52-Week High / Low

### What the number actually is A rolling window, not a calendar year. Every session it takes in the newest day and drops the day that has aged out the back, so the level moves even when price does nothing dramatic. A stock can register a new 52-week high partly because the old high from just over a year ago fell out of the window, not because it rallied hard. Prices are split-adjusted, so a 2-for-1 split halves the historical high rather than inventing a fake collapse. One practical gotcha: sources differ on whether the level is set by the *intraday* extreme or by the *closing* price. Two platforms can therefore disagree about whether a new high actually printed. Check how your feed defines it before you build any rule on top of it. ### Why funds and algorithms genuinely care This is not folklore. In [the Journal of Finance (2004)](https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-6261.2004.00695.x), Thomas George and Chuan-Yang Hwang found that a stock's nearness to its 52-week high forecasts future returns, and that it dominates and improves on past returns as a momentum signal. Returns forecast this way did not reverse in the long run, which is unusual for momentum effects and is part of why the level gets so much institutional attention. The usual explanation is **anchoring**. Traders treat the old high as roughly what the thing is worth, so when good news lands they are slow to bid above it and the information gets priced in late. When the level finally gives way, that delayed adjustment arrives all at once. That is the mechanic underneath "new highs often continue". Crossing the level also changes *who* is in the market. Huddart, Lang and Yetman (*Management Science*, 2009) found trading volume is strikingly higher when a stock's price crosses the upper or lower limit of its past trading range, and more so the longer it has been since that extreme was last reached, the smaller the firm, and the murkier the valuation. The line is not just a line. Breaching it drags in participants who were not watching an hour ago. ### How to read it inside our system Treat a 52-week high or low the way you treat any other obvious level: as a **liquidity pool** and a **POI**, never as an entry in itself. Stops and resting orders cluster just beyond it for the same reason they cluster at equal highs. It is also an HTF object, which means it belongs to bias and targets, not execution. You still need the setup timeframe, a valid entry model, the rejection candle, the RR check, and divergence where the rules demand it, before any of it becomes a trade. The mistake is buying the touch. Continuation and violent rejection happen at exactly the same price, and in the moment they look identical. The level tells you where the fight is. Your entry criteria tell you who won it. The forex equivalent is the yearly high and low on a major pair. Same anchoring, same order clustering, with no earnings report to detonate it overnight.