The order book, the tape and market speed

Everything so far was read from what has already traded. This lesson is about what is waiting — the orders resting in the book, the ones paying up to take them, and how fast that is happening.

Lesson 5 of 12. It ends with TVVR, which is the complete entry system the first four lessons have been assembling.

The order book, and who is in it

An order book is the list of every buy and sell order waiting on an instrument, with the size sitting at each price. It is where market depth comes from, and it is a picture of intention rather than of activity.

There are two kinds. A centralised book collects every order in one venue, which is what makes the information the same for everyone looking at it. A decentralised book spreads orders across a network, which is how most crypto venues work. The futures market is centralised, and that is the reason its volume is usable at all — the same argument as lesson one, arriving from the other end.

Everyone in the book is passive. They have placed a limit order and they are waiting. And here is the part that is easy to miss and changes how you read the screen: passive buyers and passive sellers have no direct connection to each other. Two queues of people waiting, on opposite sides, who cannot trade with one another. Something else has to bring them together.

A depth ladder with resting sell orders stacked above the current price and resting buy orders below it
Two queues of people waiting, and no way for them to reach each other. Nothing in this picture has traded yet. Click to enlarge
A depth ladder with resting sell orders stacked above the price and resting buy orders below it, both labelled passive
The book itself: passive sellers above, passive buyers below, and the size waiting at each price.

Market by price, and market by order

How much of the book you can see depends on which feed you have, and the difference is larger than the names suggest.

MBP — market by price. Everything resting at a price is consolidated into one line: the total quantity and the number of orders making it up. It is capped at ten price levels. You cannot tell how that total is composed, or whose order is where in the queue.

MBO — market by order. Every order individually, at full depth: its size, its position in the queue, and how long it has been sitting there. Each order gets a unique OrderID that lasts until it is filled or cancelled.

What MBO buys you is iceberg orders — large orders that display only a fraction of themselves and refresh as the visible part is taken. Exchange-held ones, which he calls native icebergs, keep the same OrderID when they refresh, so they can be followed.

His qualification is important and it is the sort of thing usually left out: not every iceberg lives on the exchange. Many are held outside it, which means they are not always trackable at all. The tool shows you the ones it can see, and you should not assume that is all of them.

One price level shown as a single total under MBP and as its individual queued orders under MBO
The same price, two feeds. One says four hundred and twenty; the other says who, how much each, and who has been waiting longest. Click to enlarge
A market by order view showing individual orders queued at each price beside the chart they belong to
What MBO shows that a price-aggregated feed cannot: every order separately, its size, its place in the queue and how long it has been there.

The aggressors

The thing that brings the two queues together is the other kind of trader.

An aggressive trader does not join the book. They buy or sell immediately at whatever is resting, crossing the spread to do it. Buy aggressively and your order is matched against the first sell order waiting, and that resting size is reduced by what you took.

This is what makes the market work. The futures market is two-sided, and traders act both ways at different moments — so it is the aggressors who connect the passive orders to each other, and that is where liquidity and flow come from.

You can watch them in the time and sales list: time, price, volume, one line per trade. But there is a better arrangement of exactly the same information — put those numbers next to the candle, at the prices they happened, and you get what he calls from here on a footprint.

The footprint is easier to read than the tape, it can be filtered and aggregated, and it makes tracking a large buyer possible. On Level 1 data, none of it exists.

A candle with the buying and selling volume printed at each price beside it, forming a footprint
The same trades the tape scrolls past, filed by the price they hit instead of by the moment they happened. Click to enlarge
A footprint chart with buying and selling volume printed at each price, beside a running time and sales list
The two views of the same thing: the tape running down the side, and the same trades filed by price as a footprint.

How the two sides actually meet

There are only two combinations, and being clear about them fixes a lot of confusion about footprint charts.

An aggressive buyer buys from a passive seller who was resting on the ask. An aggressive seller sells to a passive buyer who was resting on the bid. That is the whole mechanism, and thousands of orders move across it in both directions to produce the flow.

He adds one note that is worth having in front of you the first few times you read a footprint: it makes no difference whether the candle is red or green. The aggressive buying is always on one side of the print and the aggressive selling on the other, in the same places, regardless of what the candle ended up doing.

