Auction Market Theory

≈35 min read · Real gold futures examples with XAUUSD equivalents · Free

In this chapter: Auction Market Theory (AMT) is the framework that ties your volume measurements together into one picture of what the market is doing. You will learn why markets exist, how to tell balance from imbalance, how to judge acceptance and rejection with time and volume, how to read a TPO chart, how to define gold's Initial Balance and read range extension, the four opening types, the open's position relative to yesterday, the main day types, excess, poor highs and lows, single prints, gold's session map and reference levels, ending with a pre-New York checklist. AMT describes context; it does not predict prices.

All times in this chapter are New York time (ET) unless marked "UK". Prices are illustrative gold futures (GC) prices; where a price matters, an approximate XAUUSD (CFD) equivalent is given using a gap of about $28. The real gap moves around (recently roughly $27–30), so always measure it live on your own platform.

Why Markets Exist: Facilitating Trade

The core idea

Auction Market Theory grew out of the work of J. Peter Steidlmayer at the Chicago Board of Trade in the 1980s, and was later developed and popularised by James Dalton. Its central idea is simple and slightly surprising: the purpose of a market is to facilitate trade. The market's job is not to discover some fixed "correct" price. Its job is to find the prices at which the most business can be done.

The market works like a two-sided auction. Price moves up to find out whether sellers will become active or buyers will stop. Price moves down to test the same thing from the other side. Every move is an experiment.

Think of an auctioneer who does not know what a painting is worth. If many hands go up, they raise the price; if none do, they lower it. The auction settles where both sides keep trading. A futures market does the same, continuously.

Price advertises; time and volume answer

A useful way to remember this: price advertises, and time and volume are the market's answer to the advertisement.

• If price moves up and trading increases, the new price is acceptable to both sides. The market can stay there and do business.
• If price moves up and trading nearly stops, nobody is willing to transact at that price. It has gone too far, and the market moves back to find business elsewhere.

That is why AMT always cares about two things together: price (where) and time plus volume (how much business was done there). A high-volume area is a zone of temporary agreement. A low-volume area means either disagreement (one side refused to trade) or speed (price passed through too quickly for business to build up).

Notice what high volume does not mean. It does not mean "a strong move in one direction". Heavy volume at a price usually means agreement in that area, not direction.

Two kinds of participant

AMT divides participants by their time horizon.

• The Other Timeframe (OTF) participant has a longer horizon than a single day: funds, commercial hedgers (such as miners or jewellers protecting against price changes), and traders who hold positions for days or weeks. When OTF participants become active, they tend to move price away from the current value area, because they are willing to pay up (or sell down) to get their business done.
• The Day Timeframe participant is a short-term trader who mostly opens and closes positions within the session. Day-timeframe activity tends to rotate price inside the current range, buying low and selling high within it.

No trade carries a label saying "OTF" or "day trader"; these are categories of behaviour, inferred from price, time and volume. Still, when price leaves the established range and is accepted outside it, a longer-horizon participant is usually at work.

A worked gold example

Over five sessions on a 30-minute chart, imagine two days that both push to new highs:

• Day A: GC spikes from 4,050.0 to 4,062.0 (XAUUSD ≈ 4,022 to 4,034) in the late morning. The volume profile shows almost nothing traded above 4,058.0. Within forty minutes price is back below 4,052.0. The market advertised higher prices, and nobody answered.
• Day B: GC rises from 4,050.0 to 4,062.0 over the morning, and then spends the afternoon trading between 4,058.0 and 4,066.0, building a thick volume node around 4,062.0. The market advertised higher prices, and both sides answered.

Both printed a new high; only Day B accepted it.

What AMT is not

AMT is a descriptive framework, not a forecasting formula. It tells you what the market is doing now: searching for new value, or rotating within current value. It does not tell you where the next price will be. Many popular statements associated with AMT ("price always comes back to fill X") have never been put through a proper statistical test.

Common mistake: "The market wants to reach the right price." The market wants trade; a "right price" is not a fixed thing.

Common mistake: "AMT tells me when to enter." It tells you what context you are in. That is valuable, but it is not an entry rule.

Key idea: Price moves to find trade. Where trade happens (time and volume), the market has accepted the price; where trade does not happen, it has rejected it. AMT is the common language behind both Market Profile and Volume Profile.

Balance vs Imbalance: Brackets and Trends

The two states

AMT says a market is always in one of two states: balance or imbalance.

Balance means buyers and sellers broadly agree on a range of prices. Price rotates (swings back and forth) between an upper and lower boundary. Day after day, the value areas overlap, and a multi-day profile takes on a bell shape. The price zone where this happens is called a bracket.

Inside a bracket, traders typically act responsively: they sell near the top and buy near the bottom, expecting price to return towards the middle. Responsive activity means trading against the latest move, at the edges of value.

