Hey, Michael Lewis: Stop DEMONISING Wall Street’s SUPERHUMAN high-speed trading

HFT's NOT the free-market crusher new book says it is

What's the worst that could happen? 'Front running'

On the downside, this concentration on trading speed has enabled some market participants to see large orders before they are transacted. The US stock market include perhaps 13 or 15 actual exchanges.

Given the latency of information travel, those close to one exchange might see a large buy or sell order before other exchanges do: thus they can go and buy and sell that stock before that large order actually arrives, move the price either way by a cent or two and then sell to that large order when it arrives.

This is known as “front running” and can be thought of as being a bit naughty. Certainly, if you get the information from your mate executing the large order and do this you're guilty of insider trading. It's this bit that Lewis is decrying in his new book.

Whether this is important is another matter: most of the book is actually about an exchange that has managed to frustrate this tactic simply by having the fibre optics coiled up in certain lengths so that all market participants see the information at the same time whatever their geography. If there's such a trivial solution perhaps it's not actually a large problem.

But that front running is something entirely different from the basic concept of HFT itself. Bit like a shotgun perhaps: which can be used to get yummy peasants* for tea or to take out the troublesome ex. One isn't a problem, the other is a crime – although both use the same tool.

And the equivalent of our yummy peasants for tea here is that HFT, by its very nature, something that is going to increase the flow of traffic across exchanges. This is also known as increasing liquidity and it has an interesting effect.

There's something called the “bid-ask” or the spread. You cannot both buy and sell Microsoft at $39.95 at the same moment. You can buy at perhaps $39.97 and sell at $39.93 at any one moment: that's the spread (or the fee that you're paying to the bloke willing to do the buying and selling).

More liquidity drops these spreads – so much so that we've had a two orders of magnitude drop in the spread over the last two decades as HFT has expanded. It has plunged from perhaps 0.2 per cent of the order amount in the early '90s to 0.002 per cent today. That saves money for everyone buying or selling anything: yes, including your pension plan even if you've no stock holdings directly.

And it's pretty much a slam dunk that the savings from this collapse of spreads is larger than whatever is being picked off by that (illegal or not) front running.

What Lewis has actually done here is point to one possibly important implication of HFT: front running – which is all rather blown up into a rampage about HFT itself.

Given that HFT makes trading on a market cheaper by orders of magnitude, this might not be all that good an idea, as it confuses the benefits of the greater technology with the costs of one particular use to which people may be putting it.

But then hysterical attacks from Main Street on Wall Street are an evergreen of American public discourse, rather like Islington v The City here in the UK.

And I have seen one proposal for doing away with that front running: let's have a financial transactions tax to make it unprofitable. Something which would, presumably, widen that spread back out to 0.2 per cent again, to the cost of everyone trading anything. What was that thing about babies and bathwater again? ®

*I refer you to the errata slip at the start of 1066 And All That (see page 50 "The Pheasants revolt" et cetera).

Does Algorithmic Trading Improve Liquidity? (PDF) – Terry Hendershott, Charles M Jones and Albert J Menkveld

More on High Frequency Trading and Liquidity – Alex Tabarrok

Sponsored: 5 critical considerations for enterprise cloud backup