Module 4 Quiz: Reading the Auction
Eleven lessons, and every one of them ended at the same place: the number you were about to act on belonged to a setting you had not written down. A buffer, a base rate, a bin width, an aggressor rule, a displayed quantity, a swing lookback. Six questions, all arithmetic. The last runs the fractal over the twenty bars this module has been carrying since lesson 32 and turns the swing it finds into a stop and a payoff.
Covers: Lessons 25 to 35, and the twenty bars that lesson 32 stated in full, lesson 33 gave bodies, and lessons 34 and 35 carried out to sixty.
Every question below hands you numbers and asks for a number back. Work all 6 with a calculator before you scroll to the answers; each answer shows the arithmetic, so a wrong result tells you which step to go back to rather than only that you were wrong.
The questions
1. What the buffer buys, and what it costs
A long taken on the reclaim of a level. The entry is 0.3 average ranges above the level and the target is 4.0 above it. The method works 45 per cent of the time when it is left alone, and a probe past the level travels on average 0.4 average ranges before turning, so the chance a stop d below the level survives one is 1 − e raised to the power −d ÷ 0.4.
Ask. Put the stop 0.20 average ranges below the level, then 0.80. Give the payoff, the breakeven win rate, the survival chance and the expectancy for each.
2. What a wall read is worth on two instruments
You have a way of telling, from how a large resting order behaves, whether it will still be there when price arrives. Tested against what actually happened, it is right 75 per cent of the time when the wall holds and right 75 per cent of the time when it does not. On one instrument 8 walls in 100 hold. On another, 30 do. Take 200 tests on each.
Ask. When it says the wall will hold, what is the chance it holds, on each instrument?
3. Where the point of control is, at three settings
A session, one tick at a time.
| Price | Volume |
|---|---|
| 100.20 | 300 |
| 100.19 | 420 |
| 100.18 | 900 |
| 100.17 | 640 |
| 100.16 | 560 |
| 100.15 | 480 |
| 100.14 | 520 |
| 100.13 | 860 |
| 100.12 | 780 |
| 100.11 | 340 |
| 100.10 | 260 |
| 100.09 | 220 |
Ask. Find the point of control at one tick, at two ticks with the grid on the even cent, and at two ticks with the grid shifted one cent. How far apart are they?
4. Two footprints that cannot be told apart
Two bars. Both open at 100.05, both high at 100.65, both low at 99.85, both close at 100.35, and both trade 10,380 contracts. Here is what each did at each price.
| Price | A buys | A sells | B buys | B sells |
|---|---|---|---|---|
| 100.60 | 160 | 180 | 70 | 270 |
| 100.50 | 340 | 300 | 160 | 480 |
| 100.40 | 720 | 520 | 360 | 880 |
| 100.30 | 980 | 580 | 570 | 990 |
| 100.20 | 1,020 | 620 | 720 | 920 |
| 100.10 | 1,440 | 620 | 980 | 1,080 |
| 100.00 | 1,440 | 620 | 1,030 | 1,030 |
| 99.90 | 520 | 320 | 410 | 430 |
Ask. What is each bar’s delta, and what does the footprint settle about which one to fade?
5. What the detector could not have seen
One price level, displaying 300, tested a hundred times over a fortnight. A reserve can only reveal itself if the display is cleared and something is still there, so here is what actually happened on each test.
| What happened at the level | Tests | Refill seen |
|---|---|---|
| Under half the displayed 300 traded | 64 | 0 |
| Over half traded, display not cleared | 19 | 0 |
| Display cleared, price moved on | 5 | 0 |
| Display cleared and refilled at the price | 12 | 12 |
Ask. On how many of the hundred tests could the detector have fired at all, and what does a quiet book prove?
6. The twenty bars, one last time
The series lesson 32 stated in full, which lesson 33 gave bodies and lessons 34 and 35 carried out to sixty. The highs are 100.8, 101.6, 101.1, 102.9, 102.4, 104.2, 103.5, 103.0, 105.4, 104.7, 106.8, 106.1, 104.3, 105.9, 103.6, 105.1, 102.2, 103.4, 100.9 and 101.8. The lows are 100.0, 100.7, 100.2, 101.5, 101.3, 102.8, 102.4, 101.9, 103.8, 103.5, 105.0, 104.4, 102.9, 103.9, 102.0, 103.0, 100.6, 101.4, 99.2 and 100.1. The average bar range is 1.46 points.
