Join our community of smart investors

When style drift pays, and when it doesn't

The stats brains at quant firm AQR believe they've found a way to make style drift pay, but as the Woodford debacle illustrates, if you're not an algo-trader, not sticking to your knitting can prove very costly
When style drift pays, and when it doesn't

The fall from grace of former investment superstar Neil Woodford was the product of several missteps. While the more egregious of these are likely to continue to dominate the headlines, no account of his travails would be complete without considering the turn in fortune experienced by his traditional style of income investing. Indeed, while there’s no denying the extent to which Mr Woodfood’s portfolios drifted towards holding highly speculative illiquid investments, had his holdings in large liquid income plays delivered strong returns, events may not have proved as cataclysmic – or at least not so soon.

A sharp turn in fortune of a previously successful strategy is an inherent risk in fund management and investing generally. Managers that retire without encountering such reversals often achieve near mythical status, despite the near certainty that luck plays a noteworthy role in such outcomes as well as skill (many believe what is classed as active manager 'skill' is simply mislabeled 'luck'). In the past, Mr Woodford had actually been lauded for sticking with his income investment style while enduring significant underperformance during the dot-com boom. Then he was judged to be fully vindicated by his outperformance during the subsequent dot-com crash; this time he may not have the chance to salvage his reputation.

Changes of fortune for investment strategies beg the question of how quickly fund investors should be to act when an erstwhile star manager starts to underperform. Two research papers from the past 12 months offer some interesting, if somewhat contradictory, perspectives on this issue.

The most recent of these papers, published in February, comes from the stats brains at quantitative investment firm AQR. Building on earlier work from the firm, Tarun Gupta and Bryan Kelly tested a “momentum” investment approach that chased the market’s hottest investment-strategy trends. Quants often call investment strategies 'factors' or 'smart beta'. Such factors can be based on something as simple as buying the cheapest shares on a common measure such as the price/earnings (PE) ratio, to something a bit more complex, such as buying high-yield shares with low betas (a measure of risk).

To get quants really excited, the distinguishing features a 'factor' needs to show are that it has historically produced strong performance relative to the risk taken on – known as 'alpha' – and that it is distinct from phenomena that have previously been identified. Finance wonks have found a vast number of such factors and Mr Gupta and Kelly looked at the performance of 65 of the most established. The key focus of the study was the US market starting in the 1960s, but similar results were found for global stocks and for a strategy based on just six factors.

They found strong evidence that “persistence in factor returns is strong and ubiquitous”. In other words, if a strategy has performed well or badly in the recent past, it’s likely to continue to do so in the near future. Given this persistence, it may not come as too much of a surprise that the research showed it is possible to produce significant 'alpha' through a highly disciplined factor performance-chasing strategy. The research focused on scaling exposure to strategies based on recent performance (one-month performance worked best) and rebalancing factor exposure every month.

This so-called 'time series factor momentum' (TSFM) strategy produced superior performance to any single factor and, using its best formulation, an annualised average return of 12 per cent. What’s more, by quickly switching out of hot strategies when they blew up, TSFM avoided the gnarly bear market drawdowns experienced by momentum strategies focused on individual stocks or industries. Indeed, while a stock-centric momentum portfolio experienced a 31 per cent loss between March and May 2009, TSFM actually earned 18 per cent.

While AQR’s observations sound like potential investment gold, for private investors the fact that these factor-momentum portfolios were rebalanced monthly should give pause for thought. Outside the world of algorithm-driven trading, deliberations about portfolio overhauls, especially investment-strategy overhauls, tend to drag out for much longer and be more nuanced. Indeed, there is a body of evidence that in the real world, performance chasing is more likely to work to investors’ detriment than to their gain, which brings us to the second paper being considered by this column.

Late last year another quant firm, Research Affiliates, published a paper by John West and Jason Hsu that highlighted studies that had sought to quantify the cost of poor timing when switching between actively managed funds. Observations cited ranged from the tendency of fund managers dropped by pension funds to outperform their newly hired rivals by an average of 1.4 per cent a year for the next three years1, to findings of a 1.9 per cent negative impact from poorly timed US equity fund purchases and sales from the start of 1991 to mid 20132.

Indeed, these numbers suggest the issue of poor decision-making by fund investors is a much bigger issue for performance than the evidence of underperformance, on average, of active funds versus passive funds. As the paper says: “The oft-cited negative alpha of active versus passive equity management – what we’ll call the manager returns gap – obscures a far larger performance issue, the investor returns gap.” In fact, it is hard to see the two issues as being unrelated given the propensity for withdrawals from active funds experiencing bouts of underperformance to lead to closures, as well as the clustering of new fund launches around 'hot' investment styles.

Like AQR, Research Affiliates looked at a number of popular factor strategies – 29 in total – and found that since 1968 they had delivered a median annual outperformance of 1.5 per cent, or about 1 per cent after notional costs. Not bad, but by no means good enough to offset the 1.9 per cent lost due to the poor timing decisions. The quant’s conclusion for investors was simple: “The majority of performance-chasing [smart-beta] investors would have been better off in buy-and-hold cap-weighted indices… Is this a rock solid argument for going passive? Hardly. It’s a rock solid argument for not chasing past performance. Investors should pick a strategy they believe will add value – and has added value using live assets – and stick with it.”

For those not seeking passive exposure, this conclusion rings of common sense. Few investors are able to regularly reinvent their investment approach, whether they be trend-followers of the TSFM ilk, or the kind of dogged pursuer of dividend that Mr Woodford built his reputation as. Indeed, from the perspective of investment style, the key lesson from the Woodford debacle centres on the perils of so-called 'style drift', which has the potential to both introduce underperforming investments to a portfolio while also distracting attention from applying a tried-and-tested approach.

.