The FX markets will never bore those who trade them. In many ways, they are an excellent example of a ‘pure’ market, in that they are deep and liquid for many common currencies, and any inefficiencies and arbitrage opportunities are very small and short-lived. However, currency risk is different from other frequently analysed risk sources, in ways which are not always appreciated.
Consider the issue of expected return. When an investor buys an equity, for example, he or she has a reasonable expectation that the equity will increase in value. A quick glance at the major equity indices over the past 20 years reveals that in general this is indeed the case – they show an exponential increase over a long timescale, though there is volatility in the shorter term. Thus, an investor expects an equity to increase in value, to compensate him for the risk of holding on to it.
A currency, in contrast, will probably give him just as much risk with no expectation of an increase in value. Equities are asymmetric; their value cannot fall below zero, and they are expected to rise in value. Currencies are symmetric; there is no limit to how low one currency can go relative to another, and there is no expectation that they will either rise or fall. This is because in a world where gold standards etc no longer exist, a currency is not a store of intrinsic value, but is simply a rate of exchange, and therefore it has no limit beyond which it cannot fall. Thus, an investor who expects an increase in value to compensate him or her for currency risk is in for an unpleasant surprise.
The assumption is sometimes made that currencies offset other risks, and that this correlation effect in some way compensates for currency risk. While a superficial examination of some data series may appear to indicate significant degrees of correlation or anti-correlation between currencies and other financial variables like equities, a more in-depth study will show that such correlations are highly variable, can range from positive to negative, and are not to be relied upon. Thus currency risk cannot be said to be compensated in this way either.
Another comparison which highlights the uniqueness of currency risk can be made between currencies and interest rates. Interest rates in established economies exhibit a type of behaviour known as mean reversion. They have a stable level to which they are attracted, which is somewhere between 2% and 15% for most currencies. While it is not impossible for them to move away from this level, they experience a strong ‘pull’ back towards it, which usually increases the further that they move away in an elastic fashion. Also, interest rates have a ‘barrier’ at or very slightly below zero, beyond which they do not go.
Currencies have none of these features. They have no attraction back to any mean value, and can always become more or less valuable. The pound/dollar rate illustrates this very well. Several decades ago it was apparently stable at about $4 to the pound. Now, it appears to be stable at about $1.50 to the pound. No one is suggesting that there is any attraction which will draw sterling back to its old levels. Some rates, like the Italian lira before EMU, have reached levels of several thousand to the dollar, due to inflationary conditions, and there is no reason for them ever to return to levels of one or two to the dollar.
Another important difference between interest rates and currencies is that of hyperinflation. Interest rates can, under extreme circumstances, hyperinflate and effectively shoot off to infinity. While no one denies that currencies can exhibit highly volatile behaviour, and move a long way in a very short time, they do not undergo classic hyperinflation behaviour such as that experienced by Germany earlier this century. However, if a hyperinflation occurs, in a country, its currency will show a huge decrease in value due purely to the hyperinflation itself.
What drives currency movements? There are many factors which can be highly influential. These include interest rate differentials, economic conditions, political events, natural events like extreme weather conditions, contagion from other currencies, and – of course – supply and demand.
However, all of these currency drivers can fade in comparison to internal effects. The vast majority of currency trades are made more for speculation than hedging or cash transfer reasons, and thus many moves are driven by so-called ‘technical’ signals, like trends and channels. These can be somewhat self-fulfilling; essentially, if a majority of traders believe that an FX rate will fall, they will sell that currency. This will in turn make the rate fall, so the reason for their belief becomes unimportant!
All in all, it is safe to say that there are a large number of underlying drivers of currency movements, and that it is difficult at any one time to tell which of the various candidates is dominant.
A number of assumptions are often made in the currency markets. One of the most common is that the forward FX rate, which is just the current rate adjusted for interest rate effects, is a good predictor of the rate in the future. In fact, it has no predictive power at all, and does not even contain information about the direction of future currency moves, let alone the size of the move. In fact, on average, betting against the forward rates will make money over a long timescale, though the returns are too small and volatile to form a viable trading strategy.
Another common assumption relates to the idea that currencies are heavily influenced by interest rate differentials. If one currency offers a higher rate of interest than another, the immediate temptation to the investor is to borrow in the low interest one while simultaneously lending in the high interest rate one. This will increase demand for the higher interest-bearing currency at the expense of others, which will, all other factors being equal, drive the FX rate up so that the higher interest bearing currency is more valuable. However, this does not always happen, because the factors which can lead to high interest rates are often those which would give the investor less confidence in the currency. So, when the UK experienced the ERM crisis, and the UK government raised rates to very high levels in a very short time frame, the currency continued to weaken, because investors had no confidence in the currency and not even the high interest rate levels could tempt them to buy it.
The medium and long-term effects of higher interest rates can be opposite to the short term ones. High interest rates are usually associated with high inflation. High inflation leads to high levels of seigniorage, where additional money supply is generated by the government to ensure that the total value of circulating money remains the same year on year. Thus, when seigniorage is high, the value of a fixed sum of money decreases. This means that high interest rates make investors wary of holding on to a currency for too long, as it will be likely to decrease in value. Thus, the assumption that interest rate differentials drive FX rates is not a good one to make – the situation is usually far more complex.
The last common assumption which is worth discussing is that economic fundamentals like current account balances control currency rates. In a world without volatility or speculative trading it might well be the case that economic drivers dominate currency movements. However, that world is not this one, and in fact economic fundamentals can disagree with currency rates over long periods of time, and to a highly significant degree. Purchasing power parity, a calculation based on a basket of goods which may be purchased by a unit of currency, gives an indication of the absolute value of currency. This can lie significantly out of line with actual exchange rates, for periods of up to five or six years. Eventually trade arbitrage will act to bring the exchange rate back into line, but import duty and tax issues often extend the lifetime of the discrepancy. We can conclude that economic fundamentals only have a long term effect upon currency rates, and in the short and medium term can often be irrelevant.
It is worth considering one last widely held assumption in the currency markets – that options are expensive because of their premium cost. The most common hedging alternatives to neutralise currency risk are options and forwards. A forward locks an investor into a future rate, giving him stability at the risk of losing out on positive currency movements. Forwards are zero-cost instruments, with the only outlay to the investor being the bid-offer spread. Options on the other hand have an immediate cost – the premium, which is something like 2% for a three-month option, depending upon volatility levels. They allow the investor to profit from positive currency moves, while guaranteeing protection on the downside.
The ‘true cost’ of an option is not the premium, however. It is the premium minus the average payoff made at the end of the contract. Looking at this figure gives a different picture – once a long enough time frame is considered, and both premium and payoff are taken into account, the average ‘true cost’ of an FX option is usually very close to zero. Thus both options and forwards actually have similar expected returns, albeit with different distributions. This seems obvious with hindsight – liquid financial instruments are usually fairly priced, and so one should expect that in an efficient option market payoff and premium would tend to cancel out.
The very uniqueness of FX rates means that they are becoming more popular as a separate asset class. A good currency trading programme should have a positive expected return. There is no reason at all, however, for this expected return to be correlated with that of other asset classes like equities, and indeed, a thorough analysis shows correlations between the two to be nearly zero. This makes currency trading programmes a highly attractive addition to portfolios which are heavily weighted towards equities or bonds, and need some diversification.
Jessica James is head of research, strategic risk management advisory, Banc One in London