Leithner Letter No. 76
April 26, 2006

Let me start with a summary of my thesis. In the past, the speculative elements of a common stock resided almost exclusively in the company itself; they were due to uncertainties, or fluctuating elements, or downright weaknesses in the industry, or the corporation’s individual setup. These elements of speculation still exist, of course; but it may be said that they have been sensibly diminished by a number of long-term developments. ... But in revenge a new and major element of speculation has been introduced into the common-stock arena from outside the companies. It comes from the attitude and viewpoint of the stock-buying public and their advisers – chiefly us security analysts. This attitude may be described in a phrase: primary emphasis upon future expectations.

Benjamin Graham
“The New Speculation in Common Stocks”
The Analysts Journal (June 1958)

Most people get interested in stocks when everyone else is. The time to get interested is when no one else is. You can’t buy what is popular and do well.

Warren Buffett
Newsweek (1 April 1985)

Thierry de Montbrial, French foreign affairs expert, told me that this moment reminded him of a joke. Mikhail Gorbachev was once asked how – in one word – he would sum up the Soviet economy. “Good,” he said. Then he was asked how – in two words – he would sum up the Soviet economy. “Not good,” he said.

“A Healthy Dose of Reality”
The Age (2 February 2004)

On a cyclically adjusted basis, [Anglo-American stock markets are presently] as highly valued as in any period of the past 120 years, except the late 1920s and the late 1990s. What is needed to justify this condition is the bold assumption that [the] current exceptional profitability will endure forever. ... Large dangers of disruption exist. But markets are ignoring them. So we must recognise the danger not only that something will go wrong, but that markets will then multiply the needed corrections. Today, alas, the world is priced almost for perfection. Something close to perfection may indeed be with us in any year. But disappointment is ultimately certain.

“Readiness Should Be All in This Unknowable New Year”
The Financial Times (4 January 2006)

Heard the Bad News? The All Ords Has Breached 5,000

It was ubiquitous in the late 1990s and into 2000, disappeared briefly in 2001-2003 and has rebounded smartly during the past couple of years. And the events of the week of 20 March – when first the S&P/ASX 200 Index and then the All Ordinaries Index closed above 5,000 points for the first time – will probably intensify it. Sometimes it recurs at dinner parties, and other times beside the office water cooler or over the garden fence. Just as oceans feed a cyclone, tides of confident buyers trigger outpourings of this practice. You’ve probably noticed it. But what, exactly, is it? The rejoicing of your friends, colleagues and neighbours at the extent to which, during the past one, two or three years, the prices of their shares (or the units of their managed fund) have risen.

“Performance” – that is, the percentage change of the price of a particular security or market index during the past month, quarter, year, two years, etc.–is once more a prominent subject of conversation and even bluster; and expectations about performance during the next several years are again rosy. If the All Ordinaries Index can climb more than 80% in three years, then, say many performance-junkies, surely the future is bright. It must be bright because financial engineers, an alleged shortage of infrastructure and, above all, China’s voracious appetite for minerals will combine to make it so. All the more reason, then, to chase the “hottest” stocks and “best-performing” managed funds.

The Sunday Telegraph (26 March) told many people all they need to know. “Analysts expect the Australian share market to put on as much as another 10 per cent before the end of the year.” One analyst “believes the big market cap resource companies and banks ...will push the index towards 5,500. The fund managers are extremely focussed on a handful of stocks, whilst virtually ignoring the rest of the market. If you don’t own these big stocks, you are virtually a bystander.” Another analyst says things that “always [drive] the market higher, and that’s strong economic growth, low inflation and relatively low interest rates,” remain benignly and securely in place.

Leithner & Co.’s shareholders cannot easily participate in this chase of, and these conversations about, performance. Their results depend not upon the short-term levitation of market prices, but rather upon the flow of dividends that the Company’s investment operations generate. Further, and unlike a listed stock whose price fluctuates from one minute to the next (or a managed fund whose units are marked to market every day or week), “performance” does not readily apply to the shares of an unlisted private company. Even worse from the mainstream point of view, Leithner & Co.’s portfolio has never contained – and in the future is unlikely to hold – banks, infrastructure vehicles and major miners. Indeed, and most shocking to the mainstream, it consciously avoids what are generally regarded as “growth” investments. So when the chitchat at the dinner party, office water cooler or back garden fence turns towards investments’ recent short-term results, what can an investor – as opposed to an unwitting speculator – contribute?

