The most over-hyped job in hedge funds
If someone told you they were a trader for a hedge fund, would you be impressed? Probably. But hedge fund traders are not as important as you might think they are. On the sell side traders are right at the top of the front office food chain, but buy side traders don’t usually make big decisions about where to commit the firm’s risk capital.
Instead those decisions are made eithier by human portfolio managers, or increasingly by computerised trading strategies. Hedge fund traders are execution traders. They have one job – to execute the trades demanded by their human or robot masters. So, for example, a portfolio manager or computer strategy might decide that the fund needs to lose 20% of it’s exposure in Uber stock by the close of business. The execution trader then has to sell all those shares before the market closes, and receive the best possible price.
Without p&l responsibility, execution traders are relatively unappreciated. They are also seriously underpaid when compared to hedge fund portfolio managers. Given the hype about super-fast, fully automated, trading algorithims you might be surprised to learn that there are still humans engaged in such a mundane task.
In fact, execution traders are still a vital part of many hedge funds.
Firstly, there is quite a lot of trading that is difficult to automate. Not all products trade on markets which can be connected to via automated APIs. Some trading still takes place via manual interfaces, Bloomberg messaging, or even the thoroughly old fashioned medium of the telephone.
Secondly, there are still many markets where execution traders still do a better job of trading than automated algos. Markets that aren’t especially liquid, or have other unusual uncharacteristics, are hard for algos to get right.
For a small fund, or one that has just started up, it may be unwise or too costly to use fully automated trading systems. Writing code that is robust enough to trade fully automatically is no easy task, and is of limited value compared to time spent optimising your trading strategies. I know of several relatively large systematic trading funds that generate their positions fully automatically, but then rely on execution traders to actually submit the orders.
Of course, most funds do automate many of their trades, but execution traders still have other important roles besides actually trading. Execution traders can also help portfolio managers and trading strategy designers do a better job. They are in regular touch with the market, and can pick up valuable intelligence and market colour. Human traders can make themselves more valuable by helping firms comply with changing regulatory demands, such as temporary bans on short selling, or the best execution policy mandated by MFID-II.
This job is even more important in a systematic trading fund, where the strategy designers are often from non financial backgrounds with Phds in maths or physics but no market experience. Experienced traders can help ensure their strategies don’t do anything that is imposssible or crazy.
The deep experience of execution traders can help the trading algorithim designers to improve the performance of their algos. In the long run this will mean fewer execution traders are required, so hedge fund managers need to incentivise them properly.
Funds that do not have their own trading algos may outsource the job to banks or specialist firms like Quantitative Brokers. Experienced execution traders are often the best people to manage the relationship and ensure the fund is getting a good level of service.
With multiple routes to market (human traders, in house algos, outsourced algos, outsourced human traders) it is important to use proper benchmarking to ensure that everyone is kept honest. Benchmarking is easiest when you can compare like with like. So it may make sense to randomly allocate orders to different routes, even when this is sub-optimal. Your internal algo might be better at trading S&P 500 futures than your human execution trader, but you still might want to send 10% of your flow to the human trading desk.
As well as checking you are getting good value from your algo, this will be very helpful if your algo system ever crashes. An airline pilot who always uses the autopilot will forget how to fly manually in an emergency, so pilots are encouraged to regularly practice controlling planes the old fashioned way. Similarly, human execution traders need to keep their skills fresh.
With proper benchmarking it is relatively easy for execution traders to prove they are adding value, versus other execution options. However, it is notoriously difficult to work out whether a portfolio manager is generating statistically significant alpha. Poorly paid execution traders might want to consider this when they are negotiating their next bonus.
Robert Carver is a former head of fixed income at quantitative hedge fund AHL, where he worked with many brilliant and underpaid execution traders. In the distant past he has also worked as a sell side trader. Robert is the author of 'Systematic Trading' and 'Smart Portfolios'.