For better investment results, concentrate

(Published in the Financial Times on 29 June 2014)

As the Financial Times' John Authers rightly points out, "active share" — the portion of a fund that differs from its benchmark —should never be used in isolation to assess funds ("Active fund managers are closet index huggers", March 12). The notion that there exists a simple measure to predict performance is clearly absurd.

Furthermore, whether or not it should be used in conjunction with other tools is also contentious. Research by Martijn Cremers, professor of finance at the University of Notre Dame, and Antti Petajisto, a portfolio manager at BlackRock and former assistant professor of finance at Yale, found that, net of costs, high active share funds on average performed better than index funds, which in turn performed better than low active share funds. This was certainly an interesting conclusion, but why should it be true that high active share funds tend to perform well?

Ross Miller, finance professor at the State University of New York, devised a clever way to demonstrate why closet index funds were a bad idea. His insight was that any fund could be divided into two parts: a passive portion that correlated 100 per cent with the benchmark and an active portion that was 0 per cent correlated.

By assigning typical passive fund fees to the passive portion one could then calculate the effective fee (the actual fee less the passive fee) that one was paying for the active portion.

In the case of the benchmark huggers, the active fee was so high — in some cases as high as 7 per cent — that the chances of producing sufficient alpha, or excess returns, to cover it as well as leaving a big chunk for unit holders were negligible.

High active share funds are therefore the ones giving themselves a chance to produce sufficient alpha, but why do they also tend to succeed in this endeavour?

There are essentially two ways to get a high active share: run a fairly concentrated portfolio, taking little notice of benchmark constituent weights, or run a diversified portfolio full of stocks that are not in the benchmark. There are few if any funds that fall into the latter category, so the argument is really about why concentrated portfolios are a good idea.

Skill, and thus good performance, is about first spotting and then taking advantage of price anomalies. It is other investors — "the market" — that create the price anomalies in the first place, so good performance will always be at the expense of others. Alpha generation is after all a zero-sum game.

The behavioural trait known as herding from time to time takes prices to levels that, with analysis, can be declared cheap or expensive. Such mis-valuations — or mispricings — can be tiny ones that correct in a few seconds or hours, or bigger ones that correct over a few years. Jim Simons, founder of hedge fund Renaissance Technologies, uses a very powerful and very quick computer to spot the tiny ones before anyone else. Others are more interested in the longer ones.

Longer-term price anomalies are well documented. Sanjoy Basu, former professor of finance at McMaster University in Ontario, and Robert Shiller, professor of economics at Yale, found that stocks with high dividend yields outperform stocks with low dividend yields. US economist Eugene Fama and Kenneth French, professor of finance at Dartmouth College, showed that this predictability increased with time: while over one year the dividend yield explained 15 per cent of the variation in excess returns, over five years it explained 60 per cent.

It seems that investors systematically overestimate the extent to which growth stocks will grow. Since it is growth stocks that tend to make the headlines with stories of expansion and acquisition, it is hardly surprising that this attention can cause overvaluation.

Corporate governance is another factor that gets systematically mispriced. Paul Gompers, professor of business administration at Harvard, Joy Ishii, assistant professor of finance at Stanford, and Andrew Metrick, professor of finance at Yale, found that the stocks of companies with good corporate governance outperformed those with poor corporate governance by 8.5 per cent per annum. Given that the well-governed companies were ones that had lower levels of capital expenditure and made fewer acquisitions, this anomaly is perhaps also explained by an interest in newspaper headlines.

Armed with this evidence, and having identified potential winners, one then has to decide how much to commit to each. Here we turn to mathematician John Kelly and his work on horse betting. He devised a very simple formula, known as the Kelly Criterion, to determine what percentage of your purse to bet on a particular horse, the equivalent perhaps of how much of your portfolio to invest in a particular stock. If the market odds of a horse winning were 5o per cent but you had an edge of some sort and believed its chances were 55 per cent, the formula said you should bet 10 per cent of your purse to maximise your winnings.

The empirical findings with respect to dividend yields and corporate governance are so strong that one should consider them to represent a considerable edge. While it is impossible to put a number on precisely what that edge might be, the evidence strongly supports the case for putting 10 per cent in a thoroughly researched, well-governed, high-yielding stock, not 1 per cent.

