We Tested It: Friday's Low on Monday

"If Friday can't beat Thursday's high, Friday's low gets hit on Monday" — checked on 20 years of S&P 500 days: the real hit rate, the comparison that matters, why Friday isn't special, and why a hit is not a trade. This free lesson covers: The claim, Does it fit?, The test, Is 53% overwhelming?, Compared to what?, Measure the gap, Friday isn't special, What did the weekday test show?, A hit is not a trade, How would you use it?, Does it hold up?, Put it together.

Academy

0 XP

Strategy Playbook · Learn · Step 1 of 12 · about 4 min

The claim

Thu

Fri

Mon

A trader in a viral clip says: if Friday's high stays below Thursday's high, Friday's low gets hit on Monday — with "overwhelming" odds.

We tested it on 20 years of daily data. Here is what came back.

Going deeper

More on we tested it: friday's low on monday

This lesson is the long form of the Friday-rule Reel. The rule comes from a short clip of a trader saying that when Friday's high is below Thursday's high, Friday's low gets hit on Monday with "overwhelming" odds. We paraphrase the claim because the clip's exact wording is hard to make out.

Data: daily S&P 500 and Nasdaq Composite candles from Yahoo Finance, January 1999 – December 2018, as shipped with the `arch` Python package (5,031 trading days). A Friday counts when the session before it was a Thursday; "Monday" is the next trading day, so after a Monday holiday it is the Tuesday. "Hit" means Monday's low was at or below Friday's low. Fridays whose high equalled Thursday's (rare) count with the higher highs.

The weekday comparison uses the same rules for every day: the session before must be the previous weekday, and the next trading day must reach the day's low. After a lower high, S&P 500 figures ran from about 50% (Thursday) to 55% (Tuesday); after a higher high, from 32% (Friday) to 42% (Monday).

Common mistakes4

Quoting a hit rate without the control group — the same outcome when the pattern is absent.
Treating a touch of a level as a trade. Monday reached Friday's low about half the time but closed lower than Friday less than half the time.
Naming a pattern after one weekday when every weekday shows it.
Stretching the window ("sometime next week") until almost any nearby level gets hit.

From experience3

Use the previous day's low after a lower high as a first target or a level to watch, not as an entry signal.
When you test a rule, write down three numbers: the hit rate, the control group's hit rate, and whether the outcome made money.
Split your data into two periods. A real effect should show in both, even if the size changes.

Questions people ask2

Does this apply to US30?

The test used the S&P 500 and the Nasdaq, which moved the same way. The Dow usually moves with them day to day, but its exact figures would differ — check it on your own broker's data before relying on it.

Why stop in 2018?

That is the free daily dataset we could test reproducibly. A newer period could give different numbers, which is one more reason to treat the result as a tendency, not a rule.