Jim Hamilton Doing Recession Probabilities

Economist James Hamilton has a data-rich post up on recession probabilities. It’s worth reading all the way through, but the gist is that he uses historical data to come up a measure for the likelihood of an upcoming recession predicated on current GDP growth. Long story short, he ends up with this figure, and with updating his own probability of a recession this year to 16.9%.


  1. Although the analysis looks quite thorough, common sense makes me think that it’s not the best approach to predicting recessions. If we’re messing with mathematical tools ignoring fundamental data, I’d rather stick to convenient tech tools – stochastics.
    The 50 year old data, even casted to relative values, cannot be rightfully applied to todays environment. And if we’re still talking probability, wouldn’t it be more clever to extrapolate it from more recent data with classical overbought/oversold approach, updated for GDP data?
    Besides, a ~17% value is a pretty high bet in my view… Still, ignoring the results, fundamental give us less precise but at least equally strong signs that recession *may* be ahead of us.
    Dotcoms guys, prepare for the worst!

  2. I think the 17% value is about whether the economy is currently in a recession.
    You incorrectly wrote that it is the “probability of a recession this year”.

  3. Denis Kolesnichenko says:

    Ehr… Okapi, isn’t it the same? If we’re currently in recession then this year will be a recession year… :)

  4. Denis,
    The probability of a recession this year allows for a recession at any point this year.
    The probability the economy is currently in a recession implies the recession needs to have begun in Q1.