Adjusted vs unadjusted prices will silently break your thresholds

There are three versions of any price series and picking the wrong one is a known beginner mistake, though you would not guess it from how rarely it comes up. The three: Unadjusted, what actually printed on the day. Correct history, but a split makes the series jump. Back-adjusted — history rewritten so today's price is real and the past is scaled. Continuous, but the old prices are not what anyone paid. Forward-adjusted. The past is real and recent prices are scaled instead. Why it matters concretely: any strategy with a price-level condition breaks. "Buy under $10" on a back-adjusted series is testing a completely different universe than it would have at the time, because a stock that's $200 today after splits might show as $2 back then. Same for anything using absolute dollar stops. It also breaks volume. If prices are adjusted and volume isn't, your dollar-turnover figures are wrong by the split factor for the entire pre-split history. Practical rule: for signal logic use adjusted, for anything expressed in absolute price or matched against raw tick data use unadjusted, and never mix the two in one calculation.