In other words, interest rates – interest rate differentials – are central to the currency markets. Currency participants have always known this at some level. They know that, other things equal, a move in interest rates tends to lead a move in the currency –a rise in rates tends to cause a currency to rise and vice versa. But they also know that this rule in itself is no formula for consistently forecasting currency movements, let alone trading currencies profitably. Financial markets are never that easy. To take just one example, they have often seen efforts by governments to protect a falling currency with higher rates prove counterproductive.

In fact, if you look at the charts in this book, you will see examples of currencies regularly moving in the opposite direction to interest rates –totally ignoring increasingly favourable, or increasingly unfavourable yield differentials. The truth is that in such cases, it is the currency that is the cause of the yield movement and the yield movement which is the effect. So obviously it would be pointless to try and use yields to forecast currencies in such circumstances, though you might have a chance trying to use currency movements to forecast yields.

In any case, we cannot forecast interest rates. Or at least we can no more forecast interest rates than we can forecast currencies, and I would say much less. If we could, we could make fortunes in the T –Bill* or T –Bond* futures markets and “retire to cut coupons in the South of France” as someone once said. In our hearts, we all know this. Yet we often cannot help carrying assumptions about interest rate prospects around with us. These are nothing but the conditioned expectations that are already discounted* in the current interest price structure. They have no prophetic value. Yet time and time again you will see respected currency “experts” predicting the dollar on the basis of their interest rate expectations. In fact, many of them spend all their time combing the daily economic and financial releases for data that will help them hone their interest rate expectations. Like the Sufi mullah, they are looking in the bright sunlight for the key that is lying indoors in the shade.

Now you may say: “Hang on. There are times when we definitely can forecast interest rates. Take UK rates in 1990/1991: anyone could see they had to fall. Or German rates in 1990: everyone knew they had to rise.” OK, take UK rates in 1990/1. Everyone could see they had to fall. But did the pound fall as a result, in 1990/1 ? No. (Not to the time of writing). The pound actually strengthened against the benchmark D-Mark.

The expectation of falling interest rates was, as they say, discounted. You could see the degree of the discounting in the interest rate markets them– selves, since interest rates in the future are quoted, for the next few quarters. At any time you looked at the structure of interest rate futures in 1990 or 1991 you would find they were expected to be whole points lower 3 and 6 months ahead. In 1990 these expectations were constantly being confounded. They either didn’t fall at all, or else when they did fall, they didn’t fall as fast as people expected. They lagged expectations.

If we cannot make useful forecasts of interest rates, and if, in any case, future interest rate movements are most unreliably related to currency movements, what should we do? The answer is that we should take a big breath and undertake never, ever , to use rate forecasts as a basis for forecasting currency movements –let alone for trading currencies. It’s not that it will never work: we may occasionally hit lucky. It’s that we can never know whether it will work.

You will have spotted that when interest rates are leading a currency, the key to their effect on the currency lies in the way rate changes pan out in relation to expectations. This does give us something to go on, because we

can often gauge expectations. And if we work on the rule of thumb that expectations will be confounded, we shall be right much more often than wrong. If you spot a time when everyone is expecting US interest rates to fall, for example, buy the dollar, and you’ll make out OK. A few people did, all the way from 1980 to 1985– when US rates did indeed fall.

So the fortunate truth is: we don’t even need to forecast interest rates. All we have to do is monitor crowd expectations (in order to go contrary, on occasions ). Interest rates spend their time fluctuating, and much of the time 1hey are confounding existing expectations. People expect rates to fall and they don’t: they expect them to stay steady and they move down. Rates shift, and people think they will revert to earlier levels. For these reasons or otherwise, the fact is that when interest rates do lead currencies, they usually do so with a long lead time. The currency movement tends to lag by months, often 6 months or more (see chart opposite).

Once again, what we are looking for is the set-up conditions* for a movement in a currency. The “key” lies in the background, not in the foreground where everyone is looking. Here are some rules which seem to govern the relationship between interest rate shifts and currency movements.

1) Interest rate shifts prompted by domestic pressures (an over-heating economy, for example) tend to lead currency movements. The lead time is measured in months rather than days (e.g. the rise in the DM and SF in late 1989, and the dollar in 1981-5).

2) Interest rate shifts prompted by international (and currency) pressures tend to follow currency movements and go in the opposite direction ( e.g. the Yen in 1989, and the pound in 1988-9).

3) Following rule 2), a currency will tend to turn round and follow the lead of interest rates up: a) if/when they have risen around 4 points and especially, b) when the currency is perceived to have stabilised over several months (e.g. sterling in Spring 1985, and in 1990).

4) In this event, interest rates will peak when the currency troughs; and falling interest rates will coincide with a rising currency. Why? Because by this time everyone is expecting interest rates to fall tomorrow and they stay up longer than expected –i.e. rate cuts tend to lag expectations in these circumstances (e.g. Pound, Can $ and Aus $ in 1990-91). In other words, interest rates are higher than expected and this supports the currency.

5) When currencies are led by interest rates, their movement is directly related to the unexpectedness of interest rate levels, rather than to their rate of change (e.g. $ in 1983-4, pound and SF in 1990).

Rule 4) is an exception to rule 1) and there are others. For example falling interest yields call for rising bond prices. The result, in the case of the major investment currencies, can be a tug-of-war, when falling interest rates might discourage international investors but rising bond prices attract them –as they did in the case of the dollar in summer 1984 to spring 1985.Finally rules) is a cardinal rule. It is often the acid test of whether any given interest rate will lead a currency: it only does so when unexpected or unseen. When observers are ignoring an important shift in interest rates for any reason –because they expected it to be short-lived or because they are preoccupied with other things –you can bet it will only be a matter of time before the currency is led inexorably in the direction of the interest rate shift. What matters, as I say, is the background, not the foreground –because the foreground is where everyone is looking. What matters in financial markets –and in life? –is not what you see straight ahead but what you glimpse out of the corner of your eye.