Imbalance means one side, usually the other timeframe, is more active and pushes price out of the bracket to search for a new place to trade. The signs are:

• Value areas migrate day after day, consistently higher or consistently lower (the value migration of Chapter 6).
• Profiles become elongated.
• Low-volume areas (LVNs, or single prints on a TPO chart) are left behind price.

In imbalance, initiative activity dominates: buying above value or selling below value. Initiative activity trades with the move, outside value.

The cycle

Markets alternate: balance → breakout → imbalance (trend) → new balance at a different level, and then the cycle repeats. Neither state is "better". Problems arise when a trader uses balance tools (buying the bottom of the range, selling the top) during imbalance, or uses imbalance tools (chasing breakouts) during balance.

That makes one question the most important of the day: Are we in balance or imbalance right now?

Worked example

Over twelve sessions, suppose gold's daily value areas look like this:

Limits

A bracket's boundaries are a matter of judgement. Two analysts may draw two different brackets on the same chart. And a move out of a bracket is only a test until acceptance outside it is established (the next section explains how). Many breakouts fail and return inside: they are failed breakouts, and they are common.

Common mistake: "Every break of the high is the start of a trend." Many breakouts return inside the bracket.

Common mistake: "Balance means the market is safe." Breaks out of balance can be very fast.

Common mistake: Drawing a bracket from one candle's high and low instead of several days of overlapping value.

Key idea: Balance is rotation inside a bracket with overlapping value; imbalance is a directional search with migrating value. Responsive traders dominate balance, initiative traders dominate imbalance. Leaving a bracket without acceptance is just a test.

Acceptance vs Rejection: Time and Volume at Price

Definitions

Acceptance means the market spends time at a new price and builds volume there. Both sides are trading, and that price is turning into value.

Rejection means price reaches a level, no meaningful trading develops, and price returns quickly. Rejection usually leaves behind a tail (on a TPO chart) or a thin, low-volume area (on a volume profile).

A compact summary to remember:

Key idea: Time + volume = acceptance = new value. No time and no volume = rejection = excess.

Why it matters

When price crosses a reference level (say yesterday's value-area high, or the Initial Balance high), the only thing that separates a genuine move from a failed test is what happens after the cross. If price stays beyond the level, if subsequent bars close beyond it, and if volume builds there, the market is probably establishing new value. If price quickly returns, the level was defended.

How much time? How much volume?

There is no universal answer. Dalton, working with Market Profile, usually talks in terms of 30-minute TPO periods; for example, one full period beyond a level is often treated as initial evidence of acceptance. On a minute chart you need to write your own rule in advance, such as:

• "Three consecutive 1-minute bars close beyond the level", or
• "One 30-minute bar closes beyond the level", or
• "Price stays beyond the level for 20 minutes and a visible volume node forms there."

Then keep that rule fixed. Changing the definition after seeing the chart is self-deception.

Worked example

Yesterday's high (PDH) is 4,048.0 (XAUUSD ≈ 4,020). Consider two mornings, using the rule "acceptance = at least 30 minutes beyond the level with a visible volume node":

1. Accepted: GC crosses 4,048.0 at 09:40, trades between 4,049.0 and 4,056.0 until 10:30, and builds a node around 4,052.0. Acceptance above PDH.
2. Rejected: GC trades to 4,049.5 at 09:40, prints a few contracts there, and is back at 4,045.0 within four minutes. Nothing was built above the level. Rejection at PDH.

Limits

Acceptance is backward-looking. By the time it is confirmed, part of the move is already over. And acceptance can itself fail later: value accepted at 10:30 can be abandoned at 12:00.

Common mistake: "A wick above the level is a definite rejection." Without volume and context, a wick is just a wick.

Common mistake: Changing the time or volume threshold after seeing the result. That is hindsight, not analysis.

Common mistake: Ignoring time frame. What looks like acceptance on a 1-minute chart may still be rejection on a 30-minute chart.

TPO and Market Profile Basics vs Volume Profile

Chapter 6 introduced the difference between time-based and volume-based profiles. Here we look at how a TPO chart is actually built and read, because much of the AMT vocabulary in the rest of this chapter comes from it.

Building a TPO chart

The Market Profile was introduced by Steidlmayer at the CBOT and made publicly available in the mid-1980s. The session is cut into fixed periods, usually 30 minutes, and each period gets a letter: A, B, C and so on. During each period, the period's letter is printed next to every price that traded. Each letter at a price is one TPO (Time Price Opportunity).

