The View From the Bluff

Recent economic data has been better, but we continue to see indications that leading indicators are pointing to a potential recession. Lakshman Achuthan insists the recent data has no bearing on their recession call. Ed Harrison pinpoints the issue (Ed is a must read by the way)

Note Achuthan’s derision for the precision of economic modelling. He thinks the consensus has it wrong because of this focus. When Achuthan talks about forward-looking indicator ‘contagion’ it sounds to me like he is speaking to a fat tail amplifier type of effect, meaning that a confluence of multiple downward-pointing indicators amplifies the trajectory in that direction. Models do not capture this.

They certainly do not. The video Ed is referring to is below, and highly recommended, but let’s look at the history of consensus model driven forecasting. From James Montier of GMO:

Pretty much useless. The economics profession as a whole never sees a recession, there will always be a soft landing. The data is no better for interest rates, inflation, anything! So, we don’t forecast either except in extremis or about general economic environments. e do however asses whether risks are higher than normal, and let us just say that a recession here seems more likely than in 2007, and we thought it was pretty likely then. It should be shallower, but I thought that then as well (which is why our base case in 2007 was for the declines to be in the 35-45% range from the peak.) Then the risks were skewed to the downside and I said so vociferously. I am less pessimistic this time. I think the base case is much less likely here in the US to be exceeded by a large margin. However, our base case for declines from the April peak if we actually enter a recession are in that same range as 2007 despite the market being somewhat cheaper. Read more »