FOREX - Is it possible to accurately predict forex markets?
Think you can predict where the dollar’s heading? Forget it. “There are reasons why an individual can’t consistently predict prices in any efficient market, and FX is just one of those markets,” says Benjamin Hunt, associate professor and director of the UTS Graduate School of Business.
Zero sum game
Many markets are a “zero sum game”, he says, where consistent profits on one side of the trade are offset by consistent losses on the other.
Secondly, if there were recognisable patterns from which people could make profits, such as the market being down in October but up in November, their very activity would remove those patterns – because they would buy in October (pushing prices up) and sell in November (forcing prices down).
Counter-arguments exist, Hunt says. “A principal polemic counter-argument is technical analysis,” he says. This is the analysis of past movements to predict future movements.
Trading rooms only
Trading rooms consistently generate profits for banks, Hunt says – not so much because the predictions made by traders are correct but because those profits include commissions or payments for the provision of a market – for the provision of liquidity. “But people who work in the markets don’t believe that,” he says. “They strongly believe that they can make predictions.”
But can they? In the 1980s, Hunt and another professor then at UNSW, former Liberal leader John Hewson, would phone about 16 banks each week asking for a prediction on the value of the Australian dollar one week out.
‘Toss a coin instead’
The results, published in The Age, made two things obvious: first, the bankers tended to predict away from that day’s spot rate, “giving a bimodal distribution” (imagine two groups of values either side of the spot rate, instead of a bell-shaped curve with the spot rate as the average); and second, when the bankers’ predictions were averaged, they foretold as a group that the spot rate would be unchanged.
“We did it for two years, and when you look at the results you could have tossed a coin,” Hunt says.
On a chart showing one-year-out currency predictions for the $A against the $US between 1984 and 2005, triangles on their sides represent the range of values provided by experts. Plotted against a moving spot price, the predictive powers of the “rainmakers” look pretty dismal. Only eight of 19 ranges overlap actual spot prices, a strike rate of 42 per cent.
Delusional
If you think you can predict, “you’re deluding yourself”, says Hunt, who puts the phenomenon he and Hewson observed down to a “generate-the-trade mentality”.
“There’s plenty of evidence people can’t predict exchange rates,” he says.
The market is divided into proponents of efficient markets and those who trust technical analysis. But Hunt points out that if a client calls asking for an opinion on which way the price of gold is moving, it is the technical analyst who will turn that call into a trade.
“An efficient market analyst, when asked what will happen to the price of gold, will say: ‘I don’t know. My best guess is it won’t change’. ”
In an efficient market, the best guess about tomorrow’s price is today’s price. “But if you get a technical analyst, the response would be: ‘Have you seen these trends? By next Friday it’ll be X’.”
The nihilistic efficient market view
The person with the definite view on what the market’s going to do will more likely generate a fee than the one with the “nihilistic” efficient market view, Hunt says.
If you believe in an efficient market, an investment decision in some markets becomes the same as flipping a coin.
“You can make this trade and you’ve got a 50 per cent chance of making a profit and a 50 per cent chance of making a loss,” he says. “And after you take out the charges you will necessarily incur from making the trade, you will have an expected loss, equal to that charge.”
Some markets are not efficient, such as small stocks, where consummate efforts in research can identify undervalued or overvalued companies. Currencies, however, are too deep and liquid for prices to hide anything from anyone (SPI futures and gold are other examples he cites).
Random new information
“And [prices] will only be moved by the arrival of new information,” Hunt says. “New information, by definition, is random, otherwise it wouldn’t be new information, and so the price moves randomly.”
Hunt concedes there are cracks in the “efficient market façade” evidence of minor inconsistencies with regard to the random walk thesis, “but I can say without fear of contradiction that for a retail investor to think they can make money trying to pick directions of highly traded liquid commodities or stocks, forget it”.
Subscribe to:
Posts (Atom)