Common sense and insight. But bear in mind that friendships are valuable, civility is a virtue and discretion is the better part of valour; accordingly, remember to be polite and gentle. Yet also stand your ground. Kindly but firmly, take pity upon and express your condolences to the friends and colleagues whose investments have “outperformed” during the past couple of years. Why on earth should you adopt this stance? Isn’t performance the raison d’être of investment, and outperformance a cause for celebration? Although torrents of advertising and legions of funds managers insist that it is, the truth speaks otherwise. In a variety of endeavours including investing, tortoises surpass hares. Alas, the growing confidence and even excitement of the past couple of years has encouraged people to think and act like rabbits (or should that be sheep?) rather than turtles. Fleeting stimuli like the headlines in today’s paper and tonight’s news flatter the mainstream of impulsive and credulous speculators. But enduring factors that appear infrequently in the mass media – that is, clear premises, explicit logic and hard evidence – ultimately and decisively favour the minority of dispassionate and sceptical investors.

Figure 1
Australia’s Terms of Trade (2003-04=100)
Historically High For How Much Longer?

Figure 1

It is very likely that, bolstered by and thus extrapolating from the “performance” of the recent past, the typical participant in financial markets is presently paying unduly high prices for stocks, bonds and real estate. Perhaps he does so because a vocal chorus of journalists, funds managers, brokers and strategists sings in robust unison that the prices of Australian financial assets, relative to the streams of income they emit, are on the whole quite reasonable. But the members of this chorus somehow forget to mention a very unusual set of circumstances – namely that Australia’s terms of external trade and level of company profits have attained levels not seen in many years. The prices in world markets of the things that Australians export, namely primary resources, are high and rising; and the prices of the things they import, particularly capital and consumer goods, are low and either stable or falling. Yet you should restrain your applause. Logically and historically, these conditions are transitory; and more often than not – Australia is, after all, primarily an exporter of primary resources – they have fallen (see, for example, Long-term Patterns in Australia’s Terms of Trade).

Figure 2
Profits’ (a) Share of Total Factor Income: Can It Grow Forever?

Figure 2

But don’t bother telling that to the bullish chorus. Rightly rejoicing at China’s embrace of capitalism and trade, it also celebrates the allegedly permanent impact its development is exerting upon the prices of primary resources. Alas, it forgets (or has never learnt) that during their post-war recoveries Germany, Japan, South Korea and Taiwan also demanded huge amounts of these resources, and that as a group they bulked much larger in the world economy than China does now (see also The Coming Rebalancing of the Chinese Economy by Stephen Roach).

Figure 3
Wages’ (a) Share of Total Factor Income: How Low Can It Fall?

Figure 3

(The source of Figures 1-3 is Australian National Accounts: National Income, Expenditure and Product, Dec 2005, released on 1 March. Residents of NSW, in particular, may wish to read it carefully: their state continues to skirt close to recession.) For exporters of mineral and agricultural exports, the early 1950s and 1970s were golden interludes. But what is the norm? Today’s bulls don’t want to know because this time, they implicitly insist, it truly is different. The bullish refrain also overlooks the reality that profits, too, exhibit a very strong tendency to regress to their historical average. It is admittedly very unlikely that profits and prices will revert suddenly and simultaneously. But they did within a few years of these two booms (see “Some Friendly Reminders for Australian Investors” in Letters 72-73). So heaven help Australians if they do so again. Given the cumulative impact of Australians’ various delusions – particularly the growing conviction that The Lucky Country is exempt from the laws of economics – the prices of securities may conceivably be twice as dear as the confident chorus maintains. And that is dear, indeed.

Are Most Australians Succumbing to the Growth Trap?

It is likely that, in order to generate the recent run of stellar “performance,” Australian market participants have concentrated their attention and lavished their funds upon “hot” companies, sectors and countries. Less than a decade ago, the Internet and telecoms technologies were all the rage. In the late 1990s, America’s “New Economy” straddled the investment world like a Colossus. Australia, in sharp contrast – Telstra Corp. Ltd being the major exception – was derided as an “Old Economy” backwater. But investment fashions, like terms of trade, can change quickly and with little warning. Today, Telstra has been banished to the furthest reaches of the investment doghouse; and yesterday’s pariahs are the market’s darlings. Similarly, the attitudes of the world’s biggest financial institutions towards this country have reversed. These days they admire Australians’ ability to engineer, leverage and repackage assets. They also laud the continent’s massive deposits of coal, iron ore and uranium, and confidently project that China and India will demand more of these things than Australians can possibly supply.

To focus upon “hot” companies, sectors and countries, as most market participants routinely do, is by definition to discount and perhaps even denigrate their less prominent and unfashionable counterparts. The trouble is that sane investors, like shrewd shoppers, strive to minimise the ratio of money expended to value received; moreover, over the long term the tried and true trumps the bold and new. In other words, it is the less exalted, fashionable and exciting – and hence more modestly priced – securities that reliably generate superior results for investors. Jeremy Siegel, the author of Stocks for the Long Run (McGraw-Hill Trade, 2003) and The Future for Investors (Crown Business, 2005) has coined the phrase “the growth trap” to encapsulate the fervent, incorrect and thus damaging belief that popular assets – i.e., the ones that presently command media headlines and lofty prices – deliver superior returns to investors.