Ask. Run the fractal at two bars either side: which bars are swing highs and lows? A swing is confirmed two bars after it forms, so when could you first act on the top, and where was price by then? Then put a stop half an average range below the swing low, enter 0.4 above it, target the series high, and give the payoff.
The answers
Each one is worked in full. Where a figure comes from a lesson rather than from this page, the lesson is named.
1. What the buffer buys, and what it costs
The risk is the entry offset plus the buffer and the reward is 4.0 − 0.3 = 3.7 average ranges either way.
| Stop below the level | Risk | Payoff | Breakeven win rate | Survives the probe | Expectancy |
|---|---|---|---|---|---|
| 0.20 | 0.50 | 7.40 | 11.9% | 39.3% | +0.49R |
| 0.80 | 1.10 | 3.36 | 22.9% | 86.5% | +0.70R |
The expectancy is where the two halves meet. When the probe removes you, you lose a whole R and the trade never finds out whether it was right; when it does not, you have a 45 per cent shot at the payoff. So at 0.20 that is 0.393 × (0.45 × 7.40 − 0.55) − 0.607 = +0.49R, and at 0.80 it is 0.865 × (0.45 × 3.36 − 0.55) − 0.135 = +0.70R.
Read the first two columns of the tighter stop and it wins outright: more than twice the payoff and eleven fewer points of win rate needed. It is still the worse trade, by a fifth of an R, and the whole of the difference is the six trades in ten that never got to be right or wrong.
Answer. At 0.20 the payoff is 7.40, the breakeven 11.9 per cent, survival 39.3 per cent and the expectancy +0.49R. At 0.80 they are 3.36, 22.9 per cent, 86.5 per cent and +0.70R.
2. What a wall read is worth on two instruments
The read is only ever a multiplier on what was already true. On the first instrument 16 of the 200 hold and the detector calls 75 per cent of them, so 12 true calls; of the 184 that do not hold it wrongly calls a quarter, so 46 false ones.
| Walls that hold | Hold, out of 200 | True calls | False calls | Chance it holds when the read says so |
|---|---|---|---|---|
| 8 in 100 | 16 | 12 | 46 | 20.7% |
| 30 in 100 | 60 | 45 | 35 | 56.3% |
So 12 ÷ 58 = 20.7 per cent and 45 ÷ 80 = 56.3 per cent. The same eye, the same book, the same 75 per cent, and a read that is wrong four times in five on one instrument and right more often than not on the other.
Notice what never entered it. Not the size of the wall. Ten thousand and forty thousand give the same answer, because the two quantities that decide it are the base rate and the accuracy, and neither of them is on the screen.
Answer. 20.7 per cent on the first and 56.3 per cent on the second, from the same read.
3. Where the point of control is, at three settings
At one tick it is simply the tallest bar: 900 at 100.18. Then pair the prices up, twice, starting one cent apart.
| Bin width | Point of control | Volume in it | Runner-up | Margin |
|---|---|---|---|---|
| 1 tick | 100.18 | 900 | 860 at 100.13 | 40 |
| 2 ticks, grid on the even cent | 100.12 to 100.13 | 1,640 | 1,320 | 320 |
| 2 ticks, grid shifted one cent | 100.17 to 100.18 | 1,540 | 1,380 | 160 |
The one-tick reading and the shifted two-tick reading both land at the top of the session. The even-cent grid, which is the one most platforms ship, lands six ticks lower — not a neighbouring price, the other half of the range. It gets there by pairing 100.13 with 100.12 and separating 100.18 from 100.17, and nothing about that pairing came from the market.
The session is 6,280 contracts and the margin on the shifted grid is 160 of them, about two and a half per cent. Lesson 19 would call a difference that size unmeasured rather than measured. This is the reason lesson 30 reads a value area instead: an interval moves far less than a mode does when you change the bin width underneath it.
Answer. 100.18 at one tick, 100.12 to 100.13 on the even grid, and 100.17 to 100.18 shifted — six ticks of a twelve-tick session.
4. Two footprints that cannot be told apart
Add the columns. A bought 6,620 and sold 3,760, so its delta is +2,860. B bought 4,300 and sold 6,080, so its delta is −1,780. Both traded 10,380 contracts in total, and both printed the same four prices.
