High frequency trading – a threat to global economic stability or an opportunity for Scotland?
January 26, 2010 by Nick Clayton · 8 Comments
Computing: The future of trading?
HFT means using statistical models developed by some of the world’s best mathematicians on high-powered computers to buy and sell shares in thousandths of a second. The profit on each deal might be tiny, but multiplied millions of times the returns are very attractive.
In fact HFT now represents up to 70 per cent of the share trading volume in the US and not much less in Britain. The speed, volume and secretiveness of the companies involved worries legislators.
It’s a very recent phenomenon so nobody really knows how much HFT distorts markets or whether it provides a useful function by creating liquidity, ensuring there are always buyers and sellers of shares. Certainly, it hasn’t been blamed for exacerbating the recession.
Of course, the fact it wasn’t responsible for the last global financial crisis doesn’t mean it won’t be to blame for the next. There is no such thing as bug-free software and as these are computers trading with computers it’s not impossible to visualise a situation where one error triggers a catastrophic chain reaction.
On the other hand, there is a feeling of inevitability in the growth of HFT. Traders have been using computer programs for decades to assist in buying and selling shares, commodities and currencies. Reducing the human element and allowing computers to trade with computers is a natural development.
It also represents a trend that could benefit Scotland. Intense competition between HFT firms is squeezing profits while all are trying to squeeze the last millisecond out of their systems. That even means putting their computers in highly-expensive office space next door to stock exchanges in order to save the almost imperceptible time it takes for data to move along fibre-optic cables.
While this technological arms race is going on between HFT firms, some trading is moving away from the big exchanges. It’s just not profitable for them.
The London Stock Exchange has had even its share of FTSE 100 trading reduced from 76 per cent to 62 per cent in the last year according to the Fidessa Fragmentation Index. Much of this was a result of being squeezed by the low-cost electronic markets BATS Europe and Chi-X.
In the US it was announced last Friday that four “new” exchanges would open this year. Actually they’re not always physically new. For instance, the Nasdaq OMX is based in the old Boston Stock Exchange.
So why not do the same in Scotland? It doesn’t have to be in the Glasgow Stock Exchange which was merged with its London counterpart in 1973. In fact there are advantages in having it elsewhere where rents are lower.
The small stock exchanges in the US are based on low-cost, low-staffing models. They wouldn’t fit the SNP constructional change proposal for a separate stock exchange to provide funds for companies north of the border. Instead it could undercut London and take some of the business that’s ceasing to be profitable there.
London Stock Exchange chief executive Xavier Rolet said recently he expects consolidation to lead to there being no more than five stock markets worldwide in the next five to ten years. That was in the context of alternative exchanges taking bites out of his business.
One of those could be based in Scotland. Although whether the SNP would welcome a cut-price stock exchange designed for high-frequency trading is another question.
Related posts:
- Euro bounces back after £430bn crisis deal collective sigh can metaphorically be heard from finance ministries...
- Could Scotland throw away its chance to be a world leader in data centres? Lack of whole-hearted government support and vaguely-worded environmental legislation are...
Surely this can only be bad for the economy? It does not produce wealth (i.e. goods and serviced that increase people’s quality of life). It has to be a zero-sum game, as any ‘profits’ made by a computer on these short timescales have to be paid by someone else making a loss?
Report This Comment
The Daily Show looked at this last year. It requires not only high powered computers but computers that are on site in the exchanges in order to minimise communication delays.
In short it is not a level playing field and is ethically dubious (at best)
Report This Comment
Computer based trading is far more effective than human based prop trading. Every investment bank in the world relies on it’s computer based trading models to analyse opportunities in the market, the most famous being Goldman Sachs “black box” which employs an element of AI that makes it a fraction of a second faster than any other bank thus generating it bigger profit.
Also, in answer to eratforte, if one company is making a profit it does not mean that some poor bugger is making a loss, they just make slightly less profit.
Finally, this is a very old story, computer based trading has been around longer than I have
Report This Comment
To Bill Wallace – of course it is not level, why would one want to play on a level field? If you want a level field try communism. Capitalism is based on one company gaining an advantage over another, and to do this the best computer based trading equipment is needed along with the sharpest minds, minds that must be paid massive bonuses to retain them at my bank and be sure they do not stray to a competitor.
Report This Comment
The biggest concern to us all is that the UK economy is mainly based on financial transactions.
We make next to nothing,better to use computer technology to come with some good R&D into new products we can make to pay off the City of london spivs screw up of our economy.