The more Siegel has investigated the sources of investment results over long periods of time, the more evidence he has found that the growth trap afflicts not just individual securities but also sectors of the market and entire countries. Over the years and decades, Siegel finds, the most prominent, fastest growing and optimistically priced firms, industries and countries typically produce comparatively poor investment returns. The point is subtle but fundamental. Economic growth is vital, and for people as a whole it begets higher standards of living over time. But for investors, it is not growth alone that produces good results; rather, it is growth above the rates that market participant build into prices. Alas, the expectations people attach to “hot” investments are almost invariably excessively optimistic. Moreover, market participants usually take their sweet time to return to their senses. Accordingly, favoured countries, sectors and companies eventually disappoint, generate financial losses and create psychological pain. Then the next Sure Thing appears and the cycle begins anew. Siegel concludes that the growth trap is one of the stoutest barriers standing between investors and success.

Because it is paradoxical, the growth trap is dangerous. It prompts investors to pay unduly high prices for the very firms, industries and regions that drive innovation and spearhead economic expansion. It repeatedly encourages people to venerate the “New Economy” (i.e., motor cars and radio in the 1920s, transistors and television in the 1950s, computer chips and telecoms networks in the 1990s, etc.) and thereby to neglect well-established firms and mature industries. The mainstream’s relentless pursuit of “growth” – which manifests itself in the hype of “killer” technologies, the obsession with fast-growing countries and hence the stampede into “hot” securities and trendy market segments – occasionally rewards market participants with impressive short-term “performance.” Usually, however, the growth trap will eventually doom its victims to mediocre and even poor long-term results. The Great Bubble should, but apparently does not, provide a salutary lesson.

What, then, to do? If you concentrate your passions upon your personal life and hobbies, and if you restrict your diet in financial matters to cold logic and hard evidence, then you put yourself in a strong position to realise that remunerative investments dwell disproportionately among established, mature, boring and above all modestly-priced companies, stagnant or even shrinking industries and slower-growing or even sleepy countries (see also Peter Lynch, One Up on Wall Street, Fireside, 1989). Tomorrow’s returns are where today’s investors are not – and vice versa. In Siegel’s words, “our fixation on ‘growth’ is a snare, enticing us to place our [money] in what we think will be the next Big Thing. But the most [technologically] innovative companies are rarely the best place for investors. Technological innovation, which is blindly pursued by so many seeking to ‘beat the market,’ turns out to be a double-edged sword that spurs economic growth while repeatedly disappointing investors.”

IBM or Standard Oil of New Jersey?

To illustrate the growth trap, in Chapter 1 of The Future for Investors Siegel proposes a thought experiment. Imagine that you can undertake time travel and thus rely completely upon hindsight to make your investment decisions. This ability puts you in a remarkably lucrative position. You can travel into the past; arriving there, you are able to invest; knowing in advance what the future will bring, you invest with absolute certainty; and travelling back to the future, you effortlessly collect your substantial reward. Repeat this process a few times and a really tidy sum can accumulate.

As an example, imagine that a benefactor gives you $100,000, that you travel 50 years into the past and use the money to buy the shares of one or the other of two specified companies. Assume that you must buy and hold for fifty years, i.e., from 1956 to 2006, and that during this interval you must reinvest all of your dividends into the shares of whichever stock you buy. Also assume (mostly because the relevant data are readily available) that you travel to the U.S. and that you must choose between either IBM or Standard Oil of New Jersey (which, through various permutations, mergers and acquisitions over the years, has become ExxonMobil). After making your selection and buying $100,000 of one of these companies’ stock, you place your investment under lock and key for 50 years. Which stock, IBM or Standard Oil, should you buy?

Given that you can foresee events, upon which ones should you focus your attention? People tend strongly to ask: “which sector of the economy, technology or energy, will grow faster during the next 50 years?” As a traveller from the future, you know that in the mid-1950s tech firms in general and IBM in particular were poised for spectacular growth. You also know that, relative to the economy as a whole, energy companies faced decades of stagnation and relative contraction. Since the mid-1950s, information technology has advanced by unimaginable leaps and bounds, and its commercial and economic significance has increased vastly. As a percentage of the Standard & Poor’s 500 Index, the capitalisation of the tech sector increased from 3% in 1956 to almost 20% in 2006. In sharp contrast, although in absolute terms it has grown steadily, in relative terms the oil industry shrunk dramatically during these decades. Yes, it enjoyed an interlude of prosperity during the 1970s and is doing so again today; but it also idled during the 1960s, suffered a near-depression in the early 1980s and experienced doldrums of various extents during most of the subsequent quarter-century. Accordingly, oil and gas stocks (which comprised approximately 20% of the market capitalisation of all U.S. stocks in 1956) fell to less than 5% by 2000 (they subsequently rocketed to 6% in 2005). This shrinkage occurred despite the fact that nuclear power never remotely attained the position envisaged by its advocates in the 1950s and 1960s. Then and now, and for better or worse, oil and natural gas power much of the world.

