Breaking the ‘curse of benchmarks’
Is short-termism and a slavish preoccupation with benchmarks by institutional investors hurting markets?
Market ups and downs: a focus on short-term performance means money managers are engaging in activity that is causing bubbles and crashes
Institutional investors’ obsession with short-term performance means money managers are engaging in activity that is “not merely superfluous but wealth-destroying”, causing bubbles, crashes and financial market dysfunction.
So says former fund manager Dr Paul Woolley of the London School of Economics (LSE), co-author of a new paper that pulls no punches regarding the “absurdities” that characterise the world of institutional investment.
Equity markets are increasingly dominated by institutions. Between 1900 and 1945, institutional investors controlled roughly 5 per cent of the US stock market; by 2009, that figure had risen to 67 per cent.
Woolley knows the world of institutional investment well, having founded and run the UK arm of fund management giant GMO for 20 years. Returning to academia in 2007, he funded the LSE’s Paul Woolley Centre for the study of capital market dysfunctionality, and is especially concerned with one issue in particular – benchmarking.
Active fund managers’ performance tends to be measured against a given benchmark such as the S&P 500 or the MSCI World Index. Active managers usually seek to beat their benchmark but not to trounce it – a common instruction would be to aim for annual returns three percentage points over benchmark returns, says Woolley, while not deviating from the benchmark index by more than a predetermined amount.
Consequently, funds must keep a close eye on their benchmark: if a fund manager does not like a particular stock or sector, for example, he might underweight it rather than avoiding it entirely.
There are good reasons for benchmarking, which prevents incompetents from taking excessive risk, as well as allowing fund returns to be compared to the overall market.
Unfortunately, argues Woolley, it also creates perverse incentives that result in herding and market distortions, with fund managers needing to be especially mindful when it comes to underweight positions in high-flying large-cap stocks and sectors.
If a security doubles in price and the manager is half-weight, the mismatch compared to the index doubles. However, if the manager is double-weighted and the price halves, the mismatch also halves.
In other words, a manager will be under much more pressure if he is underweight in a risky stock that rises than if he is overweight in a risky stock that declines, incentivising him to be fully weighted in risky stocks he might otherwise avoid.
Even if the manager does opt to underweight such stocks, he will be forced to buy if his returns deviate excessively from the index.
This, says Woolley, is the plight of value managers “forced to buy bubble stocks they know to be overpriced”.
Value funds benchmarked to any index, he says, “will be a mix of carefully selected cheap stocks combined with a bunch of dear stocks bought to comply with the tracking constraint”.
The problem of overvaluation is exacerbated by momentum traders, says Woolley. Momentum investing – essentially, buying stocks that have been rising in price, frequently rebalancing so that a portfolio always consists of the stocks with the most momentum – is an age-old and successful strategy. It can also be used to exploit the predictable behaviour of benchmarked funds forced to buy rising stocks; momentum traders buy these same rising stocks and dump them as they lose momentum, leaving the benchmarked fund manager as the fall guy left holding the bag.
This “exploitation” is “stepping up to a new level of sophistication” due to “predatory” computer-driven models seeking to “game the trades of traditional players”.
Risk inversionThe net effect is that risky stocks are driven up to unrealistically high levels, ensuring future underperformance; in contrast, low-risk-stocks become too cheap, driving their future returns higher.
This seems counterintuitive. After all, investors are rewarded for taking risk, so riskier securities should, in theory, deliver higher returns, as they did between 1926 and 1968. Since then, however, a deep inversion has occurred.
Woolley cites GMO data showing the riskiest quartile of US stocks has delivered annualised returns of 7.2 per cent since 1970, while less volatile equities returned 10.6 per cent annually. More recent global data is even starker: since 1984, volatile stocks have returned just 4.1 per cent annually compared to 10.1 per cent for less risky equities.
This “risk inversion”, concludes the LSE paper, “is the inevitable consequence of benchmarking.”
Pricey stocks get pricier, then, while cheap stocks get cheaper, but the two forces do not cancel each other out: as the effect is stronger for rising stocks, this results in “secular overvaluation” – that is, permanently overpriced markets.
Many commentators have been scratching their heads for some time now as to why US stock market valuations appear to have remained well above their historical norm for almost all of the last two decades, with the most popular explanation being that Federal Reserve support has been keeping stocks artificially high ever since Alan Greenspan became chairman in 1987. However, Woolley’s analysis suggest that benchmarking, not the Fed, is the explanation.
The effects of market mispricing and the misallocation of capital are most damaging when an industry sector or entire asset class is involved, with Woolley pointing to the technology bubble of 1999-2000, the commodities boom in the mid-2000s and the Japanese equities bubble of the late 1980s. Obviously, many funds are not tied to benchmarks, although fund managers in general keep a close eye on competitors’ returns, so Woolley suggests there is an inbuilt pressure to buy into even the bubbliest of environments.
GMO, Woolley’s former fund, which is renowned for its strict adherence to value-investing principles, comes to mind in this regard. Several GMO clients actually banned the firm from its buildings during the late 1990s due to its refusal to buy into the technology bubble, with the firm’s assets declining by almost 45 per cent during that period.
QuestionsIs Woolley right? Is benchmarking distorting financial markets? Benchmarking is one explanation for the persistence of seemingly elevated valuations in the US, although bulls would argue other secular forces are also at work.
Even if one accepts that benchmarking is driving US valuations, however, one must ask: why has it not done the same in Europe, where current equity valuations appear relatively reasonable, and which was dirt cheap at various stages in recent years?
Additionally, many institutional investors would reject the argument that their actions are irredeemably compromised by benchmarking or short-term concerns about their competitors. Similarly, one could argue market bubbles and crashes are driven more by animal spirits than by benchmarking, and that behavioural factors may also explain the aforementioned low-volatility anomaly.
Woolley’s paper is ambitious – a sceptic might think it is too ambitious, and possibly a touch hyperbolic at times. Nevertheless, his CV is an impressive one, as is that of his co-author, Prof Dmitri Vayanos, who worked at Stanford and MIT prior to joining the LSE.
Additionally, it is certainly plausible that benchmarking, coupled with the huge growth in passive investing – in the US, approximately 30 per cent of investors’ money is now tied up in index funds – may well be resulting in increased herding behaviour. It’s notable that in the US, for example, stocks in the Russell 2000 index were trading at a premium of 50 per cent to similar non-index stocks at the end of 2015, according to S&P capital IQ data.
Next stepsWhat should be done? Woolley suggests fund managers commit to investing on the basis of fundamental value rather than slavishly following market benchmarks. A multinational agency such as the Bank of International Settlements or the EU Commission could promulgate a code of best practice, while tax concessions could be withdrawn from momentum funds showing high turnover.
Unless some such changes take place, he concludes, both the wider economy and asset owners will suffer courtesy of the “curse of the benchmarks”.