Volume Profile II: Shapes, Yesterday's Map and Value Migration

≈31 min read · Real gold futures examples with XAUUSD equivalents · Free

In this chapter: In Chapter 5 you learned how a volume profile is built and how to find its point of control (POC), value area (VAH and VAL), and high- and low-volume nodes. This chapter teaches you to read a finished profile. You will learn the common profile shapes (D, P, b, thin trend and double distribution) and the auction story each one is usually said to tell, with the limits of those stories spelled out. You will learn how yesterday's final profile becomes today's map, what "naked" levels are and why the claim that price "always returns" to them is folklore, how to classify value migration from one day to the next, how to describe where gold opens relative to yesterday's value (including an honest look at the famous "80% rule"), how volume profile differs from the older TPO (Market Profile), and finally the setup and reading mistakes that trip up most beginners, ending with a pre-session checklist for gold.

A quick word on prices before we begin. Every price in this chapter is an illustrative gold futures (GC) price. If you trade gold through a CFD broker, the instrument you see is usually called XAUUSD, and its price sits below the futures price by a gap that has recently been roughly $27–30 (the gap changes with interest rates, time to contract expiry and your broker's own pricing). Where a price matters, we give an approximate XAUUSD equivalent beside it. Treat those equivalents as rough, never exact: always measure the live gap on your own platform.

The Balanced Day: The D-Shaped (Normal) Profile

What it looks like

Picture a finished daily profile turned on its side so the price axis runs vertically. On a balanced day, the horizontal bars are longest in the middle of the day's range and get gradually shorter towards the top and the bottom. The outline looks like the capital letter D, or like a bell curve standing on its edge. This is called a D-shaped profile, and in Market Profile literature it is also called a normal profile (the name comes from the "normal" bell-shaped distribution in statistics).

Three features usually go together:

• One main bulge. There is a single high-volume area, roughly in the middle of the range.
• The POC sits near the middle. The point of control (the single price with the most volume) is close to the midpoint of the day's high and low.
• A roughly symmetric value area. The value area (the band holding about 70% of the day's volume) extends a similar distance above and below the POC.

The auction story

A profile is a record of where business was done. A D-shape tells a story of a two-sided auction. During the session, price moved up, but at higher prices it did not find enough new buyers to keep going, so it rotated back down. It moved down, but at lower prices it did not find enough new sellers, so it rotated back up. After several of these swings, most of the trading piled up around one central price that both sides were willing to accept.

In the language of auction theory (which Chapter 7 covers in full), this condition is called balance. Neither side had dominant control, and the market found a price that was "fair" for that window of time. The important words are for that window of time. Balance describes the past session; it is not a promise about tomorrow.

Think of a pendulum swinging between the edges of value, losing energy at each edge because there is less willingness to trade out there, and spending most of its time in the thick middle.

Companion clues

D-days rarely appear on their own; other indicators usually tell the same story:

• The daily VWAP is flat. VWAP (the volume-weighted average price from Chapter 4) barely slopes, and it tends to sit close to the POC. When the average price paid all day is close to the most-traded price, nobody pushed price far away for long.
• Price crosses VWAP and the POC many times. On a 5-minute chart you will see price wander back and forth across both levels.
• The range is often modest. A D-day's high-to-low range is usually smaller than on days with major economic news.

Example: on a quiet Monday, GC trades between 4,012.0 and 4,031.0 (XAUUSD roughly 3,984–4,003), the POC forms at 4,021.5 (XAUUSD ≈ 3,993.5), the value area runs from 4,016.0 to 4,026.5, and VWAP finishes at 4,021.8. Value is almost centred and VWAP sits on the POC: a textbook D.

What a D-shape does not tell you

A D-shape does not say that tomorrow will also be balanced. Balance always ends eventually, usually because something new arrives (a data release, a central-bank surprise, a geopolitical headline) and pushes one side into control. That shift from balance to directional movement is called imbalance.