Given your gift of perfect prescience, what other facts should inform your decision? In addition to industry growth figures, many people also rely upon three standard measures of a company’s growth (namely the growth of sales, earnings and dividends). Table 1, which summarises them, shows that IBM handily beats ExxonMobil along each dimension. Most notably, IBM’s earnings per share, which is probably Wall Street’s all-time favourite stock-picking criterion, grew a bit more than three percentage points per year more rapidly than Exxon’s. After fifty years, this seemingly modest margin generates a massive disparity of earnings. If IBM earned $1 per share in 1956, for example, and if its EPS grew at exactly 10.9% for each of the next 50 years, then it would earn $177 per share in 2006. In contrast, if ExxonMobil also earned $1 per share in 1956, and if its EPS grew by 7.5% per year for each of the subsequent 50 years, then it would earn only $37 per share a half-century later. Cumulatively, then, IBM’s earnings per share outstripped Exxon’s by a whopping 376%.

Table 1
Annual Growth Rates, IBM and ExxonMobil, 1956-2006

Growth Measure

IBM ExxonMobil Advantage
Revenue Per Share 12.2% 8.0% IBM
Dividends Per Share 9.2% 7.1% IBM
Earnings Per Share 10.9% 7.5% IBM
Sector Growth** 14.7% -14.2% IBM

** change in market share of technology and energy sectors, 1957-2003

As a time traveller to 1956, then, you would know that as a “growth” stock IBM beats ExxonMobil hands down. If you had known these particulars in 1956, in which company would you have invested? If like most people your answer is “IBM,” then you have squandered your lucrative advantage and succumbed to the growth trap. Although both companies generated good results for their owners, an investment in ExxonMobil earned an average of 14.4% per year from 1956 to 2006. In compound percentage terms, that is more than half a percentage point better than an investment in IBM (13.8%). But when you open your lockbox, the difference expressed in dollar amounts is staggering. If in 1956 you invested $100,000 in Big Oil, then the market capitalisation of your grubstake would today be $83.4 million. If you invested it in the Big Blue of the tech world, then its market cap would grow to “only” $64.1 million. That’s $19.3 million and 30% fewer dollars. As in Aesop’s Fable, so too in investing: the tortoise bests the hare.

It’s the Valuation, Silly

To choose IBM over ExxonMobil on the basis of these two pieces of information, namely macro (industry) and micro (company) rates of growth, is clearly to overlook other and much more important information. If Big Oil fell short of Big Blue in every category of growth, then why did Exxon accumulate so many more investment dollars than IBM? On simple reason: valuation – that is, the price originally paid for these companies’ earnings, the amount of dividends received and the price paid to buy more shares with dividends. Compared to ExxonMobil, the price paid for IBM’s stock was, year after year, simply too high; similarly, its dividends were too paltry. Even though the computer giant’s objective growth handily trumped Exxon’s, the subjective valuation of Exxon was much more compelling than IBM’s; and when all is said and done – and as Benjamin Graham knew in the 1930s but most people have since forgotten (or never learnt) – valuation determines investors’ ultimate (that is, accumulation of dollars) results.

Table 2, which is also derived from Siegel’s book, summarises these companies’ average valuation during these years. The price-to-earnings ratio, the price paid for $1 of a company’s earnings, is Wall Street’s fundamental yardstick of valuation. Another, once widely regarded in the U.S. but sadly much less prominent these days, is a stock’s dividend yield (i.e., the dollar amount of dividends received during a year divided by the price of the shares). Exxon’s average price-to-earnings ratio during the 50 years to 2006 was less than half of IBM’s. To buy $1 of IBM’s earnings during these years cost an average of $26.80; $1 of Exxon’s earnings, on the other hand, cost just $13.00. Each dollar used to buy a share of Exxon during these years thus bought twice as many dollars of profits as a dollar used to buy a share of IBM. Similarly, ExxonMobil’s average dividend yield was more than three percentage points higher than IBM’s. Each dollar used to buy an Exxon share thus bought more than twice as many dividends (an average of $0.052) as a dollar used to buy an IBM share ($0.022).

