Our fears are always more numerous than our dangers.`
— Lucius Annaeus Seneca
Roman Philosopher (c.4 BC-65 AD)


The world’s economies are in a poor state. The three largest markets — namely the United States (US), Europe and Japan are in recession — and the outlook for 2009 is gloomy. Starting in mid-2007, policy makers in these (and other) regions have given effect to a series of monetary and fiscal policies to repair investor confidence that has become increasingly shriveled as the events of 2008 have unraveled. Indeed, in some cases, the policies adopted take the form of extreme action, with the US Federal Reserve Bank’s Open Market Committee having taken the decision effectively to cut the key lending rate to zero percent in mid December. This action, if anything, illustrates the depth of the financial crisis as well as the degree to which policy makers are willing to go to restore economic health.

Still, the extent of the economic fallout and the depth of the financial crisis have left many investors anxious about the outlook for capital markets. Further, despite the sharp falls that have taken place in share prices, concern remains that markets may not have fully discounted the impact of the economic fallout on company earnings. In periods of such anxiety and turbulence it is useful to be reminded that, in the fullness of time, hard facts triumph over shrill sentiment. This is not a novel insight. To be sure, in the world of investment finance the point was first made more than 70 years ago, in the wake of the stock market crash of 1929, when Columbia University professors Benjamin Graham and David Dodd urged investors to focus on hard facts — including a company’s past earnings and the value of its underlying assets — rather than trying to guess what the future would bring. In fact, the argument developed by Graham and Dodd was deceptively simple: a company with a strong profit history and a relatively low stock price is probably undervalued.

Graham and Dodd developed this argument in their classic 1934 textbook, Security Analysis, in which the authors reminded investors to focus on long-term trends and not to get caught up in the moment. Unfortunately, this point has been lost on today’s participants in capital markets. For instance, when analysts talk about relative value of investments, the default metric is the price-earnings ratio which, in its most common form, is measured as a company’s stock price divided by its earnings per share over the last 12 months.

Leaving myopia aside for the moment, a price-earnings ratio tells you how much a stock costs relative to a company’s recent earnings performance. All else equal, the higher the ratio, the more expensive the stock — and, according to value investors such as Graham and Dodd, the stronger the argument that the investment will not do well going forward.



Bull markets are born on pessimism, grow on skepticism, mature on pessimism, and die on euphoria. The time of maximum pessimism is the best time to buy, and the time of maximum optimism is the best time to sell.
— Sir John Templeton, February 1994


Right now, the stocks in South Africa’s FTSE/JSE All Share Index (ALSI) have an average price-earnings ratio of about 9.4 times which, by historical standards, is quite low. Since 1980, for instance, the average trailing price-earnings ratio has been 13.1 times (see Figure 1). During the emerging market bubble of the mid-1990s, however, the ratio rose to almost 25 times. By contrast, the depressed economic environment and political uncertainty that shrouded the mid 1980s resulted in a price-earnings ratio equal to about six or seven times trailing earnings. Read against this backdrop, the JSE’s current price-earnings ratio suggests that as long as earnings are able at least to tread water through the financial crisis, then stock prices do not have much further to fall. Many forecasters, however, suggest that near-term earnings will be under pressure which would justify further stock price falls and so vindicate investors’ current aversion to equities.

Source: Datastream and Cannon Asset Managers

Whilst Graham and Dodd considered the price-earnings ratio a crucial measure, they would have had at least two problems with the above conclusion and its reliance on the simple price-earnings metric. First, the men would have advised investors against focusing on forecast earnings, to derive a forward price-earnings ratio (this matter is not dealt with further in this note; suffice it to state that our ability to forecast has not been evidenced).

Second, Graham and Dodd cautioned against putting too much emphasis on the recent past (or near future, for that matter). They realised that a few months, or even a year, of financial information could be deeply misleading. Certainly, there is a strong case that can be made in the current environment that reason has been thrown out of the window in favour of emotional reaction. In this vein, the men argued that a price-earnings ratio based on one year of data would say more about what the economy was doing at a particular moment than about a company’s long-term prospects.



