Like the yeast-based spread, much goodness might be missed by lumping all quantitative funds into one bucket all too often labelled a ‘black box’. Investors eschew these systematic traders for a variety of reasons, some of which are made on fear-based assumptions, which is ironic because one of the premises of quantitative investing is that decisions driven by emotions are eliminated from the investment process.
One of the first points to make is this: ‘quantitative investing’ is not really a strategy. The term is used to distinguish ‘how’ the strategy is created and implemented, namely a systematic method that seeks to eliminate the arbitrariness that so often pervades discretionary investing strategies. Secondly, it appears that many investors still believe the term ‘quantitative managers’ to be synonymous with trend-following CTAs/managed future funds.
Once you lift the lid on managers that adopt a systematic approach to managing money there is a significant variety not only in the strategies that underlie the generic term ‘quant’ but in the ways quantitative strategies are implemented. For example trading speed, average holding period, portfolio construction and risk management techniques are all keys to a universe of potentially accretive and non-correlated return streams.
Dispelling myths and quashing fears about systematic managers is a challenge, not least because quants are often associated with a lack of transparency – the term ‘black box’ adding to the perception of opacity.
Most quants are fiercely protective of their secrets because their proprietary datasets and research are their lifeblood and integral in their quest to maintain an ‘edge’ over the competition.
To move beyond fear to maximising potential returns, it is important to dispel the belief that this protectiveness of intellectual property means that nothing is known about how quants make money. The variety in the strategies they deploy can be seen further down below under Quant Strategy Highlights.
For some, the memory of the ‘Quant Meltdown’ of August 2007 is still fresh. In the week of 6th August 2007, a number of quantitative long/short equity hedge funds experienced losses of greater than 20%.
The true trigger of the moves is not known, but it seems likely that losses from the credit crisis led to systemic deleveraging that then saw some firms taking down market exposure across their full range of strategies. This created a negative feedback loop in the quant space causing further losses and deleveraging.
Two big lessons have been learned. The first is that the common factor risk between quants was much higher than investors and the quant funds themselves had appreciated. The second lesson revolved around excessive use of leverage. To achieve returns and volatility targets, quants had created a toxic mix of incrementally increasing leverage and assets, all chasing small opportunities. So when the shock came, the effect was amplified causing violent P&L losses.
Although many of the factors and stocks that had been so violently hit in the deleveraging quickly reversed and some quant managers saw lucrative trading opportunities in the aftermath,for most the ensuing panic changed the quant trading landscape. Post August 2007, we have witnessed a significant reduction of assets investing in longer-term quantitative equity models and a massive reduction in leverage.
Today, however, managers are far less complacent and more diversified than they were previously. An interesting side-point revolves around the recent proliferation of alternative beta and risk parity products; is the recent massive rise in ‘alternative beta’ and ‘risk-premia’ harvesting strategies sowing the seeds of an August 2007 style scenario?’ The enormous asset flow into highly commoditised areas combined with leverage is a story we have seen before.
So how do you pick a winner in the quant space? Rishi Narang, author of Inside the Black Box, provides one perspective: “the systematic implementation of trading strategies that human beings create through rigorous research.” He highlights an important aspect of quantitative investing: namely that people, not machines conduct the research, decide the strategies, and select the universe of securities to trade and what data to use.
Ironically, by investing in a quantitative manager an investor is making a highly qualitative decision. What may have worked to produce 95% up months in the fund 10 years ago is unlikely to be what they are doing today, given the constant changes witnessed in markets. Never has the adage ‘adapt or die’ been more true than when dealing with the quantitative trading space.
Given the inherent necessity for quant managers to be able to evolve, any investor is essentially taking a bet on the manager’s robust research process; ability to innovate and continue to add to alpha sources; manage risk; and adapt to ever changing and more complex fragmented global markets. So how can investors mine the nuances of the quant universe’s diversity to access a potentially important tool in the alpha harvesting armoury?
Some questions to consider. How diversified are the sources of alpha and of those, how many are unique to the manager? Is the manager diversifying across models, region and time horizon? Correlation of the alpha sources and crowding in the space are also critical elements to assess, as August 2007 highlighted. What is the capacity of the fund if many of the sub-strategies are highly capacity-constrained? This last one is the ‘million dollar question’.
