I had a fun experience at work the other week. I went over some macroeconomic forecasts in preparation for a project, and then told my boss "this bullshit is fit only to make astrology look respectable by comparison." His response was "let's keep them on the table, I'll look them over."

So we come to our weekly meeting, and he asks me if I have understood why the trends go up in the forecast when they are consistently down in the historicals. Because he finds that strange. I tell him that they are up in the forecast because they are down in the historicals. The logic being that the economy cannot indefinitely depart from the long-run trend, so whenever it has departed from trend, the best guess for next year is that it will be closer to the long-run trend.

His response was along the lines of "but that's silly." I told him that yes, yes it is. But that is how these forecasts are made - it's a fundamental property of the statistical models in use. "By our guys down in forecasting?" he asked. To which I answered "by everyone. World Bank, IMF, Treasuries all over the world, Central banks, etc., etc."

The forecasts in question did not end up featuring in the final product.

- Jake

Friends come and go. Enemies accumulate.