Table 2
Average Valuation Measures, IBM and Exxon, 1956-2006


IBM Exxon Advantage
Average P-to-E Ratio 26.8 13.0 Exxon
Average Dividend Yield 2.2% 5.2% Exxon

Why does valuation matter so much? Assume by two criteria that capital compounds over time. That is, (1) a company reliably earns a profit, pays a dividend and sensibly reinvests those earnings it does not pay to its owners as dividends, and is thus able to compound its earnings over time; and (2) an investor buys shares of such companies, prudently reinvests the dividends he receives from the shares he owns, and thereby compounds his capital over time. Under these conditions, the greater the amount of earnings and dividends that $1 invested today purchases (i.e., the lower the price and hence the better the valuation of the company’s stock), the more the investment’s capitalisation grows.

Because the price of ExxonMobil’s shares was relatively low and their stream of dividends was comparatively generous, people who bought its stock in 1956 and then reinvested their dividends every year for the next 50 years were able to accumulate almost fifteen times their starting number of shares. In sharp contrast, people who bought IBM’s stock were penalised by its initially and subsequently high price and measly dividend. Because its dividend was meagre, they had little cash with which to buy more shares; and because the price of these shares remained dear, they were able to exchange their cash for relatively few additional shares. As a result, people who bought IBM’s stock in 1956 and reinvested their dividends eventually accumulated fewer than three times their original number of shares. Fortunately, the price of ExxonMobil’s stock appreciated almost three percentage points a year less quickly than the price of IBM’s stock. Big Oil’s higher dividend yield generated more cash with which to buy more shares, and the shares’ moderate price enabled buyers, year after year, to increase the number of shares they owned relatively rapidly. The elapse of time and the power of compounding then worked their wonders.

Heard the Good News? Valuation Trumps Growth

To generalise this result, Table 3 summarises the progress of four hypothetical investments that commenced in 1956 and concluded in 2006. In order to improve their comparability, I have standardised them such that at the beginning of 1956 the shares of each sold for $1.00. The table’s top half shows the number of shares accumulated at particular junctures; and its bottom half shows each investment’s capitalisation (i.e., number of shares owned multiplied by the market price per share) at those junctures.

IBM, remember, is a “growth” investment in the sense that its earnings per share and dividend per share grow at relatively rapid average rates (10.9% and 9.2% per year respectively). Probably because IBM grows so quickly, Mr Market bids the price of its stock to a very high level (equivalent, on average, to almost 27 times its EPS). Exxon Mobil, on the other hand, does not attain the exalted status of “growth” investment. Its EPS and DPS grow at average rates of 7.5% and 7.1% per year respectively. Underling their disparate rates of growth, note that IBM’s dividends grow more rapidly than Exxon’s earnings. Perhaps because its earnings grow at a more modest pace, market participants price Exxon’s stock more moderately (i.e., at an average of 13 times its EPS). And perhaps because the petroleum industry is growing slowly and shrinking relative to others, Exxon is able to locate few opportunities to invest its earnings. Hence it returns an average of almost 95% (7.1/7.5 = 0.95) of its earnings to shareholders as dividends.

Table 3

No. of Shares Accumulated from Reinvestment
of Dividends
IBM ExxonMobil X Ltd Y Ltd
1956 102,201 105,195 106,500 101,000
1966 124,869 176,197 197,459 108,622
1976 148,290 289,740 358,162 112,854
1986 171,853 468,057 636,041 115,142
1996 195,027 743,261 1,106,657 116,361
2006 217,376 1,160,903 1,887,861 117,005
1956 $102,164 $105,300 $106,500 $101,003
2006 $351,249 $349,530 $406,969 $337,372
1976 $1,173,775 $1,184,617 $1,521,418 $1,088,653
1986 $3,827,786 $3,944,150 $5,568,514 $3,449,723
1996 $12,223,640 $12,908,640 $19,968,780 $10,827,720
2006 $38,338,280 $41,554,660 $70,209,090 $33,815,530
Rate of Return
12.6% 12.7% 14.3% 12.0%

Unlike IBM and Exxon, X Ltd is an imaginary company – but one with attributes that are realistic and attractive enough to prompt investors like Leithner & Co. to search high and low and long and hard. Like ExxonMobil, X Ltd is no “growth” investment. Indeed, like Exxon its EPS and DPS grow at 7.5% and 7.1% respectively per year. Yet in one critical respect, X differs from Exxon (and even more from IBM). Perhaps because it is never the flavour of the month or year, and because market participants never attach high expectations to it or laud the industry in which it operates, and perhaps because they never pay much attention to it at all, market participants depress the price of X’s stock to a level equivalent, on average, to just 10 times its EPS. Per dollar of earnings, in other words, the shares of X Ltd sell for roughly one-third of IBM’s. Finally, like X Ltd, Y Ltd is an imaginary company – but one that unfortunately has counterparts on the ASX. Unlike X, however, Y’s attributes are so repellent that Leithner & Co. runs from it as quickly as its legs can carry it. Even more than IBM, Y Ltd is a “growth” investment: its EPS and DPS grow at 12.0% and 7.5% respectively per year, its dividend yield is a mere 1% and its PE ratio is a lofty 32.5.

