HFT: the latest scam devised by Wall Street and the City

Felix Martin discusses Flash Boys by the American financial writer Michael Lewis, which examines high-frequency trading (HFT).

the New York Stock Exchange reopens after the Easter holiday, 21 April. Photo: Getty
Stock image: the New York Stock Exchange reopens after the Easter holiday, 21 April. Photo: Getty

Flash Boys, the new book by Michael Lewis, America’s explainer- in-chief of all things financial, is an account of “high-frequency trading” (HFT) – a technique developed by financial firms that deploys vast computing power to trade electronically on the world’s stock exchanges at extreme speed.

That may sound pretty esoteric. However, the book is generating an enormous amount of attention because it argues that HFT is the latest in the litany of scams that Wall Street and the City have devised to relieve unwitting investors of their money.

Whenever you hit Enter to buy shares through an online brokerage, Lewis shows, your order does not go straight to the stock exchange as you might think. Instead, HFT firms get a look-in first – and they use their superior speed to “front-run” your order by buying the shares ahead of time and then offloading them into the market at a marginally higher price. The resulting profits are tiny on any individual order but they run into the billions when you add them up. And they are made at your expense. Given how many people have a stake in the stock market these days with their Isas and their Sipps, this is certainly a disturbing revelation. Lewis deserves all the praise he is getting for exposing it.

Yet, to my mind, Flash Boys is even more important than this. For it exposes HFT as a prime example of one of the major problems of our age: the unintended consequences of technological innovation. Technologists, regardless of their political bent, tend to be idealists – it probably requires a healthy dose of idealism to take the risks required to innovate. But all too often, idealism can slip into naivety. The unstated assumption is that if new technology can be used to better the lot of the individual, it will. Everything will be OK so long as you “don’t be evil”.

Unfortunately, it doesn’t always work like that in the real world. The new technologies developed by well-intentioned young geeks in Silicon Valley and Old Street get grafted on to an economy that is still dominated by big, profit-seeking corporations run by shrewd old-economy dinosaurs. Innovation is driven by the admirable belief that new technology is a tool for the emancipation of human creativity and self-fulfilment. Less thought is given to what might happen after, say, News International buys your app.

The point is more general than just the compromises that come with commercialisation by big business. What the technologists are missing is the crucial importance of the social context in which new technology is deployed and, above all, the role of that most reliable of social scientific regularities, the law of unintended consequences.

The canonical problem is that we design some new technology to solve a problem but in doing so we make a crucial assumption: that everything else will remain unchanged and in particular the way that people interact, the social context, will be unaltered. What happens is that behaviour adapts. The technology succeeds – the old problem is eliminated – but new problems arise.

An example that is almost guaranteed to have infuriated anyone reading this at some time or other is the computerisation of personal credit scoring. Companies such as Experian or Equifax apply information technology to the problem of deciding who should and should not get loans.

In an economy where mortgages and mobile-phone contracts are considered essentials, the decisions that their computers churn out are important. Their claim is that their algorithms are not just cheaper than the Captain Mainwaring-style bank manager of old but also more objective and therefore fairer.

If it were true that people’s behaviour had remained constant after the introduction of computerised credit scoring systems, that might be the case. But in reality, people game the system. Personal finance articles and chatrooms warn them that cappuccinos and city breaks flag them for a downgrade, so they take a breather for three months before applying for a mortgage – and then they start up again as soon as the ink on the contract is dry.

It is no different from the snag that the Soviet Union discovered with a planned economy. You could solve the problem of low productivity – at least as the bean-counters captured it – with more demanding targets. The underlying disease of demotivation proved more resilient, however. As an aphorism of the period had it: “They pretend to pay us and we pretend to work.”

The story that Lewis tells of HFT is a perfect example of the law of unintended consequences at work in the technological transformation of the stock market, one of the most basic institutions of our capitalist economies. The computerisation of stock exchanges that began in 1986 promised to make them simpler and more efficient. The world of barrow-boy traders bellowing at one another in the pit and the old-boys network of City stockbrokers was abolished in favour of anonymous electronic trading on a virtual exchange.

The intention was to stop investors being ripped off by an uncompetitive industry. However, this assumed that behaviour would not adapt. The stockbrokers and pit traders did hang up their red braces and garish blazers but a new generation of rent-seekers emerged in their place. As Flash Boys documents, the fixed commissions levied by the stockbrokers of yesterday were replaced by the cuts taken by the HFT firms of today.

So, what is the lesson to be learned from Lewis’s latest blockbuster? Well: this past week, the government’s ambassador for digital industries announced that schoolchildren should learn less French and more code. Maybe. But the lesson of the burgeoning HFT scandal is that the naive application of technology can be a uniquely dangerous force. We should be teaching our budding technologists not just code – but the law of unintended consequences.