There is a popular view in auction theory that long periods of balance tend to be followed by directional moves ("balance leads to imbalance"). As a general description of how markets alternate, that is reasonable. But the profile does not tell you when the break will come or in which direction. Two related claims circulate widely:

• "After a D-day, a trend day always follows." This is a popular claim, not proven.
• "On a D-day, always fade the edges" (that is, always trade against moves to the value-area edges). This is also a popular claim, not proven. It describes what balanced days look like in hindsight, not a rule that holds in advance.

The time-frame warning

A profile's shape depends on the window it covers. The same day can look like a D on a profile built only from regular trading hours (RTH, roughly 08:20–13:30 New York time for gold) and look quite different on a profile built from the full 23-hour Globex session. Several D-days can also merge into one larger D on a multi-day composite (Chapter 5). Whenever you name a shape, name the window too: "RTH profile, D-shaped", not just "it's a D".

Common mistake: Naming the shape before the session has ended. A developing profile at 10:00 a.m. can still turn into a P, a b or a trend profile by the close. Shape names belong to finished profiles.

Key idea: A D-shaped profile describes a two-sided, rotational auction in which the market found a temporary fair price. It tells you what happened, not what will happen next.

P and b Profiles: Short-Covering and Long-Liquidation Stories

The two shapes

A P-shaped profile has its volume bulge near the top of the day's range and a thin "tail" of low volume below it, so the outline resembles the letter P. Reading it as an auction record: in the lower part of the range, price moved up quickly, on little volume, and then, near the top, it slowed down and built volume.

A b-shaped profile is the mirror image: the bulge is near the bottom of the range, with a thin tail above. Price fell quickly through the upper part of the range and then built volume near the lows.

The thin part of each shape is the fast part of the day. The thick part is where the market slowed down and accepted prices (did lots of business there).

The classic interpretation

Market Profile literature, especially the books of James Dalton and his co-authors, offers a well-known reading of these shapes:

• A P-shape is often linked to short covering. Traders who had sold short earlier (bet on a fall) buy back to close their positions. Their buying lifts price quickly. Once they have finished covering, there are not many new buyers arriving, so the market stalls and rotates near the top.
• A b-shape is often linked to long liquidation. Traders who had bought earlier sell to close their positions. Their selling pushes price down quickly. Once the liquidation is done, there are not many new sellers, so the market rotates near the lows.

The logic is about fuel. A move driven by old positions being closed may have less fuel to continue than a move driven by new money entering (what auction theory calls initiative activity). When the closing is done, the push is done.

Honest limits

This story is plausible, but you should hold it loosely, for three reasons.

1. A volume profile cannot see whether trades opened or closed positions. A contract bought to close a short looks exactly the same in volume as a contract bought to open a new long. To get even a rough idea, you would need to look at changes in open interest (the total number of futures contracts outstanding, which the exchange publishes once a day). Even then, open interest is a blunt, end-of-day number and cannot tell you what happened in a particular hour.
2. P and b can simply be continuation. In a steady uptrend, a P can mean nothing more than "price moved up, then consolidated at its new level", which is exactly what healthy trends do.
3. Shape depends on the window. As with D-days, an RTH P-shape may look different on a Globex profile or a multi-day composite.

So treat P and b as a hypothesis about the day's story, then test that hypothesis with order flow: did delta and cumulative delta (Chapter 3) confirm aggressive buying during the fast leg? Did heavy selling get absorbed in the bulge? Later chapters give you the tools to ask those questions.

A worked gold example

Imagine an 08:30 New York data release that comes in weaker than expected. GC jumps from 4,018.0 to 4,034.0 (XAUUSD roughly 3,990 to 4,006) within fifteen minutes, with only a few contracts traded at each price on the way up. For the rest of the session, price rotates between 4,030.0 and 4,040.0 and builds heavy volume there, with the POC at 4,035.5 (XAUUSD ≈ 4,007.5). The finished RTH profile is a P: a thin tail from 4,018 to 4,030 and a fat bulge above.

Now ask the honest questions. Was the fast move covering or new buying? The profile cannot answer. If cumulative delta rose strongly during the jump and then flattened in the bulge, that fits "aggressive buyers lifted price, then two-sided trade took over", which suits both the covering story and a continuation story. On some days like this the move continues; on others it fades. The shape alone does not decide.

