The 50-day and 200-day moving averages are the most-drawn lines in this market. "Above the 200-day is bullish, below is bearish" appears in nearly every course sold in Pakistan.
We tested it on 67,098 observations.
| TRAIN sample | touched +20% | TEST sample | touched +20% | |
|---|---|---|---|---|
| Above the 50-day average | 16,971 | 1.08× | 17,720 | 1.00× |
| Below the 50-day average | 19,079 | 0.93× | 13,328 | 1.00× |
| Above the 200-day average | 18,051 | 1.10× | 20,873 | 0.98× |
| Below the 200-day average | 17,999 | 0.90× | 10,175 | 1.04× |
Read the held-out column. 1.00×, 1.00×, 0.98×, 1.04×. Whether a PSX stock was above or below its 200-day moving average told you essentially nothing about whether it would move 20% in the next month.
Worse, it inverted. In 2016-2022 being above the 200-day looked mildly better (1.10× vs 0.90×). In 2023-2026 being below it looked mildly better (1.04× vs 0.98×). A signal that reverses sign between periods is not a weak signal. It is an absent one, and the apparent edge in the first period was the era, not the rule.
So why does the 200-day appear in the winning combination?
Because it works as a filter on a signal, not as a signal itself. From the last lesson:
- Breakout alone: 1.42× / 1.20×
- Breakout and above the 200-day: 1.55× / 1.25×
The moving average added roughly a tenth of a turn, only once something else had already fired. On its own it added nothing. This is the difference between a trigger and a qualifier, and it is worth more than any individual indicator in this course:
A qualifier improves a setup you already have. It cannot create one. Most people use moving averages as triggers, which is why the line does no work for them.
Do this now, practical activity
- Open SuperCharts, load any stock, and add both the 50-day and 200-day moving averages.
- Count how many times in the last year price crossed the 200-day line.
- For each crossing, mark whether the next month was up or down.
What you will almost certainly find: roughly half and half. That is the 1.00× above, rediscovered on a single chart. If your stock happens to show a clean pattern, that is what 20,873 observations exist to correct.
Measured: in the held-out years 2023-2026, stocks above their 200-day average touched +20% at 0.98× the base rate (n=20,873) and stocks below it at 1.04× (n=10,175). No usable edge in either direction, and the sign flipped from the earlier period.