Hindsight, however, is dangerous. It invites us to believe that because something is easily explicable after the fact, it must be predictable beforehand. This inference, though, is wrong, as recent research shows.
First, some background. The standard view in economics - though for obvious reasons it doesn't get much publicity - is that exchange rates are generally unpredictable. In 1983 Ken Rogoff and Dick Meese, two US academics, wrote a famous paper showing that conventional economic models couldn't forecast currency moves - a finding that was corroborated in 2005 by Menzie Chinn at the University of California Santa Cruz.
And, says Barbara Rossi of Barcelona's University of Pompeu Fabra, this is still largely true. "The empirical evidence is not favourable to traditional economic predictors," she concludes. Things such as interest rates, prices, output and money stocks can't much forecast currencies.
There are, though, some exceptions to this.
First, Lucio Sarno at Cass Business School has shown that net foreign assets - cumulated current account balances - have some predictive power, albeit only at horizons of three years or more for the dollar-sterling rate.
Secondly, says Ms Rossi, there's some evidence that it was possible before the financial crisis to predict exchange rates by using Taylor rules, simple equations that predict interest rates - the point being that forecasting interest rates could give us an edge in forecasting currencies.
Thirdly, Tanya Molodtsova at Emory University and David Popell at Houston University have estimated that, during the financial crisis, the euro-sterling rate was partly predictable by using unemployment rates and indicators of financial stress such as spreads between interbank rates and overnight indexed swap rates.
These, though, are exceptions, not the pattern. Generally, says Ms Rossi, forecasting ability "is an occasional and short-lived phenomenon". The famous Meese-Rogoff finding "does not seem to be entirely and convincingly overturned".
Does this mean investors should forget about foreign currency as an asset?
It does, if you intend to use it for speculative purposes, betting on currency moves.
For some of us, though, there is still a role for FX. If you're planning on retiring overseas, or buying a foreign holiday home, you face exchange rate risk - the danger that the euro or dollar will rise so much that your pounds won't buy you as much as you're hoping. This risk is significant. Over the last 20 years, the annualized volatility of weekly moves in the dollar-sterling rate has been 7.4 percentage points, and that in the euro-sterling rate has been 8.4 percentage points. These numbers imply that there's roughly a one-in-seven chance of them rising by 20 per cent or more over the next five years. That would put a big dent in the spending power of a sterling-based investor. The obvious and easy solution to this is to hold the currency you're planning on needing in future.
Otherwise, though, FX is rather like elephants - interesting enough to look at, but not worth owning.
