On the limitations of the yield curve as a recession predictor

The yield curve has an amazing track record at predicting U.S. recessions: it basically predicted all of them, giving only a few false signals which still turned out to foresee slowdowns. In the current environment however, its predictive ability may be seriously lowered : Wright's model B tells us we would need to see a yield of 0.7% or less on the ten-year T-Bond to have a recession probability of at least 50%... I would bet we could fall back in a recession without ever getting even close to that yield level. Ask the Japanese.