Monday Is the Worst Trading Day — Right?
Many investors have the feeling: Mondays tend to go poorly on the stock market. On Fridays people sell. But is that really true — or just a gut feeling?
This is exactly where statistics come in. With SeasonAlpha you can look up for any stock or index which weekdays have historically performed better or worse — and, most importantly: whether the difference is statistically robust or pure chance.
Using the example of Siemens stock, we show how it works.
What Is a Significance Test?
Imagine you flip a coin 10 times and get heads 7 times. Is the coin rigged? Maybe — but 10 flips are not enough to say for sure. Only with significantly more flips and a clear pattern can you speak of a genuine effect.
It is the same on the stock market. If Siemens has risen an average of +0.17% on 213 Mondays — is that genuine strength or random noise?
The t-test is the statistical tool for exactly this question. It checks whether an observed return is large enough that it is very probably not due to chance.
The key metrics involved:
- t-value: The larger the absolute value, the stronger the signal. Rule of thumb: |t| > 2 is a first indication of significance.
- p-value: The probability that the observed pattern is pure chance. p < 0.05 is the threshold for statistical significance — meaning: less than 5% probability that it is coincidence.
- n: The number of trading days evaluated. More data = more stable conclusions.
- Win rate: How often was the day positive? 54% means more than every other day showed a gain.
Siemens Weekdays Under the Microscope
Let us look at the results from SeasonAlpha for Siemens stock:
| Weekday | Avg Return | Win Rate | t-value | p-value | Significant? |
|---|---|---|---|---|---|
| Monday | +0.17% | 54% | 1.15 | 0.2531 | No |
| Tuesday | +0.46% | 53% | 1.99 | 0.0476 | Yes |
| Wednesday | +0.00% | 54% | 0.02 | 0.9846 | No |
| Thursday | -0.17% | 49% | -0.94 | 0.3489 | No |
| Friday | +0.78% | 49% | 1.15 | 0.2520 | No |
The result is surprisingly clear: Four out of five weekdays show no statistically robust return. Only Tuesday passes the test — with a p-value of 0.0476, just below the 5% threshold.
What Does This Mean in Practice?
A few important points for context:
Tuesday is statistically significant — but that does not mean Siemens rises every Tuesday. It means: over many hundreds of trading days, Tuesday was systematically stronger than chance could explain. The t-value of 1.99 is just above the critical mark of 2.
Friday has the highest average return (+0.78%) — sounds impressive, but statistically it is not. The p-value of 0.2520 shows: the dispersion of Friday returns is too large to speak of a genuine pattern. This is a classic example of "large but not significant".
Wednesday is completely neutral: t=0.02 and p=0.9846 — statistically barely distinguishable from zero. No pattern, no signal.
Thursday is slightly negative (-0.17%, win rate only 49%) — but here too the data is not clear enough to derive a stable conclusion.
How Do You Use This in Practice?
Significance tests are not a trading system — but they help you separate gut feelings from genuine patterns.
A few ways to think about it:
- Long-term investors can use significance tests to identify favorable entry windows — not as a timing tool, but as an additional filter when planning purchases.
- Compare different stocks: Is the Tuesday effect unique to Siemens, or does it also appear in other DAX stocks? SeasonAlpha makes exactly this comparison possible with a few clicks.
- Combine with other filters: On the Weekdays page in SeasonAlpha you can add indicator filters (e.g., RSI, SMA) — and check whether the Tuesday effect only appears in certain market phases.
How to do it in SeasonAlpha:
- Open the "Weekdays" page
- Enter the ticker in the sidebar (e.g.,
SIE.DE) - Expand the "Statistical Significance of Weekday Effects" section
- The five gauges show you at a glance: score, t-value, p-value, average return, and significance
Conclusion: Statistics Protects Against Costly Mistakes
The weekday effect is real — but only when it is statistically robust. For Siemens stock, that is only the case on Tuesday. Four other days may look interesting visually but do not pass the test.
This is the strength of SeasonAlpha: not simply displaying average values, but showing whether you can actually trust them. Because on the stock market, a pattern that looks good without a statistical foundation is often more dangerous than no pattern at all.
Try it yourself at seasonalpha.ai — choose your favorite stock and see which weekday statistically really counts for it.