The game of picking stocks is mirrored in a game of picking managers who pick stocks. The performance effects of playing the active management game are the sum of behavioural impacts at both levels: fund managers and the ‘end investor’ or his/her agent. 

Because most investors playing this game are doing so because they believe past performance is predictive of future performance, rather than random, they will naturally tend to select new holdings from the sample of managers/funds that have performed better than average over some recent period.  If, on the other hand, it really is random (or even much less predictable than they thought), there is a very high chance of disappointment – where disappointment is defined as the ‘unexpected’ slippage of the fund through the rankings table. Consistent with the beliefs they held when they bought, they will now tend to sell, because they will assume that they made a mistake or that the manager in question has lost his/her touch – in other words the new performance is predicting more of the same. Because they have not changed their beliefs, they then go through the same exercise to select the replacement fund. And so it goes on, turning random underperformance of holdings into a non-random string of portfolio underperformance – until they realise it is their beliefs that are wrong. 

This second behavioural effect is under-researched in academic studies, perhaps because the key insight itself has not attracted nearly as much attention as the effects at manager level. However, there are a number of industry firms that survey money-weighted returns in mutual funds (measuring client-experienced average returns): Dalbar, Vanguard and Morningstar. This analysis suggests playing the loser’s game costs up to 6% pa – far more than just the incremental costs of active funds (which in the UK we put at between 0.6 and 2% depending on the buyer’s agency relationships). 

The data available to researchers is fund performance (obviously) and fund flows. But when making a connection between the two there are some problems:

• Both are absolute – so if people sell after poor performance they could be making a market timing decision or a switching decision based on poor relative return but the data won’t tell them which. The two may of course be positively correlated. Given the evidence (in Dalbar’s annual surveys, for instance) that net flows are positively rather than inversely associated with market movements, it seems likely most of the effect is due to poor market timing rather than switching, so a reaction to absolute rather than relative performance.
• Flows are partly idiosyncratic decisions and partly regular contributions/withdrawals but the two are not distinguished.
• Money-weighted return calculations are affected by the sequential pattern of returns and flows – although I’m not convinced this methodology point ‘explains’ the apparent behaviour effect.

In ‘Mutual Fund Flows and Investor Returns: An Empirical Examination of Fund Investor Timing Ability’ (Friesen and Travis, 2007) the authors say this about the relative-return  impulses to trade funds:

In testimony to the United States Senate, investment guru John Bogle (2003) argued that the Dalbar (2003) study ignores an additional “selection cost” borne by investors, whereby investors place a disproportionate amount of money into actively managed funds that subsequently underperform the S&P 500 index. Bogle suggests that after accounting for this selection cost, the average mutual fund investor return over the 1984-2002 period is actually negative.

Whilst it may be difficult to measure actual effects for a universe of mutual fund investors, it is possible to model the behavioural effects to test for the likely scale of impact of the relative-return impulse, by applying some simple decision rules, based on rank orders, for both buying and selling.  This is a project for future research.

It is, however, possible to infer something today about the prevalence of this impact just from observation of a limited sample of portfolios. We have taken in new clients from a wide range of advisers/banks/wealth managers over the past six years. Because our clients have well above-average wealth, their previous agents are typically respected and popular firms, so we would expect them to be less prone than the average to systematic wealth-destroying behaviour. We also take on clients who are more experienced than average and so many have previously been wholly or partly self-directed. Our initial financial planning process for all new clients includes a critical appraisal of existing investment arrangements. We observe, in most cases, the same destructive behaviour by both agents and self-directed investors.

Even in the case of investors with agents, we suspect the investor’s own ‘predicted’ behaviour effectively encourages the agents to sell, because the agents think individuals think that poor performance tells them something about the agent’s skill.

When looking at agent behaviours, it is always sensible to consider whether incentives or game theory might explain them. Whilst I believe this is realistic in the case of fund managers, whose active-management payoffs are quite different from their investors, I do not think gaming validates the behaviour of agents selecting funds. Because both these agents and their clients are acting consistent with a set of beliefs (however false) about skill, it looks more like a case of the blind leading the blind.