Picture four columns of letters. Period A traded from 4,030 to 4,036, so A is printed at every price in that range. Period B traded from 4,033 to 4,040, so B is printed at those prices. C traded 4,034–4,038, D traded 4,035–4,039. Now push all the letters to the left so they stack in rows. The price rows from 4,035 to 4,036 have four letters (A, B, C, D); the rows at 4,030–4,032 have only A. The result is a "time histogram": the longer a price's row, the more time the market spent there.

From this chart come familiar concepts:

• TPO POC: the price with the most TPOs (the longest row).
• Value area: usually the range around the POC containing about 70% of all TPOs.
• Initial Balance: the range of periods A and B together, that is, the first hour.
• Single prints: prices that have only one letter.

Time versus volume

A volume profile asks "how many contracts traded here?" instead of "how many periods visited here?". The two usually look similar but can differ. A news release can produce a huge burst of volume at one price in three minutes, moving the volume POC there, while the market spent hardly any time at that price, so the TPO POC stays elsewhere.

Why this book leads with volume on gold

In a 23-hour market, TPO counts a quiet Asian half-hour exactly the same as a hectic half-hour after a New York data release, over-weighting the low-participation hours. A volume profile shows the real weight of participation, though it is sensitive to sudden volume spikes. So volume profile is our main tool, and TPO supplies the language: Initial Balance, periods, single prints and poor highs.

When the TPO POC and the volume POC are far apart, ask why. The answer usually tells you something about the day, typically that a short burst of news-driven trading carried a lot of volume.

Common mistake: Comparing TPO charts from two platforms that use different session templates (full Globex versus RTH). They are measuring different days.

Common mistake: "The 70% value area is a law of nature." It is a convention, inspired by the bell-shaped normal distribution in statistics, not a natural constant.

Initial Balance on Gold: Which Clock?

The classic definition

In classic Market Profile, the Initial Balance (IB) is the price range of the first two 30-minute periods of the session, that is, the first hour. Its top is the IB High and its bottom is the IB Low.

The idea behind it: during the first hour, day-timeframe traders establish an opening range. Whether the market later breaks out of that range (range extension, the next section) tells you whether the other timeframe has become active.

Gold's problem: when does the session start?

For a stock-index future, the answer is easy: the stock market opens at 09:30 New York time. Gold has no such single moment. There are two common conventions:

1. The Globex open at 18:00 New York time. This is the true start of the CME trading day.

• Advantage: no part of the day is excluded.
• Disadvantage: the first Globex hour is usually thin, with little participation. Its IB is often very narrow and gets broken almost every day, so it carries little information.

2. 08:20 New York time. This was the opening time of the old COMEX trading floor (the "pit"), and a surge in volume is still visible around that time. It marks the start of gold's regular trading hours (RTH, roughly 08:20–13:30).

• Advantage: the IB is built when U.S. participants are active and London is still open, so it reflects real participation.
• Disadvantage: the overnight hours (Asia and London) are left out of the IB, so you must track the overnight high and low (ONH and ONL) separately.

Other conventions exist too: some traders use 09:30 to match the stock market, and some use a 90-minute IB.

In this book, gold's IB is 08:20–09:20 New York time. Which convention you choose matters less than staying consistent, and when you read someone else's statistics or lessons about the IB, ask which definition they used.

Worked example

On one day, the first Globex hour (18:00–19:00) trades between 4,031.0 and 4,033.5, a range of just $2.50. By morning, gold has moved through that range many times; its high and low mean almost nothing. The 08:20–09:20 hour, by contrast, trades between 4,036.0 and 4,046.0 (XAUUSD ≈ 4,008–4,018), a $10 range built on heavy participation. When price eventually breaks above 4,046.0 at 10:15 and stays there, that break carries information. The Globex IB's "break" carried none.

A daylight-saving detail

The United States and the United Kingdom do not change their clocks on the same dates. For a few weeks each year, in March and in October/November, the gap between London and New York shifts by one hour. An IB defined in New York time stays fixed in New York time, but its relationship to the London session changes during those weeks. Keep this in mind when comparing days around those dates.

Common mistake: Copying "IB = 09:30–10:30" from stock-index lessons and applying it to gold.

Common mistake: "The IB is broken on X% of days." Popular claims of this kind are not proven for your market and your definition: the figure depends entirely on how the IB is defined and on the market. Do not accept any such figure without testing it on your own data.

Gold note: Many platforms define session templates in Chicago time (CT), because CME is based in Chicago. 08:20 New York time is 07:20 Chicago time. If your IB looks wrong, check the time zone of your session template first.

Range Extension: When the Other Timeframe Shows Up

Definition

Range extension (RE) means price trades above the IB High or below the IB Low after the IB period has ended.

The Market Profile reasoning goes like this. Day-timeframe traders tend to rotate within the IB. For price to leave the IB with conviction, larger or longer-horizon participants (the OTF) usually have to get involved. So range extension can be seen as a possible footprint of the OTF. That is why the CBOT's Market Profile study material and Dalton's books note that the absence of range extension usually means short-term traders are in control.