As with discretionary managers, strong references are clearly essential, however, having experienced allocators, can give an investor a significant information edge. Aurum has been investing in quant strategies for 20 years and the CIO and analysts have developed long-standing relationships with leading thinkers and practitioners in this space.
One very important factor when looking for quant teams is to make sure they have a balance of analytical/mathematical skills in combination with experience of having traded real markets, preferably over several cycles, testing their approach through extreme periods and tail events: a portfolio manager with a pure academic approach and no ‘real-world’ experience is a ticking time bomb.
When it comes to the assessment of historical performance, investors should also look at returns on the gross book deployed and take into account of the use of leverage. Here one also needs to look at how frequently the manager is turning over the book and how many independent ‘bets’ are made over any given time period.
Every quantitative investment strategy can be explained without giving away the ‘secret sauce’, but knowing how they manage risk and what protocols they have in place for sudden market dislocations such as August 2007 is paramount.
We believe that if one takes the time to lift the lid on the so-called quant ‘black boxes’, understanding them is not so daunting a task as it may first appear. It can ultimately give investors access to a whole new ‘tool-kit’ and can open up new sources of diversifying returns that in today’s world of ultra-low interest rates is worth its weight in gold.
Quant Strategy Highlights
Quantitative equity market neutral: Alpha is predominantly provided by fundamental factors that drive stock selection; not entirely dissimilar to the way equity analysts normally look at stocks – also sometimes known as quantitative long/short. Here quant managers indicate that they have a significant advantage in being able to crunch infinitely more data than a single human being could ever process and make a lot more independent ‘bets’ where they believe they have a slight ‘edge’. The process is stable and repeatable and theoretically should deliver a higher Sharpe ratio over the long run at lower risk as these portfolios are typically constructed to be neutral to the broader markets. More sophisticated approaches tend to go even further, attempting to constrain the portfolio’s exposure to other generic ‘risk-factors’, such as market cap, geography, value or momentum biases etc. Others attempt to ‘time’ or dynamically adjust exposure to such risk factors in the hope it will be accretive to returns.
Statistical arbitrage: Typically where the inputs are price based and/or other technical signals, e.g. where one would expect certain pairs or clusters of stocks to behave in a similar fashion/predictable manner. Statistical arbitrage uses various mathematical techniques to identify significant explanatory relationships between the movements of different instruments; they then construct trading strategies to benefit from discrepancies between model predictions and the observed market. The most ubiquitous models are variants of mean reversion and momentum strategies.
Event-driven: The identification of certain market participant actions/events in the market that typically results in prices moving in a predictable manner, e.g. price action of stocks around the announcement of index rebalancing, merger targets, analyst earnings estimate revision announcements, tax-year end dates, option expiry dates, futures rolls, etc.
With the exception of trend-following, futures and specialist FX managers may trade more technical relative value strategies between instruments over time frames ranging from fractions of a second to a day; they use countertrend, carry and econometric models; and global asset allocation strategies that use both technical and fundamental information.
FX managers may also use econometric techniques as well as various carry models and fundamental analysis of countries’ macroeconomic data
Quant macro: This includes various combinations of the above, plus cross asset relative value and may add other areas like volatility trading, fixed income curve models and fundamental commodity models using supply/demand metrics and predictors.
High frequency trading and intra-day trading: (HFT): No quant article would be complete without mentioning HFT; but it is important to distinguish between the above strategies and HFT, which has had a lot of a ‘bad press’, as well as ultra high frequency traders who simply make money by being the fastest, and those whose strategy uses low latency execution technology infrastructure in order to exploit temporary anomalies in pricing. HFTs are here to stay as they ultimately get paid an economic rent to provide liquidity to markets. In the strictest definition of the word pure high frequency trading arbitrage strategies is the domain of proprietary trading shops; while there are a handful of sophisticated hedge funds that are trying to identify and capture very short term anomalies (not HFT). This whole space is highly capital/capacity constrained, and while it is associated with providing very high Sharpe returns, it is often very difficult for typical allocators to quant to gain access.
Source: Aurum Funds
