The dollar proves less than almighty

Almost all currency forecasters expected the dollar to sweep all before it this year as the Federal Reserve raised American interest rates, and as German and other European rates continued to fall. But although monetary policy has changed in the expected direction - albeit very slowly in Germany - the dollar has so far refused to rise.

As in other financial markets, there is no doubt that the hedge funds and other short-term traders were heavily overweight in dollars some weeks ago. As these investors have moved to reduce their overall position size in order to effect deleveraging, dollars have been dumped into the market. This is just one more example of a promising trade that has come to grief as excessive speculative positions have been washed out of the system.

If that were all there was to it, then some easy money could be made in the long term simply by buying the assets the hedge funds are abandoning. In some cases, however, recent moves may be telling us something about long-term fundamentals, and I am beginning to wonder if this may be true in the case of the dollar.

The bullish consensus among currency forecasters about the dollar has been based on a congruence of long-term and short-term factors. In the very long term, most exchange rate models require that a currency should return to its 'purchasing power parity' (PPP). This is the level of the exchange rate which ensures that the same goods should sell for the same price in all markets. In other words, the dollar price of a Lexus in the US should be the same as the yen price in Japan, translated into dollars at the ruling exchange rate.

Virtually all direct measures of selling prices for equivalent goods in the US and elsewhere indicate that, on this basis, the dollar is still substantially undervalued. For example, the graph shows that, on Goldman Sachs' calculations, the US currency is still trading about 25 per cent below its PPP rate. Similar calculations by the Organisation for Economic Co-operation and Development indicate that the PPP rates for the dollar against the main alternative currencies are DM/dollars 2.12 and Y/dollars 188, suggesting that on this basis the dollar is extremely undervalued.

The trouble with these calculations is that the dollar has barely ever been able actually to trade at PPP at any time in the past 20 years. As the graph makes clear, the only time the US unit exceeded its PPP rate was in 1984-85, when the combination of President Reagan's profligate budgetary policy with the tight monetary policy of the Federal Reserve under Paul Volcker temporarily pushed the dollar to unsustainable highs.

Does it really make sense to use a dollar PPP rate as an analytical or forecasting tool when that rate has been achieved only during the most exceptional couple of years in the past 20? It seems highly doubtful. If indeed prices in the US are cheaper than those of equivalent goods elsewhere, perhaps they need to be relatively low in order to offset the poor non-price competitiveness of American goods. After all, the American economy seems to be stuck with a chronic long-term trade deficit, which is hardly the most common characteristic of a seriously undervalued currency.

An alternative way of estimating the long-run sustainable equilibrium for an exchange rate is simply to observe what level that rate attains on trend, after adjusting for differentials in price inflation. The chart shows that on this basis the dollar looks anything but undervalued.

In fact, the dollar's real effective rate has recently been trading slightly above its trend level, and it is only fractionally below its 5 and 10-year averages.

Although American goods are cheap, they are therefore no more cheap than usual. In this light, the chances of a sustained dollar appreciation look much less compelling than is often imagined. Certainly, shorter-term cyclical forces, such as GDP growth and monetary policy, could still push the dollar temporarily above its sustainable equilibrium - say to around DM/dollars 1.85 later this year - but the US unit may have trouble sustaining these valuations once the European economies start to recover in earnest.

There is a world of difference between a situation in which a currency is being pushed by cyclical forces towards its long-term equilibrium, and one in which it is being pushed away from its equilibrium - in fact, the difference between pushing a rock up or downhill. At present, the dollar begins to look as if it is being pushed uphill, and with some difficulty as well.

In theory, the dollar should approach its sustainable equilibrium on a path determined by two factors - the difference in real interest rates between the US and other countries, and the risk premium needed to encourage markets to hold dollar or non-dollar assets. At present, the real bond yields in the US and Germany are similar, so this force is roughly neutral. However - and this may come a shock to most people - the markets appear to need a positive risk premium to encourage them to hold dollars at present.

How do we know this? One way of estimating the risk premium is to compare the consensus economic forecast for the dollar over the year ahead with the interest rate gap between the US and Germany. In the past few months, the consensus forecast has consistently shown the dollar appreciating by 2 to 3 per cent more than is sufficient to offset the interest rate gap.

In other words, the markets need to expect a 2 to 3 per cent excess return on dollar assets, otherwise they will not hold the dollar. The US currency needs permanently to trade at low enough levels to allow the markets to expect it to appreciate by 2 to 3 per annum more than implied by per cent interest rate differentials.

In the 1980s, the reverse situation applied - the return expected on the dollar was generally below that explained by the interest rate gap, so there was a negative risk premium on dollar assets. This boosted the value of the dollar, whereas the same factor is now working the other way.

The main reason for this change is probably the big swing in the capital account of the US balance of payments. In the 1980s, there was a huge and persistent inflow as Japanese investors acquired American assets. Now, by contrast, the Japanese inflow has dried up, and the American investor has finally caught the bug for overseas markets.

There is currently an outflow of about dollars 100bn a year from America to foreign bond and equity markets, on top of the US current account deficit of dollars 120bn a year. Consequently, the US needs to attract about dollars 220bn a year of short-term capital inflows just to leave the dollar unchanged - and that could prove increasingly difficult over time.

The dollar may still move temporarily higher as the Federal Reserve tightens monetary policy this year, but it would be surprising if it could sustain these higher levels for very long. And for the struggling European economies, any relief from a rising dollar will be short-lived - unless the Bundesbank and other central banks are willing to ease monetary policy much more aggressively than they have done so far.

(Graph omitted)