Measuring in IB multiples

To compare days fairly, use the IB's own range as the unit. If the IB is $8 wide and price extends to $12 above the IB Low, the day's range is 1.5 × IB. IB multiples make a quiet day and a busy day comparable.

Worked numbers: IB Low 4,036.0, IB High 4,044.0 (an $8 IB). Price later reaches 4,052.0 (XAUUSD ≈ 4,024). The range from IB Low is now $16, or 2 × IB.

These multiples are useful for comparing days. They are not price targets.

Three important situations

1. One-sided RE with acceptance. Price leaves the IB on one side, stays out, and builds volume. The structure starts to resemble a normal variation day or a trend day (see day types below).
2. RE that quickly returns inside the IB. This is a failed extension. Sometimes the market then travels to the other side of the IB.
3. Two-sided RE. Price extends above and below the IB. This produces a neutral structure: neither side has complete control.

The gold caveat: news inside the IB

Major U.S. economic releases arrive at 08:30, which falls inside an IB starting at 08:20. On those days, a single news spike can make the IB very wide, and any later range extension means less. Always ask: was today's IB a normal IB or a news IB?

Remember also a purely mechanical effect: a narrow IB is easier to extend than a wide one, because price has less distance to travel to leave it.

Common mistake: "The IB almost always extends to 2×" or "1.5 × IB is the target." These are popular claims, not proven.

Common mistake: Automatically buying a break of the IB High without checking for acceptance and context.

Key idea: Range extension is a possible sign that longer-horizon participants are active. Measure it in IB multiples, check for acceptance, and remember that news inside the IB changes what it means.

Opening Types: How the Session Starts

Dalton's four types

Dalton classifies the start of the session into four types, ordered from the most to the least conviction (how strongly and clearly the market chooses a direction):

1. Open-Drive (OD). Price moves strongly in one direction right from the open and does not return to the opening price. Interpretation: the other timeframe made its decision before the open.
2. Open-Test-Drive (OTD). Price first tests a nearby reference level in one direction (for example yesterday's high or the overnight high), finds no business there, and then drives strongly in the opposite direction.
3. Open-Rejection-Reverse (ORR). Price moves in one direction, meets opposition, and reverses back through the opening price to the other side. Conviction is lower, and the day's direction is not yet clear.
4. Open-Auction (OA). Price rotates on both sides of the open with no clear direction. This usually signals balance and a rotational day, especially when the open is inside yesterday's value.

The conviction ladder is: OD > OTD > ORR > OA.

Picturing them

Draw the opening price as a horizontal line and sketch the first 30–60 minutes:

• OD: a nearly straight line away from the open.
• OTD: a small hook in one direction, then a straight line the other way, never returning to the open.
• ORR: a move one way, then back across the open and beyond it.
• OA: a zigzag around the open.

Worked example

Gold opens the RTH session at 4,040.0. The overnight high (ONH) is 4,044.0 (XAUUSD ≈ 4,016).

• Price rises to 4,043.8 by 08:28, stalls with almost no trading near the ONH, then falls steadily to 4,028.0 by 09:15 without returning to 4,040.0. That is an Open-Test-Drive: a test of the overnight high, no business found, then a drive lower.
• If instead price had fallen from 4,040.0 straight to 4,028.0 without any initial test upward, it would be an Open-Drive.
• If price had risen to 4,044.5, then fallen back through 4,040.0 to 4,034.0 and kept swinging, that would be closer to an Open-Rejection-Reverse.

How to use opening types

An opening type gives you an initial hypothesis about the day. OD and OTD make a directional, elongated day more likely; OA inside value makes rotation between known levels more likely. That hypothesis must then be confirmed or rejected by the rest of the day: range extension, acceptance, and so on.

Gold's caveat

Which moment is "the open"? This book uses 08:20 New York time. But gold has been trading since 18:00 the evening before, and the 08:20 open is really a continuation of a move that may have started in London. Opening-type labels are also partly subjective: one person's OD is another's OTD. A practical fix is to write a quantitative criterion, for example: "OD = in the first 30 minutes, price never traded more than N ticks on the opposite side of the opening price", choosing N in advance and keeping it.

Common mistake: "OD means a guaranteed trend day." It only raises the likelihood; open-drives can fail.

Common mistake: Labelling the opening type after seeing the whole day (hindsight labelling).

Common mistake: Ignoring where the open sits relative to yesterday's value and range, the subject of the next section.

The Open vs Prior Value and Range

Chapter 6 introduced Dalton's three opening situations from the volume-profile side. Here we place them inside the AMT framework and turn each into scenarios.

