic episodes of reasonable valuations when passion and bad judgment is so prevalent. Perhaps the elite corps of rational investors cancel the effects of their irrational brethren by exploiting the latter's mistakes? Indeed, the EMH invokes such arbitrage trading to deal with the glaring fact that not everyone lives up to its assumption of utility maximizing behavior. BF, however, declines this tack, arguing instead that the rational face constraints on arbitrage. Once prices move away from correct levels, no guarantee exists that the deviation will not further widen and persist under the sway of irrational traders. Value arbitrageurs thus expose themselves to the danger of having to carry a losing position over a long period during which the intrinsic value of the security undergoes an adverse change or paper losses grow to the point of inducing financial stress. Recognizing this, rational investors will either resist the urge to trade against the irrational, or perhaps even try to join them for as long as they are in control of the price movement, thereby reinforcing the divergence from correct values. As Keynes originally put this BF claim: "Investment based on genuine long-term expectation is so difficult to-day as to be scarcely practicable. He who attempts it must surely lead much more laborious days and run greater risks than he who tries guess better than the crowd how the crowd will behave" (Keynes, 1964, p. 157).

Where BF goes beyond Keynes is in specifying the types of individual biases that combine to mislead the crowd. To cite the more notable findings, BF scholars conclude that investors, particularly male, overestimate their investment abilities (the overconfidence bias); they lag in updating their beliefs to new evidence (the conservatism bias); filter information that corroborates their existing beliefs (the confirmatory bias); surmise that runs within a series of events must soon reverse (gambler's fallacy); deduce that a repeated occurrence of events portends a larger trend (clustering illusion); and overly rely on easily accessible memories or ideas in rendering probability judgments (the availability bias). So too, BF observes that investors are more sensitive to losses than to gains of the same amount of money, overweight small probabilities and underweight large ones, and change their decisions about the same probability scenarios when these are framed differently (Barberis and Thaler, 2002; Shefrin, 2000).

3. HAHN'S CRITIQUE OF MARKET EFFICIENCY

L.A. Hahn wades into the debate about financial markets because it illuminates the relationship between expectations and facts in economic life--that is, the extent to which human subjectivity affects the decisions that people make relative to the necessities imposed on individual choices by the objective realm. Entranced by the prospect of replicating the success of physics in the human realm, the dominant streams of economics over the last two centuries--whether classical or neo-classical--have tended to emphasize the determinative role of objective variables in accounting for economic phenomena. The mathematical techniques employed by orthodox economists, with all their equations depicting a given state of affairs as a function of various conditions, presuppose the priority of the objective over the subjective. But this epistemological preference, Hahn points out, falls afoul of the fact that businesspersons do not simply react to changes in the economic data taking place now, but must form some estimate of what shifts might occur in the future. This is because of the interval existing between the initial decision to deploy resources in the production of a good or service and the time it is ready for sale. Between these two points, input costs might change, as may consumer tastes and the competitive environment, either of which can significantly affect the profitability of one's business projects. With businesspersons thus compelled to become forecasters, as the Austrian economists recognize (Mises, 1963, pp. 105-106), the subjective factor assumes an influential role, precisely insofar as the future is ultimately incomprehensible. From this it follows that the future cannot determine anything in the present and that the mind is liberated to conceive numerous scenarios in line with its psychological propensities.

That such expectations constitute the basic stuff of financial markets is obvious from even the most cursory observation of trading activity. When a company reports its earnings or the government releases the latest employment numbers, it is not so much what the data itself reveals that moves markets, but more so how the newly divulged information compares to expectations. Record profits may be announced and the unemployment rate fall dramatically, but stock prices may still drop if the good news does not accord with forecasts. In evaluating the efficiency of markets, the question thus becomes: do the expectations have a tendency to be on the mark? Prediction errors are inevitable, of course, but if these turn out to be normally distributed around realized levels then the argument for market rationality is greatly strengthened. It is precisely the contention of the EMH that the market's forecasting mistakes are normally distributed. As such, the subjective element in expectations is rendered mathematically tractable by the application of statistical techniques. The result is that expectations are objectified, so to speak, by virtue of being construed as a mechanism reflecting the real probabilities of events. But if expectations err universally and systematically, then human subjectivity cannot be viewed simply as a mirror to the objective world, and must instead take on the character of a truly independent cause of market phenomena. This is exactly Hahn's point.