‘It’s a miracle! I can see!’ cries Eddie Murphy in Trading Places when rumbled as a ‘blind’ beggar. In investing, miracles do happen and when an investor truly can see, their rumbled agents are unlikely to get away with it. They, rather than their clients, are the ones who should have known better.

Note: This article (in draft) was extensively quoted by Matthew Vincent, FT Money editor, in his Serious Money column on 30th July. It followed a piece the previous week about the first, manager, level of the active-management game titled How fund managers get paid for winning the lottery. You may need to be an FT subscriber to view these articles.

Is he right?

In this increasingly popular view, the actions by governments to pump money into the stalling financial system are seen as giving stocks a reprieve before they reached the sort of valuation levels that properly discounted the post-credit crunch realities facing economies and businesses. Hence the anxiety now as the patient looks to be coming off life support. 

John Authers’ preferred measures of value, using replacement value of assets (Tobins q) and a trailing 10-year average of earnings, supports the story that US equity investors, like the banks, got lucky and were bailed out before reality had sunk in to share prices.

I don’t think we can agree with that. Our own measure of value suggests that the last move down in the S&P at the start of 2009 took the market to an extreme level with only a few historical precedents in any major equity markets, all equally important as long-term buying opportunities.

Whilst we attach a great deal of importance to our measure of value as a basis for forecasting long holding-period return probabilities that can be used in asset allocation, it still does not mean there is a negligeable chance that the S&P will fall through the earlier low. Though it would be unprecedented for the US market, it happened in the UK in 1974 and in Japan in the 90s and again in 2003. What can happen in one place can happen in another.

There are also several historical examples where extremely high predicted value has persisted for a decade or more, Japan being one but Continental Europe in the late 70s and early 80s is a less familiar one, being disguised at the time by high inflation.

Though the FT piece was about the market most likely to influence the general direction of global equities, it is worth noting more parochially that on our measure the FTSE All Share Index is the least undervalued of all the major markets and did indeed recover well before reaching a historically exceptional valuation.

Lateral thinking about equity valuation

We agree with John Authers that Tobin’s q and trailing 10-year earnings will dampen the effects of extreme cyclicality in current fundamental inputs to valuation, particularly reported earnings. As an accompanying chart in his column showed, reported earnings in real terms reached a low point last seen in the 1930s although clearly this is not what has happened to ‘earnings power’ for American business.  The q ratio dampens variance simply because replacement value of assets is very stable, so the significant information comes from price change. Averaging earnings over several years will also dampen the valuation variance but is transparently arbitrary and may also leave price variance as the important information. These may therefore be better than conventional price/earnings ratios when earnings are disturbed but they are still not rigorous indicators of extreme valuation, high or low.

As market historians, Chris Drew and I have since the late 1990s adopted a lateral approach to equity market valuation, formalised in a return-generating model called Lambda. We ignore fundamental inputs, because of the measurement problems, in favour of tracking the path of achieved real total returns from each market and comparing it with its own long-term trend, using ’best fit’ regressions. The model has not been invalidated by subsequent return behaviour and has indeed handled several episodes of extreme valuation very well.

On this measure, the recent low for the S&P was as low as the previous extremes of the early 30s and 40s as well as 1974 which was revisited (in real but not money terms) in 1982.

Fig 1 shows the S&P continuously-compounded logarithmic returns with gross dividends reinvested and deflated by the CPI from 1925, as an index. The whole-history regression trend (colour) is 6.8% pa but the hindsight-free trend using data up to each point in the series (colour) was above or below this as a function of the particular period. In my book I showed a Wilshire series for US equities for nearly two centuries that revealed a similar trend in both centuries, in spite of massive contextual differences. Data for other major equity markets also suggests about 5-7% pa as a trend, with much narrower dispersion of the differences than economic context and culture might suggest. Global evidence suggests regression trends are not just a statistical accident but an economically sustainable quantity that is a core element of the equity return process. We believe there are strong theoretical underpinnings for this, based on both economics and risk aversion.

