Global Pensions asked a panel of experts to share their thoughts on quantitative strategies and how they are performing in the current economic climate
Why were quantitative strategies hurt so badly during the credit crunch?
Mike Arone: It’s not a case of quantitative strategies being hurt badly during the credit crunch. Rather, the disciplined, bottom-up, stock picking approach employed by most quantitative and many fundamental managers has not been rewarded during the financial crisis.
For quant managers valuation and sentiment are two of the key determinants of an investment’s attractiveness. Value investing identifies assets that are inexpensive relative to their financial fundamentals and comparable investment opportunities. Sentiment or momentum investing exploits the tendency for rising asset prices to rise further. Yet value and sentiment investing are not the sole preserve of quant managers.
A fundamental stock picker seeking reasonably priced stocks with a catalyst for improved performance utilises a similar investment approach, but more subjectively. Whether employed quantitatively or qualitatively this active investment approach has performed badly during the credit crunch.
For investment managers seeking to add value through disciplined, bottom-up, stock picking, performance has fluctuated violently as macroeconomic expectations have changed. Successful investing has been driven more by profitable industry and style timing, which has depended significantly on being on the right side of those changing macro expectations. The good news for quant managers is that stock returns are becoming more differentiated within industries and industry performance is becoming less dispersed. Stock selection should therefore start to outperform industry and style timing in the near future.
Diane Miller: To understand what went wrong, we need some background. Quant techniques are used across a wide spectrum of products, from long only equities to currencies to hedge funds. Growth of quant has been helped by improved computing power, academic research into suitable factors and ready availability of data. Also, as short selling became more acceptable, long only quant managers appeared ideally placed to extend their processes to sell investments that scored poorly.
Like other investment trends, quant became a victim of its own success and excess returns fell as more competition was attracted. The prolonged period of benign conditions lulled many into a sense of false security, even as managers struggled to repeat their past success. Use of standard optimisation techniques developed for a world that followed the theoretical models left many strategies on autopilot. As volatility fell, the models increased position sizes to achieve risk and return targets.
All this left many managers vulnerable to the dislocation in markets, when reality was no longer anywhere near the theory. Although some had realised that the real world was different, they failed to anticipate just how dangerous the difference could be.
Most models were developed for a world that behaved in a certain way, for example reacting positively to earnings upgrades, not one totally focused on risk aversion. Problems were exacerbated when volatility spiked and position sizes calibrated for a more clement environment were now too large.
Risk models were often too slow to adjust. In the worst cases, for strategies that used leverage and short selling, losses caused prime brokers to reduce borrowing limits, forcing a liquidation of investments into difficult markets. This was compounded for those who faced redemptions, often from hedge funds of funds that needed cash to meet their own redemptions.
Have these strategies recovered? If so, why? If not, why not?
Mike Arone: The recovery of quantitative strategies has been mixed. Quant managers that have successfully recovered have struck the right balance between evolving their investment approach for the new market environment while sticking to the core components that have worked successfully for more than 80 years.
Quant managers that surrendered to the ‘you must radically change or die’ pressure would have missed out on the positive returns from the value components of their approach during long stretches in 2009. Those that evolved their investment approach by identifying new distinct factors, diversified sources of data and/or improvements in dynamically tackling changes in market sentiment have recovered strongly.
Anecdotal evidence and positive performance results suggest that quant managers focused on the eurozone have done a better job of striking that balance compared to their quant brethren elsewhere.
Diane Miller: Not all quant strategies suffered. Some did very well, especially trend following strategies that ignored valuation. Others were managed conservatively with an awareness of what could go wrong or avoided ‘crowded factors’ – those that led managers to favour similar investments for the same reasons. Managers with a strong capital preservation ethos also recognised the warning signs and reduced risk, often overriding their models.
Recovery is difficult for those strategies that experienced the most severe losses, if they are still operating. Where managers had problems but avoided the worst, then the challenge has been to rebuild performance. This will take time but now that we have passed the eye of the storm, markets seem to be paying more attention to factors such as valuation and growth again. This should allow the better quant strategies to add value once more.
How did pension funds react to the underperformance?
