Sticky Luck

Nine:

Why the Market-Expected Return

Matters

The problems of pursuing superior returns and attempting market timing.  The illusive expected term of the market's expectations.  Reasons why it's probably less than two years.  No bounce-back mechanism built into the market.

Earlier, we talked about how the emergence of underlying reality, in terms of data that carries a probability signal, can lead to lengthy market trends with the constant possibility that the trend will break and take a steep dive or rise.  In the last chapter, we discussed a model which showed how all these probabilities can balance out to a single, common expected return.  A probable good return can be balanced by the possibility of debacle; a probable poor return can be balanced by the possibility of the jackpot, and so on.  The variety of combinations can create any manner of price-chart shapes, from the NASDAQ run of the last forty years to the Dow during the Great Depression through World War II. 

The reason why all market-expected returns are the same is because the market constantly adjusts prices to give that result.  The reason why this single return is equal to the risk-free return is because that is the one return we can calculate with certainty.   The reason why no investor can be expected to beat the market is because they cannot know more about the past than the market knows, and because they know nothing about the future that is not a personally chosen projection of the past.

We define the market view as the aggregate of the market’s buy/sell decisions weighted by money transacted.  The sheer scale of data analysis backing the big-money decisions makes any one person’s analysis look very small.  Moreover, the complete market is a money-weighted balance of all conflicting perspectives.  Though all that analysis still can’t know the next piece of new information, it fully exhausts all probability analysis that can be undertaken with past information.

We can draw one important conclusion from the analysis that should be clear and should matter to all investors.  The sticky luck of a fund manager or advisor should not be rewarded with fees that include any assumption of continued sticky luck.  The legitimate role of a fund manager is to define, clarify and manage risk, as I will discuss later.  The financial rewards for such services are likely to be well below those associated with the chase for superior performance.  Moreover, those asked to pursue superior performance, based upon the unintended nature of their incentives, will consciously or unconsciously take on risk which may well be wholly inappropriate for the unwary investor.  The investor may well receive a service that actually drives against his or her long-term interests, at much too high a fee.

Another conclusion is the futility of market-timing for superior performance.  I will later discuss bubbles and the avoidance of their negative consequences, but market-timing is not a legitimate strategy to take advantage of bubbles.  This conclusion is essentially an extension of my first one about superior performance, as market-timing is just another way of claiming that the individual has more wisdom than the market as a whole.  As we will discuss, market-timing can only be justified as a risk-avoidance strategy, and its use under such circumstances may be very limited.

After almost two centuries of equity premiums, what is the market telling us with its risk-free return?  It is clearly well below the 9-12% figures that history has given us, and which is now commonly quoted by industry professionals as the “expected” return for the equity market.  How can we account for this paradox?

One reason is the bias of professional opinion.  The equity industry pays its advisors and managers well, and the source of that pay is the cash that flow into equity funds.  It would be strange if such opinion did not reference the equity market’s performance over almost 200 years and suggest it was a long enough record to project confidently into the future.  In most other walks of life, a perspective that gives consistently good results measured over 25, 50, 100 and 200 years would be considered solid indeed.  So it is not surprising that there has generally not been a serious challenge to the impartiality of the advice, provided it allows sufficiently for medium-term volatility.

The whole-market view differs from the advisors and fund managers, in that its investors are investing their own money.  The market is therefore paid only by the success of its investment strategy, not by the convincing nature of arguments.  We hear it only through the price movements of securities – information it does not even willingly give to us, but is an inevitable consequence of its activity.  Why should a large number of those investors have such a pessimistic view of the equity market, dragging down the weighted-average expectation to the risk-free return?

The Buy and Sell Equation

We first restate the fact that these investors evidently exist because of the fact that as many people sell equities as buy them, at least in the case of freely-traded securities for which we consider only the secondary market (i.e. exclude new issue).  No matter what the sellers expect to buy with their proceeds, their sells are “negative votes” for some stock and any “positive vote” as a buyer is balanced by some other seller.  So though the individual seller may be focused only on the prospect of one stock, the sum total of all sellers is a powerful negative vote for the entire equity market, fully balancing the positive vote by all buyers for the same market – regardless of how many of investors are buyers and sellers at the same time.

