The profile key: why 'is ATR high?' has no answer without three coordinates.

The relativity problem

Eyeballing ATR is a judgment with no denominator.

The manual states the problem exactly: a live ATR can look large or small by eye, but that judgment is weak unless compared against the correct asset, timeframe, and session — GBPJPY 5M during London, GBPJPY 5M during Asia, XAUUSD 5M during New York, and NAS100 15M in a news window are different volatility environments. Every one of those differences matters to a stop: a distance that gives a London runner ordinary breathing room may be triple what the Asia session needs, and a coefficient chosen against the wrong baseline is wrong in whichever direction the mismatch runs. The eyeball has no access to any of this. The baseline database does.

The key construction

One string, one lookup, one correct comparison population.

The profile key is deliberately simple machinery: Asset, Timeframe, and Session (or ALL) concatenated into a single identifier that every lookup runs through. Each key owns its own accumulated sample of ATR readings and its own derived baseline statistics — minimum, average, median, maximum, and the percentile thresholds. The baseline mode selector chooses the comparison population: All Sessions for the asset-timeframe's general character, or Session-Specific for the sharper question of what this hour of the trading day usually looks like. The construction guarantees the one property eyeballing can't: the live reading is always being compared to history that was actually its peer — same instrument, same bar size, same slice of the day.

The zone read

Five zones from four percentile fences.

Against its key's baseline, a live ATR classifies into the official five-zone scheme via the percentile thresholds — P20, P40, P60, P80 — yielding a placement from the quiet bottom zone through normal to the extreme top. Alongside the zone, the matrix reports the ATR Ratio (live divided by historical average — the expansion or compression multiple) and the Intensity Score (the ratio mapped to a one-to-ten context scale). The three reads answer the same question at three granularities: the zone for classification, the ratio for magnitude, the intensity for quick communication. All three inherit their meaning entirely from the key — which is why the matrix's first QA question about any strange reading is always whether the profile key resolved to the population you intended.

• Percentile fences move per key: London's P80 and Asia's P80 are different pip values describing the same rarity.
• The ratio is the bridge to intuition — '1.6× normal for this key' communicates instantly what a raw pip value can't.
• Zone placement feeds the coefficient spectrum: the same rung translates to different distances as the key's ATR changes.

Who consumes this

Mostly the trail trades — and deliberately not the Normal branch.

The matrix's scoping doctrine follows branch architecture. Its readings matter most where runner survival and continuation capture are the trade — Trend Partial, Trend No-Partial, and any ATR-trail-managed position — because those are the structures whose management parameter the volatility context actually prices. The Normal branch is explicitly walled off: its static 2R model has no trail to price, and the guidance is blunt about not over-engineering it — Normal trades shouldn't be routed through coefficient analysis unless the plan explicitly calls for a runner review. The matrix should reduce hesitation, not create cockpit clutter: three practical questions — what is live ATR in pips, is it low or normal or high or extreme relative to baseline, and what does each coefficient translate to in actual distance — and out.

The key idea

Context is a lookup, if you built the table.

The profile key system is unglamorous infrastructure with an outsized payoff: it converts the hardest-to-defend judgment in trade management — 'volatility seems high' — into a percentile fact with a named comparison population. The cost is discipline at capture time, feeding the sample log so every key's baseline hardens. The return is that every trail decision, forever after, starts from a placement instead of an impression — and placements, unlike impressions, can be audited when the coefficient-versus-outcome review comes around.