Do any blog readers routinely trade on the basis of Twitter comments? Or more specifically, do any trade within milliseconds of receiving a tweet? The answer of course is “no”, as readers have no ability to trade in milliseconds.
But last week the computers did just that. As the Financial Times chart shows, the US S&P 500 fell 150 points in 2 minutes after a rogue tweet claimed the White House had been attacked.
The reason is that stock exchanges have encouraged the growth of high-frequency trading (HFT), which operates via what is called ‘machine readable news’. The trading companies involved pay large fees to the exchanges to allow their computers to be installed next to the exchange computer.
And then the machine trades as fast as it can, based on its ability to read words such as “rise”, “fall” and “warn”.
It would be funny if it were not true. As the Wall Street Journal notes, the markets have completely lost sight of their key role, to help companies raise capital to grow their businesses.
As trading in oil markets confirms, the exchanges have also abandoned their role of enabling price discovery to take place. Last week’s mini-crash, like the 600 point fall on the Dow in May 2010, is yet another reminder of the fragility of the global financial system. It highlights yet again that no single market now knows what it is pricing.
Another worry, as Andy Haldane of the Bank of England has warned, is that “high-frequency trading adds liquidity in a monsoon and absorbs it in a drought”. But in spite of his efforts, other regulators are still only talking about placing speed limits on trading – and so far only for foreign exchange markets.