Common mistake: "P means bullish, b means bearish." This is a popular claim, not proven, and in some material you will even see it stated the other way round (P as a sign of exhausted buying). Neither rule holds reliably. Likewise, "a P is always short covering" is a popular claim, not proven.

Key idea: The thin part of a profile shows where price moved fast; the thick part shows where it was accepted. Who was trading, and whether they were opening or closing positions, is something the profile cannot see.

Trend Days and Double Distributions

The thin, elongated trend profile

On some days the profile has no dominant bulge at all. Instead it is long and thin: volume is spread thinly across a wide range of prices, with no single area where the market settled. This is the thin or elongated profile, typical of a trend day.

The auction story is a one-sided auction. The market kept probing new prices in one direction, and at each new price it failed to find enough opposite-side business to stop it. No fair price was established during the day, because the price kept moving on before it could become accepted.

Companion clues:

• VWAP slopes steadily in the direction of the move.
• Price spends most of the day on one side of VWAP, sometimes riding the outer VWAP bands (Chapter 4).
• The value area is wide and not very meaningful. On a trend day, "the band containing 70% of volume" is mostly a stretch of price that was traded through, not a zone of agreement.

Because nothing was accepted, tools built on "returning to the average" tend to struggle on trend days. That is a structural observation, not a statistic: if the market never returns to its average, a method that waits for a return keeps waiting.

The double distribution

A double distribution is a profile with two separate bulges (two high-volume nodes, HVNs) joined by a thin neck of low volume (a low-volume node, LVN). The story: the market was balanced in one area, then something knocked it out of balance (on gold, often an 08:30 New York data release such as CPI), price moved quickly to a new area, and there it found a new balance. One day, two small auctions.

Each of the two distributions has its own POC. The POC of the whole day may simply be the POC of the larger one, which hides the other completely. If you report only the day's POC and value area, you might draw a value area that sits right across the thin neck, an area where very little trading actually happened. That is misleading.

Many charting platforms offer a split profile function, which lets you break a session into separate profiles so each distribution gets its own POC and value area. On a double-distribution day, splitting the profile usually gives a much more honest picture.

The LVN in the middle

In profile literature, the thin neck between two distributions is the boundary the market crossed in a hurry. When price comes back to it on later days, traders often watch it as a decision area: either price is turned away and stays in its current distribution, or it passes through and travels back towards the other one. That is a popular claim, not proven. It is a descriptive tendency some traders report, not a rule.

Worked example

Suppose, before an 08:30 CPI release, GC balances between 4,040.0 and 4,048.0 with a local POC at 4,044.0. The number comes in hot, and in four minutes GC falls to 4,022.0, passing 4,026–4,038 on very light volume. For the rest of the session it balances between 4,015.0 and 4,026.0 with a local POC at 4,020.5. The whole-day POC lands at 4,020.5, because the afternoon distribution was bigger, and the morning POC at 4,044.0 (XAUUSD ≈ 4,016) is invisible unless you split the profile. The neck at roughly 4,026–4,038 (XAUUSD ≈ 3,998–4,010) is the LVN between the two auctions.

Common mistake: Fading (trading against) a trend day just because "price is outside the value area". On a trend day, the value area means little and price outside it is normal.

Common mistake: Assuming the LVN between two distributions "must be filled". That is a popular claim, not proven.

Key idea: A thin profile means no agreement was reached; a double distribution means the day contained two separate agreements. Split the profile so each gets its own POC and value area.

Yesterday's Profile as Today's Map

Why use yesterday?

Today's profile is still being built. Its POC and value area move as the day goes on (Chapter 5 called this the developing profile). If you use today's developing levels as references, you are aiming at a moving target. Yesterday's final profile, by contrast, is fixed: it will not change. That is why it is the most common reference map for the current session.

The standard levels

Before the session, traders usually mark these levels from the prior session's final profile:

• pPOC (prior POC): yesterday's most-traded price.
• pVAH and pVAL: the upper and lower edges of yesterday's value area.
• pHigh and pLow: yesterday's high and low, the edges of yesterday's range.
• pVWAP: yesterday's final VWAP (optional but useful).
• The most prominent HVNs and LVNs from yesterday's profile.