Skilled managers do exist

(Published in the Financial Times on 27 January 2013)

John Authers asserts (Quest for test of investment skill persists, January 11) that nobody can tell which investors are more skilled than others and that active fund managers are unaware of their skills. 

These assertions are wrong. I suggest - with statistics on my side - that Berkshire Hathaway’s Warren Buffett and Renaissance Technologies’ Jim Simons are more skilled than most and furthermore that there are others who, with a little work, can be identified. I also know that at Aberdeen we are aware where we possess skill and - more pertinently - where we not.

Knowing where you possess skill however requires knowing what it is. Simply, investment skill involves identifying non-randomness in financial markets, then profiting from it. It is about knowing the difference between what is predictable and what is not. It is about understanding where and when one has an edge and then swooping in for the kill.

The sad fact is that the vast majority of investors waste their energies and capital guessing the equivalent of which side a coin will land, often supported by written and verbal justification that is both sincere and articulate.

The key to investing is knowing when the odds of a coin landing heads have increased to, say, 70 per cent or 80 per cent. And this is where serial mean reversion comes in. 

Occasionally, random walks take asset prices so far away from their mean or trend that they are pulled back towards it rather than continuing in a random fashion. If a coin lands heads 10 times in a row, the odds of a tail on the next throw are still 50 per cent. In the world of investing however, it may be 70 or 80 per cent. Put another way, despite the ubiquitous disclaimer, past performance can sometimes be a guide to the future.

Identifying pattern is very much what the two aforementioned maestros do in their own different ways. Mr Buffett’s edge, in my humble opinion, has been his remarkable understanding of human nature, both its strengths and its weaknesses, combined with discipline, patience, honesty and a very good grasp of statistics. Mr Simons, on the other hand, is a brilliant mathematician and has smarter and faster computers than anyone else. While Mr Buffett is the king of predicting share prices over 10 years, Mr Simons is unrivalled over 10 minutes.

Mr Buffett and Mr Simons are rare birds yet many still believe themselves to be good investors when the facts may tell a blatantly different story. The reason for this is that we humans evolved a survival mechanism to believe that we are better than we actually are. A timid approach to facing down a sabre-toothed tiger or attracting a cave mate would have been disastrous. Furthermore, we have an asymmetrical ability to blame our failures on (bad) luck but to attribute our successes to skill, a bias termed the fundamental attribution error. Thus you only need a couple of successes among all the failures to think you are a skilful investor.

If you have correctly identified an edge, the next step is to know how to use it. Question: if you have a biased coin that you know has a 6o per cent chance of landing heads and you are playing with someone who does not know the coin is biased, what percentage of your bankroll should you bet each round in order to increase your wealth over time?

If you bet nothing, you are wasting your edge and your wealth will remain the same. If you bet your entire purse, there is a 40 per cent chance the coin will come up tails, and you will lose everything and be out of the game (even with the bias, the chance of there being one tail in 10 tosses is 99 per cent.) So the optimal percentage must be somewhere between 0 per cent and 100 per cent. 

The answer, in fact, is 20 per cent. Bet 21 per cent or 19 per cent and over time you will end up less wealthy than if you bet 20 per cent. If you want to know the formula, google the term Kelly betting criterion.

How does this apply to investing? In Mr Buffett’s case, he of course understands that to put all one’s eggs in one basket is foolish, but also that being overloaded with baskets will wear you out. The efficient market hypothesis asserts that you should diversify as much as possible to eliminate stock specific risks. Mr Buffett on the other hand actively seeks out stock specific risk because he knows that is where his edge lies. As he has noted: Wide diversification is for people who do not know what they are doing.

Does Mr Buffett know precisely what his odds are? Of course not. What he does know is that he has a good feel for where a company will be in 10 or 20 years’ time, giving him the confidence to run a concentrated portfolio. There is much we can learn from him, though I am not the first to suggest that.

Affiliates offer way into emerging markets

(Published in the Financial Times on 26 Feb 2012)

Some domestic US fund managers argue that investors need not bother with emerging markets: they can obtain exposure to emerging market growth through developed market companies with substantial overseas sales. Since active managers in the emerging world underperform, the argument continues, why not buy a portfolio of multinationals?

