As a young financial journalist in the late 1980s, I went to visit the Chicago Board of Trade when it remained an entirely open outcry market. The impact of the noise when trading began after the opening bell, a physical reaction in the pit of the stomach, has stayed with me all these years. In those days, of course, the computer revolution was in its infancy, and all the major markets were open outcry. I don’t think many people outside the financial world appreciate how minor a role humans now play in the financial markets.
Computers do about three-quarters of the trading on US stock markets, a quarter in Europe, and rapidly growing proportions in other asset markets, with this algorithmic trading occurring at intervals of fractions of a second – so fast that the speed of light is a material barrier to trade, and high frequency traders co-locate their computer servers in the same building as the servers of the exchanges. Asset prices display extreme volatility. The ‘Flash Crash’ was one, but research suggests there are thousands of them, just so fast we don’t notice. The average holding period for shares has dropped from about 4 years in the 1950s to 22 seconds now. The number of companies listing on US stock markets has declined from 530 a year in the 1990s to 126 a year by the late 2000s, and the number of listed companies has fallen from 8,200 in 1997 to about 4,000 by the end of 2010.
[amazon_link id=”1847940978″ target=”_blank” ]Dark Pools[/amazon_link] by Scott Patterson tells the story of the rise of algorithmic trading, and it’s both a fascinating and a completely terrifying tale. The development of computerised trading began as a sort of revenge of the nerds and mavericks against the insiders of the NYSE and Nasdaq, albeit seasoned with the classic greed of the financial markets. The book describes the motivation of Josh Levine, the developer of one of the key early electronic trading systems, as a classic hacker desire for transparency and information. He wanted a fairer, freer market, and the rise of the algorithmic traders was indeed greatly assisted by anti-trust actions against the main markets.
Levine, though, has left finance now. For the titans of the financial markets did what they do when upstarts begin to make headway against them: bought them out. The NYSE and Nasdaq, and major banks like Goldmans and Credit Suisse, acquired the most successful of the computerised traders.
What’s more, as the book explains, successful algorithms trade so well that either they engage in an arms race among themselves for minuscule advantage, or they need to be fed a constant diet of less successful traders from whom they can profit. That would be me and you, either as retail investors or through our pension and mutual funds. The rules of the market enable computers to jump to the head of the queue for trading a particular stock; so if a big investor has placed an order, the algorithmic investor will see it and take advantage of its pre-cognition. The computerised traders are also paid ‘make’ fees by exchanges for providing liquidity, while the ‘takers’ (i.e. you and me again) pay for it. The computers can finally jump around from exchange to exchange, the result of well-intentioned regulation, and use their unique speed to their advantage as well.
In short, this is a rigged market, ineptly regulated, in which the owners of the main electronic trading systems steadily milk money from savers. And algo trading is spreading from stock markets to other assets. There is some evidence (although it is challenged) that the creation of commodity indices traded electronically has contributed to spiking food prices. Today’s FT suggests Goldmans and other mega-banks now want to play an active role in the oil market. I hope the regulators turn them down, but expect they won’t.
There is a fundamental problem of trying to achieve information transparency in an asset market, inherently characterised by at least an asymmetry of perception and probably often an actual asymmetry of information. It makes the competitive structure of the market all the more important, and [amazon_link id=”1847940978″ target=”_blank” ]Dark Pools[/amazon_link] suggests to me that the anti-trust authorities need to step in again. But that is a slow and uncertain process, and meanwhile the financial system is just not pooling savings to place into productive investments, which will generate a return for savers. On the contrary, it has become a global video game, Final Fantasy Finance, with unfortunate real life side effects. The best bet may be, as Andrew Haldane hinted in one of his recent speeches, to work around the dysfunctional financial ‘markets’ and build new financial institutions such as the peer-to-peer lenders. It will be very important for regulators not to strangle these by throwing up impossible barriers to their growth.
Meanwhile, I highly recommend [amazon_link id=”1847940978″ target=”_blank” ]Dark Pools[/amazon_link]. It is a terrific read, both for a history of high frequency trading well told, and for a different kind of perspective on what’s gone so wrong with finance. If you really want to scare yourself, pair it with Robert Harris’s [amazon_link id=”0099553260″ target=”_blank” ]The Fear Index[/amazon_link], and to think about what to do, turn to the speeches of the Bank of England’s Andrew Haldane and Robert Jenkins.
[amazon_image id=”1847940978″ link=”true” target=”_blank” size=”medium” ]Dark Pools: The rise of A.I. trading machines and the looming threat to Wall Street[/amazon_image]
A guy goes into Barclays and says “I want to buy 200 shares in Shell” The cashier says “I’ll get the securities clerk, sir” The securities clerk, a well spruced guy of around 25, arrives, fills out a form and says “Thank you very much, we’ll be in touch”
The securities clerk consults the customer’s account in a large heavy paper ledger (written up in pen and ink) and then places the order on the Branch Manager’s desk. The Branch Manager returns from visiting a local farmer about an hour later, reads the form and then notices it’s a couple of minutes past one pm, so he goes upstairs into the Bank flat, where he lives, and sits down to a large steak pie his wife has prepared for his lunch complete with garden peas and potatoes.
At 2.10, he returns to his office. He picks up the order form, walks into the ledger room and ensures the customer has the money to buy 200 Shell shares. He goes back into his office and rings the local broker who then goes onto the floor of the local stock exchange and buys the shares. He rings the bank manager who tells the securities clerk to phone the customer and tell him he will need to settle the deal at the end of the trading period which is 11 days away. About a month later a letter arrives from the stockbroker which includes a share certificate signed by the company registrar of Shell
That is how it worked when I was a securities clerk in Barclays. We have progressed since then, but I suggest we have also gone backwards. Algorithms and computers have lead investment into the world of http://globalthermonuclearwar.net/ where no one wins but the dealer
The retail investor loses out by relying on OEIC and pension plans exactly because computer trading ramps up the dealing cost but does not show in the TER. The way to go is to look for yourself at the shares/bonds the top guys are investing in, buy the things yourself, and avoid the churn that devalues your investment
It’s very much like computers themselves. When I built my first, it was all DOS. Mr Gates changed that. The last one I built with W7 worked like lightning. But now those at the front end, including some well known to you, are back to Raspberry Pi
Dark Pools obviously got missed by Sir Merve, just as Steady Eddie missed a similar warning about BCCI. Trouble is, it’s the investors and customers that loose out, not the regulators. But the Harris book is the one that rings my bell
Pingback: Frankenstein’s monster at large | The Enlightened Economist