Table 3, then, summarises the progress of one investment, Y Ltd, whose stock pays the meanest dividend and sells at the highest multiple of its earnings; a second, IBM, whose stock pays a somewhat less meagre dividend and sells at somewhat less exalted price; a third, Exxon, whose stock pays a reasonable dividend and sells for a moderate price; and a fourth, X Ltd, that consistently pays the highest dividend (relative to the market price of its stock) and sells for the lowest price (relative to the company’s EPS). If the ultimate objective is to fructify the initial amount invested to the greatest possible extent, which of these companies is the most attractive?

The top half of the table provides a vital clue. Regardless of the duration of time, short-term or long term, one year or 50 years, the reinvestment of X Ltd’s dividends generates the largest number of shares. After 50 years, this reinvestment increases one’s shareholding almost 18-fold. The reinvestment of Exxon’s dividends, on the other hand, increases one’s number of shares 11-fold; the reinvestment of IBM’s merely doubles one’s original stake in the company; and the reinvestment of Y’s dividends barely budges the number of shares owned. Does this vastly different ability to increase the number of shares owned affect the capitalisation of one’s investment? The lower half of the Table shows that it does.

The shares of X Ltd sell for ten times its EPS, and after 50 years its EPS has grown from $0.10 to $3.72. At ten times earnings, that’s a share price of $37.19; multiplied by 1,887,861 shares, that’s an end capitalisation of $70.2 million. After 50 years, Exxon’s EPS has grown from $0.077 to $2.86, its the shares sell for $35.80; and multiplied by 1,160,903 shares, that’s a final capitalisation of $41.6 million. IBM’s concluding capitalisation is $38.3 million and Y Ltd’s is $33.8 million. For any investor worthy of the name, this result is fundamental. The ugly duckling, relative straggler and unloved company (in terms of its growth and valuation), X Ltd, accumulates $28.7 million extra (that is, 69% more) dollars than Exxon. It also amasses $31.9 million (that is, 83% more) dollars than the “growth stock” IBM, and $36.4 million extra and more than twice as many dollars as Y Ltd. By this criterion, X Ltd is clearly the best investment. More generally – that is, over both “short” intervals like one year and extraordinarily long ones like 50 years – reasonably run and modestly priced companies trump world-beaters commanding nosebleed prices. Notice that every stage of this race, the quiet and unfashionable turtle trumps the flashy and trendy hare. So mark these words: low prices (relative to earnings and dividends) and the reinvestment of dividends – and not high prices and short-term “performance” – make good investments.

Sharp-eyed readers will notice that the ending capitalisations of IBM and Exxon stemming from Table 1 ($64.1 million and $83.4 million respectively) differ from the final capitalisations in Table 3 ($38.3 million and $41.6 million). Why the disparities? The figures derived from Table 1 are actual historical results. Those emerging from Table 3, on the other hand, are stylised. They assume, for example, that IBM sold for exactly 26.8 times its earnings every year for 50 years, and that its earnings grew at precisely 10.9% per year, year after year, etc. Obviously, this assumption does not correspond to reality. Why, then, are the end capitalisations derived from Table 1 greater than those in Table 3? Particularly in the first half of the period under consideration, IBM often sold for less than 26.8 times its earnings (and Standard Oil for less than 13 times its earnings). During these early years, earnings and dividend yields were therefore higher than the 50-year averages. Given this early boost, both IBM and Standard Oil placed more cash in their owners’ hands, and owners who reinvested their dividends were able to buy more shares per $1 of dividend.

The fundamental point, however, remains unaffected. Valuation matters so much because the better the valuation – that is, lower a security’s price relative to its earnings and dividends – the more substantial is the impact of the reinvestment of dividends. Companies must successfully reinvest any earnings not paid as dividends, and owners of shares must successfully reinvest their dividends. If companies and shareholders can do so these two things consistently over the years, then the stage is set for the passage of time and the power of compounding do the heavy lifting. Table 4, which disaggregates the stylised returns of these four potential investments, makes this explicit. IBM’s total compound return, for example, is the sum of a share price appreciation component (10.4%) and a dividend component (2.2%). The table shows that the more compelling the investment, the more significant is its dividend as a source of its total return. Dividends comprise almost half (6.5/14.3 = 45%) of X Ltd’s annualised compound return. They comprise a bit less (41%) of Exxon’s, 17% of IBM’s and only 9% of Y Ltd’s.