Why these levels matter

These levels summarise yesterday's agreement: where the market did most of its business and where it stopped. Today the market must, in effect, answer one question: Is yesterday's value still valid, or is the market looking for new value?

• If price keeps rotating inside yesterday's value area, the market is still accepting yesterday's prices.
• If price moves outside yesterday's value area and is accepted there (it spends time and builds volume), the market is looking for new value.

Those two possibilities are exactly what the next sections (value migration and the open relative to value) describe in more detail.

What does "yesterday" mean for gold?

Gold trades almost 23 hours a day, so "yesterday" needs a precise definition, and it must match the session setting you chose for your profile (Chapter 4 discussed session settings for VWAP; the same logic applies here):

• If you use Globex sessions, "yesterday" means the session that ran from 18:00 New York time the evening before last until 17:00 New York time yesterday.
• If you use RTH sessions, "yesterday" means roughly 08:20–13:30 New York time yesterday. The overnight hours then form a separate overnight profile, which supplies its own reference levels: the ONH (overnight high) and ONL (overnight low).

Do not mix the two. A Globex profile from yesterday combined with RTH logic today produces levels that mean nothing consistent.

A simple pre-session routine

1. Load yesterday's final profile (in the session type you always use).
2. Draw pPOC, pVAH, pVAL, pHigh and pLow, extending them across today's chart.
3. Note where current price is: inside yesterday's value area, outside value but inside yesterday's range, or outside yesterday's range.
4. Add weekly levels (Chapter 5) as a second layer, if they are nearby.

Worked example: yesterday's RTH profile had pHigh 4,052.0, pVAH 4,045.5, pPOC 4,041.0, pVAL 4,036.5 and pLow 4,029.0. At 08:00 New York time this morning, GC trades at 4,048.0. The table below sets out the map, with approximate XAUUSD equivalents assuming a gap of about $28.

Current price sits above yesterday's value area but below yesterday's high: outside value, inside range. That single observation already frames the morning: will the market accept these higher prices, or rotate back into yesterday's value?

This is a map, not a forecast. It tells you where to pay close attention, not what will happen there.

Common mistake: Drawing too many levels. With dozens of lines on the chart, every wiggle in price can be "explained" by some line, and nothing can ever be tested.

Common mistake: "Today's price will definitely test the pPOC." This is a popular claim, not proven.

Try it: For five consecutive days, before the session begins, draw only the five standard levels from the prior day's final profile, and write one sentence about where price is relative to them. After the close, note which levels price reached and what happened there. Do not add any other lines for those five days.

Naked (Virgin) POC, VAH/VAL and VWAP

Definitions

A naked POC (also called a virgin POC, often abbreviated nPOC) is the final POC of a past session that price has not traded at again since that session ended. "Naked" simply means "not yet revisited". The same idea applies to other levels:

• Naked VAH / naked VAL: a past session's value-area edge that has not been revisited.
• Naked VWAP: a past session's final VWAP that has not been revisited.

When price later trades at the level, it is no longer naked; traders say it has been "filled" or "covered".

The popular story

The usual narrative goes like this: the POC was where the most business was done; if the market left that price quickly and never came back, there is "unfinished business" there; therefore price tends to return and test it. You will often hear this stated as a law: "price always returns to the naked POC."

That statement is a popular claim, not proven. It is folklore, and it has some serious problems.

A reality check

1. A random walk also comes back. Imagine a price series that moves completely at random, with no memory and no magnets. Such a series still revisits nearby prices very often, sooner or later. So when a naked POC that was $3 away eventually gets touched, that alone proves nothing about magnetism. The right question is: are naked POCs touched more often than random levels at the same distance and age? Answering that requires a proper statistical test comparing naked levels with randomly placed control levels. We have not presented such a test here, and most material you will read on the topic has not either.

2. Selection bias. Naked POCs that do get filled make memorable chart examples. Naked POCs that stay untouched for months (which is common during strong gold trends) quietly disappear from the conversation. If you only ever see the successes, the idea will always look better than it is.

3. Touch is not reaction. Even when price does reach a naked level, nothing special may happen there. Price can slide straight through it. "It was touched" and "it mattered" are two different claims.

