Cyclical fair value for the dollar

Our economists have consistently argued that the market is split between a cyclical way of looking at the world and a structural one. Hence the market buys the dollar on any signs that rates will be tightened more than current expectations, but it also sells it when worsening trade or current account data are released.

Before we pit the value of the dollar on a structural basis against its cyclical basis counterpart, it is worth taking a neutral approach and look at the OECD calculation of purchasing power parity (PPP) for different currencies. Given the nature and problems with PPP, one usually estimates that a currency is over- and undervalued if it is a minimum of 10% away from its fair value. On the current OECD estimates the dollar is shown to be reasonably valued.

Using the weights as derived by the Fed's major trade-weighted index, the euro, yen and Canadian dollar have weights of 29%, 26% and 30%, respectively. These three currencies account therefore for roughly 85% of the total major index.

On an OECD PPP basis the dollar is undervalued by around 10% against the euro, 25% against the yen but overvalued by around 11% against the Canadian dollar.

Given the weights, this would mean that using just these three currencies the dollar is undervalued by around 7%. Adding in sterling and the Swiss franc, the dollar becomes around 8.5% undervalued, so the latest moves in the dollar compared with January have taken the dollar from around undervaluation territory to within the fair value range.

Yet the current account deficit continues to deteriorate, as US growth outperforms, so the question our economists asked themselves last year was: what kind of move would they need to see in the dollar to correct the structural imbalances?

Last year our economists posed the question: how low can the dollar go? On that occasion they looked how the dollar would have to fall to start to see an improvement in the current account balance. The reason this methodology came up with very big falls in the dollar is that they continued to expect the US economy to grow significantly faster than the rest of the industrial world. Thus any prospect of a narrowing current account imbalance would require the dollar to do all the work.

On a structural analysis they estimated it would take a further 20% decline in the dollar's trade weighted index to generate a narrowing of the current account deficit by 0.5% of GDP in 2006. The levels for the euro and yen implied by a further 20% decline in the dollar index appear to be extreme, and they would hesitate to forecast them. However, the structural analysis suggests that the risks for the dollar continue to be skewed heavily to the downside.

Historically for a given move in the trade-weighted index the euro moved by more than the yen, and both moved by more than the trade-weighted index. If the dollar index were to fall by a further 20% and the euro and yen were to move in the same way as they have on previous occasions, this would imply a E-US$ rate above 1.60 and a US$-¥ rate below 90. This was never a forecast, but gave an indication of the pressures when the FX markets focused on structural factors alone. The question now is how do they calculate the magnitude of the pressures on the dollar when cyclical forces dominate?

On a cyclical basis interest rate differentials are what matter. In a structural world, the dollar needs to fall to rectify the external imbalances. In a cyclical world the market buys growth and the higher rates of return represented by higher interest rates.

A factor that can undermine the power of interest rate differentials is if a central bank is behind the curve and chasing an inflation threat rather than responding to strong growth. One way to see if this is happening is to look at how inflation-linked bonds are performing.

If one takes conventional bonds minus the real rate implied by index linked bonds they are left with the 'break even inflation rate'. That is the implied inflation rates over the long term as well a risk premium. Now the issue of the risk premium is difficult, but for the purposes of this analysis our economists assume that from the beginning of this year all the risk premiums have moved in tandem with each other in the major markets. Thus, what they are comparing is changes in inflation expectations rather than changes in risk premiums.

The level of break-even rates does not suggest any impending inflation problem. The strong cyclical upswing has seen US break-even rates rise around 35 basis points this year. Hence, given the combination of strong growth and an implied rise in inflation it is not surprising that the market now expects US rates to rise substantially.

Nevertheless, both the level and the changes in break-even rates can not be construed as an inflation threat. Thus for the moment it is safe to assume that the rises in interest rates will create a greater return that will not be eroded by inflation.

In a cyclical world where interest rate differentials are a prime driver it is worth constructing an interest rate differential chart of US against its trade-weighted partners. This is then compared to the trade-weighted dollar. One can clearly see that interest rate differentials have a substantive influence on the trade-weighted dollar.

What is also clear is that the dollar fell 10% in trade-weighted terms following the Dubai G7 without any associated change in interest rate differentials. It was this decoupling of rate expectations and the dollar that signalled the market was de-rating the dollar on structural grounds. However, after the upside surprise in the non-farm payrolls in April, the market began to price in rate rises in earnest.

At this point the market started to trade on a cyclical basis and the relationship between interest rate differentials and the dollar resumed.

According to our economists, US rates need to rise by more than the rest of its major trading partners. Thus it will come as no surprise that trade-weighted dollar needs to rise in a normal 'neutral' world. The question is by how much.

This report has been compiled by HSBC Bank Malta plc on the basis of economic research carried out by HSBC International Bank's team of economists and financial analysis.