Three situations, three sets of scenarios

Before the open, you have two things from yesterday: its value area (VAH to VAL, about 70% of volume) and its range (high to low). Where the open falls relative to these is your first clue about balance or imbalance.

1. Open inside yesterday's value. The market still accepts yesterday's prices. The base case is rotation and responsive behaviour; a day similar to yesterday is more likely. To leave value, the market has to find a reason.

2. Open outside value but inside range. A mild imbalance. Two main scenarios:

• Price returns into yesterday's value and is accepted there, which means the prices outside value have been rejected; or
• New value is built outside yesterday's value area.

3. Open outside yesterday's range. The strongest possible sign of imbalance. If price stays outside the range (acceptance), directional movement is more likely. If it returns inside the range, the move outside has failed, and traders who positioned in the direction of the opening move are left in a weaker position.

Worked example with scenarios

Yesterday's RTH: high 4,052.0, VAH 4,045.5, POC 4,041.0, VAL 4,036.5, low 4,029.0. Today at 08:20, GC is at 4,049.0 (XAUUSD ≈ 4,021): outside value, inside range.

• Scenario A: price falls back below 4,045.5 and is accepted inside yesterday's value. The higher prices were rejected; yesterday's POC at 4,041.0 becomes the next reference.
• Scenario B: price holds above 4,045.5, builds volume there, and later accepts above yesterday's high at 4,052.0. The market is building new value higher.

You do not choose between them in advance; you write them down so that you recognise each one as it unfolds.

Gold is not a stock gap

In equities, a "gap" is the jump between yesterday's close and today's open after a night with no trading. Gold's 08:20 "open" comes after about fourteen hours of overnight trading, so a gap in the stock-market sense does not exist. What we compare is the price at 08:20 relative to the value and range of yesterday's reference session. You must also decide whether yesterday's value comes from the whole Globex session or from RTH only; that choice changes the answer, so make it once and keep it.

The 80% rule, briefly

As Chapter 6 explained, the "80% rule" claims that when the market opens outside value, returns inside it and stays there for two consecutive 30-minute periods, it has an 80% chance of crossing to the other side of value. This is a popular claim, not proven. No transparent, repeatable public test exists for gold, and the result is sensitive to how value and the session are defined. Only your own carefully designed backtest could examine it. Until then, treat it as folklore.

Common mistake: "An open outside the range means a trend day." Many opens outside the range return inside.

Common mistake: Computing value from RTH on one day and from Globex on the next.

Key idea: The further the open is from yesterday's value, the greater the chance of imbalance, but never a certainty. A return into value with acceptance there means the outside prices were rejected.

Day Types in Brief: Normal, Variation, Trend, Double Distribution, Neutral

The classification

Market Profile classifies finished days mainly by two things: the size of the IB and how much range extension occurred, and on which side.

• Normal day. A wide IB that covers nearly the whole day's range; little or no extension. The market found both sides in the first hour and rotated for the rest of the day.
• Normal variation day. A moderate IB, then significant extension to one side (in common descriptions, up to about twice the IB). The other timeframe entered on one side, but without trend-day intensity.
• Trend day. A relatively narrow IB, then persistent extension to one side. Price barely retraces, leaves single prints behind, and closes near its extreme.
• Double distribution trend day. A quiet start with a narrow IB and a first small distribution; then a fast move and a second distribution at a different level, with a zone of single prints between the two.
• Neutral day. Extension on both sides of the IB. If the day closes near one of its extremes, it is a neutral extreme day (one side won by the close). If it closes in the middle, it is a neutral center day (uncertainty).

Dalton also describes other types, such as the non-trend day: a narrow IB with no extension, often seen ahead of major news.

Worked example

IB 08:20–09:20: 4,036.0–4,044.0 (an $8 IB). Consider three ways the day might finish:

• Day range 4,033.0–4,046.0 by the close: barely extended. Closer to a normal day.
• Price extends to 4,056.0 (XAUUSD ≈ 4,028), about 2.5 × IB from the IB Low, never returns to the IB, and closes at 4,055.0. Closer to a trend day.
• Price extends to 4,049.0 in the morning, then to 4,031.0 in the afternoon, and closes at 4,040.0. Extension on both sides with a central close: neutral center.

What day types are really for

Day types are mostly a post-session review tool. During the day you can only say "so far, the structure looks like X" and update that view each period. Labelling is subjective and the boundaries (normal variation versus trend, for example) are fuzzy: a different definition could label the same day differently.

About gold

You may come across claims that a very high share of gold's RTH sessions are trend days. That is a popular claim, not proven: without a precise definition and a test on real data, it should not be repeated. The general idea behind it, that gold may produce directional days more often than stock indices do, is worth investigating, but it is an idea to test, not a fact.

Common mistake: Trying to "predict" the day type in the first ten minutes.

