Week 6 Recap: Officially In The Green
Six weeks in, and for the first time since the very first day of the experiment, all four portfolios are positive since launch. I am not going to do a victory lap, because I have been doing this long enough to know the market will punish anyone who does, but I am allowed to be quietly pleased.
The S&P 500 clawed its way back to a razor-thin +0.11% since launch. The portfolios are sitting between +5% and +10%. That is a much bigger spread than I expected this early, and I will be very interested to see how much of it sticks.
Week 6 Performance
| Dates | S&P 500 | Ensemble | Growth Eval | ML Model #1 | ML Model #2 |
|---|---|---|---|---|---|
| Week 6 (4/6 – 4/10) | +3.57% | +8.40% | +7.27% | +9.07% | +9.90% |
Daily Breakdown
| Date | S&P 500 | Ensemble | Growth Eval | ML Model #1 | ML Model #2 |
|---|---|---|---|---|---|
| 4/6 | +0.47% | +1.12% | +1.04% | +1.43% | +1.74% |
| 4/7 | +0.04% | +0.69% | +0.61% | +0.82% | +0.75% |
| 4/8 | +2.55% | +5.17% | +4.49% | +5.31% | +5.33% |
| 4/9 | +0.58% | +1.16% | +0.96% | +1.19% | +0.96% |
| 4/10 | -0.07% | +0.26% | +0.17% | +0.32% | +1.12% |
Cumulative Since Launch (3/2)
| S&P 500 | Ensemble | Growth Eval | ML Model #1 | ML Model #2 | |
|---|---|---|---|---|---|
| Total | +0.11% | +6.15% | +5.71% | +7.70% | +10.46% |
Tuesday 4/8 did most of the heavy lifting. The S&P 500 was up 2.55%, and the portfolios were up more than 5%. That is the second big bounce day in two weeks, and it is the same pattern we saw in Week 2, the portfolios hold higher-beta names, so they catch more of the upside on days like this. Of course they also catch more of the downside on days like Week 4's Thursday, so I am trying to stay even-keeled about it.
Every single day this week was green for at least four of the five tracked lines, including Friday where the S&P 500 barely dipped but the portfolios squeezed out small gains. That is the kind of quiet consistency I like to see.
ML Model #2 is leading at +10.46% since launch. That is a six-week number on a very small sample, and I would be lying if I said I was not watching it closely. But I wrote this same warning back in Week 2, the sample size is tiny, and the true test of any of these models is months of data, not weeks. If ML Model #2 is still leading in July, then we can talk.
Q1 earnings start rolling in next week in earnest. That should tell us whether this rally has legs or whether it was just a relief bounce from the late-March sell-off. Either way, I will be back here next weekend with whatever the market decided to hand me.
Thanks for reading!