An aggressive buyer taking from a passive seller and an aggressive seller hitting a passive buyer, shown as the only two ways a trade happens
Two combinations, and no others. The colour of the candle is a summary of the result, not of who did what. Click to enlarge
A footprint and an order book side by side with arrows showing aggressive buyers taking from passive sellers and the reverse
The four roles drawn onto real windows — and his note that the positions do not change with the colour of the candle.

What your own orders make you

A short section, and a genuinely useful reframing.

When you enter a long, you have not placed one order. You have placed three: the buy, and two sells — one as the stop and one as the target. Usually the stop is aggressive and the target is a limit, though either can be either depending on conditions. A short is the mirror: one sell and two buys.

The consequence is structural. In futures you cannot hold two opposing positions — no hedging. If you open an opposite position of the same size, both close, because of exactly the order mechanics above.

And that gives the market a character worth knowing. It makes futures mean-reverting: large institutions, which have the resources not to use stops, work by averaging into their positions rather than being taken out of them. You can see it happening when price crosses the averages on rising volume.

A long position shown as one buy and two later sells, the stop and the target, with the same for a short
Your stop is a sell order sitting in the book with everyone else's. That is worth remembering before you place it somewhere obvious. Click to enlarge
A long position and a short position each drawn as an entry with a stop and a target on the opposite side
Your own orders, drawn out. Every long is also two sells; every short is also two buys.

The hourly signal candles

Now the entries. The first timeframe that produces a signal is the hourly chart, and there are three candle shapes he uses.

What qualifies them is where they close relative to the significant volume area — the same reading as lesson four. Found in an hourly correction, and regardless of their delta, they indicate buying or selling pressure.

He ranks them, and the ranking is his: the red candle first, the yellow second, the blue last. That is their order of strength in revealing a reversal.

The entry is mechanical once one appears:

  • Wait for the candle to close. There is no signal until it has.
  • Place a limit order at the value edge — VH for a long, VL for a short.
  • The stop goes behind the candle.
  • Part of the order can be placed on the POC instead.
Three candle shapes ranked by strength, with the entry placed at the value edge and the stop behind the candle
Three shapes, ranked. One of them can also appear at the end of a movement — which is why the next section exists. Click to enlarge
Six candle shapes with their delta conditions and value edges marked, beside two charts showing where the signal position sits
The signal candles with their delta conditions, and the two places on a trend where they count.

The signal is worth nothing in the wrong place

This is the section to read twice.

Signals appear wherever there is buying or selling pressure, which is to say constantly. The only ones that matter are the ones inside an area you had already identified. The hierarchy is explicit and it does not bend: trend first, then the area, then the signal.

His own example is a downtrend where the correction reaches Midas VWAP, sellers arrive, and the signal candle forms there. The candle is not what makes that a trade — the location is, and the candle only confirms it.

He notes that when a signal candle coincides with fishing in the area it can be strong enough that the next candle never retests the value edge at all, in which case the close itself can be used for entry.

Three restrictions, and all three are his:

  • Do not use a signal against the trend, and do not use one generated against the fundamental data.
  • If a signal forms before the American session and the market has not moved, expect the bottom of the signal candle to be fished when that session opens. Better to be out before it and use a fishing entry instead.
  • If the signal candle is very large, leave it. The stop is too far away to be worth the trade.
Three gates in order: the trend direction, the statistical area, and only then the signal candle
A signal before price has reached the area is not an early entry. It is not an entry. Click to enlarge
A downtrend correcting into the weighted average, with the signal candle it produced shown enlarged
His own example: the correction reaches Midas VWAP, sellers arrive, and the signal candle forms there rather than anywhere.

Untested points of control

The POC is the price in a candle where the most volume traded, so later candles usually come back and retest it. The interesting case is when they do not.

A POC that stays untested stayed that way because of buying or selling pressure — something stopped price returning. Those are high-potential reversal points, and there is a decay rule attached: they are valid for two days at most. A POC from several weeks ago is not one of these.

His conditions:

  • It must be in the direction of the trend. No volume factor outranks structure — the same rule as lesson four.
  • It should carry significant volume, usually about half the first limit filter.
  • A tested POC is finished. It is not usable again.
  • The best one is the previous hour's, and at most a few hours back. One of the best entries available is in the direction of the previous hour's POC.
  • An untested POC against the trend is not an entry. It can be used as a partial exit.
A point of control from an earlier hour that price has not returned to, marked as still live, beside one that has already been retested
A price nobody came back to, and a reason they did not. It stops being interesting after two days. Click to enlarge
A chart carrying several horizontal point of control levels with price reacting to the ones it has not yet returned to
Filtered untested POCs on a live chart, with the reactions where price came back to them.