He establishes it, first, by presenting a straightforward model of stock prices based on investor behavior. Noting that most individuals who buy shares do so with the aim of earning dividends in the future, Hahn infers the correct value of a stock as equal to the present value of that projected income. Since people value a dollar more today than a dollar to be had in the future, the present value of estimated dividends must be a discounted at a rate corresponding to the passage of time until their receipt. What Hahn thus arrives at is a discounted cash flow (DCF) model of stock prices:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Where S is the stock price, D is the dividend amount, n is the number of time periods over which dividends are being paid, while t refers to the nth time period, and r to the discount rate. Assuming the dividends are secure, this rate must equal the yield on long-term bonds, for otherwise investors would have an incentive to switch between bonds and stocks to whichever offered the higher return. Hahn does not spell out what happens if the dividend payments are not as certain as the bonds, but the obvious consequence is that the discount rate on the stock will then contain a risk premium.

One might counter that investors actually focus on earnings, rather than dividends, and that they often pin their hopes on selling at a higher price than where they bought. Even so, as Hahn observes, investors cannot really care about earnings per se, since they do not necessarily receive them as income, but only insofar as these signal the firm's capacity to eventually issue dividends. Investors intent on capital gains must also concern themselves with dividends inasmuch as the stock price can only rise if the prospect of their payment and growth increase, as proxied by earnings (Hahn, 1956, p. 197). Given that financial markets are not populated solely by short-term traders, the latter will eventually have to deal with someone having a longer-term horizon. Such an investor will not purchase shares from the short-term trader, and help close the latter's position at a profit, unless the outlook for dividends justifies the price (Hahn, 1956, p. 208).

Having put forward a DCF model, Hahn proceeded to test its predictions of intrinsic value against market prices. By proceeding in this fashion, Hahn anticipates the centerpiece of Robert Shiller's (2000, pp. 184-190) brief against market efficiency. But unlike Shiller, a leading figure in the BF school, Hahn did not use the model to generate point estimates of the present value of future dividends, probably because of the difficulty of forecasting those numbers the further one goes out into the future. Another factor, arguably, is Hahn's (1956) distrust of the mathematical methods that have come to dominate economics: "the mathematical language of modern economics has led many an economist to describe not so much what actually happens but what possibly could happen ... any resemblance of these descriptions and explanations to reality seems to me purely accidental" (p. xi). Consequently, Hahn only uses the model as a tool to interpret price movements.

The model thus tells us to expect stock prices to vary directly with dividends and to move inversely with bond yields. During the 1929-April 1956 period that Hahn investigated, the corporate bond yields that he used to represent the discount rate changed little from 1937 forward. Before then, stock prices accorded with changes in the bond yield, with equities declining between 1929 and 1932 as corporate bond yields rose, and equities increasing from 1933 to 1937 as yields fell. Afterwards, stock prices movements can only be accounted for on the basis of dividends. Here is how Hahn broke down the relationships amid the bull and bear markets that took place:

Three of the sub-periods, covering 12 of the 26M years in Hahn's data set (almost half), show the stock market behaving irrationally. Once again foreshadowing Shiller's analysis, Hahn goes further in discerning that, even when the two moved in the same direction, prices were noticeably more volatile than dividends. An objection might be raised here that this lack of correspondence exists because investors follow earnings instead of dividends. Yet, as Hahn notes, earnings and dividends generally move in tandem, just as one would expect from the former serving as a proxy for the latter. Earnings, indeed, are more volatile, since corporate boards prefer to keep dividend payouts steady knowing that profitability is subject to gyration from one year to the next. Stock prices, as a result, track earnings more closely.

Before we can affirm Hahn's conclusion that markets are inefficient, we cannot forget that more than fifty years of additional data has become available since he wrote. Let us see, then, how his analysis has stood the test of time using his own interpretive method. To this end, we employ the Standard and Poor's Composite Index (S&P 500) as our barometer of stock market performance, as well as the dividends and earnings per share of its constituent firms. We follow Hahn in adopting corporate bond yields as our proxy for the discount rate, and, more specifically, those rated Aaa by Moody's. Below are six charts depicting the S&P 500 index, dividends, earnings, and corporate bond yields over the 1956-1968, 1968-1982, and 1982-2010 time frames. This breakdown separates three broad trends that are discernible between 1956-2010. The 1956-1968 interval saw the continuation of the post-World War II bull market. From 1968 to 1982, the combination of inflation and slower economic growth led to a sideways range in stock prices. Between 1982 and 2010, the stock market experienced a historically unprecedented bull phase from which it has been correcting since 2000.

[FIGURE 2 OMITTED]

In Figure 2, it can be seen that dividends and stock prices broadly tracked each other, though the latter exhibited more volatility, just as it did in the 1929-1956 time frame that Hahn examined. This greater volatility seems due to the fact that investors were swayed by the vicissitudes of earnings.

