An annual summary of Model B’s results is below.
I created Model B to see if it could provide a useful indication of whether the market is more likely than average to be higher in 4 weeks than it currently is. The QQQ ETF is used as a proxy for the market.
A ‘Positive’ model output means that in the past, under similar conditions, there was a greater than average chance that the market would be higher in 4 weeks than it was when the output was made.
A ‘Negative’ model output means that in the past, under similar conditions, the model did not provide any statistically useful information about likely market direction over the next 4 weeks.
The table below is an annual summary of Model B’s performance. For each year the table lists:
The number of ‘Positive’ signals it gave
The percentage of ‘Positive’ signals where the market was higher 4 weeks later
The average percentage change in the market on a ‘Positive’ signal 4 weeks later
The number of ‘Negative’ signals it gave
The percentage of ‘Negative’ signals where the market was higher 4 weeks later
The average percentage change in the market on a ‘Negative’ signal 4 weeks later
I encourage you to look at the weekly results as well as the annual summary. Averages are useful, but they hide a lot of important information.
I’m primarily interested in the ‘Positive’ signal statistics. I include the ‘Negative’ signal statistics for comparison purposes.
| Year | Positive Count | Positive Up Percent | Positive Avg Change | Negative Count | Negative Up Percent | Negative Avg Change |
|---|---|---|---|---|---|---|
| 2017 | 9 | 77.8% | +2.1% | 43 | 86.4% | +2.2% |
| 2018 | 10 | 80.0% | +2.8% | 42 | 47.6% | -0.8% |
| 2019 | 21 | 66.7% | +2.4% | 31 | 80.6% | +2.5% |
| 2020 |
Keep in mind that Model B was created based on past market behavior, but future market behavior may be very different than past behavior.