Table 4
Four Investments’ Sources of Stylised Returns, 1956-2006

Source Y Ltd IBM Exxon X Ltd Advantage
Appreciation of Share Price 11.0% 10.4% 7.5% 7.8% Y Ltd
Dividend 1.0% 2.2% 5.2% 6.5% X Ltd
Total Compound Return 12.0% 12.6% 12.7% 14.3% X Ltd

The Growth Trap and Today’s Unrealistic Expectations

In 2000-2003, when the Great Bubble burst and many investors’ results were preceded by minus signs, one reassuring notion helped to console the despondent. Over the years, the unrepentant reminded the disconsolate, the annualised compound return of Australian stocks has averaged approximately 10%. The unapologetic also insisted that all stocks always reward patient people. How to interpret this base rate? As if it were a passage in the Bible that promises a life span of three score and ten years. Investment base rates and life expectancies, in other words, are rough guidelines that describe a broad pattern: they are not invariable rules that apply to every time and place. Bearing in mind the hazards and follies of prediction, it is also sensible to expect that, going forward, stocks’ average return will fall short of this base rate. The “outperformance” of the past couple of years, in other words, has created stiff headwinds that might take years to overcome. To evaluate this prospect, it is important to understand both the specific process that generated the stylised results in Table 2 and the general process that has produced the 10% base rate.

From 1954 to 2004, the All Ordinaries Index generated an annualised compound return of roughly 11%. In the U.S., valid and reliable data extend further into the past. From 1926 through 2004, the Standard & Poor’s 500 Index (the American counterpart of the All Ordinaries Index) generated an annualised compound return of almost 11%. Given that Australian government paper yielded ca. 7% and U.S. Treasuries closer to 5%, stocks seemed to be the preferable asset class (see also Chap. 10 of Chris Leithner, The Intelligent Australian Investor, John Wiley & Sons, 2005).

But before drawing this conclusion, consider the components of these overall results. According to Ibbotson Associates, 4.6 percentage points of the S&P’s compound rate of return from 1926 to 2000, which they estimate at 10.7%, derived from the receipt, growth and reinvestment of dividends. If as a thought experiment you regard the S&P 500 as a stock, then in terms of the figures in Table 4 the “dividend” exceeds IBM’s but falls short of Exxon’s. Another 3.1% of the S&P’s long-term rate of return stems from the wind that the CPI allegedly blows into the sails of corporate profits. Dividends and the supposedly salutary effect of price inflation thus delivered more than 70 cents of every $1 produced by American stocks in the 75 years to 2000. The remaining three percentage points of the total compound return – that is, “performance” – can be disaggregated in several ways. In research published in 2002, Roger Ibbotson and Peng Chen described several methods. One estimates that the CPI-adjusted growth of corporate earnings is 1.75% per year. By this calculation, the remaining 1.25 percentage points of total return comes from increases in stocks’ average price-to-earnings ratio. (But this estimate of 1.75%, it is worth noting, may be generous. According to Robert Arnott of First Quadrant, the overall rate of growth of real per-share earnings and dividends in the U.S. since 1965 has been – a roll of the drums, please – little more than zero.)

Other analysts have otherwise specified how the ownership of stocks has rewarded investors. But the essence is clear. Dividends and CPI both matter most, and earnings growth and changes in stocks’ valuation are significant but distinctly less important. The trouble for the confident people at dinner parties, office water coolers and back garden fences is that the biggest benefits enjoyed by their predecessors, namely fat dividend yields and attractive prices, are presently absent.

Today (20 March), the All Ords’ dividend yield is 3.7% (that’s three quarters of its long-term historical average) and 10-year Commonwealth bonds yield 5.3% (which is also well below its average). Never mind the breezy comments in that day’s Australian Financial Review: this disparity of yields disadvantages stocks. The CPI, it seems to me (and still bearing in mind the follies of prophesy!), will average at least 3% in the future. Alas, Australian equities have generated much of their edge over bonds during periods when the CPI exceeded its fifty-year average (6.3%). Why? A galloping CPI punishes corporate earnings but absolutely shreds the purchasing power of fixed-coupon bonds. Another factor that once put much wind into stocks’ sails has now abated. At the beginning of the Australian series of data in 1954, stocks’ average price-to-earnings ratio was roughly 11 (that is, their earnings yield was 9% and the dividend yield was ca. 7%) – considerably more attractive than the yield of government bonds (3.6%). Accordingly, the P-to-E ratio could rise significantly before the attractiveness of stocks relative to bonds disappeared. And lo, from the mid-1950s until the early 1970s stocks did indeed rise smartly.

Never mind, say the Australian market’s many boosters, that corporate profits are presently at generational highs; and be glad that the All Ords’ current P-to-E ratio (15.5) is undemanding. The growth of earnings, many boosters will insist, can continue to accelerate because financial engineering, China’s insatiable appetite for Australian mineral and energy resources, and perhaps a dash of enlightened management and sheer luck, will combine to boost them. The problem is something that the bulls seldom acknowledge: the rate at which corporate earnings grow has been remarkably steady over long spans of history. Further, spurts of growth – like the one experienced since 2003 – are usually followed by much weaker rates of growth (or even decreases of profitability).