Mike Arone: Pension funds’ heightened sensitivity to the liquidity of investment portfolios and cash flow needs as well as a notable shift from active to passive mandates has contributed to a small reduction of market share for quantitatively focused managers. At the peak of the crisis many pension schemes were forced to sell their liquid investments into weak markets to fund cash needs and to meet prior commitments to investment funds. In addition, approximately 20% of the institutions surveyed recently by Greenwich Associates1 shifted assets from active to passive mandates over the past 12 months, and a similar proportion expect to do so in the coming year.
Clients invested in quantitative strategies, and their consultants, have significantly increased their efforts to better understand quantitative investment processes and the drivers of performance. The rise in frequency, quality and intensity of communication between managers, clients and consultants has benefited all involved. For some, this process has re-confirmed their approach to investing in a well-diversified portfolio of investment managers, geographies and investment disciplines. For others, it has highlighted that opportunities to add value through sound quantitative investing have never been greater. As a result, some clients have committed more assets to existing quantitative mandates.
Diane Miller: Their reaction depends on how bad the experience was and whether they still have confidence in their managers.
In most cases where performance has not been too bad, clients have been prepared to give managers the benefit of the doubt and regard recent markets as atypical. Whether they continue to do so depends on how quickly managers can resume outperformance. Some strategies have done better recently, but the evidence is still short term.
It has long been recognised that quant strategies struggle at market turning points and the 2008 turning point was unusually violent in its impact. However, pension funds understand that the advantages that helped quant in the past are still there – namely, the focus on factors that have been proven to add value when applied in a systematic manner, the ability to collect and process a large amount of information from different sources, and the exploitation of some behavioural biases and avoidance of others.
How have quantitative managers changed their processes to respond to the underperformance? For example, have they begun to include more fundamental factors in their overall investment analysis?
Mike Arone: All active investment managers – quant or not – must constantly evolve their investment approach to maintain an information advantage in the market. While quantitative investment strategies have a strong long-term track record of delivering risk-managed returns even the most successful are not immune to periods of underperformance. Many quant managers believe that dramatically changing your investment process after short-term underperformance means making changes at the worst possible moment.
Every behavioural bias that makes quantitative investing work over the long term also makes us want to ‘fix it’ when it is temporarily out of favour. Quant managers have therefore spent the last 18 months fighting the urge to surrender to their own behavioural biases because they think any changes would be poorly timed and because the opportunities for more traditional quantitative models are currently very attractive relative to their history.
However, quant managers do spend considerable time on research. Many are developing more sophisticated models that incorporate less widely known signals, data sources and implementation techniques. Managers are also formulating dynamic, factor timing strategies that enhance return or reduce risk.
Henry Ford once described failure as “the opportunity to start again with better information”. The recent challenges for quantitative investing approaches have set the stage for its continued evolution.
Diane Miller: The first step for managers is to understand where things have gone wrong and whether it is a one-off event or the start of a new paradigm. Many have tightened up on their risk management. The unquestioning reliance on standard risk models will never be the same, although the next problem is bound to come from a different direction anyway. The importance previously attached to following a strictly defined process may be tempered by the realisation that flexibility can be an advantage at times.
This could encourage development of more dynamic processes and greater use of judgement to adapt models or reduce risk. The search for new ways to exploit sources of added value will continue and the better practitioners will take care to avoid exposure to crowded factors and attempt to collate proprietary data. We are already seeing managers become more protective of their intellectual capital and reluctant to divulge too much detail about their process. That conflicts with demands for greater transparency from investors, although a middle ground may be increased position disclosure.
Most respondents in this week's Pensions Buzz do not think businesses should be able suspend AE contributions if in financial distress.
Former BHS owner Dominic Chappell has lost the appeal against his section 72 conviction and sentence for failing to hand over information to The Pensions Regulator (TPR).
This week's top stories include Marsh and McLennan Companies agreeing to buy JLT, and the home secretary calling for AE to be scrapped in a no-deal Brexit scenario.
Lesley Titcomb says the watchdog wants closer interactions with pension funds to spot problems sooner and act before having to use its more stringent powers