The answer is: because they would miss great returns from emerging markets companies that outperform.

Over time, stock markets should track economic performance. Although this has held true in developed markets, it is much less evident in the emerging world.

Over the past two decades, real GDP growth of emerging Asia has been around 4 percentage points higher than that of the US. However, since the end of 1987 — the inception of the index — the MSCI Emerging Asia index has substantially lagged the MSCI US index.

Similar patterns can be seen at a country level. From 1992 to 2010 China's economy expanded by seven times in real terms but the MSCI China index went nowhere. While the average Chinese listed company may have created value for employees, suppliers, bankers, and local and national governments, it destroyed value for shareholders.

Paradoxically, the strong growth of emerging economies may cause the relatively poor performance of their companies. Many corporations, transfixed by the growth opportunities, invest new capacity, often outside their core area of expertise. A focus on top line growth rather than profitability tended to result in destruction of shareholder value.

It is therefore right to be wary of the average emerging markets company — but not all of them. Those disciplined about expansion tended to be better performers over the long term. But discipline requires a strong corporate culture that can take decades to build, unless there is a parent company with a strong culture.

In a recent research paper* supported by Aberdeen, Martijn Cremers, associate professor of finance at Yale School of Management, analysed the performance of listed emerging markets affiliates of multinationals. Unilever, for example, has listed affiliates in India, Indonesia and Pakistan in which it owns stakes of 37 per cent, 85 per cent and 75 per cent respectively. There were 92 such companies across the emerging world, 24 of them in Asia, 46 in EMEA and 22 in Latin America.

Prof Cremers found the share price performance of listed affiliates was vastly better than that of both emerging and developed markets broadly, as well as their own local markets.

Over the 13 years from June 1998 to June 2011, a period chosen for its balance between sample size and history length, an equally weighted index of the 92 listed affiliates returned 2,229 per cent. This compared with total returns of parents, local markets and parents' markets of 407 per cent, 1,157 per cent and 147 per cent respectively. 

The pattern of outperformance was consistent across regions too. Affiliates in Latin America, Emea and Asia outperformed their local indices by 41, 134 and 50 percentage points respectively. Adjusted for volatility the affiliates' performance was even better, as many of them demonstrated defensive qualities during the 2008-09 financial crisis.

Prof Cremers says strong performance is attributable neither to changes in affiliates' valuation (price-to-book ratio) nor to them being in better performing industries. Instead he suggests two explanations.

First, the affiliates had adopted their parents' business culture and corporate governance practices, and with them a focus on shareholder returns. Second, financial support from the parents served the affiliates well in good times but particularly during the difficult times.

While there will be homegrown success stories, they are hard to identify other than with hindsight, and picking losers hurts. Listed multinational affiliates, on the other hand, may offer more reliable pickings. A strong corporate culture is hard to build so the competitive advantage companies in this select group enjoy may be a sustainable one. 

A bit of maths will help clarify fund fee picture

(Published in the South China Morning Post on 12 Dec 2010)

How should you judge whether a fund’s fees are worth paying? This can be a very hard question to answer in an industry where the performance of products is so varied, either because markets are behaving strangely, as is their nature, or because the fund manager is performing poorly, as is the case more often than not.

Furthermore, while the results that funds produce vary hugely, management fees for actively managed equity mutual funds generally fall within a narrow band of 1.5 to 1.75 per cent per annum.

This is a little like a car industry in which all cars are priced the same but buyers are unsure whether they’ve bought a Ferrari or a Trabant, at least until it’s too late.

But then if fees are so standardised, why should buyers believe there is any way to differentiate between funds? To most it must seem impossible to judge a fund other than on the basis of whether it has performed well in the past, often a bad idea as the disclaimers remind us.

However, hope may be at hand for unit trust buyers seeking a more precise way to assess funds. Research by Ross Miller at the State University of New York titled Measuring the True Cost of Active Management by Mutual Funds identifies a clever way to measure whether a fund manager is pulling his weight and thus deserves a reward or is charging active fees for what is effectively a passive service or worse.