Improbable Expectations Sow the Seeds of Disappointment

So sure, stocks can compound at 10% a year – provided that the various parts of the process that generates this base rate do not depart significantly from their historical averages. Alas, these days several of these parts are exceeding their averages. Hence my suspicion that, going forward, this base rate is optimistic. Still, many people rely upon the historical record in order to persuade themselves and others that, because stocks have generally been good things for so long, they’ll always remain so. But it is precisely when its components are out of kilter, or when a spanner is unexpectedly thrown into the works, that the past becomes a poor predictor of the future. William Bernstein, an investment adviser and author, cites an epochal example. “For the 50 years from 1932 to 1981,” he has said, “the return [of U.S. Treasury bonds] was only 2.95%, almost a full percentage point less than the [CPI] of 3.8%. Yet in 1981 common sense dictated that the bond yield of 15% was more predictive of its future return than was the historical data.” During the next quarter-century, 10-year Treasuries generated an annualised compound return of more than 13% – every bit as good but not nearly as widely known as stocks.

A high “risk free” yield, like the one available in the early 1980s, pointed rather clearly to a relatively high long-term rate of return; and a rather low (and riskier) dividend yield, presently 3.7% in Australia, suggests (albeit less clearly) to a lower base rate of return. So what might Australians reasonably expect from Australian stocks, considered as a whole, in the next 5-10 years? Following the logic of Ibbotson Associates, we start with today’s dividend yield, add to it 3% for the CPI, 1.8% for real earnings growth and grant no aid from a higher price-to-earnings ratio. That implies a base rate of 3.7 + 3 + 1.8 = 8.5% per year – which falls short of the historical average and the stellar “performance” of the past couple of years. If this new base rate emerges, it might sorely disappoint the discussants at today’s dinner parties, office water coolers and back garden fences.

In short, abnormally high short-term “performance” is not good news for investors worthy of the name because it depresses the results that they might otherwise expect over longer periods. Another prospect – exemplified by the infamous “Dow 5,000” prediction of Pimco’s Bill Gross, the manager of the world’s biggest bond fund – is that the prices of stocks must sink dramatically and quickly in order to restore the historical base rate of return. A third and less disconcerting possibility is that investors have misinterpreted the base rate. The figure of 10%, in other words, signalled that during most of the 20th century, particularly its first half, the prices of stocks were reasonably close to justifiable estimates of their values. If so, and if today’s valuations are indeed stubbornly high, then Australian investors must be willing to accept more modest results. Perhaps an average of 8-9% a year is reasonable when 10-year Treasury bonds yield 5.3%

But will this average result satisfy people whose expectations, based upon their experiences since 2003, are probably significantly more ambitious? What happens if and when the investment world’s thermostat, the yield of allegedly “risk free” U.S. Government bonds, rises? The bottom line, by my way of thinking, is that stocks like the stylised “Exxon” and X Ltd will be decent propositions over the years. But “IBM” and Y Ltd will be growth traps: their prices exceed their values, and market participants’ attitudes towards them will eventually change from excitement to disillusionment. Henceforth from that point, their owners will experience either mediocre results or outright losses. Telstra, whose underlying economics have changed surprisingly little since the Great Bubble, provides an example – and a warning. Given that a mere dozen securities concentrated within two sectors have underwritten more than half of the All Ords’ meteoric rise since 2003, it is not fanciful to think that disappointment triggered by mean regression (or who-knows-what-X-factor) can during e next several years beget either a stagnant or a sharply lower Index.

As food for thought, consider two prescient headlines fromThe Wall Street Journal: “Dow 10,000 Means It’s Time to Prepare for the Hangover” (23 March 1999) and “Forget the Party Hats: Why Dow 10,000 Should Be No Cause For Celebration” (10 December 2003). It is worth noting that from 23 March 1999 to 23 March 2006, the Dow Jones Industrial Average returned a whopping 1.6% compound per annum (1.9% including dividends). During these years, many cash deposits – the slowest of the investment world’s tortoises – have outpaced the Dow. Is that startling enough to quell the present excitement at Australian dinner parties, office water coolers and back garden fences? Not when confidence rules the roost and the herd is stampeding. But it is precisely during unusual times – in the sense that expectations are unusually high or low – that perspective, evidence and reasoning are most valuable. With this downward revision to the long-term base rate in mind, during the week of 20 March I mentally rewrote the jubilant Australian headlines to read “The All Ordinaries Index at 5,000 Provides Grounds Not to Grin But to Chagrin.”

Chris Leithner


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