Honest use

Keep naked levels on your map as reference levels: places where, if price arrives, you will watch the order flow carefully. Always write down two facts next to each one:

• Its distance from current price.
• Its age (how many sessions ago it was formed).

For example: "nPOC from four sessions ago, 4,064.0 (XAUUSD ≈ 4,036), $12 above current price." A two-week-old naked POC $60 away is a very different object from a one-day-old naked POC $4 away, and writing the numbers forces you to notice that.

Keep the list short. A trader who keeps dozens of old naked levels on the chart will find that every move ends near one of them, which is a sign the method has stopped measuring anything.

Common mistake: "A naked POC is always filled." A popular claim, not proven. "A naked POC is the strongest support or resistance." Also a popular claim, not proven.

Key idea: Naked levels are worth keeping on a map, but their famous "magnet" effect has not been shown to beat simple chance. Write down distance and age, and judge the level by what order flow does when price arrives, not by the fact that it arrived.

Value Migration: Is Value Moving Higher, Lower or Nowhere?

The idea

Price moves around all day. Value, the area where most trading actually happened, moves more slowly and more meaningfully. Value migration means comparing the position of one session's value area with the previous session's. James Dalton treats this comparison as one of the most important pieces of evidence about direction in Market Profile analysis: price can spike up and down, but the relocation of value shows where business was really being done.

The categories

Call today's value area VA(today) and yesterday's VA(yesterday). The standard categories are:

The no-overlap categories (higher and lower value) carry the strongest descriptive message. The overlapping categories show a lean with less conviction. Inside days show contraction, which often accompanies waiting before an important event. Outside days show expansion: both sides were active and the market explored widely without settling.

Worked numbers

Yesterday: VAL 4,030.0, VAH 4,042.0.

• Today: VAL 4,044.5, VAH 4,056.0. Today's VAL is above yesterday's VAH. Higher value.
• Today: VAL 4,035.0, VAH 4,047.0. They overlap; both edges are higher. Overlapping-higher.
• Today: VAL 4,032.5, VAH 4,039.5. Entirely inside yesterday's area. Inside value.
• Today: VAL 4,024.0, VAH 4,050.0. Contains yesterday's area entirely. Outside value.

(XAUUSD equivalents simply shift every number down by roughly $28; the relationships between the areas are what matter, and they are identical on both instruments.)

Price is not value

Here is why this matters. Imagine a day on which gold makes a new high, a dollar above yesterday's high, yet the day's value area is overlapping-lower. Price poked higher briefly, but most of the trading happened at lower prices than yesterday. The headline says "gold hits new high"; the auction record says "the market did not accept higher prices." Compare that with a day that has no new high at all but shows higher value: the market moved its whole centre of business upward.

Limits

1. Sensitivity to settings. The category depends on your session definition (Globex or RTH) and your value-area calculation method. Change either and a day can change category.
2. "Unchanged" needs a threshold. How close counts as "essentially the same"? Any answer is a convention, and you must choose one and keep it.
3. Migration in progress is unstable. During the day you can only see the developing value area, which keeps moving until the close. A classification made at 10:00 a.m. is provisional.
4. Description, not destiny. Value migration describes what happened. It does not guarantee that the migration will continue; migrations can fail, and value can return to where it was.

Common mistake: "Higher value today means higher tomorrow." A popular claim, not proven.

Common mistake: Comparing value areas built from different sessions, for example a Globex area for one day against an RTH area for the next.

Key idea: Price shows where the market visited; value shows where it did business. Classifying value migration day by day gives you a disciplined description of direction that a single new high or low cannot.

Where Did Gold Open Relative to Value?

Three opening situations

In Mind Over Markets, Dalton classifies the open relative to the previous day's value and range into three situations:

1. Open within value (inside yesterday's value area). The market is, for now, agreeing with yesterday's value. The context is more balanced, and rotational trading is more likely than on the other two types.
2. Open outside value but within range (above the VAH or below the VAL, but still inside yesterday's high-low range). A mild imbalance. The key question is whether the market will accept prices outside value or return into it.
3. Open outside range (above yesterday's high or below yesterday's low). A stronger imbalance. There is more potential for directional movement, or for a quick failure and return if the outside prices are not accepted.