I have no use whatsoever for projections or forecasts. They create an illusion of apparent precision. The more meticulous they are, the more concerned you should be. We never look at projections, but we care very much about track records.
— Warren Buffett, September 1995
Chairman of Berkshire Hathaway Inc.


Thus, Graham and Dodd argued that price-earnings ratios should not be based on only one year’s worth of earnings. It is much better, they wrote in Security Analysis (Graham and Dodd, 1934, 452) to look at profits for ‘not less than five years, preferably seven or ten years’. This advice has been largely lost to history. For one thing, collecting a decade’s worth of earnings data can be time consuming. It also seems a little strange to look so far into the past when your goal is to predict future returns.

However, some analysts have remembered the advice of Graham and Dodd. In the case of the US equity market, John Campbell and Robert Shiller (of Irrational Exuberance fame) have been calculating long-term price-earnings ratios for many years. When they were invited to a make a presentation to Alan Greenspan in 1996, they used the so-called Graham and Dodd price-earnings ratio based on the past 10 years’ earnings to argue that stocks were greatly overvalued. The calculation of this ratio is relatively straightforward: earnings for the previous 10 years are inflated using the consumer price index to current prices. From this an average earnings figure is calculated. This allows current price to be divided by average real earnings, where it is argued that, by virtue of the fact that the earnings figure is derived from 10 years’ data and not a single year’s data, the resultant price-earnings ratio gives a more sober result and so a more sensible perspective on value.

Mathematics aside, a few days after Mr Campbell and Mr Shiller had presented their view, Mr Greenspan touched off a brief worldwide sell-off by wondering aloud whether ‘irrational exuberance’ was infecting the markets.



Nothing is easier than self-deceit. For what each man wishes, that he also believes to be true.
— Demosthenes
Greek Statesman (384 BC-322 BC)


Today, the Graham-Dodd approach produces a sharply different picture from the one that Wall Street offered at the turn of the century when Campbell and Shiller were most vocal. As shown in Figure 2, based on average profits over the last 10 years, the Graham and Dodd price-earnings ratio has fallen to 15 times versus the average of 16.4 times recorded since 1881. As an aside, it is worth noting that the ratio has been as high as 44 during the bubble years of the late 1990s and as low as five during the Great Depression. Further, in the past 25 years the ratio has averaged around 23.3 (or 19.7 times if we ignore the bubble years). Given this backdrop, using Graham and Dodd’s long-term valuation tool, equities in the US are at their cheapest level in a generation.

Source: Adapted by Cannon Asset Managers from www.irraitonalexuberance.com

However, just because equities are cheap, it does not mean they will not get cheaper. Prices or earnings (or prices and earnings) could fall from current levels despite the concerted policy efforts noted earlier. That said, it must be stressed that whilst valuations do not really matter for short-run returns, they are a primary determinant of long-run returns. As evidence of this, Figure 3 shows the average 10-year real returns achieved in the case of the S&P500 over the past 125 years depending on the starting Graham and Dodd price-earnings ratio.

Source: Adapted from Montier (2008)

The current ratio of 15 in the US puts the market in the third column from the left, or the second cheapest quartile. This suggests that over the next 10 years investors in the S&P500 should not expect returns that flow from bargain-basement levels (ratios of less than 12), but that they can expect above-average returns. This is an interesting result given the extremely cautious stance currently being taken by investors with regard to US equities.


Unfortunately, poor data availability in South Africa shortens the period over which a Graham and Dodd price-earnings series can be generated as reliable data are available only from 1980. Although this does not undermine the strength of the tool, to increase the number of observations, we use a seven-year Graham and Dodd price-earnings ratio in the case of the domestic market to inform our view on value. This caveat aside, by bringing the Graham and Dodd tool into the South African setting, useful insights are offered into the prospect for domestic equities given the gloomy global backdrop.