Fig 1 Trend and deviations from trend in real equity returns: S&P 500

This idea of sustainable real equity returns over long holding periods is really important. It conflates into a single measure two crucial influences on returns: i) fundamentals (like earnings and dividends) and ii) investor behaviour changes (like risk aversion, sentiment and momentum following). Historically, even over long holding periods, the second set of uncertainty sources has usually dominated the first, which makes the traditional focus on fundamental inputs look odd. If jointly the two sources of uncertainty lead to bounded deviations from trend, as a further core element of the return process, we do not need to disaggregate them to model either returns or risk.

Fig 2 expresses the level of the real total returns index as a ratio of its own trend, or detrended, which we call the Market Value Ratio. We use this as a basis for adjusting upwards or downwards the mean expected returns extrapolated naively from the trend alone. Low price denotes high value and high future returns and vice versa. By making some assumption (itself based on historical analysis) of the time-dependent reversion to trend, we can estimate real return probabilities for every holding period as often as we gather new historical return data. The actual frequency is monthly.

Fig 2 Market Value Ratio: S&P 500

What can we do with value?

Fig 2 makes the point of this post that the S&P low in March last year was extreme. This observation was sufficient to justify exceptionally high expected future returns, provided long holding periods could be assumed. For investors with naturally long horizons, this amounts to a lot. It is why we added with confidence to clients’ total equity exposures up to the low point, while others were selling, and reduced them either side of this recent year end. See Is your manager doing a good job? (22nd January). 

Observing extremely low value by March 2009 may even have been enough to warrant assigning a very low probability to further downside risk over short horizons, which would also amount to a lot if, it could be relied on.  We do not believe it can. We prefer to treat short holding-period returns as a ‘random walk’. That was not the basis of our decision to buy.

An aside for more numerate readers: short-run returns, whether deflated or not, also have the highly inconvenient statistical characteristic of fat tails, or extreme outliers that would not be predicted if assuming a normal distribution. Long holding-period real returns, on the other hand, appear to be normally distributed. 

We sympathise with John Authers who is obliged (as a good journalist rather than a good story teller) to remind us that even a ’true’ valuation of the market tells us very little about the chance of the market falling or rising or staying the same over the next few years. This might read as ’sitting on the fence’ but it is more insightful to see that good valuation measures simply have restricted uses.

For us, as financial planners and even as portfolio managers, using them to project long holding-period real outcomes is actually more important than using them to try to time markets. Imagine yourself in 20 or 30 years time looking back and deciding you got the outcomes you wanted and planned for. Is it most likely you will credit the many market timing decisions that made up the path your money took or rather making the right long term decisions? We have no doubt and we don’t think our clients do either.

26th October 2006 “Market Commentary”
In this client newsletter, posted on the blog, we set the scene like this.

“The focus is on the risks facing financial markets in the late stages of an unusually long expansionary phase of the business cycle. These are conditions in which investors with low risk aversion can still make good short-term returns but unwary investors with lower risk aversion tend to get caught out with excessively risky exposures when the cycle turns. And the timing is never predictable. If you thought the business cycle had been condemned to the history books and that monetary and fiscal authorities are working in sophisticated harmony to ensure well-balanced growth, you might find these ideas challenging.”

The consistent theme of our economic analysis up to this point had been unsustainable sector financial balances. This comes from an approach to cyclical analysis which has a long historical pedigree. Either memories are short or market participants too young because it was quite wrongly overlooked during the ‘Great Moderation’ or ‘non-inflationary consistent expansion’. We explained this in this piece.

“We can speculate what it is that might bring the bull phase to an end from a level at which equities are not greatly overvalued….The most likely cause is related to the cumulative economic ‘sector balances’ that have developed after some 15 years without a full recession – in other words a recession in which the balances between the main economic sectors (personal, corporate, government and foreign) are restored to more normal relationships. This might have happened at the start of the new century had 9/11 not triggered a massive injection of stabilising liquidity by the Federal Reserve Bank, thereby extending the growth phase of the cycle which at that stage was already a decade old. The cumulative effect is not so much apparent in the level of inflation as in the record government and foreign surpluses, lower personal surplus and exceptionally flush corporate sector. These US sector balances are mirrored in the UK.”

