Good article.
Yield curve inversions for market timing seems quite overrated. Powel is raising one of their flaws, that it can just downright fail to predict a recession because not all yield curve inversions are the same (10% short-term rates with long term rates at 7% and 8% inflation are one thing, 3% fed funds with 2.5% long-term rates with 2.5% inflation are another)
But even if they somehow were super predictive, then you got the issue of timing. It takes an avg/median of roughly 20 months from inversion to recession, in the meantime stocks will be rising. Lets say they are up by 12% on avg from one point to another (the article I showed does not provide the figure).This seems a resonable return for a 20 month time frame from historical data.
Then the recession hits and markets drop about 20-30%. So the market timer had to miss out on 12% of gains in order to avoid a 25% drop, his net 'gain' (amount of losses he avoided) was 16 points (112 - 25% = 84 | 100 - 84 = 16). But that would also trigger taxes+transaction costs. Factoring that in, maybe the net valued added was around 10%.
But it gets worse because that assumed the market timer got the exact bottom, which he most likely won't. And if the market bottom out at -20% but he expected -25% (or -21% for that matter), the market might just started soaring without him, what does he do, buys or waits out for a bottom retest? Or maybe new lows?
But it gets even worse because it assumes the signal will ALWAYS work, which it wont. Sometimes it will fail (happened twice, using 3mTbill and 10yNote according to the Cleveland Fed and I'm sure it will happen at some point with 2ynote 10y note data, its a 5 sample size "research evidence" after all. A 5 sample size in a extremistan data set is not much evidence of anything) and when it fails, it will fail exponentially. That is, the market will continue to go higher exponentially for quite some time and the market timer will be scratching his head wondering when to get back in.
The problem of the failure of the signal is a significant, if yield curve inversions lead to recessions and recessions lead to market dips which are great buys, what happens when the yield curve inverts and there is no recession? When do you get back in? When the yield curve 'disinverts'? How much higher will be the stock market be by the time the yield curve disinvert? I suspect it will be quite a bit higher, perhaps 10-20%, but it could be 30-50%. Stock markets are unbounded on the upside and exponential in nature. All of that should go into the equation of the market timer
In sum, bailing out of stocks because the yield curve inverted is quite likely to be on average a bad decision. Bailing out when the curve is very flat and it MIGHT invert (like David Rosenberg seems to be suggesting his readers do) is just downright silly. Its not a surprise the David makes his money off writing and not by taking risk. Real risk takers get punished by markets quite quickly and learn that more robust forms of investing/trading are necessary