A bit of maths is involved but the concept is simple, namely to measure the portion of a fund that is being actively managed – termed the active portion. And once you do that, you can also calculate the effective returns and fees on that portion and thus better judge whether a fund offers value for money.

But what is active management?

To many it is about the amount of activity, or number of transactions within a fund – indeed this is how it was defined by Nobel economics science prize winner William Sharpe.

But to define it so would be to label the great buy-and-hold investors such as Warren Buffett, Peter Lynch and Philip Fisher as passive, an obviously absurd notion.

Rather, active management must be defined in terms of the extent to which a fund is different to those in the pack, not in terms of how frequently a fund manager buys and sells. It may be stating the obvious but for a fund to beat its benchmark, it must not look like the benchmark.

The fact that many do look like their benchmark is a sad indictment of the fund industry; it seems the business risk of getting it wrong is felt by many to outweigh the investment reward of getting it right.

Rather than getting bogged down with the calculation of active portion (those who would like to, can Google the aforementioned research paper), it is perhaps better to stay in the realm of the conceptual and grasp the notion that any fund in essence is a combination of an index fund, which is 100 per cent correlated with its benchmark, and a market-neutral hedge fund, which has zero correlation.

Precisely how you divide the two is a little complex but the concept should be clear. If not, try to understand that part of a fund’s return is due to the performance of the market – the “index” portion – and the rest due to what the fund manager is doing, the active portion. Miller’s formulae determine precisely which parts of the fund are doing which.

Now, let’s propose you should pay the same for the “index” portion as you would for an index fund, i.e. between 15 basis points for US equities and 70 basis points for Asian and emerging markets, and thus the rest of the fees are in lieu of the active portion, where the real effort is taking place, successfully or  otherwise. By isolating the index portion and the fees that relate to it – and thus arriving at the active portion and its fees – one can begin to get a good sense of whether one is getting good value for money.

Miller terms the fee that one pays for the active portion the active expense ratio (AER), and on this basis many of the US equity funds to which he applied his analysis looked very expensive, with AERs as high as 7 per cent of assets for some of them.

Of course, if a fund is capable of producing reliable outperformance, or alpha, then one should not begrudge paying a portion of this to the fund manager. Indeed two of the largest investment consultants in Australia recently suggested that fund managers should receive between a quarter and a third of alpha generated, but charge much lower fixed fees.

How can you use all this information?

Perhaps you could ask your financial adviser to download Miller’s paper – which contains his formulae – and then calculate which funds’ fees are worth paying. This may also prove a good test of whether he’s worth it.

Study finds links between portfolios' profiles and future performance

(Published in the South China Morning Post on 19 Sep 2010)

Is it possible to find a simple measure that identifies skilful active managers and thus funds that will produce good returns?

This is what the information ratio – a measure of risk-adjusted return – was meant to do, but several studies have shown there is no relationship between it and future fund performance.

However, research by Martijn Cremers and Antti Petajisto of the Yale School of Management, recently endorsed by Morningstar, finds a strong link between active share – the extent to which a fund’s holdings differ from its benchmark’s – and performance.

Unlike the information ratio, which measures performance per unit of volatility, Cremers and Petajisto’s research looks at what is inside portfolios, then finds robust links between portfolios’ profiles and future performance.

But what is active management?

Commonly it is understood to refer to the level of activity in a portfolio; the more transactions, the more active the manager. Even Nobel laureate William F. Sharpe defined it such. But were we to use this definition, then some of the world’s most successful investors would be considered more passive than active. Warren Buffett, Philip Fisher, Peter Lynch and many like them have produced great results by buying and holding, but are considered inactive by the common standard.

Successful investing arguably is about being different from the herd, or rather about finding the right opportunities to be different. Unless you are a high-frequency quant trader like Jim Simons at Renaissance, producing superior returns does not necessitate lots of portfolio activity.

Buffett, Fisher and Lynch all understood that one couldn’t predict share prices over short periods – at least not by using fundamental analysis – but that one could – sometimes – predict business performance over long ones. And, if you could do that, then you could predict share prices over long time frames.