Common mistake: Quoting precise frequencies of day types without stating the definition and the data. These are popular claims, not proven. And using thresholds designed for stock-index futures on gold without testing them.

Excess, Poor Highs and Lows, and Unfinished Auctions

How auctions end

When an auction in one direction is complete, it usually ends with excess. Price goes a little too far, meets strong opposition, and comes back quickly. On a TPO chart, excess appears as a tail: several prices at the extreme of the day, each with only one letter (in Dalton's common description, at least two single TPOs). On a volume profile, excess shows up as volume dropping off sharply towards zero at the edge of the profile.

The interpretation: "At that price, the opposite side found the market so attractive that they stepped in hard. The auction in that direction is finished."

Poor highs and poor lows

The opposite case is a poor high or poor low: an extreme with no tail. The profile is "flat" at the edge. For example, several periods traded at the exact same high price without price going higher and being rejected. This suggests the auction in that direction did not reach a conclusion: buyers became tired, rather than being pushed back by strong sellers. This is also called an unfinished auction.

Worked illustration: three 30-minute periods (D, E and F) all touch 4,061.0 exactly, and none trades at 4,061.1 or above. The high has three TPOs at the same price and no tail. That is a textbook poor high.

Why poor extremes are marked

Poor highs and lows are kept as reference levels for the future, based on the idea that the market will return at some point to "finish the auction". That is a popular claim, not proven: "a poor high always gets retested" is folklore. Even educational sources that teach the concept stress that a revisit is not guaranteed; in long trends, a poor extreme may never be revisited. Their real value is knowing that the level is "untidy", and that resting orders (such as stop orders placed just beyond an obvious high) may have accumulated nearby.

A note on vocabulary

The term "unfinished auction" is also used on footprint charts (Chapter 9) for a related but different idea at the level of a single bar. Keep the two apart: here we are talking about the extremes of a whole session's profile.

Common mistake: "Every poor high gets filled or tested." A popular claim, not proven.

Common mistake: Treating every long candle wick as excess. Excess is defined on the profile (tails of single prints, volume dropping sharply at the edge), not by a candle's shape.

Common mistake: Forgetting the session definition. A poor high on an RTH profile may have a perfectly good tail on the full Globex profile.

Key idea: Excess (a tail) marks a decisive end to an auction and is usually a strong reference level. A poor extreme (a flat edge) marks an unfinished auction: a weaker, untidy level worth marking but never worth assuming will be revisited.

Single Prints and Tails

Definition

On a TPO chart, a single print is a price where only one letter appears: only one 30-minute period traded there. When several single prints are stacked one above another, the market passed through that zone quickly in one period and never came back. One side was so dominant that the other side had no opportunity to trade.

Two important places

1. At the day's extreme: a run of single prints is a tail, the same thing as the excess described above. It marks the rejection of price and the end of the auction in that direction.
2. In the middle of the profile: a single-print section, typically seen on trend days and double-distribution days. It acts as a "bridge" between two value areas. Its volume-profile equivalent is a low-volume node (LVN).

Why they matter

These are zones where the market did not build any agreement. If price returns to them later, two scenarios are plausible:

• The zone is crossed quickly again, because still nobody wants to trade there; or
• The market finds support or resistance there, because the side that dominated originally steps in to defend it.

That is why single prints are marked as reference levels, not as signals. The claim that "single prints always get filled" is a popular claim, not proven. Whether they are filled depends on context, and no reliable fixed percentage has been established.

Technical limits

A single print depends completely on the period length. With 30-minute periods, a zone might show single prints; with 15-minute periods, the same zone might have two letters. And in gold's 23-hour market, a full Globex profile may already have filled in single prints that appear on the RTH profile. As with every profile tool, fix your settings and keep them constant.

Worked illustration: on a double-distribution day, period C drops gold from 4,046.0 to 4,030.0 (XAUUSD ≈ 4,018 to 4,002) in thirty minutes, and no later period trades between 4,032.0 and 4,042.0. Those ten dollars are a single-print bridge between the morning and afternoon distributions, and the volume profile of the same day shows an LVN in the same place.

Common mistake: Comparing single prints across platforms with different period lengths.

Common mistake: Equating a single print with every LVN, without considering how and when it was created.

Gold's Session Map: Asia, London, New York

Why split the day?

Gold trades almost 23 hours a day, but participation is far from uniform. To turn a continuous price line into an auction map, we divide the day into windows and record the high and low of each one. Exact window definitions differ between platforms and traders, so the most important rule is to write yours down. The definitions used in this book are:

London is the centre of the physical gold market; the twice-daily LBMA gold price benchmark auctions take place at about 10:30 and 15:00 London time. The New York window holds three landmark times: the 08:20 RTH open, the 08:30 U.S. data, and the 13:29–13:30 window from which CME calculates gold's daily settlement price.