Using our earlier map (pHigh 4,052.0, pVAH 4,045.5, pVAL 4,036.5, pLow 4,029.0), an open at 4,040.0 is within value; an open at 4,049.0 is outside value but within range; an open at 4,056.0 (XAUUSD ≈ 4,028) is outside range.

Gold's problem: which open?

GC trades around 23 hours a day. Its only true "opening" in the technical sense comes after the daily halt, at 18:00 New York time. But an 18:00 open happens in the quietest part of the day and says little about how the main session will develop.

Profile analysts on gold therefore usually use the RTH open, around 08:20 New York time (the old COMEX trading-floor opening time), and compare it with yesterday's RTH value and range, while treating the overnight hours separately through ONH and ONL. Whichever convention you choose, state it explicitly and use it consistently. Chapter 7 returns to this question in detail.

The "80% rule", honestly

You will meet the 80% rule sooner or later. It is usually stated like this: if the market opens outside yesterday's value area, then moves back inside it and stays inside for two consecutive 30-minute periods, there is an 80% probability that it will travel across the whole value area to the opposite edge.

What you should know about it:

• Its origin. It comes from Market Profile commentary of the 1980s and 1990s associated with the Dalton school, and it was originally defined on the TPO value area (built from time, explained later in this chapter), not on the volume value area.
• Its status. The 80% figure is a historical claim. It is a popular claim, not proven. Some informal public tests on stock-index futures have reported success rates well below 80%, and we are not aware of any rigorous test on gold futures.
• How we treat it in this book. As a historical idea to test, never as a statistic to rely on.

To examine it, define every element before looking at any chart (session, value-area method, what counts as "back inside" and "reaching the other side") and compare the result with a sensible baseline. Chapter 12 explains how.

Honest use

The position of the open relative to value frames the day's scenarios. It tells you which questions are most relevant this morning. It does not tell you which way price will go.

Common mistake: "The 80% rule means an 80% win rate." A popular claim, not proven, and in any case a probability of price reaching a level is not the same thing as the outcome of a trade, which also depends on costs, timing and where you would have been wrong.

Common mistake: Applying the 80% rule to a volume value area with arbitrary time periods and expecting the historical claim to carry over. Or using the 18:00 Globex open with RTH logic.

Volume Profile vs TPO (Market Profile)

Two ways of counting

The Market Profile, developed by J. Peter Steidlmayer at the Chicago Board of Trade in the 1980s, did not originally count volume at all. It counted time.

Its unit is the TPO, short for Time Price Opportunity. The session is divided into fixed periods, usually 30 minutes long, and each period is given a letter (A for the first, B for the second, and so on). Whenever price trades at a given level during a period, that period's letter is printed at that price. Stack all the letters against the price axis and you get a histogram, just like a volume profile, except that the length of each row means how many periods the market spent at that price, not how many contracts traded there.

A volume profile builds the same kind of histogram using volume instead.

Key differences

• Different POCs. The TPO POC is the price where the market spent the most time. The volume POC (sometimes written VPOC) is the price with the most volume. They are often close, but not always. If a news release produces an enormous burst of trading at one price within a few minutes, the volume POC may land there, while the TPO POC, which only counts how many 30-minute periods visited each price, may sit somewhere else entirely.
• Same 70% logic, different unit. Both methods build a value area around their POC containing about 70% of the total, but one counts TPOs and the other counts contracts.
• Time-based concepts. Some important Market Profile ideas are defined by time, and they are not directly visible on a volume profile:
• Initial Balance (IB): the range of the first hour, i.e. the first two 30-minute periods.
• Single prints: prices where only one period's letter appears, meaning price moved through them quickly.
• Tails: single prints at the very top or bottom of the day's profile.

The nearest volume-profile equivalent of a run of single prints is an LVN. Chapter 7 develops all of these concepts.

Which is better?

Neither. A TPO profile shows acceptance in time; a volume profile shows acceptance in volume. Many analysts look at both side by side. Because gold futures give us genuine, exchange-reported volume, this book leads with the volume profile. But you need the TPO vocabulary to understand Dalton's literature, much of which is written in time-based terms.

