A US recession indicator

"Interesting chart from Ruslan Bikbov at BofA Merrill Lynch.



As you see the yield curve has proved to be a very powerful recession indicator. So why is Bikbov drawing our attention to it now when the 2s5s curve is 74bps – 29 bps above its long term average?
The answer runs as follows.

We have often heard that the rates market is not priced for a recession yet because the curve remains historically steep. However, this argument ignores the fact that the Fed is at a zero bound. Because the policy rate cannot go negative, plausible paths of the future rate are either flat (Fed on hold) or rising (Fed hikes). As a result, the curve must be structurally steep relative to historical experience when the Fed had room to cut.

So Bikbov and his colleagues have created an adjusted curve, which is flatter than a pancake and flashing recession.

Because of the embedded optionality, the curve is steeper relative to what it could be if the policy rate were not already at zero. We constructed an adjusted curve by subtracting an embedded option premium from each forward rate. The curve adjusted in this way is the curve that could be observed today if the Fed were not at zero bound. As a result, the adjusted curve can be directly compared to historical experience and therefore is likely to be a better recession indicator than the unadjusted curve.
We have found that the curve adjusted for the zero bound effect has flattened to about 15bp. This is largely indistinguishable from a flat curve, especially given model uncertainty. As a result, we believe the curve could already be priced for a recession

Much like it did in Japan.

The Japan experience with rates at zero bound confirms our main findings. The BoJ had the overnight target rate at 6% at the beginning of the 1991-1993 recession. Consistent with the typical US experience, that recession was preceded by an inverted 2s5s JGB curve (Chart 4). The policy rate was cut to 50bp by October 1995 and has never exceeded this level since then, effectively bringing Japan to a permanent zero-bound regime. As a result, the curve has always remained steep despite subsequent recession episodes (Chart 4).

However, the JGB curve notably flattened before each subsequent recession. We believe that the 1997-1999 recession is especially relevant for our analysis because it followed the major 1991-1993 recession caused by the collapse of the housing and equity markets. So, if the 1991-1993 Japan recession corresponds to the 2008-2009 US recession, the next US recession might very well resemble the 1997-1999 Japan recession. The JGB curve flattened to about 90bp in the beginning of that episode. With the USD 2s5s curve already at 74bp, this provides additional evidence that the market may be already positioned for a recession.

Pass the tin hat."

at http://ftalphaville.ft.com/blog/2011/10/03/691266/a-us-recession-indicator/