[FIGURE 3 OMITTED]

At odds with the market's ascent is the steady rise in nominal yields from 1956-1968 depicted in Figure 3. It is not out of the question, of course, that the market's upward movement was justified by the increase in dividends outweighing the higher discount rates. Still, the climb in nominal rates was both significant--doubling during the period--and persistent.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

Beginning in 1968, stock prices (Figure 4) finally succumbed to the increase in bond yields, as the longer term trend flattened with shorter-term moves exhibiting a zig-zag pattern. As Figure 5 shows, the ascent in yields continued into the early 1980s, helping account for the generally lackluster performance of stocks that characterized the period. Dividends trended higher during this period, even as the market was essentially flat, though the ascent was sharpest after 1975 as stocks began to show signs of re-establishing the secular uptrend. Once again, stock prices were more volatile than dividends, with earnings correlating more strongly with fluctuations in the S&P 500.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

With figures 6 and 7 we see that the bull market from 1982-2000--interrupted by only three brief downturns in 1987, 1990, and 1998--was directionally in accord with both rising dividends and falling nominal rates. Yet if the trend was justified, the slope of the price move was much steeper than that of dividends, especially from the early 1990s to the 2000 high when the fall in corporate bond yields had stabilized. Partly accounting for this, it is true, was the increasing prevalence during the 1980s and 1990s of firms using available cash to buy back shares in lieu of raising dividends. Even so, like dividends, these share repurchases are generally funded out of earnings, and the latter also rose at a slower pace relative to stock prices. Correcting for these divergences, stock prices fell sharply from the 2000 high even as dividends only gently declined. In 2007, the S&P index merely approached its 2000 peak, even as dividends were establishing all-time highs.

Imitating Hahn's method, the table below offers a closer look of the 1956-2010 period, categorized by bull and bear markets. A bear market is defined as occurring upon a minimum fall of 20 percent in the daily closing price of the nominal S&P 500 index from an earlier peak. A bull market takes place upon a minimum 20 percent rise in that index from a low point. Yields are categorized as flat if there was a 50 basis point change or less during the bull or bear market in question. Dividends, in turn, are designated as flat whenever a nominal rise was canceled by inflation during the relevant period as measured by changes in the US consumer price index.

Compared to Hahn's 1929-1956 analysis, fewer of the subperiods--7 of the 21 intervals representing just under 9 of the 55 total years--exhibit stock price movements in violation of the DCF model. In this instance as well, the inclusion of corporate yields generated 4 inconclusive findings, though these only constituted approximately 9 % years of the data set. Here, it should be kept in mind that during the longest of the inconclusive periods, Oct. 1974-Nov. 1980, dividends only slightly rose after factoring in inflation, while bond yields shot up. Dividends merely increased 1.9 percent in real terms over that entire 6 year interval as compared to the 5.9 percent equivalent figure that would have been expected had the average monthly growth rate of dividends from 1956-2010 prevailed. More importantly, the table above only takes the direction of the variables into account, not their respective fluctuations. To repeat, stock prices were significantly more volatile than the DCF model inputs, dividends in particular, throughout the entire 1956-2010 time frame. And, again, the upward slope in stock prices was noticeably more pronounced than that of dividends (or even earnings) in 1990-2000. On balance, therefore, Hahn's conclusion of market inefficiency on the basis of the 1929-1956 experience is borne out by the subsequent data.

4. THE INTERPLAY OF SUBJECTIVE AND OBJECTIVE FORCES

At the same time, it cannot be ignored that the market sometimes gets it right and that, even when it does stray from intrinsic value, it ultimately finds its way back. Hahn points out that the relationship between dividends and stocks becomes much closer once the oscillations created by bull and bear markets are diminished by smoothing prices through the calculation of a moving average. Figure 9 checks to see whether this is also the case for the 1956-2010 period using a 60 month moving average of the S&P 500 index.

[FIGURE 9 OMITTED]

The correlation is indeed tighter, except that the price average fails to track the dividend increases from 2003 to 2009. These reflect legislation passed by the Bush Administration in 2003 that lowered the taxation of dividends. While this change was subsequently extended to 2010, it was then set to expire in 2011 unless the U.S. government chose to renew the extension, which it subsequently did for another two years. In view of the uncertainty that surrounded the duration of this policy, the markets likely factored in the possibility of a return to the previous tax treatment of dividends. "It must be borne in mind," as Hahn (1956) rightly says, "that if any future change is to lead to changes in the valuation of a share--or of all shares--it must be expected to be permanent" (p. 200).

