No. of Recommendations: 8
Jim (yes, Mungo's) last backtest post in 2019 - https://yorickm.com/Alas, poor Yorick. I knew his backtests, Horatio.
I haven't tested it, but there are some quite reasonable small substitutions one might make without ruining the original spirit of his methods.
First, one must appreciate that the original system was built on three models, each of which was bullish about 2/3 of the time and bearish 1/3 of the time. So any substitutions should honour that.
For P/E, a switch to CAPE would make reasonable sense. The cutoff would be whatever level was bullish 2/3 of the last 25 years, something like that.
The others are reasonably good. Maybe I would consider a self-promoting 99-day rule instead of the moving average? Very similar, fewer signal changes per year than the moving average test, works as well or better.
Still, kudos to the fellow. Not crazy.
Of course, having an omen about whether the broad US market is likely to well versus badly in the next while is only of interest if you invest in it or your portfolio correlates very strongly with it. To the extent that your portfolio is independent of that (I wouldn't buy SPY if you paid me right now), it's just a game to play.
Jim