A few bankers making huge profits will not cure the ills of the ordinary man in the street.
As we are finding out the “trickle down effect” only works when the bankers make huge losses.
Report This Comment
HFT does sound pretty dubious – the problem is that some market participants receive information and the opportunity to trade on that information before others. However, there’s some argument for saying this has always been the case anyway – there have always been different tiers of delay for the FTSE for example, and you get what you pay for.
But when it comes to “runaway chain reactions” caused by HFT this is a bit unlikely. There was such a “chain reaction” back in the 80s with black monday – at least some people have tried to pin that on program trading. But if HFT does cause a chain reaction decrease, it would merely mean that the market temporarily deviated from the value actually put on it by human beings, and this would be swiftly corrected – it wouldn’t be a systemic problem. Nowadays though – after Black Monday – there’s all kinds of safeguards in place to ensure that program trades switch off in the event of huge movements.
Report This Comment
The aim of this piece was to stimulate discussion. I agree HFT is just a variation on a series of old themes. You can see it as simply automated market making or arbitrage, and those have been around for centuries. I didn’t go into those sort of technicalities as explaining them would have made the article unwieldy.
Despite its historic roots, however, I don’t think you can dismiss the dangers of HFT. After all, the credit crunch came from mortgage lending, which is hardly new. The concern with HFT really comes down to the speed of its growth. In five years it’s grown from representing perhaps 15 per cent of volume to as much as 70 per cent in the US. The figures are vague because much of this trading goes on in ‘dark pools’ which are, as you might guess, not very transparent. The sheer scale of HFT could make it vulnerable to a systemic crisis. But at the moment there is very little research into its impact on markets or what risks are involved.
Given the appetite of governments for financial regulation at the moment, it does seem likely that some of the activities of HFT firms will be curtailed, such as naked trading and flash trading. That doesn’t mean HFT will end. There are opportunities. They won’t directly produce many jobs, although a large proportion would be high quality. And, as with so many financial services, their contribution to ‘real’ wealth creation is debatable. After all stock markets exist to sell second-hand shares and insurance companies operate in a way that’s not so different from a backstreet bookie.
Report This Comment
Great article Nick. It is good to see that you have stirred discussion.
It is a controversial subject of course and I can see why people are skeptical.
However, I believe that a lot of the hostility to high frequency trading is based on fear of the unknown and that fear has been heightened by the financial crisis of 2008. The roots to that crisis were in Washington DC where Fannie Mae and Freddie Mac gave quasi US government protection with attendant high credit ratings to huge bundles of very poor mortgages that should have had very very low credit ratings. In fact most of the sub prime mortgages should not have even been given.
Wall St firms and London firms happily bought and traded the mortgage bonds and earned pleasant fees and trading profits until the economy turned and revealed the bundles to be potentially worthless. None of this had anything to do with High Frequency Trading. It had a lot to do with the Federal Reserve keeping interest rates too low. It also had a lot to do with politicians on the Board of Fannie Mae and Freddie Mac aiding and abetting the extension of the government umbrella way beyond its proper role.
As for computerized trading, my view is that people should be allowed to use calculators to beat slide rules and computers to beat people in straightforward arbitrage calculations. If you can connect to two different exchanges and see differences in price on the same security, then why not buy the overpriced shares and sell quickly on the other exchange. This will keep the prices in line.
Does this add to the grand sum of human wealth? Not necessarily. But it does keep prices in line at different places and it keeps costs down for you and me if we want to buy shares. It keeps costs down because the competing exchanges keep their fees down as they compete for traders. Also the greater amount of trading that occurs, the smaller the spread between the bid and ask price.
In the old days you would have to go to a broker who would charge some outlandish commission and he would call an LSE member firm and they would send a runner to the floor. He would ask a jobber the price. Guess what- it cost a lot to trade and would take ages to happen. Plenty of time and opportunity to get ripped off. The LSE had a monopoly. Now we have competition and automation, we should use it.
Someone in Scotland should seek out the old Glasgow Stock Exchange banner and license, open the exchange in a data center.
Give the Scottish investment funds (Standard Life et al.) a stake and a reason to trade on the Scottish trade matching engine. Run it remotely from Glasgow (the trade matching engine itself should be near to other systems) and good things will happen. Quality jobs for smart technologists. Leadership jobs. And a key role in the modern capital markets. It will be a much easier job starting up fresh with a few good people than defending the incumbent position that the LSE is in.
Report This Comment