For clients with goal-based ‘Defined Outcome’ portfolios, which are limited to those assets with the best evidence of long-term real returns, it is a quantitative model that determines the scope to hold equities rather than risk free assets and it is individual goal-specific risk tolerance that determines the actual equity allocations. The model calls for a gradualist approach rather than big market timing bets. But for clients with multi-asset class ‘Defined Path’ portfolios, where liquidity often does not allow gradualism, we have to take bets and we have to rely on our own judgement. In October 2006, here’s how.

“Our judgement is that for property and private equity (particularly leveraged buy-outs) the expected returns are already cyclically very low and that the premium for illiquidity is inadequate.

“UK commercial property has fully made up its underperformance in the 1990s and yields are discounting unrealistic rental growth…

“Because we reject the notion that private equity returns, pre-leverage, are largely highly-incentivised returns to management skill, in favour of the view that they are mainly a leveraged bet on a change in the terms set in the public market for buying and selling equities, our cyclical view is bound to make us cautious.

“We are avoiding bets much above or below the strategic allocation to commodities [20% is the ‘neutral’ weight, by the way]. Energy and metals have enjoyed a bull market for several years…[but] we are also mindful of a more secular view of trends in real (or relative) prices of industrial inputs which offers the possibility of a reversal of the 20-30 year decline in the terms of trade between producing and consuming nations. Both secular and cyclical trends in raw materials prices underpin the empirical evidence of low correlations between commodities and equities. This ‘hedge’ argument is important in meeting an overriding objective of the diversified strategy, which is smoothing the nominal short-term path of…returns.

“One aim of these portfolios that follows from that objective is to hold assets or strategies that offer equity-type upside but with appropriately-timed downside protection. It is this aspiration, and the fact it was generally well met in the last bear market, that makes hedge funds popular with both private and institutional investors. Putting clients into a hedge fund can look indistinguishable from us delegating the timing decisions to a third party but it is actually more than that: we look for strong trading disciplines learnt over many years that include stop-loss rules. As ever, we retain a healthy skepticism about market timing skills – ours or anyone else’s.

“Both [types of No Monkey Business] portfolio are holding more in risk free assets than ‘normal’. If markets continue to rise, we will cut back exposure to risky assets further, in the first case consistent with the client’s goal-based model and in the second as a matter of judgement.”

24th March 2007: “Drowning in debt: can it be true?”
Our focus on credit analysis informed several of Stuart’s posts before the credit crisis broke, including this.

“We have enjoyed a long period of much stronger growth in money and credit than the economy as a whole. One of the features of this massive monetary expansion is that credit creation has not been constrained by bank capital.

“Until this present cycle, and in all the previous manuals on central banking, asset growth on bank balance sheets was assumed to be constrained by the rate of increase in their shareholders’ funds. In a ‘modern’ world where loans (or separate elements of the risk exposure) can be packaged and passed on to the likes of hedge funds and insurance companies, liquidity outside the banking system (itself expanded by bank credit growth) becomes an almost unlimited base of equity to support new loans.

“This process has been great for borrowers and for financial institutions eager to find new assets with different income streams. It has been great for financial intermediaries able to bundle the highest risks in an obfuscated product and flog it to wealthy individuals (not our clients, of course). But the process is a nightmare for central banks and bank regulators because it weakens their control. The banana skin they most fear is always the one they may not be able to do much about.

“We are living through a period of unprecedented debt-driven liquidity growth that is making a lot of us a lot of money, even if many trying to keep up have failed in trying and should never have tried. Debt-fuelled expansion contains the seeds of its own reversal. We do not know when or why but we will surely benefit from being alert to the consequences. These are directly related to the level of accumulated debt, even if only in particular pockets.”

04/09/2007: “Drowning in debt: the American way”
Some of the most important pockets affected are in America.

“America’s insecurity after 9/11 showed in a desperate attempt to prevent recession by cheap and easy credit. Now the chickens are coming home to roost. An unprecedented house price boom was in large measure fuelled by this monetary expansion but it was also aided by lax lending standards from mortgage bankers and by new sources of liquidity from hedge fund investors – now as big as banks. In recent weeks we have seen the first signs of this foolhardy liquidity drying up, with several lenders to the dodgiest borrowers going bust and others shutting up shop.