Furthermore, they all appreciated the principle of putting your money where your mouth is, believing that, as Buffett put it, “wide diversification is for people who don’t know what they’re doing”.

A second misconception relates to risk, the generally accepted measure of which is volatility. But does volatility really measure risk?

We think important risk relates to permanent loss of capital such as would be incurred by the bankruptcy of a company or the gradual deterioration of a company’s business rather than normal month-to-month undulations of share prices that should be viewed as temporary losses – and gains – of capital. But there is no neat measure of the potential for permanent loss of capital so we are left with one that is easy to calculate but which is also deeply misleading.

Good performance involves putting yourself in a position to produce good performance.

And while it would be nice to construct a fund that returned 2 per cent per month, month in month out, the reality is that the great long-term performers like Berkshire Hathaway come with a cost, namely high levels of short-term volatility, at least relative to the relevant index.

But this is precisely my point: short-term volatility is not a risk to be avoided but simply the cost of good long-term performance. And once one appreciates the power of compounding – the difference say, between 8 per cent per annum and 10 per cent per annum over 20 years – one can appreciate this is in fact a small cost.

Which brings me back to active share. Cremers and Petajisto looked at 2,647 funds from 1990 to 2003. They categorised funds according to both their tracking error and their active share. It should be noted that while there is clearly a link between tracking error and active share, both measure different things.

Tracking error measures a fund’s net exposure to systemic factors such as sector or size while active share is more a measure of non-systemic – or stock-specific – factors.

Thus funds that have high tracking error but low active share are “factor bet” funds and funds that have high active share but low tracking error are “diversified stock pickers”. Funds with both high active share and high tracking error are concentrated stock pickers”.

The two researchers divided the 2,647 funds into quintiles by both active share and tracking error, then looked at the performance relative to benchmark of each of the 25 groups.

What they found was that funds with low active share underperformed by around 1 per cent per annum regardless of whether they had low tracking error (benchmark huggers) or high tracking error (factor bets).

Diversified stock pickers produced good gross returns but fees wiped them out. But funds with high active share and tracking error that was not excessive produced excellent returns net of fees.

These results have huge implications. If investors know their fund’s holdings and those in its benchmark, they can tell whether their fund is managed properly or by a closet indexer claiming to be active and thus charging for a service he is not providing.

Vanguard’s John Bogle built his multibillion-dollar passive empire by pointing out that the average active fund performed poorly. However, Cremers and Petajisto’s research makes it clear that one should not tar all actives with the same brush. Bogle’s point may be correct, but when was anyone interested in the average runner in a marathon?

Chasing the elusive alpha

(Published in the Financial Times on 30 April 2011)

Last week's FTfm reported on a fund industry "overpaid by $1,300bn", citing an unpublished report by the IBM Institute for Business Value.

According to the IBM draft report, $834bn of fees annually are paid to actively managed long-only funds that fail to beat their benchmark, though FTfm believed the figure had been revised down to $300bn. According to the report, "alpha generation" or the ability to deliver index-beating returns, was "pitiful", despite the huge sums paid.

The fact is the ability for the industry as a whole to generate alpha will always be pitiful. Why? In the same spirit as Newton's third law of motion, every outperformer requires an equal and opposite underperformer. Put another way, alpha generation is a zero sum game.

The only way for alpha generation to be positive for the industry as a whole would be for other equity investors such as corporations or governments to underperform. This is possible over short periods but almost certainly impossible over longer ones. Thus it must be accepted that paying fees to the active management industry is in some respects like paying people to toss coins with each other. On the face of it active management is a destroyer not a creator of value.

So why does it survive?

First, its alternative, passive investing, seems inherently un-Darwinian. Being average has never been associated with progress. The active management industry also plays an important role in asset pricing and capital allocation, thus helping to raise beta, the overall market return. And some coin tossers are skilful. With some relatively simple analysis, it is possible to identify them.

Which brings us to the efficient market hypothesis and whether it is possible to beat the market. It is surprising to me that the question of whether market efficiency is strong, semi-strong, weak, or otherwise is always framed in objective terms. Markets are efficient, or not; markets are strongly efficient, or not.