Each session runs its own small auction

Each window can hold its own small auction: Asia often builds a narrow bracket, London tests or breaks it, and New York decides which move is accepted. Each window's high and low mark where its auction stopped, making them reference levels for the next session.

Worked illustration: Asia trades 4,030.0–4,036.0 (XAUUSD ≈ 4,002–4,008) on light volume. London breaks above to 4,043.0 and returns to 4,038.0. By 08:00 New York time, the overnight range is 4,030.0–4,043.0. New York's first question is simple: will the market accept London's higher prices, or go back into the Asian range?

Technical details

• Daylight saving. As noted for the IB, UK and U.S. clocks change on different dates. Defining each window in its own local time (London windows in London time, New York windows in New York time) keeps "London" aligned with London's real business hours throughout the year.
• Holidays and half-days. On holidays and shortened sessions, windows are cut short. Use the high and low of an incomplete window with caution, if at all.

Common mistake: Assuming every platform defines the "Asia session" the same way.

Common mistake: "London always makes the high or low of the day." A popular claim, not proven.

Common mistake: Treating Asia's high and low as strong levels without noting how little volume built them.

Reference Levels: PDH/PDL, ONH/ONL and How to Use Them

What makes a reference level?

A reference level is a price where the auction has already done something important: it stopped there, was rejected there, or built value there. The most important ones for gold:

• PDH / PDL: the prior day's high and low (always using the same session definition).
• ONH / ONL: the overnight high and low, up to the New York session.
• Asia and London highs and lows: the boundaries of the overnight's smaller auctions.
• Prior VAH, VAL and POC, and any naked POC (a POC not yet revisited), from Chapters 5 and 6.
• Poor highs and lows, and single prints (earlier in this chapter).
• IB High and IB Low, once the IB has formed.

How to use them

A level is a place to ask a question, not a place for an automatic order. When price reaches a level, the question is: What is the market doing here?

• If price crosses the level and builds time and volume beyond it (acceptance), the level has been broken, and attention moves to the next reference level.
• If price comes back quickly (rejection), the level has been defended.

This is exactly where footprint charts (Chapters 8 and 9) help: they show who was aggressive at the moment of the touch, and whether that effort achieved anything.

For each level on your map, write two scenarios in advance:

Two practical rules

1. Fewer is better. A chart with 25 lines makes every price look "near a level", and therefore tells you nothing.
2. Merge nearby levels into zones. If the ONH is at 4,047.5 and the PDH at 4,048.0, they are effectively one zone: 4,047.5–4,048.0.

An important limit

"Price reacted at the level" almost always looks true, because price always reacts somewhere: it never moves in a straight line. Without comparing behaviour at your levels with behaviour at random prices (a baseline), you cannot say that a level "works". Chapter 12 shows how to make such comparisons.

Common mistake: Drawing dozens of lines and "finding" a reaction at each one. This is confirmation bias: seeing what confirms what you already believed.

Common mistake: "A naked POC almost always gets touched." A popular claim, not proven.

Common mistake: Using PDH and PDL calculated from different sessions on different days.

Putting AMT Together: A Pre-New York Checklist for Gold

Why a checklist?

AMT becomes useful when it turns into a fixed routine. The goal of this checklist is not to predict direction. It is to know, before the main session opens, what it would mean if X happened. It takes about ten minutes before 08:20 New York time and moves from the big picture down to the detail.

The nine steps

1. Big picture (several days). Have recent value areas been overlapping (balance, a bracket) or migrating (imbalance)? Where are the bracket's edges?
2. Yesterday. What was the day type? Did it close near an extreme or in the middle? Did its extremes show excess, or a poor high or poor low?
3. Overnight. Mark the ONH and ONL, and the Asia and London highs and lows. Did the overnight session stay inside yesterday's value or leave it? Was overnight volume heavier or lighter than usual?
4. Where are we opening? Where is price around 08:20 relative to yesterday's value and range? (Inside value / outside value but inside range / outside range.)
5. Calendar. Is there U.S. data at 08:30 or 10:00? If so, the IB may become a "news IB".
6. Levels. At most six to eight levels or zones, each with an approximate XAUUSD equivalent if you follow the CFD.
7. If-then scenarios. For example: "If price is accepted above the ONH, yesterday's value has been rejected; watch the PDH. If the ONH is rejected and price returns into value, the rotation scenario towards yesterday's POC is in play." Always write at least two scenarios.
8. After the IB. What was the opening type? Was the IB wide or narrow compared with its recent average (for example, the average of the last 20 sessions)? Update your day-type hypothesis.
9. After the session. Compare your sheet with what actually happened. Where was the scenario right, and where wrong? This journal becomes the data from which you can later build your own real statistics.

