DJI 1D: does the best calendar month beat a random calendar?Dow Jones Industrial Average IndexTVC:DJIGuilleMustafaTHE QUESTION Sort 130 years of Dow Jones months by their average return and July comes out on top. Does that lead hold up against a calendar whose month labels were dealt out at random? WHAT I MEASURED The open-source indicator "Seasonality vs Shuffled Calendar: Is the Best Month Real?" on this daily chart, default settings: one close-to-close return per completed month (1,560 months) against 500 shuffled calendars, seed 12345. WHAT THE TABLE SHOWS The average month returned +0.60%. July leads at +1.44% (131 months, 64.9% up, z +1.83), ahead of November at +1.23% and April at +1.18%. The chance that a random calendar's best month reaches +1.44% is 32.1%: "within random range". September is last at -1.10%, up in 42.3% of its 130 months, z -3.67. READING A z of +1.83 falls short of the usual 1.96 even before allowing for the pick of the best of 12 months. With about one random calendar in three doing as well, the test cannot separate July's lead from the best month of a random calendar. LIMITS Only the best month gets a shuffle test, so September's -3.67, the most extreme z on the board, has no p-value here. Under the shuffle any return can land in any month, so trends and clusters of calm or volatile years are ignored. The index is price only, without dividends. TRY IT The indicator is open source and on this chart: add it to your own symbol, or switch it to day of the week. Not financial advice.