What it watches
Some price moves tend to repeat at the same point in the calendar, each month or each year. Each row is one pattern: whether it is in season today, and whether the evidence holds up.
Three examples
- Gold, August to September. The idea is that buying ahead of India's wedding season lifts the price. Tested 5 times, it averaged about 6% each time. Five is far too few to trust.
- Stocks at the turn of each month. Tested on 52 month-ends, it averaged about a third of a percent each time. Plenty of cases, and close to nothing.
- Cicadas and orange juice. The idea is that cicada swarms stress citrus crops and move orange juice prices. The next swarms are due in 2027 and 2028, so there is nothing to test it on until then.
The day-to-day screen
Why show it when nothing passes? A tool that called every idea a winner would be
broken. This one checks each idea against price history, plus weather and crop records where the idea
depends on them, and says not proven when the evidence is not there.