“Two weeks ago I wrote about UK consumer debt. In America, the position is even worse: US consumers are overstretched and overconfident. But it is not just credit card debt that looks vulnerable: over-extended mortgage debt is the bigger problem. There is a significant risk of contagion spreading from dodgy mortgages to Wall Street investors via debt-based hedge-fund strategies. This is scary stuff – and not just for holders of American investments. The American authorities may have bought growth but only at the expense of a much deeper recession later.”

12th August 2007: “Update on the real house price cycle”
UK house prices have been a consistent focus of the No Monkey Business blog. Consistent with the British obsession with housing, this is the most frequent search string directing traffic to the blog.

“Our way of measuring house prices focuses on house prices as a cycle in real terms, in common with our approach to other assets…Both ratios show extreme overvaluation and have been rising rapidly for the past four quarters.

“The data through June was still reflecting buoyant lending conditions… This may change very quickly, now that bank balance sheets and wholesale capital markets are so globally integrated. What the real house price cycle chart shows is the extent of the potential fall in real terms, given previous cycles. We have repeatedly argued that the previous down cycles have been concealed by high general inflation. This time round, ‘real’ will mean ‘for real’.”

3rd March 2008: “Japan’s ‘lost decade’: the new bogeyman for post credit-crisis stockmarkets”. As recession fears started to take hold, Japan’s ‘lost decade’ of the1990s became a popular theme in the press as well as among investment professionals.

“The same fate supposedly awaits the US (and UK too) as payday for a decade of binge banking and mass delusion by bankers and their customers about house prices. In a new position paper, wealth manager No Monkey Business examines the relevance of Japan’s dire experience of the ’90s for other countries today.

“The key distinction is the duration of the underperformance against expected growth in economies and equity real returns, rather than the cumulative degree. Prolonged underperformance is possible in Anglo-Saxon markets because of the specific effects of a bursting housing bubble, even if Japan is not a close parallel.

“For wealth managers, the lesson of Japan is that both the degree and duration of underperformance against rationally-expected returns need always to be a part of the realistic outcomes we tell clients about: they are more likely than people think.

“For Japan itself, the lesson is that the rational ‘mean reversion’ model of real equity returns is not broken and because of low expectations it now offers the best long-term returns of the major markets. And the yen looks cheap against sterling.”

The last word
Because of our skepticism about forecasting in general, it is unusual for us to be so focused on economics. In our last client newsletter (not on the blog) we felt this needed an explanation.

“Our recent newsletters have been so dominated by economic commentary that we may start to look like all the other pundits trying to use forecasting as a means of placing winning bets in financial markets, something we think plays to investors’ (and advisers’) natural weaknesses rather than their strengths.

“The justification for focusing on economic developments is not that they have made markets more predictable… Rather, heightened economic risks make it more important to be sure that they lie within the tolerances agreed with clients or, in a modelled approach, are ‘in the model’.

“The desire to avoid risks also needs to be balanced by awareness of the cost of that protection. Being realistic about the true uncertainties of the payoffs to risk-taking and the true cost (in returns and resources required) of eliminating uncertainty are things we think we are good at.”

This may make it a trader’s market. It may also be a market for a systematic, agnostic long-term approach, adding to risky assets as prices fall and selling as prices rise, but always keeping the exposure consistent with the agreed tolerance of long-term real outcomes. Both have the virtue of a discipline and discipline may be the key to avoiding being whipsawed. Thinking about what whipsawing really means, remember that one symptom is that missing the best ten days in a period as long as 40 years would have cost you half the equity return premium over cash. Failed market timing really does hurt.

In this new position paper from No Monkey Business Limited, we put the claim to a ‘no nonsense’ test. Absolute-return investing is faddish. It is not the new paradigm. It is mostly not even absolute. But a hybrid form of absolute and relative investing adopted by many hedge funds, which can be used as the basis of multi-asset class portfolios, is meeting real investor needs.

Rather than try to prove intrinsic superiority between absolute and relative approaches, we aim to make the differences clear enough for investors to choose between them (or even combine them). They can do this by relating them either to suitability for their own tasks or to persistent biases in their risk preferences. But to do this, they have to get beyond the sales.