Who gains from algorithmic trading?


Published on:

Investment banks need to think carefully about which institutions they market their services to.

The head of algorithmic trading at a leading investment bank recently expressed the following view: “If algorithmic trading is so fundamentally good because it gets such great results, then in any other industry there would have already been a massive swing towards it. The reason why it does not happen in investment banking is because customers just aren’t savvy enough. It is only the smarter customers, the more demanding ones, who are prepared to pay for it. It’s the stupid ones for whom price is everything.”

Is this just a condescending and self-serving argument – or does it have some truth?

Investment banks clearly have something to gain from algorithmic trading or they would not be pushing it so hard. They have spent millions over the years developing execution algorithms to increase automation within their trading desks, and fight off the effects of relentless commission squeezes. Offering some of these directly to clients helps them to recoup some of their investment.

Most of the clients that are in a position to use directly the algorithms that brokers are offering already do some trading for themselves, through direct market access. Offering algorithms to these clients is an attempt by banks to win back some of the revenue they are missing out on from these clients.

But has a dramatic swing towards algorithmic trading failed to take place because clients are too stupid to see its benefits or because it’s simply not worth doing?

Useful evidence about the effectiveness of algorithmic trading is scarce and most fund managers are sceptical about the measurement tools offered by brokers.

It is only the biggest traditional fund managers and hedge funds, generally accepted as the most sophisticated client groups, which are big users of the technology. But then again, it doesn’t make sense for most small fund managers to do lots of their own trading, as most have neither the scale, nor the expertise, nor the technology needed to make it worthwhile.

Few fund managers are as sophisticated as investment banks. Long-only fund managers, in particular, always seem to be at least one step behind.

Apart from issues of scale, fund managers rarely had to think about the value of the sell-side’s services as its clients footed the bill. Recent regulatory pressure towards unbundling means they now have to care.

If the swing towards algorithmic trading has not yet been as great as investment banks had hoped, it could perhaps be because the case is not as overwhelmingly obvious as they make out.