Excel Wankers and Recession Averages

Averaging pretty much anything to do with historical recessions & related bear markets is a meaningless exercise in spreadsheet-based thumb-sucking. Leaving aside my "financial history is bunk” rant, it doesn’t matter whether you’re averaging recession length, severity, or whatever, it’s all pretty much spurious and bullshit.

A quick example: There have been nine recession-induced bear markets in the U.S. since the Depression. They have ranged from just under 100 days long, to a little over 900 days. The average (there’s that dangerous word) is 486 days, with a whopping standard deviation (another dangerous word) of 250 days. In other words, based on this tiny data set — and that is a big part of the problem — we could expect (ahem) recessions to last anywhere from 0 (okay, -12 actually) to 1,000 days. But it’s all crap, from applying averages, to applying standard deviations, to sample sizes, but people pretend it isn’t.

The upshot? We have no flipping ideas how long recession-induced bear markets last, and anyone who pretends otherwise is an Excel wanker.