Anchored VWAP

The arithmetic is trivial and the interpretation is the whole point, so it is worth being precise about what the number represents.

Take every trade since your anchor. Multiply each price by the size that traded there, add them all up, and divide by the total size. What you get is not "the average price over the period" — a simple moving average is that, and it weights a quiet Tuesday the same as a panic. What you get is the price the average share changed hands at.

Which means: if price is above the line, the typical participant who bought since the anchor is in profit. If price is below it, they are underwater. The line is a live scoreboard for a specific, identifiable group of people.

Every other line on your chart is a statement about price. This one is a statement about the people who own it.

Why that produces real support and resistance

Three mechanisms, and unlike most level-drawing they are concrete:

1. Breakeven behaviour. A cohort that is collectively underwater sells into strength as price approaches their average — the disposition effect operating at scale. That is resistance made of people, not a line.
2. Adding at your own average. A cohort that is collectively in profit treats a dip back to its cost basis as a gift, because it is the price at which they already decided to own it. That is support.
3. Institutions are literally benchmarked to it. Execution desks are measured on whether they filled better than VWAP, and a very large share of institutional order flow is worked by algorithms targeting it. Few other lines on a chart have actual orders pointed at them by mandate.

Exhibit 1 · Anchored at the capitulation lowPrice against the cohort's cost basis

Note what the line does at the pullback: it does not stop price because it is a line. It stops price because the people who bought this advance have their break-even there, and a great many of them would rather add than watch a winner become a loser.

A VWAP anchored to nothing in particular tells you about nobody in particular. This is the part people skip, and it is the only part that requires judgement.

The default session VWAP that most platforms draw is anchored to the exchange's opening bell. That is a real anchor for a day trader, because the session genuinely resets the population. For anything longer it is arbitrary — it answers "how is everyone who bought since 9:30 this morning doing," which is rarely the question.

Anchor instead to the event that created the current population of holders. The same chart with three different anchors gives three different lines and answers three different questions:

Exhibit 2 · Three anchors, one chartThree different questions

Anchors worth using

• An earnings gap or a major announcement. The single cleanest anchor there is: everything before it was priced on different information, so the holders before and after are genuinely different populations.
• The all-time or cycle high. Defines the trapped cohort. Its line is where overhead supply lives, and it usually falls as time passes, which is why old resistance gets easier to clear.
• A capitulation low on climax volume. Defines the new cohort. Heavy volume at the anchor makes the line sticky, because a large share of the total weight is set right at the start.
• An IPO, index inclusion or a listing. The moment the instrument's holder base was created or transformed.
• Not: a round number of days back. "200-day VWAP" anchors to whatever happened to be going on 200 days ago, which is a property of the calendar rather than the market.

The most useful configuration is not one line. It is two lines converging.

The VWAP anchored at the high falls over time, because every share traded below it drags the average down. The VWAP anchored at the low rises, for the mirror reason. Eventually they meet — and the price where they meet is a price at which the trapped cohort and the new cohort are simultaneously at break-even.

That is an unusually loaded price. Everyone who bought the top can get out flat; everyone who bought the bottom is about to give back all of their gain. Both populations have a reason to act, in opposite directions, at the same number.

Exhibit 3 · ConvergenceTwo anchors, one decision zone

• It needs real volume. On instruments that report tick counts instead of size — much of spot FX — the weighting is fictional and the line is close to a moving average with extra steps.
• The anchor is discretionary, which makes it fittable. You can almost always find an anchor whose line touches the lows you wanted it to touch. Choose the anchor from the event, before you look at where the line lands, or you are drawing a conclusion you already had.
• It is path-dependent and it never resets. The line today depends on every bar since the anchor, so a single enormous volume day can dominate it for months. That is sometimes exactly what you want and sometimes a distortion — check where the weight came from.
• It gets less sensitive with age. Early on, each new bar moves the line noticeably. A year in, the accumulated volume is so large that new trading barely shifts it. Old anchors produce lines that are nearly flat and correspondingly less informative about anything current.
• It says nothing about direction. Like everything else in this series. It tells you where a population's pain is, not which way that population will resolve it.

• Pick the anchor from an event, and name the event out loud. If you cannot say what happened on that bar, it is not an anchor.
• Run two at once — one from the most recent significant high, one from the most recent significant low. The pair is more informative than either alone.
• Watch for convergence. Two anchored lines closing on each other marks a price where two populations are both at break-even, and those resolve decisively.
• Read the first touch differently from the fifth. A cohort defends its cost basis most strongly the first time it is tested. Repeated tests consume that population exactly the way repeated tests consume a triangle's ceiling.
• Check where the volume weight sits. A line dominated by one climax bar behaves differently from one built evenly. The profile page is the tool for that.
• Use it for position management, not entries. "Am I on the right side of the average buyer since the event that matters" is a better question than "is this a buy signal," and it is the question this instrument actually answers.

A description of market mechanics, not trading advice.