Given how the DCF model looks against a long-term moving average of stock prices, Hahn concludes that objective reality ultimately acts as a magnet drawing markets back from the errors of its shorter and medium-term ways that continually recur because of the force of human subjectivity. Usually, Hahn posits, a complete market cycle proceeds as follows: bull markets emerge out of the depths of a bear market when dividend yields are high as a result of a pervasive gloom leading investors to expect further declines in earnings and dividends. That no bargain hunters enter the marketplace to take advantage of the high dividend yields, and thereby raise prices, only reinforces the negative sentiment. This is what Hahn calls the exaggeration phase of the bear market. The bull market commences as an adjustment phase of this overextended move as indications slowly materialize that the economy is on the mend and dividends are set to rise.

Then, the market enters a normal phase in which the investing public is neither exuberant nor disconsolate about the future. Share prices now merely obey the upward trajectory of dividends. But then another exaggeration stage ensues in which the very fact that dividends and earnings have been rising generates expectations that these will continue to rise. Accordingly, stock prices move ahead of dividends. As prices do not meet resistance from sellers, and short sellers shy away from exploiting the excessive valuations signaled by the low dividend yields, investors gain reassurance. Finally, a few alert investors notice that the market's exalted levels are unsustainable. These sell and prices start to reverse. As the general optimism fades, the selling accelerates and we enter another adjustment phase that launches a bear market. This, too, subsequently goes into a normal stage during which prices decline in lockstep with dividends. With time, an exaggeration phase to the downside transpires and the cycle starts again. Throughout this entire sequence, Hahn adds, a mental inertia operates that renders investors incapable of changing their outlook until the evidence to the contrary becomes dramatically obvious. The exaggeration phases, especially, become resilient to incongruous news items as a result, helping explain how prices can diverge from intrinsic value for extended periods. A graphic representation of this cycle is given in Figure 10.

[FIGURE 10 OMITTED]

Embedded in this account of stock market cycles are a number of psychological claims that show Hahn foreshadowing elements of BF. By asserting that investors rely on recent trends in forming their expectations, and thus project the recent past onto the future, Hahn is alluding to what cognitive psychologists nowadays refer to as the recency effect, which itself is a variation of the availability bias so much talked about in the BF literature. The mental inertia, too, that Hahn invokes is equivalent to the conservatism bias. When he proceeds to outline the implications of his market theory to investing strategy, he cites a third psychological trait, namely the individual's subjection to mass opinion. "It engulfs not only those who easily succumb to foreign influences but even those with normally detached views and sober judgment. An almost superhuman effort is needed to evade the influence of mass opinion" (ibid, p. 212). What this groupthink does, clearly, is to magnify the predominant trend that the Zeitgeist of the period happens to be buttressing. Continuing in this Tocquevillean vein, Hahn even suggests that the democratization of the stock market enhances this dynamic, insofar as the widespread dissemination of prices enables investors to quickly assess what the majority is thinking (Tocqueville, 1969, pp. 254-259). Were Hahn alive to see the Internet, and all the websites offering free quotes and news, he would surely have concluded that it has augmented the mental dominance of the crowd.

Why, fundamentally, do these subjective factors, these human thought processes, assume the role that they do in so often mispricing securities? Why is it, in other words, that investors are incapable of thinking about the markets in ways that avoid systematic errors? The very thing, it turns out, that opens up the economic realm to the play of subjective forces is that which conduces to market inefficiency: we cannot know the future. The freedom from present exigencies that this gives to act on the basis of our idiosyncratic predictions also dictates our getting those wrong often. As Hahn correctly notes, if the future could be known, then securities prices would immediately reflect that information--a point that would later be stressed by EMH advocates. The future would then effectively become the present and cease to exist as a distinct temporal category. At best, according to Hahn, we can make probability judgments about the near future. As to the distant future, "it lies, shrouded in a mist, beyond the horizon of time" (ibid, p. 203). Market practitioners acknowledge this, he insinuates, by restricting the discounting of future developments to twelve months--which surely does occur, if the predominance of one year earnings estimates among stock analysts over longer-term forecasts is any indication.

Even to the extent that an investor is able to make probability judgments, these are not of the kind described as rationally utility maximizing in the standard textbook treatments of finance. To be sure, Hahn concedes that a stock price can, in theory at least, be viewed as the summed value of various scenarios for the firm, each weighted by its probability. For example, if there is a 30 percent that company PQR will, over the next year, report earnings that correspond to a share price of $50, and a 70 percent chance its eventual income will be such as to correlate with a $40 per share figure, then PQR stock will trade at $43 (0.3 x $50 + 0.7 x $40). In reality, the number of scenarios is greater than two and more complicated to delineate, so that the stock price ends up at the point where the chances of it going up or down seem to be equal, rather than whatever is dictated by the calculation of some complex equation.