A small example. Suppose gold spends five of the session's eleven 30-minute periods trading around 4,030.0, but at 08:30 an inflation release causes a burst of 6,000 contracts within three minutes at 4,041.0, after which price drifts back. The TPO POC may be 4,030.0 (most time), while the volume POC may be 4,041.0 (most contracts). Both are "correct"; they answer different questions.

Common mistake: "Market Profile and Volume Profile are the same thing." One measures time, the other volume.

Common mistake: Applying TPO-based rules (like the 80% rule) to a volume profile without checking whether they still hold. And the claim that one of the two methods has proven superiority is a popular claim, not proven.

Common Volume Profile Mistakes and a Pre-Session Checklist for Gold

This final section collects the errors that cause the most confusion, then turns everything from Chapters 5 and 6 into a short routine.

Configuration mistakes

1. An unspecified session. Is your profile built on Globex or RTH? Every level (POC, VAH, VAL, high, low) changes depending on the answer. If you do not know which session your platform uses, find out before you use any level.
2. Inconsistent row size. The row size (how many ticks are grouped into each horizontal bar) affects where the POC and value area land. If you change it from day to day, you cannot compare days. Pick one and keep it.
3. Bar data instead of tick data, without knowing it. Some platforms build profiles from candles rather than individual trades, spreading each candle's volume across its range. That is an approximation. Know which one your platform uses.
4. Ignoring the contract roll. Gold futures expire, and traders move ("roll") from one contract month to the next. Composite and monthly profiles that span a roll can mix two different contracts with different prices. Check how your platform handles this.
5. Using a visible-range profile as a reference. A profile that covers "whatever is on the screen" changes every time you scroll or zoom. It cannot serve as a fixed reference level.

Reading mistakes

1. Look-ahead. Using a final profile to judge decisions that would have been made during the day, when only the developing profile existed (Chapter 5). This is the most common error in educational chart examples, and it makes every level look far better than it really was.
2. Treating every level as a wall. Profile levels are places to pay attention, not barriers that price must respect.
3. Treating shape interpretation as law. "P means bullish" is an interpretation, not a rule.
4. Repeating folklore without a label. "Naked POCs always fill", "the 80% rule", "LVNs are always crossed quickly": each of these is a popular claim, not proven.
5. Too many levels. With thirty lines on the chart, any reaction has an explanation and nothing is ever tested. Fewer lines give you a clearer, more honest read.

Two more claims deserve special mention. "Volume profile shows where institutions bought" is a popular claim, not proven: the profile shows where volume traded, not who traded it. "Volume profile is a trading edge" is also a popular claim, not proven. It gives you more information about the auction, which is valuable but is not profit; in our own testing, order-flow signals on their own did not reliably predict direction once trading costs were included.

A pre-session checklist for gold

Use this before the session. Fill in the blanks the same way every day.

• Session: ____ (Globex 18:00–17:00 New York time, or RTH about 08:20–13:30 New York time). Kept constant.
• Row size: ____ ; data: tick or bar-based.
• Yesterday's final profile: pPOC ____ / pVAH ____ / pVAL ____ / pHigh ____ / pLow ____ (with XAUUSD equivalents if you trade the CFD).
• Overnight (if you use RTH): ONH ____ / ONL ____.
• Value migration, yesterday versus the day before: ____ (higher, lower, overlapping-higher, overlapping-lower, inside, outside, unchanged).
• Weekly and naked levels nearby: at most two or three, each with its distance and age.
• VWAP: daily and weekly, with bands.
• News calendar for gold (CPI, the jobs report, Federal Reserve decisions): exact times in New York time.
• Where current price sits: inside value / outside value but inside range / outside range.

This checklist builds a map. What happens at each level, whether price is accepted or rejected, whether aggressive buyers or sellers are absorbed, has to be read with order flow, which is the subject of the chapters ahead.

Try it: Run the checklist on a real gold morning, starting from an empty chart. Stop when you have six to eight labelled levels at most. Take a screenshot, then take another after the close, and compare. Do this for two weeks before you add anything else to your routine.

