Rotating inside the prior day's value area (buy the value-area low, sell the value-area high, target the point of control) did not hold up as a mechanical strategy on NQ. None of 84 fade variants, including sweep entries and a midday-only window, was profitable out of sample, and a two-close acceptance rule did no better. Value-area levels did help as confluence for other setups.
The idea
The value area is the price range that held about 70% of the prior session's volume, with the point of control (POC) at its busiest price. A popular auction-market-theory play fades the edges: when price reaches the value-area low (VAL) or high (VAH), trade back toward the POC, because the market "accepted" that range yesterday.
What we tested
- Touch entries at VAH/VAL with targets at the POC or the far edge, on 1- and 5-minute bars: 36 variants.
- Sweep-and-reclaim entries (price pokes beyond the edge, then closes back inside) and a midday-only window, when ranges tend to hold: 48 more.
- The acceptance rule from order-flow forums: two consecutive 15-minute closes back inside value before entering, or beyond it for a continuation trade.
Each variant was fitted on one period and checked on a later one it had never seen (out of sample).
Results
Not one of the 84 fade variants was profitable out of sample. The best in-sample results had a profit factor of only 1.02 to 1.06 (gross wins barely above gross losses), and the out-of-sample periods lost. The two-close acceptance rule fared no better: its best configuration was break-even in-sample, and the few out-of-sample winners came from samples of about 70 trades, small enough to be luck.
Why: a mechanical fade can't tell a value edge that will hold from one price is about to accept through. Every breakout through value is a stopped-out fade.
Where value levels did help
In a separate study of supply and demand zones, zones that contained the prior day's POC did much better (profit factor 2.0, against a losing result for zones overall), and value-area edges were among the few levels that improved results. The value area works as a reason to trust a setup you already have, not as a setup by itself.
Simulated results on historical data, after commissions. They don't guarantee future results; see the risk disclosure and CFTC Rule 4.41.