The most significant clue to this balancing of historic performance with future expectation was mentioned in the last chapter – our inability to know the term (period of time) over which the market expects its return.  We might poll our investors, but this may not help as it is the theoretical rational investors we need to know about.  Our arguably less-rational, real-life investors may not consciously know the implicit term, just as they do not consciously know the return they expect.  Yet there must be an implicit term, or otherwise we could not define a risk-free rate of return.  Without such a reference point, there would be no way for the theoretical, rational investor to measure risk and decide upon appropriate reward for that risk.

In the prior chapter I described the one-to-one relationship between the non-rational and rational investors, which I used to explain why the real market was indistinguishable from a set of rational investors who always measured their expected returns.  This one-to-one relationship applies for a given, market-expected term.  For a different market-expected term, there would be a different one-to-one relationship.  Generally, the longer the term, the greater the expected return of any given equity investor.  It follows that, at market equilibrium, there must be a single, market-expected term to correspond to the single, market-expected return. 

Yet many actual investors are not only intuitive and emotional, but their plans are conditional and uncertain.  We are focused only on their expected success, not on their actual success, and most investors hope or expect to sell whenever it makes best sense to, responding more or less appropriately to price changes and market signals.  Whether or not market-timing makes sense, it is commonly attempted.  Few would see it as an exact science, but probably most engaged investors believe they can do better than being wholly blind to market-timing.

The Expected Term of Market Expectations

I believe we can put some reasonable boundaries around the fuzzy area of expected term.  It cannot sensibly exceed thirty years, as we have no free market in risk-free securities longer than that.  In fact, there is not a great deal to be gained usually looking beyond ten years.  Risk-free yields for longer terms vary relatively little, and would seem like a rounding error when faced with the ranges of investors’ expectations.  It is also reasonable to assume that most buyers would view their mean expected term (i.e. each time period weighted by a probability of sale) to be no longer than ten years.  From a tax perspective, most buyers currently have a strong incentive to retain stocks for at least one year.  Therefore, under a normal market environment, we would expect even our typical retail buyers to be thinking in terms of one to ten years, even if not consciously.  There will be a significant portion of traders – especially those who are leaders in moving the market – who will operate on a short-term basis of any length.

In a bull market, sellers expectations are typically much shorter, and therefore greatly reduce the market-expected term.  Most sellers may be concerned only about the next year or less.  I do not think it is sensible to simply average the sellers’ terms with the buyers.  If that average were beyond a year and no seller had a perspective beyond a year, we have no useful market consensus beyond a year.  Happily, a buyer’s long-term expectation is consistent with a short-term expectation of the same annual rate, so we can shrink the buyers’ term to match the sellers’ term. 

We conclude that the market is making no significant statement about expected returns beyond a time-frame of at most one year.  There is therefore no contradiction between the market outlook and its historic experience.  The one-year-or-less horizon for the market also makes nonsense of any market forecast that assumes the continuation of current interest-rate patterns.   By adopting market interest rates, the forecaster is acknowledging the primacy of the market.  By assuming they continue according to any pattern into the future beyond the futures market itself, the forecaster is then imposing a personal view that is wholly unsupported by the market, as the market cannot be assumed to take a view beyond six months.

The reason that the market does not expect future returns above the risk-free rate is, obviously, because of the constant risks inherent in equity investment.  Though down-markets have been rare in the past, there is no reason why they should be rare in the future.  The market’s progress over 200 years is also the product of sticky luck.  That is because the excellent equity returns have rested on improved productivity, and most of that improvement has followed from a single source – controlling the flow and storage of electrons, first in electrical equipment and then in electronics.  Looking back from our current position of “electron management,” it is easy to see how we have constantly exceeded prior generations’ expectations.  In our tree analogy, we have traveled deep within a forest where before no one knew about forests.  Yet our forest could at any time start thinning or disappearing.  Current expectations for productivity improvement are high and entrenched, but progress is not inevitable.  Future technological progress may not translate into better productivity – at least at the level to which we have become accustomed.

