Yield curves continue to flag trouble accurately

Serious Money: Anticipating turning points in the business cycle can lead to exceptional investment returns.

Prof Jeremy Siegel, author of the widely acclaimed Stocks for the Long Run, calculated the excess returns an investor could have earned relative to a buy-and-hold strategy in the post-war era if they had correctly predicted turning points in the US economic cycle.

The implementation of a simple market-timing strategy whereby an investor moved out of stocks and into cash four months before a business peak and switched back into stocks four months in advance of the trough would have yielded additional returns of almost five percentage points per annum or a more than threefold increase in wealth over a 30-year period.

Unfortunately, as Nils Bohr, a Nobel laureate in physics, observes "prediction is very difficult, especially if it's about the future".

Professional forecasters have an extremely poor record when it comes to predicting the business cycle. Alfred Cowles concluded an article entitled "Can stock market forecasters forecast?" in 1933 with the words "It is doubtful". More recently, monthly publication Blue Chip Economic Indicatorsnoted in July 1990 that "the year-ago consensus forecast of a soft landing in 1990 remains intact" and that "the economy in 1991 is projected to be a shade better". The economic expansion peaked that very month.

Similarly, in March 2001, 95 per cent of forecasters were predicting that there would be no recession that year. An economic recession began just days later. The experts' unfortunate record through time gave rise to the quip that "an economist is an expert who will know tomorrow why the things he predicted yesterday didn't happen today".

If investors cannot be confident in professional forecasters' ability to anticipate turning points, perhaps they would be better served to look to the message impounded in financial market prices. Indeed, it has long been recognised that the yield curve, which charts the price of short-term money relative to long-term money, is the most reliable indicator of an imminent economic downturn.

An inverted yield curve, where long-term interest rates are below short-term rates, has preceded every economic recession over the past 40 years and the accompanying decline in corporate profits with an average lead time of 15 months. Furthermore, the curve has also managed to predict the duration and severity of each subsequent downturn. The inversion that began in 2000, for example, persisted for eight months and was not particularly pronounced, the recession that followed lasted for two quarters and was relatively mild.

Despite the yield curve's proven track record, some commentators continue to dismiss its significance on the grounds that no good explanation for the relationship exists. There is, in fact, an abundance of reasonable explanations. Monetary policy, for example, can influence the slope of the yield curve. Tighter monetary policy via higher short-term rates is typically implemented in order to reduce inflationary pressures. Lower rates are expected to follow when the economy slows and those pressures subside. Since long-term interest rates reflect long-term expectations, they usually rise by less than short rates over a tightening cycle. Thus, a tightening cycle may result in a flat or even an inverted yield curve in advance of an economic slowdown.

A further explanation can be made on Prof Irving Fisher's arguments in his influential book The Rate of Interestin 1907 that, in equilibrium, the one-year interest rate reflects the marginal value of income today in relation to its marginal value next year. Fisher's argument rests on the premise that individuals attempt to smooth their consumption over an economic cycle. Some individuals will save and shift income from today to tomorrow, while others will borrow and give up some income tomorrow in order to have more income today. The equilibrium interest rate equates the supply and demand of this income shifting.

Individuals are risk averse and consequently, they receive more benefit from a dollar in a recession when their consumption levels are low, than at a peak in the economic cycle when their consumption levels are high. Thus, the desire for a hedge that will enable individuals to smooth consumption levels in the future inevitably grows as expectations of an economic downturn increase. The desire to hedge will cause individuals to shift short-term cash balances into longer-term instruments that deliver pay-offs during the downturn.

The sale of short-term instruments will cause a fall in prices - a rise in short-term interest rates - while the purchase of longer-term instruments will see prices rise - a decline in long-term interest rates. Consequently, the yield curve will flatten or even invert.

The ability of the yield curve to predict recessions gives rise to an obvious question - can it be used to time the stock market? The historical evidence suggests that it can. An inverted yield curve has preceded every serious decline in stock prices over the past 40 years apart from the infamous crash of 1987. Furthermore, examining the stock market's performance in each month that immediately followed a month in which the yield curve was inverted on average reveals that the stock market incurred a mean annualised price decline of more than 6 per cent during these months over the past 50 years.

The yield curve is currently inverted but most commentators argue that it is no longer a reliable indicator of future growth due to the recycling of current account surpluses into dollars by price-insensitive central banks in Asia and the oil-producing countries.

Like-sounding arguments have been made before. The experts argued in 2000 that large budget surpluses had led to a shortage of long-term bonds and, consequently, the message emanating from the bond market should be ignored. They were wrong. They could be wrong again.

Charlie Fell is an independent consultant and lectures in finance and investment in UCD and the Institute of Bankers in Ireland.