Volume above the average

A price that traded far more than the prices around it marks where the large participants were working. Land that on a support or resistance area and you have an entry.

Two details in how it is measured matter. First, buying and selling are added together, not netted — this is about total activity, not about direction. Second, on the hourly chart volumes are looked at in aggregate rather than one at a time.

What produces these is algorithmic: either averaging down an existing position or opening a new one.

And the same restriction as everywhere else: if the unusual volume occurs outside your technical ranges, do not use it. His reason is specific rather than general — it is the averaging nature of the futures market. Large volume appears in a lot of places for reasons that have nothing to do with a level.

A run of ordinary volume bars with one far larger than the rest, sitting at a marked level
Unusual only means anything against what is usual. The measurement is the average of the session it sits in. Click to enlarge
A price ladder with one unusually large combined volume highlighted, beside the fall that followed it
Buying and selling added together at one price, the unusual total marked, and what price did afterwards.
Two charts with the unusual volumes boxed where they land on statistical support and resistance
The same reading at statistical levels — which is the only place he uses it.

What a tick actually carries

Every order that enters the market is a tick, and each one carries three things: the time, the quantity, and whether it was a buy or a sell. Everything in this section is arithmetic on those three fields.

ATS — average trade size. The volume divided by the number of orders that made it. A high ATS means each individual tick was large, which points at algorithms. His thresholds: above 1.8 for a candle, above 3.5 at a single price level.

Trade size. One tick that is very large on its own — fifty contracts in a single order, say. Computed per price rather than across a candle, and the figure that counts as large varies with the instrument.

Big trade. The most interesting of the four, and his explanation of it is the clearest thing in the range. Tick times are recorded to the millisecond. The fastest human reaction — a blink — takes 100 to 150 milliseconds. So when more than a hundred orders arrive inside a single millisecond, no person did that. It is high frequency trading, and the aggregation of volume within equal milliseconds is what the measurement reports.

AVB. For the algorithms that deliberately avoid the filters above by breaking an order into small pieces. AVB looks for small blocks entering in a particular sequence. Candles with a high AVB tend to act as order blocks themselves, and price reacts when it returns to them.

Four more sit alongside them: bar volume for candles carrying unusual activity, average bar which normalises across the day so the small ones become visible, bar delta — and he notes that a candle whose colour opposes its delta is a good fishing candidate — and the volume filter, used hourly for levels and on ten minutes or less for entries.

Last, the delta line: tick-by-tick delta accumulated from the start of the day. Divergences between it and price warn that a move is unsupported or artificial, and he calls it extremely important for scalping.

One tick broken into time, size and side, and the four measurements computed from a stream of them
Three fields per tick. Everything above is arithmetic on them, and none of it exists in a Level 1 feed. Click to enlarge
A footprint chart with ATS, AVB, average bar, trade size and big trade each labelled where they occur
All of the tick measurements printed onto the chart at the prices they happened, instead of scrolling past in the tape.

Down the timeframes

All of the above works on any timeframe. The hourly signals are the one exception — the value edges are much weaker below an hour, so the signals do not transfer down.

The hourly chart is used two ways: as an entry in its own right, and as the reference for the nearest support and resistance — provided the hourly level itself sits at a major one. It suits trades held more than a couple of days.

To shorten the stop you go down: ten minutes, five, one. This course enters on the one-minute chart, which brings the stop to roughly eight pips from entry depending on the instrument — instead of the thirty or forty a higher timeframe would need. Each timeframe needs its own calculations, which is what his volume filter detector is for.

Two honest caveats he attaches, and both are worth repeating:

  • One minute is fast. If it is too fast, start at five or ten and work down as you get quicker. Higher timeframes simply mean larger stops.
  • Do not trade the one-minute chart on its own. Because the large participants average into positions all over the place, one-minute activity appears constantly — and only the activity inside an area you already identified counts. Trading all of it does not work.
The hourly chart choosing the area and the one minute chart timing the entry, with the stop distance shrinking as the timeframe does
The higher chart picks the place. The lower one picks the moment. Neither does the other's job. Click to enlarge

The tape, and market speed

Watching the tape directly can be done, but he is blunt that it takes a long time to master and that the numbers scrolling past can obscure the overall view rather than reveal it. So the measurements above are printed onto the chart at the prices they occurred instead, which also gives you a history the tape does not keep.