So bar B closed thirty cents above its open on net aggressive selling. One school calls that absorption and buys it: sellers hit the bid all the way up and price rose anyway, so somebody large was taking the other side. The other calls it exhaustion and fades it: the buying that lifted the bar came from nowhere the tape can see, and a close near the high on negative delta is a bar that ran out of fuel.
Both readings are internally consistent, both are widely taught, and the footprint contains nothing that decides between them. It is a second measurement with its own error rate, not the truth behind the candle. And the delta itself is a classification rather than a count — lesson 8 bounded how wide that interval gets when the trades between the quotes are signed by a rule.
Answer. +2,860 for A and −1,780 for B, on identical bars. The footprint settles nothing.
5. What the detector could not have seen
Only the last two rows are tests. On the first 83 the displayed quantity was never exhausted, so a reserve of any size sitting behind it would have gone unobserved — the measurement was never made.
Of the 17 tests that could have fired, 12 did: the display cleared and was replaced at the same price. That is 70.6 per cent of the tests that count and 12 per cent of all of them, and the second figure is the one people quote.
Which makes hidden size a different object from lesson 26’s wall. When the detector fires it is an observation, not a read, and no posterior is needed: something refilled, so something was there. What it cannot do is the negative. A book that never fires is 83 tests that were never run, and every measurement of how much was hidden is a floor rather than a figure.
Twelve episodes is also a small number in lesson 19’s sense. It is enough to establish that reserves exist on this level and nothing like enough to say how often they hold.
Answer. Seventeen. It fired on twelve of those. A quiet book proves nothing, because 83 tests could not have shown a reserve however much was hidden.
6. The twenty bars, one last time
A bar is a swing high if its high beats the two on each side of it. Bar 6 at 104.2 beats 102.9, 102.4, 103.5 and 103.0. Bar 11 at 106.8 beats everything near it. Nothing else survives: bar 9 at 105.4 loses to bar 11, bar 14 at 105.9 loses to bar 12, and bar 16 at 105.1 loses to bar 14. On the lows only bar 8 at 101.9 beats its four neighbours.
| Swing rule | Swing highs | Swing lows | At bars |
|---|---|---|---|
| 1 bar either side | 8 | 8 | every other turn |
| 2 bars either side | 2 | 1 | highs 6 and 11, low 8 |
| 3 bars either side | 1 | 0 | high 11 only |
Now the lag, which is the part that does not appear on anybody’s marked-up chart. The high at bar 11 is not a swing high until bar 13 has closed, so the earliest you could act on it is bar 13 — and bar 13’s low is 102.9, which is 3.9 points below the 106.8 you would be reacting to. The chart in the screenshot and the chart you were looking at are two bars apart.
Then the trade. Lesson 32 said the stop belongs beyond the swing that defined the break, and lesson 25 priced how far beyond. Half an average range below 101.9 is 101.9 − 0.73 = 101.17, and 0.4 above it is 102.48, so the risk is 1.31 points. The target at 106.8 is 4.32 away, so the payoff is 4.32 ÷ 1.31 = 3.28 and the breakeven win rate is 1 ÷ 4.28 = 23.3 per cent.
Every number in that paragraph came from a rule with a setting in it. Change the lookback to three and there is no swing low at all, so there is no stop, so there is no trade. That is the module in one line: the level was never on the chart, it was in the settings.
Answer. Swing highs at bars 6 and 11, one swing low at bar 8. The top is actionable at bar 13, by which point price is 3.9 points off it. The trade is 102.48 with a stop at 101.17, a payoff of 3.28 and a 23.3 per cent breakeven.
What this quiz was testing
Whether you can say which of your numbers is a measurement and which is a choice. The volume at a price is a count and every venue agrees on it; the point of control computed from it moved six ticks when a grid shifted one cent. A refill is an observation; a wall read is a posterior that swings from 21 to 56 per cent on the base rate alone. A swing high looks like a thing on the chart and is the output of a number nobody says out loud. None of that makes the tools useless. It makes the setting part of the claim.
Module 5 asks the question this one has been carefully not asking: whether any of it predicts anything, and it starts by finding that the same instrument is two different instruments depending on which regime it is in.
Educational only. Trading involves substantial risk of loss. Not financial advice. Past performance does not guarantee future results.