More critically, however, the probabilities imputed are not mathematical in plotting the frequency of similar events in the past. In the world of investing, there is no equivalent of a billion throws of a die to consider, no large samples of essentially identical phenomena. As Hahn observes, a market event taking place in one cycle is always different, in some decisive respect, from an analogous occurrence in another cycle. An analyst might note, for example, that the stock market has experienced higher returns under Democratic presidents, as opposed to Republican ones, but one cannot simply infer from this that the pattern will recur. Not only were previous presidents of the same party often distinctive in their ideological mindsets and policy approaches, a multitude of other factors were driving stock prices--whose operation, for all we know, may subsequently combine to overwhelm the relevance of whether the White House is being occupied by a Democrat or a Republican. Besides the lack of homogeneity in the slices of history, the brute fact remains that too few of them are going to repeat themselves over an investor's lifetime to enable him or her to depend upon the law of large numbers.

Consequently, though the playing of chances that investing entails means it can be likened to gambling at a craps table, no one can proceed in the buying and selling of securities the way a casino does in operating its games--that is, by continually playing across numerous locations according to the same rules on the expectation that, over time, the expected frequencies will assert themselves. Since his or her number of plays is much shorter than that of a casino, an investor's risk is significantly higher--the variance of their potential outcomes is far greater--than what a historical sense of the probabilities might suggest. This offers an explanation as to why the risk models that Wall Street employed so spectacularly failed in the recent financial crisis. Instead of reflecting the mistaken specification of a normal distribution (Dowd and Hutchinson, 2010; Triana, 2009; Mandelbrot and Hudson, 2006), or the input of insufficient historical data, the problem lied in thinking that numerical probabilities could even be assigned at all.

Here is how Hahn (1956) aptly puts it:

 The case is comparable not to that of the bank in Monte Carlo,
 which can and does rely on red and black turning up equally often
 in the long run, but rather to that of the individual player, who
 cannot know whether the ball will stop on red or on black. He has
 to take his chance. He may be playing red ten times in succession,
 although black may win ten times (pp. 204-205). (3)

Indeed, Hahn goes so far as to invoke David Hume's (1978, pp. 127-130) contention that probability assessments are subjective mental acts. One simply feels inclined in favor of one outcome rather than another, with the level of intensity felt varying roughly with the preponderance of that outcome relative to other scenarios in one's previous experience.

That brings us to the objective factors moving stock prices, by which Hahn means all market relevant phenomena operating externally to investor minds compelling their rational faculties towards similar evaluations. Despite the thoroughgoing value subjectivism of Austrian economics, Hahn nevertheless echoes that tradition in the objectivist side of his theory. As we have seen, he argues both that stock prices exhibit cyclical behavior and that, as per the DCF model, those prices are a function of expected dividends and prevailing interest rates. Now since dividends come out of profits, and these in turn fluctuate with the vicissitudes of the economy, it follows that the generality of stocks, though their long-term trajectory will follow secular trends, are nevertheless affected by the business cycles that have been a feature of capitalist economies since the 19th century. The upshot is that the question of what objectively drives the stock market is necessarily connected to the riddle of business cycles.

For Hahn, the puzzle here is figuring out why the demand for goods and services sometimes runs below production and other times above it. It is not that he thinks Say's law is wrong. When the business cycle is viewed as a whole, demand and supply tend to balance. It is just that Say's law does not prevail at all times. Finding underconsumption (Keynesian) and overconsumption accounts of the business cycle wanting, Hahn is brought to Knut Wicksell's (1936) natural interest rate theory, which grounds the Austrian understanding of business cycles. According to Wicksell, the free play of supply and demand forces in the credit market work to establish the natural rate of interest or, if you will, the market rate. However, the central bank, with the commercial banks assisting, can exercise their monopoly power over currency issuance to drive interest rates away from the natural or market rate. Where the rate is higher than the market, fewer loans are sought to finance consumption and investment expenditures, thus lowering demand for goods and services. Prices fall and the economic activity declines. Where the rate is lower than the market, more loans are sought to finance consumption and investment, thus raising the demand for goods and services. Prices increase and economic activity rises. Accordingly, business cycles are the result of central bank policies.

