Hacker News new | ask | show | jobs
by ajross 905 days ago
> It's best not to conflate actual measures being tentatively interpreted on the basis of a century of history with some guy with a camera and a YouTube channel

The commenter upthread "penciled in" a recession for Q3 of next year and cited your favorite metric. I think that's closer to a TikTok hit than a "tentatively interpreted" bit of pop economics.

But to treat with your actual point: it still sounds like bunk to me. I had to dig, but FRED does indeed have a chart for this (https://fred.stlouisfed.org/series/T10Y2Y) and sure, if you squint, it looks like it predicts. Except that the time between an inversion and the predicted recession is all over the map. It looks like the 1988 recession took two years (!) to actually arrive, while the 1980 recession jumped the gun. And 2008 seems to refute the theory, because the inversion had corrected itself almost a year before the financial crisis (which pretty clearly had nothing to do with bond rates anyway). Also the magnitude of the inversion doesn't seem to have any correlation with the recession, the inversion swung way lower in the late 70's than it did any other time, but that recession was actually pretty mild. And the inversion of 2006 was barely an inversion at all.

Yeah, this is wrong. No serious economics seem to be pushing this.

We have an inversion right now because the Fed has been swinging its hammer like crazy and the market is responding to the fact that they think rates are going to drop rapidly RSN (which is a higher risk for longer term bonds, obviously). That explanation makes a ton more sense than some handwaving about a "predictive metric".

1 comments

> while the 1980 recession jumped the gun.

No it didn't, it inverted in 1978. 1-2 years after the inversion starts is pretty consistent.

And yes, it is actually used as an indicator: https://fortune.com/recommends/investing/the-inverted-yield-...