High frequency trading is the very soul of the Devil, apparently

By Tim Worstall , written on April 2, 2014

From The News Desk

High frequency trading is evil. At least to those who have been reading Michael Lewis' new book, "Flash Boys," it is.

The entire stock market's rigged in favour of the Big Boys, and the little guy just gets the shaft -- which is a fun bit of polemic, but it doesn't really quite cover the entirety of what has been going on. Oh sure, someone's getting the shaft, but it ain't the little guy. In fact one of the identifiable groups that is getting that broom handle to the backside appears to be Goldman Sachs, something I'm sure we can all cheer on.

And here is that cheering piece of news:

Goldman bought specialist firm Spear Leads & Kellogg for $6.5 billion in 2000. It’s now worth a mere $30 million.

The Wall Street firm has already lined up a buyer for the unit, according to reports.

The news comes, ironically, when the technological advances that put specialists out of business are coming under fire in a new book by Michael Lewis called “Flash Boys.”

Electronic trading has largely done away with the role once played by market makers, with individuals buying and selling shares of publicly traded companies. Just a little canter back through history: time was when you wanted to buy or sell a stock you told your broker, who then bought or sold that stock from a market maker on your behalf. The market maker was sitting there, promising to buy or sell at clearly marked prices while the market was open and charge some amount for that service. That is called the bid ask, or the spread. Back in the 1990s this was around 0.2 percent of the order value crossing the market makers' desks -- a bit smaller in large deals ("in size") and a bit larger in very large deals ("over size").

Maybe that was a good way to do it, and maybe it wasn't, but it was the way. Then along came the ability for software and computers to do the trading themselves. This is what all of you guys do for a living, working out how to mechanize these boring tasks. And obviously, the computers could do all of this faster than humans. So we've now got a system where at least 50 percent of stock market trading is done on computers and with algos just spoofing each other and trading whatever in pursuit of profits of a fraction of a cent per deal. That's all pretty boring, and it could be seen as a waste of time and effort and even a misapplication of the skills of those building those machines.

However, what has also happened is that all these bots trading with each other has increased the liquidity in the market on the normal day. Pretty obviously really, if software is buying ABC Corp to hold it for 11 seconds (one estimation of the average time a stock is held for these days) then there's going to be more stock trading passing back and forth across the exchanges. This is what we mean by liquidity. And increased liquidity has an interesting effect. It reduces that spread, that bid ask that is the market makers' fee. In fact, average NYSE spreads are down from that 1990s 0.2 percent to some 0.002 percent today. That's two orders of magnitude reduction.

This in turn means that anyone buying or selling for "real" purposes -- my wanting to buy one share of Youngs' Brewery, for example, in order to attend the blindingly good annual general meeting (this is indeed the usual reason that people buy stock in this Brewery... no, really) where they try out the new brews for free on the shareholders -- this means that I am paying a much smaller fee to that market maker for the privilege. For which Hurrah! etc.

So, all users of the market benefit from the way in which HFT has brought dealing costs down. The markets are more efficient, even if we can't see all that much point in the activity that causes this beneficial effect.

Now, there are things going wrong here, too. People are also using HFT to do something called "front running." That is, when there's a large order about to move through the system some computers are picking up that information fractions of second before others can. This allows the really fast computers to buy or sell appropriately, ahead of this big order, and make a fraction of a penny or two by doing so. That's actually what the protagonists of Lewis' book are shouting about (and the problem they're solving, too). Which is not HFT itself, but one of the uses to which the technology is put.

This front running doesn't happen with small orders -- you or me or our 401 (k)s. We're way too much small fry for anyone to do this to us. And we are the people who get some of the benefit of that smaller spread.

I'm not defending the front running, only pointing out that overall HFT has been of great benefit to everyone trading on the markets. Well, okay, almost everyone. We all get lower trading costs. The HFT firms are making money doing this. But it's worth noting that they're making less than the older market making firms were. Which means that we can identify whose hide our lower trading costs are coming out of.

For example, Goldman Sachs, which appears to have just lost $5,970,000,000.00 on the collapse of profit margins in traditional market making.

Shame, eh?

[image via ComicBookPlus]