Torsten Slok, chief economist at Apollo Global Management, delivered the gorgeous chart above. The Fed’s forecast for interest rates has performed similarly. The same goes for “market forecasts” embedded in the yield curve, which typically slope upwards.

Torsten’s conclusion:

The economics community’s track record for forecasting 10-year rates is not particularly impressive, see chart [above]. Economists and strategists involved have systematically and incorrectly predicted higher long-term interest rates since the Philadelphia Fed began surveying professional forecasters 20 years ago. Their latest version has the same prediction.

good. Like the famous broken clock that gets right twice a day, note that the forecast is “correct” when interest rates are higher. So don’t necessarily run out and buy bonds today.

Could it be true that professional forecasters just act stupidly, refuse to learn, and the agencies that employ them refuse to hire more rational people?

My favorite alternative (I’ll admit, I’ve progressed a few times on this blog): When a survey asks people “What do you ‘expect’?” people don’t answer using the true measurement conditional mean. By giving answers very close to the real yield curve, these forecasters report numbers close to the risk-neutral conditional mean, i.e. not (sum_spi(s)x(s)), but (sum_s u ‘[c(s)]pi(s)x(s)). The risk-neutral mean is a better sufficient statistic for decision making. Pay attention not only to how likely events are, but also how painful it is to lose money in those events. Don’t make a mistake on a day when the company loses a lot of money. Weather services also tend to overstate wind forecasts. I interpret the 30 mph wind forecast as “if you go out and capsize your boat in a gust of 30 mph, don’t blame us.” “If you buy bonds and they fail, don’t blame us”, This is certainly part of the “forecast” of the financial industry.

If the risk-neutral mean is equal to the market price, do nothing. Doing nothing has to be advice to the average investor. There is some logic to directly reporting the forecast implied by the yield curve.

Economists are reacting too quickly, I think, and people are being asked “what do you expect?” surveys. Lamenting that the answer does not represent a conditioned means of true measurement, and blaming it on stupidity. It seems like anyone answering this question has the most vague idea of ??what the definitions of mean, median, mode, condition, and true and risk-neutral measures are. We might be a little more humble: they gave us a sensible answer, a question asked in English, but since we asked the question in a foreign language, it was an answer to a different question.

(Teaching is good for you. Most of my students don’t understand that “risk” can mean you make more money at the beginning of the course than you “expect”. I hope they get it by the end! But they are not wrong In English, “risk” means downside risk. In portfolio analysis, it means variance. Know what the word means.)

But that also means, don’t interpret the answer as a true measure of conditional means!

This graph is particularly challenging because it only deals with 10-year interest rates. Similar charts of short-term interest rates also show a consistent tendency to predict higher rates that won’t happen. But this is more forgivable as a risk-neutral mean, since the yield curve was upward sloping in the first few years. The projected rise in 10-year yields corresponds to a slope from 10-year to longer maturities, which is usually smaller. Torsten, if you’re listening, it would be very interesting to compare with the forecast implied by the forward curves for each date!