Global: Recession Lessons

Investors have been fretting about a US recession for months, but I don't think it's priced into markets. We may soon find out. Here's what I see as the implications of a US recession:

First, the Fed always cuts aggressively in a recession. The real Fed funds rate over the past 50 years has fallen to zero or below in every recession. That suggests to me that the minimum target for the funds rate in this cycle is 3%.

Second, market forecasts for Fed policy are usually wrong, and the error tends to be larger in an easing cycle than in a tightening cycle. I think that short-end futures are again being too conservative.

Third, long-end Treasury yields usually fall in recession - although the impact isn't as dramatic as for short rates. Moreover, real long-end yields are already at low levels, suggesting that long-end Treasuries are priced for sharply slower growth, if not outright recession. My own hunch is that in this cycle long-end yields have further to fall, driven not only by the usual cyclical forces (lower short rates, lower inflation expectations) but also by a larger-than-usual safe-haven bid as large parts of the non-Treasury debt universe are de-rated (relative to Treasuries).

Fourth, as short rates respond more aggressively than long rates to recession, it's usual for the yield curve to steepen. This cycle should be no different. The yield curve (e.g., the spread between 10- and 2-year Treasury yields) follows the Fed. The curve always steepens when the Fed cuts. If the real funds rate falls to zero, then the 10-2 year spread should widen to around 200 basis points. Our rate strategists see the story the same way - indeed, they are forecasting steepening curves around the world. (See Jim Caron, “2008 Global Interest Rate Outlook: Year Without a Consensus,” 4 December.)

The fifth lesson from past recessions: Earnings always get hit. As Morgan Stanley US strategist Abhijit Chakrabortti notes, the median peak-to-trough decline in S&P 500 earnings in recession over the past 50 years has been 16% (see “Atonement - Navigating a US Recession,” 10 December).

Over the long term, real S&P 500 earnings have fallen to below-trend levels in every recession in the past 50 years, except the brief 1980 downturn (but then they got clobbered in 1982). Earnings are now 62% above trend. If history repeats - and I see no reason why it shouldn't -there is a huge earnings shock coming. To get real earnings to (the rising) trend by end-2008 would require a 38% decline.

Sixth, the consensus never forecasts a decline in earnings - at least not in the 22 years for which we have IBES data. There is certainly no sense of looming weakness in current forecasts. The point, however, is that consensus forecasts should never be taken as a leading indicator of the economic outlook or, indeed, earnings.

The final lesson is that as often as not, the US$ rallies in recessions. Stephen Jen, our global head of FX, thinks this pattern will be repeated, and I agree (see “The Dollar Smiles in a Recession,” 10 December).

By Gerard Minack Sydney
Morgan Stanley
December 14, 2007