Gold note: Gold's most important scheduled U.S. data usually arrives at 08:30 New York time, ten minutes after the RTH open. A profile built from the RTH session therefore often contains a news spike in its first hour. Keep that in mind when you read shapes and value areas: a single burst of news-driven volume can move the POC and change the shape of the whole day.

Chapter summary

• A D-shaped (normal) profile has one central bulge, a POC near the middle and a roughly symmetric value area. It describes a balanced, two-sided auction, usually with a flat VWAP near the POC. It says nothing about tomorrow.
• P and b profiles show a fast, thin leg and a fat area of acceptance. They are often read as short covering (P) and long liquidation (b), but the profile cannot see whether positions were opened or closed, and either shape can simply be continuation.
• A thin trend profile shows a one-sided auction with no accepted price. A double distribution shows two separate balances joined by an LVN; split the profile so each has its own POC and value area.
• Yesterday's final profile is today's fixed map: pPOC, pVAH, pVAL, pHigh, pLow, plus optional pVWAP and key nodes. "Yesterday" must match your session setting.
• Naked levels are past levels that have not been revisited. The claim that price always returns to them is folklore; random price paths also revisit nearby levels, and selection bias inflates the examples you see.
• Value migration compares today's value area with yesterday's: higher, lower, overlapping-higher, overlapping-lower, inside, outside or unchanged. Price and value can tell different stories.
• The open relative to value (inside value, outside value but inside range, outside range) frames the day's scenarios. The "80% rule" is a historical TPO-based claim, not a proven statistic.
• TPO profiles count time; volume profiles count contracts. Their POCs can differ. TPO gives us concepts such as Initial Balance, single prints and tails.
• Consistent settings, no look-ahead, labelled folklore and a short list of levels are the foundation of honest profile reading.

Checklist

• I always say which session (Globex or RTH) and which window a profile covers before naming its shape.
• I name shapes only on finished profiles, never on developing ones.
• I treat P and b stories as hypotheses to check with order flow.
• On double-distribution days, I split the profile.
• Before each session I draw pPOC, pVAH, pVAL, pHigh and pLow from yesterday's final profile, with XAUUSD equivalents if I trade the CFD.
• I note every naked level's distance and age, and keep no more than two or three.
• I classify value migration with the same session and value-area method every day.
• I write down where gold opened relative to yesterday's value and range, using the 08:20 New York RTH open (or my own stated convention).
• I never repeat the 80% rule, naked-POC magnetism or LVN-fill claims without labelling them as unproven.
• I keep my chart to six to eight labelled levels.

Quiz

1. A finished RTH profile has one central bulge, a POC near the middle of the range, and a flat VWAP sitting close to the POC. What kind of auction does this describe?
2. Why can a P-shaped profile not prove that short covering took place?
3. Today's VAL is above yesterday's VAH. Which value-migration category is this, and what does it describe?
4. A naked POC $2 away from current price is touched later in the session. What does that prove about naked POCs, and why?
5. How do the TPO POC and the volume POC differ, and why might they land at different prices on a news day?

Quiz answers

1. A balanced, two-sided auction. Price rotated around a temporary fair price, with neither side in control during that session. It does not predict tomorrow's behaviour.
2. Because a volume profile records only how much traded at each price. It cannot tell whether a trade opened a new position or closed an existing one. At best, daily changes in open interest give a rough, end-of-day hint. A P can also simply reflect continuation in an uptrend.
3. Higher value. There is no overlap, and the whole area of business has moved above yesterday's, which is a clear description of acceptance at higher prices. It is a description, not a guarantee of further migration.
4. Very little on its own. Even a random price path frequently revisits nearby levels, so a touch at a short distance is expected by chance. To show a real effect, you would need to compare touch rates with random control levels at the same distance and age. And a touch is not the same as a meaningful reaction.
5. The TPO POC is the price where the market spent the most time (the most 30-minute periods); the volume POC is the price where the most contracts traded. A news release can concentrate heavy volume at one price in a few minutes. That moves the volume POC there, while the time-based TPO POC stays where the market spent most of the session.

Education only. Not financial advice. Trading involves substantial risk of loss.