His worked example is a good one. The chart shows a trade size of 10 at a price where the tape shows 12. Both are right: twelve contracts traded there, and ten of them were a single order. The flag beside the number is what tells you so.

But large orders are only part of it — algorithms do not always trade in size. Which leaves the question of what all the small numbers mean, and the answer is market speed. It measures how fast orders and volume are arriving from each side, and shows which side is currently winning that race.

The panel has four parts:

  • Session delta — the overall delta line from the start of the session.
  • Market speed — the speed of buy and sell orders, in three forms: by volume, by order count, and by average trade size.
  • Tape delta — the delta of the rows currently passing, for a real-time read.
  • Order status analysis — what has happened since the open, by order count and by volume.
Two speed dials for buying and selling beside a session delta line, showing which side is currently moving faster
The interface for all the small orders the tape shows you too quickly to read. Speed, not size. Click to enlarge
A chart with trade size and big trade marked on both sides, beside the time and sales list the numbers came from
His worked example: a trade size of 10 printed on the chart where the tape shows 12 — ten of the twelve were a single order.
Two speed dials for buying and selling beside a session delta line and an order status panel
The market speed panel: buy and sell speed, the session delta, the tape delta and the order analysis.

TVVR, in his nine steps

Here is where the course arrives. TVVR is his complete entry system, and the letters are the four stages: T for the daily and hourly trend, V for the statistical levels, V for the algorithmic volume on the hourly and one-minute charts, and R for the retest.

He describes it as the best way to capture a large movement with a small stop, and the nine steps are worth reading as a summary of the previous four lessons rather than as something new:

  • Look at the fundamental position of the instrument — which phase of the cycle it is in, and what is on the economic calendar in the next few days.
  • Read the daily trend for direction: beginning of a trend, end of a movement, start of a correction.
  • Move to the hourly chart and decide whether this is rotation, movement or correction. This is where long or short is decided.
  • Measure the slopes and the deltas for strength and weakness, and identify the fishing points if there are any.
  • With the direction settled, approach the statistical areas — pivots, VAH and VAL.
  • Wait at those areas for the algorithms to react, on the one-minute chart.
  • If activity appears in that minute, expect price to move in the intended direction.
  • Measure the delta of that movement to confirm aggressive traders actually entered.
  • Enter on the retest — when price comes back to the algorithmic area.
Four stages in order: trend, statistical volume levels, algorithmic volume on the one minute chart, and the retest that is the entry
Nine steps, and eight of them happen before anything is bought. The last one is the only one that costs money. Click to enlarge
A trend diagram with the value area edges marked and the one minute activity at the level shown enlarged
The shape of the setup: the level, and the one-minute activity inside it magnified.
The daily chart, the hourly chart and two magnified one minute views showing the aggressive entry and the retest
All four stages on one slide — daily trend, hourly structure, the one-minute volume, and the retest that is the entry.
Three variations of the same setup showing where the one minute volume sits relative to the statistical level
Three placements of the same one-minute activity against the level, and what each one means.

What to take from this one

Three things, in order of how much they matter:

  • The book is what is waiting; the tape is what was done. Passive orders cannot reach each other — an aggressor has to cross the spread, and only that moves price.
  • Trend, then area, then signal. The range says it three times about three different subjects: a signal before price arrives is not a trade, an untested POC against the trend is a partial exit, and the one-minute chart alone produces nothing.
  • The entry is the retest, not the reaction. You watch the algorithms act at your level, confirm it with delta, and enter when price comes back — which is the whole of the R in TVVR.
Three panels: passive against aggressive, the hierarchy of trend then area then signal, and the retest as the entry
The three in the order they matter. The last one is what keeps the stop small enough for the rest to be worth doing. Click to enlarge

Next: stop hunting, and entering after it

This lesson kept mentioning fishing — the spike that takes the obvious stops before the move happens. Lesson six is that on its own: why price so often takes yesterday's low by a few points and turns, and the entry that follows.

Back to the twelve lessons