What is 'quant' or 'algorithmic' trading

The word 'quant' or 'mathematics' as an approach to investing rarely causes any excitement, rather, it causes some investors to walk away and others to either fall asleep or at best raise their eyebrows. We do not think this is justified, as quant is not just bean counting. We believe that a quant approach to investment is superior to a traditional discretionary approach and to show this, we list some of the main advantages of quant.

- In a quant investment process, focus is on the scientific approach as strategies are researched and tested before they are implemented. Indeed, only empirically proven strategies should be given a chance in the real world. The longer the testing period, the better, as the empirical evidence becomes more reliable. Quantitative investment processes are often perceived as lacking human judgment. This, in fact, is not true as human judgment indeed plays a critical part during the model building stage. The main difference between a quant and a traditional manager therefore is in the timing and not the presence of the judgmental input. Intuition, experience and understanding are important in research but not a feature in implementation.

- This leads to another advantage of quant. The implementation is model driven and therefore emotionless. And we know well that emotions and trading don't mix well. In fact, emotions often cause behavioural price anomalies (such as overreaction) that can be exploited readily by quant models. At the same time quant models can trade more cheaply and more efficiently. In this area program trading and algorithms are already being recognised as adding value compared to a traditional discretionary approach and their penetration is now widespread amongst quants and non-quants alike.

- Better timeliness is achieved by being able computer processing power. In contrast, discretionary managers can hold a rather more limited number of 'ideas' to analyse new information more quickly and simultaneously for thousands of stocks due to ever improving at any one time.

- More accurate volatility targeting is provided by quants who would typically exploit a linear relationship between leverage and volatility. Discretionary managers would typically seek higher volatility through concentration but as this relationship with volatility is more complex (i.e. quadratic), volatility targeting is much more difficult. This yields better Sharpe ratios for quants and it also implies better and easier structuring of leverage and portable alpha products for institutional clients.

- A further advantage concerns growth of the business. Better assessment of capacity is achievable through simulations in an environment that is already designed and built on simulations.

- In addition, the same simulation friendly environment yields easier and better risk testing. Again this is no longer thought of as a quant only feature. Few discretionary managers will admit to not using quant at all for risk measurement. However risk measurement is one thing and risk management another -merely measuring the risks might not lead to any action even in the case of excessive and/or unfamiliar risk taking. As in the case of some investors who 'do not care about Sharpe Ratios' many discretionary managers may not care either: 'If I look after the return, then the risk looks after itself.'

For all these reasons we don't think that you have to be a quant to understand the advantages of quant. Hardly surprising then that many of the biggest and the best performing fund managers are quant.

Why should Quants improve with time?

Recently, there has been a tendency amongst hedge fund investors to pursue hot new managers. This is justified on the basis that managers are more exciting early on as they are hungrier for returns and that their competitive edge declines with time. This is presumably because managers expand too much and lose alpha as they get too big at the same time during which they lose their competitive edge. But the obvious problem is that in absence of hard evidence, selection needs to be made on 'gut feel' and other soft criteria such as reputation (often in a different field!!!) or following the herd.

Not surprisingly, such an approach is not likely to bear fruit for many. Often start ups have failed to perform to expectations and we are aware that in the recent days at least one multi billion dollar quant start up is closing down after not much longer than a year.

We are not surprised with the difficulty in picking early stage quants, as good quant managers are unlikely to do better earlier rather than later on for the simple reason that good models do not decline but improve with time. As more and more data becomes available, better calibration can be made of the existing models and new strategies can be added to the process. This includes models that rely on data that can only be obtained in real time, as no historical databases exist for it. Such models are often called 'the Insider Models' as they do not rely on publicly available information but the manager's own databases. And the more such data exists, the better. Furthermore, the stronger the investment process, the more it makes sense to increase return and volatility targets. This is what comes with time.

The message is clear - 'Past performance is not indicative of the future performance. In the case of a good quant, it should get better!' Ignore established and successful quants at your peril.

Hedgeweek