Day of the week rule

The code

The script does what an honest test should: it finds the pattern in the first half of the prices, then tests it on the second half, which played no part in the choice.

Data the script reads: invented-prices.csv. Invented values, never a real market. Save it in one folder and run the script from that folder.

The file that was run: day_of_week.py

# A day of the week rule, used to show how easily a pattern is found by chance.
# Prices come from invented-prices.csv: a random walk made by a script, never a real market.
# The script that made them treats every weekday the same. Any weekday pattern below is luck.
import csv
from datetime import date

NAMES = ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday"]

weekday, closes = [], []
with open("invented-prices.csv", newline="") as f:
    for row in csv.DictReader(f):
        weekday.append(date.fromisoformat(row["date"]).weekday())    # 0 is Monday
        closes.append(float(row["close"]))


def average(values):
    return sum(values) / len(values)


def changes_by_weekday(first, last):
    """Group each day's change from the previous close under its weekday."""
    groups = {d: [] for d in range(5)}
    for day in range(max(first, 1), last):
        groups[weekday[day]].append(closes[day] - closes[day - 1])
    return groups


COST = 0.10              # points paid on every completed trade, a stand in for the spread
HALF = len(closes) // 2

# Step 1. Look at the first half of the prices and find the weekday with the highest average change.
first_half = changes_by_weekday(0, HALF)
print("First half, average change by weekday:")
for d in range(5):
    print(f"  {NAMES[d]:10} {average(first_half[d]):+.3f} over {len(first_half[d])} days")
pick = max(range(5), key=lambda d: average(first_half[d]))
print("Highest in the first half:", NAMES[pick])

# Step 2. Test that weekday on the second half, which played no part in choosing it.
# The rule: buy at the close before that weekday, sell at its close, pay the cost each time.
second_half = changes_by_weekday(HALF, len(closes))
trades = second_half[pick]
print("Second half, the same weekday:")
print("  Average change:", f"{average(trades):+.3f}", "over", len(trades), "days")
print("  Trades:", len(trades))
print("  Total points before costs:", round(sum(trades), 2))
print("  Total points after costs:", round(sum(trades) - COST * len(trades), 2))
print("  Simply holding over the second half:", round(closes[-1] - closes[HALF] - COST, 2))

# Step 3. Where did that weekday come among the five in the second half?
order = sorted(range(5), key=lambda d: average(second_half[d]), reverse=True)
print("  Its place among the five weekdays in the second half:", order.index(pick) + 1)

What the run printed

First half, average change by weekday:
  Monday     -0.169 over 74 days
  Tuesday    -0.014 over 75 days
  Wednesday  +0.022 over 75 days
  Thursday   +0.096 over 75 days
  Friday     +0.077 over 75 days
Highest in the first half: Thursday
Second half, the same weekday:
  Average change: -0.111 over 75 days
  Trades: 75
  Total points before costs: -8.33
  Total points after costs: -15.83
  Simply holding over the second half: -8.55
  Its place among the five weekdays in the second half: 4

Output of the run on 9 Oct 2026, copied as printed. Every input to this run was invented, so the result says nothing about any real market.

How to read it

In the first half the weekday averages are spread out, with Thursday highest and Monday lowest. Shown only this table, most people would start to think of reasons why Mondays are weak. There are none. The file has no weekday effect in it.

In the second half the same weekday had an average change below zero and came fourth of the five. The pattern did not weaken. It was never there.

A rule that acted on it paid a cost on every one of its trades and ended well below zero, and below simply holding. Had the test stopped at the first half, none of that would have been seen. It says nothing about any real market.