It is very important to note, however, that Hahn supplements this essentially Austrian account with a psychological theory influenced by A.C. Pigou's (1929) Industrial Fluctuations. Hahn does so to address an objection that would later be made by the rational expectations school, namely that the central bank cannot take the economy up and down, unless people err by not accounting for its actions. Who would invest in a new capital project well into an upswing, if they can see that the Federal Reserve is eventually going to raise interest rates to stop the economy from overheating? Who is going to shy away from a big investment amidst a slump if the central bank is committed to a very loose monetary policy? In Hahn's view, such mistakes can only be explained by a pro-cyclical psychological dynamic in which people are carried by excitement in prosperous times and sunk by pessimism in recessionary periods. His analysis of the subjective forces operating in financial markets is meant to corroborate this point.

Beyond influencing the level of profits, and hence dividends, that stock prices incorporate, the central bank's activities obviously impinge on the discount rate applied by investors to shares. Everything else remaining equal, an easing of monetary conditions, by lowering the discount rate, will raise stock prices and vice-versa. Hahn cites that easing which takes place during a recession to explain how bear markets manage to end even as everyone is pessimistic about the economy's ability to generate profits and dividends. Once these begin to revive, the bull market gains momentum, especially as the central banks keeps interest rates from immediately jumping to ensure economic recovery. As this bull phase matures, rates do begin to rise, with the central bank less willing to accommodate increased credit demand, but the ascent in profits and dividends outweighs the higher discount rate. The bull market ends as interest rates continue to push higher, with the ensuing bear market strengthening in the wake of falling profits and dividends. All this, it must be said, is not much different from the story often told in Wall Street and the City. But, as Hahn (1956) says, "the world does not consist of economists who know and business men who err" (p. 166).

5. CONCLUDING REMARKS

Using a DCF model of stock prices, Hahn maintains that the stock market is inefficient, doing so on the basis of 1929-1956 data. Updating his mode of analysis to the end of 2010, we substantiate his conclusion that dividends are only loosely correlated with stock prices, even after taking discount rates into account. In place of an efficient markets model, Hahn describes the stock market as subject to recurring cycles in which subjective and objective factors combine to set prices. Though the subjective forces of psychology, consisting of mental inertia and dependence on mass opinion, regularly take prices either above, or below, levels dictated by the objective facts, the latter do act as a magnet checking the movements of the former. The objective factors, insofar as these reflect business cycles, are decisively influenced by central bank practices of maintaining interest rates at non-market rates.

By arguing that financial markets are inefficient, Hahn ends up on the side of investment practitioners who subscribe to either FA or TA as well as the academic school, BF, most reflective of the conventional wisdom in the financial community. His more precise stance, though, with respect to these three approaches comes to sight in the personal investing counsel that he draws from his theory of the stock market. While agreeing with BF that psychological variables create divergences between stock prices and rational prices, and while sympathizing with FA that market values do not always equal intrinsic values, Hahn's proposed strategy is surprisingly aligned with TA. Granted, Hahn does not refer to trend lines, moving average crossovers, head and shoulder formations, or any of the other constructs of TA. In the one instance he does allude to TA, he is dismissive. This happens when he briefly discusses the Dow theory, a century old concept that seeks to identify the market's current trend by comparing movements in the Dow Jones Industrial Average to those of the Dow Transportation Average (Rhea, 1932). Hahn contends the Dow theory has proved to be of limited use and that its value is negligible in any case because it is widely known.

That said, Hahn figures that the optimal strategy is to buy shares at the intersection of the exaggeration phase of the bear market and the adjustment phase of the incoming bull market. Here, the wise investor must go against mass opinion, which is overwhelmingly pessimistic at this stage when the subjective forces of psychology are in control. From here, though, one must become willing to travel with mass opinion. For the wise investor is then supposed to hold on to their shares as the subjective element is brought back to objective reality. Furthermore, he or she is to maintain their position afterwards when the two diverge again as the bull market reaches its most enthusiastic phase. This is, of course, the moment in which one must again oppose mass opinion. Thus, the wise investor sells (or short sells), and stays out (or remains short) throughout the normal stage of the bear market when the objective and subjective orders are reunited once again, only re-entering (and covering) when these two diverge at the exaggeration phase of the bear market. "Thus it is as wrong always to oppose the prevailing tendency as it is always to follow it. In a nutshell, the right rule is: first against the tendency, then with it, and finally against it" (Hahn 1956, p. 214). Heeding this advice, one will spend quite a bit of time following the trend, precisely as TA counsels. Even Hahn speaks of the change in trend occurring over a "moment" (ibid.). Yet it must be conceded that Hahn does not think such changes can be scientifically predicted. To this extent, and only to this extent, is Hahn in accord with EMH.