A worked context sheet

Notice what the sheet does not contain: no prediction, no entry, no target.

Limits

This sheet does not remove risk and does not guarantee an edge. Its value lies in consistency and in reducing emotional decisions: when something happens, you have already thought about what it means.

Common mistake: Writing only one scenario. That is directional bias dressed up as analysis.

Common mistake: Moving levels during the session to "fit" the move.

Common mistake: Believing that a completed checklist means a trade must follow. Many days, the honest conclusion is that nothing on the map was resolved clearly.

Try it: Fill in the context sheet before 08:20 New York time for ten consecutive sessions. Take a screenshot of your chart at 08:10 and another at 13:30. After the tenth day, count how often each of your written scenarios played out, without changing anything you wrote in advance.

Chapter summary

• Markets exist to facilitate trade. Price advertises; time and volume show whether the market accepts the advertisement. The other timeframe moves price to new value; the day timeframe rotates it within current value.
• Balance is rotation inside a bracket with overlapping value and responsive trading. Imbalance is a directional search with migrating value and initiative trading. Markets cycle between them.
• Acceptance is time plus volume at a new price; rejection is a quick return on little volume. Write your acceptance rule before you look at the chart, and remember that confirmation always comes late.
• A TPO chart prints one letter per 30-minute period at each traded price, measuring time. It supplies key vocabulary (IB, single prints, tails), but on 23-hour gold it over-weights quiet hours, so this book leads with volume.
• Gold's Initial Balance in this book is 08:20–09:20 New York time. Consistency of definition matters more than the choice itself.
• Range extension beyond the IB hints at other-timeframe activity. Measure it in IB multiples, check for acceptance, and watch for 08:30 news inside the IB.
• Opening types (Open-Drive, Open-Test-Drive, Open-Rejection-Reverse, Open-Auction) and the open relative to prior value and range give initial hypotheses, not signals. The 80% rule is folklore until tested.
• Day types (normal, normal variation, trend, double distribution, neutral) are mainly review tools. Labels depend on definitions.
• Excess (tails) marks a completed auction; poor highs and lows mark unfinished ones. Single prints mark fast, one-sided movement. None of them is guaranteed to be revisited.
• Gold's session map (overnight, Asia, London, New York, RTH) gives each window its own high and low, which become reference levels along with PDH/PDL, ONH/ONL, prior value, naked POCs and poor extremes.
• A pre-New York checklist turns AMT into a routine of context, levels and if-then scenarios, followed by an honest review.

Checklist

• I can say whether the market is currently in balance or imbalance, and why.
• I have a written acceptance rule (time and volume) that I do not change after seeing the chart.
• I know my platform's session template and time zone, and my IB is 08:20–09:20 New York time (or another written convention I keep).
• I measure range extension in IB multiples (never as targets) and note whether 08:30 data fell inside the IB.
• I record the opening type with a written criterion, not by feel after the fact.
• I classify the 08:20 price relative to yesterday's value and range, and label day types only after the session.
• I mark poor highs, poor lows and single prints as references, never as certainties.
• I keep six to eight levels or zones at most, merging levels that sit close together, with XAUUSD equivalents where needed.
• For each level I have written an acceptance and a rejection scenario, and I review the sheet after the close.

Quiz

1. According to Auction Market Theory, what is the main purpose of a market, and what do time and volume tell you about a price?
2. Daily value areas have overlapped for five consecutive days. Which market state is this most likely to be, and which type of trading activity usually dominates it?
3. Price breaks above yesterday's high, spends 40 minutes there and builds a volume node. How would you describe this, and why?
4. Price tests the overnight high just after 08:20, finds almost no trading there, then drives lower without ever returning to the opening price. Which opening type is this?
5. The day's high shows three TPO periods at exactly the same price and no tail. What is this called, and what is the honest way to use it?

Quiz answers

1. To facilitate trade: to find the prices at which the most business can be done. Time and volume are the market's response to a price: lots of time and volume means the price is accepted (agreement), while little of either means it is rejected or was passed through quickly.
2. Balance, with price rotating inside a bracket. Responsive activity usually dominates: selling near the top of the range and buying near the bottom.
3. Acceptance above the prior day's high. Price spent time beyond the level and built volume there, which is the definition of acceptance (time plus volume). It suggests the market may be building new value higher, though acceptance can still fail later.
4. Open-Test-Drive: a test of a nearby reference level (the overnight high) that found no business, followed by a strong drive in the opposite direction without returning to the open.
5. A poor high, an unfinished auction. Mark it as a reference level for future sessions, knowing it is an "untidy" level where resting orders may have gathered. Do not assume it will be revisited: "poor highs always get retested" is a popular claim, not proven.

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