Considering the evidence, it is interesting to note that the average Graham and Dodd price-earnings ratio for the South African market since 1980 is 16.7 times. Coincidentally, this accords closely with the long-term average for the US. However, South Africa’s maximum of 26.3, observed during the recent resource boom, is substantially lower than the US’s maximum of 44 observed during the dot.com era. Also, our minimum of 8.8 is higher than the US’s minimum of 4.8. This result probably is explained by the fact that the US observation occurred during the Great Depression; the South African data set only extends back to 1980. Interestingly, if we truncate the US data set to run from 1980 to present, then the minimum reading in the US is 13.3. This result suggests that South African equities have experienced periods of being particularly cheap compared to US equities since 1980. One such example would be the period immediately after the 1987 crash, when the South African minimum of 8.8 was observed whilst the US market reading was about 60 percent higher at that time.

Leaving history aside, it is instructive that the current South African reading of 13.2 times is well-below recent peaks, and about 20 percent below the long-term average. This simple result suggests that, despite the gloomy mood, the domestic equity market is attractively priced. The full South African series is shown in Figure 4 below.

Source: Cannon Asset Managers from Datastream

What then of the prospect for equity investors? Figure 5 shows the average one-year, five-year and seven-year real returns achieved in the case of the JSE over the past 20 years depending on the starting Graham and Dodd price-earnings ratio. Quartiles are used to sort the starting ratio, with quartile one capturing the highest price-earnings ratios (expensive markets) and quartile four capturing the lowest price-earnings ratios (cheap markets) .

Source: Cannon Asset Managers

The current ratio of 13.2 puts the South African market in the set of columns furthest to the right, or the fourth quartile, representing a cheap market. Based on historical performances, a starting Graham and Dodd price-earnings ratio in the cheapest quartile correlates with an average annual real return of 20.0 percent over the next year, 13.1 percent per annum over the next five years and 31.6 percent per annum over seven years. In short, from a future returns perspective, there is no better starting point. In other words, South African equity valuations currently are in bargain basement territory.



If stocks are attractive and you don't buy, you don't just look like an idiot, you are an idiot.
— Jeremy Grantham, October 2008
Chairman of Grantham, Mayo, Van Otterloo & Co


Despite these potentially highly attractive returns, investors remain shy of equities. Instead, economic and market uncertainty means that investors prefer the hypothetical safe havens of cash or government bonds. There might be some sense in this stance. As already noted, just because equities are cheap, does not mean they cannot get cheaper. Based on historical analysis, the worst performance equities have given over one year when ratings have started in quartile four is minus 33.5 percent in real terms. This would be an awful result to add to the sharp decline recorded in 2008 and would vindicate the current equity-shy stance. Arguably, however, this is a worst case scenario as the average real return over one year from such a starting valuation is 20.0 percent, whilst the greatest real return recorded over one year is 63.0 percent. On a one-year stance, then, investors could anticipate a return that ranges by as much as 100 percentage points (63.0 percent to minus 33.5 percent).

But when we move beyond a one-year view, and adopt a stance that Graham and Dodd would agree with, returns become more attractive and less wide ranging. For instance, when the Graham and Dodd ratio has been in the cheapest quartile, the average real return over five-years is 13.1 percent whilst the minimum is minus 1.2 percent and the maximum is 44.4 percent. Under this scenario, the likelihood of earning negative real returns is low (about a two-in-fifty chance, or less than five percent likelihood).

The investment argument favouring equities is reinforced when returns over seven years are considered. When equities are priced in the cheapest Graham and Dodd price-earnings bin, the average annual real return over seven years is 31.6 percent. Under this scenario, the value of a portfolio increases more than seven-fold in inflation-adjusted terms over the period. The minimum real return equals 10.9 percent, whilst the maximum real return is 47.6 percent.

Viewed against this canvas, it is difficult to escape the conclusion that South African equities are in bargain-basement territory and represent compelling value for investors who have the temerity to push aside the gloomy forecasts for 2009 and look beyond the noise of today.

7. Sources:
Shiller, Robert J. (2000) Irrational Exuberance. Princeton: Princeton University Press.
Montier, James (2008) The Road to Revulsion and the Creation of Value, Mind Matters, Societe Generale, 25 November 2008.
Graham, Benjamin and Dodd, David L. (1934) Security Analysis: Principles and Techniques. New York: McGraw Hill.