REFERENCES

Ackert, Lucy F., and Richard Deaves. 2010. Behavioral Finance: Psychology, Decision-Making, and Markets. Mason, Ohio: South-Western, Cengage Learning.

Bachelier, Louis. 2006. Louis Bachelier's Theory of Speculation. Mark Davis and Alison Etheridge, trans. Princeton: Princeton University Press.

Baker, H. Kent and John R. Nofsinger, eds. 2010. Behavioral Finance: Investors, Corporations, and Markets. Hoboken, N.J.: John Wiley and Sons.

Barberis, Nicholas, and Richard Thaler. 2002. "A Survey of Behavioral Finance," National Bureau of Economic Research Working Paper. Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=327880.

Bloomfield, Robert. 2010. "Traditional versus Behavioral Finance" in H. Kent Baker and John R. Nofsinger, eds, Behavioral Finance: Investors, Corporations, and Markets. Hoboken, N.J.: John Wiley and Sons.

Boettke, Peter J. 2010. "What Happened to Efficient Markets?" The Independent Review, 14, no. 3: 363-375.

Cottle, Sidney, Roger Murray, and Frank Block. 1988. Graham and Dodd's Security Analysis. New York: McGraw-Hill.

Cowles, Alfred. 1933. "Can Stock Market Forecasters Forecast?" Econo metrica 1: 309-324.

--. 1944. "Stock Market Forecasting," Econometrica 12, no. 3: 206-214.

Dowd, Kevin & Hutchinson, Martin. 2010. Alchemists of Loss. New York: John Wiley and Sons.

Fama, Eugene. 1970. "Efficient Capital Markets: A Review of Theory and Empirical Work," Journal of Finance 25: 383-417.

Fernandez, Pablo. 2001. "Valuation Using Multiples: How Do Analysts Reach Their Conclusions," IESE Business School Working Paper. Available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=274972.

Keynes, John Maynard. 1964. The General Theory of Employment, Interest, and Money. San Diego: Harcourt.

Hahn, Lucien Albert. 1921. Economic Theory of Bank Credit. Publisher unknown.

--. 1949. The Economics of Illusion. New York: New York Institute of Finance.

--. 1956. Common Sense Economics. New York: Abelard-Schuman Ltd.

Hume, David. 1978. A Treatise of Human Nature, 2nd ed. Peter Nidditch, ed. Oxford: Oxford University Press.

Kahneman, Daniel, and Amos Tversky. 1979. "Prospect Theory: An Analysis of Decisions under Risk," Econometrica 47: 313-327

Kahneman, Daniel, Paul Slovic, and Amos Tversky. 1982. Judgment under Uncertainty: Heuristics and Biases. New York: Cambridge University Press.

Leeson, Robert. 1997. "Influence (or the Lack of It) in the Economics Profession: The Case of Lucien Albert Hahn," History of Political Economy 29, no. 4: 635-638.

Malkiel, Burton. 2007. A Random Walk down Wall Street. New York: W.W. Norton.

Mandelbrot, Benoit, and Richard L. Hudson. 2006. The Misbehavior of Markets: A Fractal View of Financial Turbulence. New York: Basic Books.

Mises, Ludwig von. 1963. Human Action: A Treatise on Economics. San Francisco: Fox and Wilkes.

Murphy, John. 1986. Technical Analysis of the Futures Markets. New York: New York Institute of Finance.

Pigou, Arthur C. 1929. Industrial Fluctuations, 2nd ed. New York: A.M. Kelley.

Samuelson, Paul. 1965. "Proof that Properly Anticipated Prices Fluctuate Randomly," Industrial Management Review 6: 41-49.

Schleifer, Andrei. 2004. Inefficient Markets: An Introduction to Behavioral Finance. New York: Oxford University Press.

Selgin, George A., and Donald J. Boudreaux. 1990. "L. Albert Hahn: A Precursor of Keynesianism and the Monetarist Counterrevolution," History of Political Economy 22, no. 2: 261-279

Shefrin, Hersh. 2000. Beyond Greed and Fear. Boston: Harvard Business School Press.

Shiller, Robert J. 2000. Irrational Exuberance. New York: Broadway Books.

--. 2010. Online Data. Available at http://www.econ.yale.edu/~shiller/ data.htm.

St. Louis Fed. 2010. Economic Data. Available at http://research.stlouisfed. org/fred2/series/AAA?cid=119.

Stowe, John D., Thomas R. Robinson, Gerald E. Pinto, and Dennis W. McLeavey. 2007. Equity Asset Valuation. New York: John Wiley and Sons.

Triana, Paolo. 2009. Lecturing Birds on Flying: Can Mathematical Theories Destroy the Financial Markets? New York: John Wiley and Sons.