Yet to accept the SLMH we do not even have to acknowledge the dangers of slower progress.  A downturn in the stock market could easily occur over any five- or ten-year period because of unexpected events.  The events that led to the 1930s and 1970s bear markets were not even inevitable as causes – they appear to be simply devastating losses of confidence rather than any fundamental economic problem beyond the world’s ability to resolve.  Now compare the impact of severe energy crises brought on by global conflict, an uncontrollable influenza strain, unexpected drug resistance of other diseases, or terrorist WMDs.  It is not surprising that the market would make large adjustments to reflect such contingencies.  Yet perhaps the market’s biggest worry may simply be our elevated expectations against which loss of confidence may have even greater and longer effect than in the 1930s.  Back then, equities were perceived to be more risky and P/E ratios reflected that fact.  It took the “E” (earnings) to largely disappear to topple the “P” to one-tenth of its original value.  As in the 1920s, we have witnessed a surge in corporate earnings which are assumed, implicit in the current P/E ratios, not only to be maintained but to continue to grow. 

We will illustrate with some simple arithmetic.  Suppose the more optimistic 20% of investors project equity returns on average of 9.5% per year for the next five years.  A more cautious 20% project returns of 7.5%.  Some 40% expect, on average, to do not better or worse than the risk-free return.  Some 20% expect, on average, a market decline of some 10% over the whole period.  If the risk-free return is 5.5%, then the market’s expectations average out to that same return.

A Horizon of More Than Two Years Unlikely

Experimenting with such arithmetic, it would seem most plausible that investors are, on average, not looking out much further than five years.  For example, if we go our ten years, we have to balance the optimists with some much deeper market collapses over that longer period.  We do know that there is a significant number of optimists who expect equity returns to maintain their historic averages, but it would be surprising if an equal number of pessimists would (on average) expect a fall of 50% or so to be sustained by the end of the period.  It is no so much that such falls are not possible.  But it is probably the sellers who have shorter-term horizon, perhaps even in terms of months, and who draw down the average term of the expected return compared with equity buyers, and who are also shorter-term buyers of debt.  A negative horizon is easier to see over a period of a few years at most, rather than over a decade.

Does this short horizon for market-expected returns give us greater confidence that, provided we hold for the long term, we are in good shape?  I do not think we should be so reassured.  The market itself tells us nothing beyond its horizon.  Should we fall to one of the pessimists’ lower levels, our expected return will remain at the pedestrian level of the risk-free return.  In other words, if we follow the SLMH, there is no mechanism that increases our chances of a “bounce back.”  This may seem unreasonable to the optimists, but a little analysis tells us that they cannot have it both ways.  The history of the market has been a “bounce up” of spectacularly unexpected proportions.  The SLMH does not join the pessimists in calling for an overdue fall.  Should one occur, however, for example reducing the long-term equity premium from 5% to 2% (which would be a huge fall) the optimists can hardly argue that the prior premium of 5% was “natural” and that the revised premium of 2% was not.  There are only two consistent approaches:  either to believe that history is an authoritative guide to the future, or that it is not.

History is No Guide

The evident market dominance of sticky luck tells us that history is no guide at all.  No one could possibly have known it in Michael Faraday’s time but, when Mankind embarked on its taming of the electron, it had stumbled on a path which, by virtue of our specie’s fortuitous ingenuity, was likely to lead to secure on-line shopping.   Without taking away from the genius of all the inventors who got us here it is safe to say that, had none of them been born, others would have taken their places.  (In most cases these individuals narrowly beat rivals to their discovery.)  Looking back, as we do with all patterns, it is easy to see how the combination of the human mind and the electron’s inherent properties got us so far in two hundred years.  However, as we were discovering those properties, it was natural for us to have doubts about where the next step would take us.  A family that gambled its estate on purchasing an exclusive franchise on “all things electron related” back in 1800 would have been richly rewarded in most years since then.   Let us say that franchise is now up for sale - divided into plenty of shares so that there is a freely traded market in them.  Are they a good buy?

The answer depends upon whether or not the progress of electronic goods, and their income generation, meets current expectations.  We cannot estimate that probability better than 50-50, because we cannot know more than they money-weighted opinion of the market.  If it does not meet expectations, then the spectacular 200-year run of the franchise, in terms of delivering value to its owner, will have come to an end.  How can it be that a track record of 200 years can be broken, just like that?  The simple answer is that the market has reset the price so there is nothing to choose between gain or loss.  The more complex concept to grasp is that, if we had marketed the franchise every year for the past two hundred years, in most years the free-market purchasers would have been rolling in it.  But all of that past franchise activity, and even whether or not it occurred, has no influence upon the crap-shoot of the current franchise opportunity.  This is sticky luck at its most deceptive.  And this is the wonder of the free market.