A better way to understand efficiency is to frame the issue in subjective terms. In other words, whether markets are efficient depends on who you are. For most, markets are and always will be 100 per cent efficient. But a privileged few have the skill required to spot inefficiencies big enough to profit from. Indeed the harsh reality is that this privileged group is profiting at the expense of the aforementioned majority.

Of course, all active managers believe they are skilful even though many are not. Such delusion can be explained by the natural human behavioural bias known as the fundamental attribution error, in which we attribute gains to skill and losses to (bad) luck.

Successful investing is about being different. The skilful managers tend to be those who understand that to perform well, and leave enough for clients once fees have been deducted, you need to put yourself in a position to perform well. That means being different from the benchmark.

In theory, a fund that departs significantly from its benchmark has as much chance of lagging it as beating it, but evidence suggests otherwise. According to research by Martin Cremers and Antti Petajisto of Yale University, US mutual funds with high active share (very different to their benchmark) outperformed those with low active share (closet indexers) by 2.5 percentage points a year (from 1990 to 2003).

But if it is that simple, why doesn't everyone simply run high active share funds? Perhaps because such funds are likely to exhibit volatile short-term performance, in some quarters substantially underperforming their benchmark.

The fact is that tracking error is used as a measure of risk and thus seen as something to be minimised. Instead, it should be seen as the cost of good long-term performance, and a small one at that. In sport, winning requires guts. It may be the same in fund management.

Don't tar all actives with the same brush

(Published in the Financial Times on 30 May 2010)

The active versus passive management debate is intensifying. In FTfm the two sides have traded increasingly hostile blows, with the most recent attack (passive on active) by Richard Stott (Twisting the facts on active management, May 9 2010).

Aberdeen does not feel particularly threatened by the world of indexing — after all, both camps believe in diversification, just to varying degrees — but we do get concerned when the engagement gets a bit nasty and puts the game itself in jeopardy.

In combat it is important to understand one's enemy. This, however, requires you first to have identified him correctly. And this is where the active world makes its first, and rather basic, error. As the passive world's arrows rain down, most wrapped with the message that the average actively-managed fund underperforms, the natural response of the active camp is to assume that the enemy is the side firing the arrows. This, we think, is wrong.

It is not the passive world that is the enemy but those in the active world who make us a target. In other words, the two-thirds of managers who consistently underperform and in the process give the rest of us a bad name. In the medical profession, underperformers get disbarred. Not so in the fund management world.

But while we do not take issue with the concept of index investing — it makes sense for those who have not the time, inclination nor ability to identify a skilful active manager — we do take issue with the arguments used, the main one being that the average actively-managed fund underperforms its benchmark.

This may be true, but lumping us all together in the same pot is absurd. We all have different investment strategies — long term/short term, value/growth, concentrated/ diversified, etc — and different track records.

While the average fund may underperform, about a third of active managers do better than a passive fund.

The question is, can these skilful managers be identified? The indexers, quoting the deeply flawed efficient markets hypothesis, claim not, but research by Martijn Cremers and Antti Petajisto of the Yale School of Management, looking at fund performance from 1990 to 2003, suggests otherwise.

Their research maps fund performance by tracking error (the extent to which fund performance deviates from that of the benchmark) and active share (the extent to which a fund's holdings differ from benchmark constituents). Their results showed the following.

Funds with low active share underperformed regardless of whether their tracking error was low (benchmark huggers) or high (top-down strategies/factor bets). Funds with high active share but low tracking error (diversified stock pickers) were able to outperform but this out-performance got wiped out by fees. However, funds with high tracking error and high active share (concentrated stock pickers) outperformed by 1.0-2.8 per cent per annum, even after fees.

These results may shock those who think constraining tracking error is a good idea but they are completely logical. Funds that do well are those that take no notice of what is in the benchmark. Those cowards who stick close to the benchmark, or those who try to predict systemic factors, appear destined, like the roulette player who stays at the table, to fail.

Albert Einstein allegedly referred to compounding as the most powerful force in the universe. Once it is appreciated what annual outperformance of 1.0-2.8 per cent does to a fund over 10 years, it becomes clear that high tracking error is a small price to pay and that it might perhaps be worth gambling on there being something to this active management lark, or, rather, certain parts of it.