Tocqueville, Alexis de. 1969. Democracy in America. George Lawrence, trans. Garden City, New York: Anchor Books.

Wicksell, Knut. 1936. Interest and Prices: A Study of the Causes Regulating the Value of Money. London: Macmillan and Co.

(1) Describing BF's position as one in which the human mind is said to be, "the servant of sub-rational forces" is no exaggeration. BF proponents often refer to the impact of cognitive biases as "systematic" rather than episodic or occasional (for example, see Barberis and Thaler, 2002, p. 11). The sheer number of biases and empirical deviations from the EMH put forward by BF advocates also suggest they view the influence of the irrational as being pervasive. BF advocates, too, resist the EMH description of the evidence mustered against market rationality as a collection of mere anomalies. Thus, Richard Thaler, a leading spokesman for BF, hopes for a future in which the discipline of finance will no longer be divided between the supporters and detractors of rational choice theory. "In their Enlightenment," he says, "economists will routinely incorporate as much "behaviour" into their models as they observe in the real world."(Thaler, cited by Bloomfield, 2010, p. 36). In other words, the irrational side of human nature will be acknowledged as a universal cause of human conduct.

(2) Pablo Fernandez (2001, pp. 2-3) surveyed equity analysts at Morgan Stanley Dean Witter and found that more than 50 percent used the P/E ratio, well above the slightly more than 30 percent applying the EV/EBITDA (Enterprise Value divided by Earnings before Interest, Taxes, Depreciation, and Amortization) metric. For a comprehensive treatment of share valuation can be found in Stowe, Robinson, Pinto and McLeavey (2007).

(3) What Hahn basically describes here is the distinction that Ludwig von Mises (1963, pp. 107-115) drew between class and case probability. Class probability refers to situations in which all the factors relevant to the production of numerous events sharing a set of characteristics are known. The probability that this set, or class, will occur can be mathematically calculated. Falling under this category are the chances of a seven arising from the throw of two dice. Case probability, by contrast, deals with circumstances in which some, but not all, of the causal variables are known and the event in question cannot be classified within a class. The event is unique and its probability is, therefore, not subject to a mathematical determination. Hahn, like Mises, places the game of investing under the heading of case probability.

George Bragues (gbragues@uoguelph.ca) is Assistant Vice-Provost and head of the business program at the University of Guelph-Humber, Toronto, Ontario.

Table 1. Stock Prices vs. Dividends, 1929-April 1956

Time Period Dividends Corporate Stock Markets
 Yields Prices vs. Model

1929-1932 Down Up Down Consistent
1933-1937 Up Down Up Consistent
 Up Flat Down Not Consistent
1942-1946 Flat Flat Up Not Consistent
1946-1949 Up Flat Down Not Consistent
1949-1952 Up Flat Up Consistent
Jan 1953-Sep 1953 Down Flat Down Consistent
1953-Apr 1956 Up Flat Up Consistent

Source: (Hahn, 1956, pp. 193-195)

Table 8. Stock Prices vs. Dividends and Yields, 1956-2010

Time Period Dividends Corporate Stock Markets
 Yields Prices vs. Model

Jan 1956-Jul 1957 Flat Up Up Inconsistent
Jul 1957-Oct 1957 Up Flat Down Inconsistent
Oct 1957-Dec 1961 Up Flat Up Consistent
Dec 1961-Jun 1962 Up Flat Down Inconsistent
Jun 1962-Feb 1966 Up Flat Up Consistent
Feb 1966-Oct 1966 Up Up Down Not Conclusive
Oct 1966-Nov 1968 Flat Up Up Inconsistent
Nov 1968-May 1970 Flat Up Down Consistent
May 1970-Jan 1973 Down Down Up Not Conclusive
Jan 1973-Oct 1974 Flat Up Down Consistent
Oct 1974-Nov 1980 Up Up Up Not Conclusive
Nov 1980-Aug 1982 Flat Up Down Consistent
Aug 1982-Aug 1987 Up Down Up Consistent
Aug 1987-Dec 1987 Up Up Down Not Conclusive
Aug 1987-Mar 2000 Up Down Up Consistent
Mar 2000-Sep 2001 Down Down Down Not Conclusive
Sep 2001-Jan 2002 Up Flat Up Consistent
Jan 2002-Oct 2002 Flat Flat Down Inconsistent
Oct 2002-Oct 2007 Up Down Up Consistent
Oct 2007-Mar 2009 Flat Flat Down Inconsistent
Mar 2009-Dec 2010 Down Flat Up Inconsistent