Naked POC: what counts as a fill, and why the same record gives different lists
By José Luis Rodríguez Author of ORDER FLOW — Market Mechanics
Ask four sources when a naked POC stops being naked and you get four answers. One indicator retires the level as soon as any bar's range contains it. Another also retires it when a bar closes within a tenth of its own range of the level, without touching it. A long guide on a futures forum says a tick through the level does not count: price has to stay for a few prints or a few seconds. A session-levels script ignores every touch outside regular trading hours.
Each rule is defensible, and each produces a different list from the same record. Underneath the line sit five decisions: which POC, what counts as a fill, which hours, which contract, and when a level expires.
What a naked POC is
The point of control is the price with the most volume in a volume profile. If price moves away in later sessions and never trades a session's POC again, that POC is naked, or virgin, and traders extend it to the right as an open reference: the session's price of greatest agreement, left behind and never retested. The first return is watched for a reaction, which is reasonable; open POCs are also treated as destinations, which is a prediction, and we are not aware of a published, checkable measurement behind it.
Which POC: three prices from one session
Before a POC can be naked, someone has to decide which price is the POC. That depends on what is counted, volume or time, and on how prices are grouped.
Counted tick by tick, the POC is 21,375.00: a single price where 3,132 contracts traded. Grouped into one-point rows, that price is diluted: its row totals 3,552 and the POC moves to the row from 21,394.00 to 21,394.75, with 8,897. Counted by time instead, as Market Profile does, the POC is the row at 21,424.00, traded in 8 of the 13 half hours. The first and the last are 49.00 points apart. Each is the right answer for its definition, and each starts a different naked line.
A profile built from bars rather than trades also has to spread each bar's volume across its prices, and the spreading rule moves the POC. The window matters too: the book builds the session profile on regular hours, 9:30 to 16:00 New York time, because that is where the volume is, and a full-session profile gives another POC. Even the end of regular hours is a setting: some templates still stop at 15:15 CT, and that changes the prior-day high a chart plots.
What counts as a fill
The rules in use fall into three families: proximity, which counts a near miss; touch, which counts any trade at the price; and acceptance, which asks price to stay, by time, prints or closes.
The naked POC sits at 21,412.50. In the first return, price stops 0.25 points short: with a 2-tick tolerance the level is filled, under the touch rule it is still naked. In the second, a wick goes 0.75 points through and the candle closes 1.75 points below the level: filled by touch, still naked if the rule asks for 2 consecutive closes at or above it. Only the third, with 3 consecutive closes above, counts as a fill under all three.
None of the three is the true one. Touch is objective and easy to check; acceptance is closer to what the auction story says matters. The one indefensible rule is the one chosen after the fact.
Which hours count, and which contract
The third decision is which sessions are allowed to fill a level. When the profile is built on regular hours, some traders count only regular-hours returns, on the argument that like should be compared with like. Others count any trade at any hour.
With the touch rule in both panels, the list changes with the hours. Of the 11 POCs that have at least one later session, 3 stay naked if any session counts and 5 if only regular hours count. The 2 that differ, from sessions 6 and 11, were touched overnight and by no regular session. It is the same kind of choice as picking which price counts as yesterday's close, which on NQ has more than one candidate because the settlement is not the close.
The contract adds one more. NQ expires every quarter, and the next contract trades at a different price. A back-adjusted continuous chart shifts all history before each roll, so its old POCs sit at prices that never traded; an unadjusted one compares one contract's POC with another's prices. Either way, naked POCs can appear that nobody traded in the contract on screen.
When a level expires
Some tools keep an unfilled POC forever, some retire it after a number of days, some keep a fixed number of levels. In the schematic, a 5-session expiry leaves 1 if any session counts and 2 if only regular hours count.
Without an expiry, the destination claim cannot fail: if a level can be filled today or in three years, every return confirms it and no absence counts against it.
What the book weighs instead
Chapter 5 of the book does not use the idea of a naked POC, but it sets two criteria that place one. The first is memory: “The limitation of the session profile is its lack of memory,” and “The four-week composite POC carries more memory than yesterday's session POC.” Naked POCs are usually marked from single sessions.
The second is test frequency, and it points the other way from the naked-POC intuition: “A level that price has tested three times in the last week is more relevant than one that formed a month ago and hasn't been touched.” The two fit together: the naked-POC tradition asks whether price will go to a level; the book asks how much a level weighs once price is there. Under those filters, a single-session POC with no composite level behind it belongs with the book's second-order levels, alongside “a daily-profile HVN with no composite backing”: marked as a reference, with a higher bar for confirmation on the footprint. The book does call composite high-volume nodes magnets, but in another sense: “when price reaches a composite HVN, it tends to slow down and go sideways.” Friction on arrival, not a pull from a distance.
Limitations and false positives
Fill rates are not comparable. Percentages of naked POCs that get filled circulate without the fill rule, the hours, the contract, the expiry or the distance behind them. Two such figures cannot be compared with each other or with your own.
Distance explains many fills. In the schematic, under the touch rule with any session counting, price fills some naked POC in 8 of the 11 sessions after the first, and all 8 fills are of the immediately preceding session's POC. A level that close gets traded by an ordinary day's range. As with unfinished auctions, showing a pull would need naked POCs compared against arbitrary prices at the same distance and age. And with enough open levels on the chart, almost any price is near one.
Trend days pile them up. On a trend day the market leaves prior value instead of returning to it. In the schematic, every POC still naked is one the rally left behind. Using them as targets mid-trend applies a return-to-value idea exactly when value is being rejected.
Rules changed after the fact. If a level “didn't work” and the fill rule or the hours are adjusted until it does, the list stops recording and starts confirming what you already wanted to see. The settings are fixed before the session starts.