LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
The firm reverse-engineers a funded trader the host will not name, called Trader X, from internal account data and from email. The host's figures, presented as a case study and not a forecast, are a monthly payout range from about forty-seven thousand three hundred dollars to one hundred two thousand three hundred fifty, similar amounts across multiple funded accounts, and over the last twenty-five days a fifty point one five percent return, a six point eight four payoff ratio, a sixty-one point nine percent win rate, and four point two one to one reward against risk. He contrasts that with a more ordinary sixty-two percent win rate sitting near one to one. The strategy he extracted is a selective breakout, built on a stricter version of the NR4. These numbers are his reading of one account's recent book, not an audit and not a forecast of the next twenty-five days.
Why it matters
A traditional NR4 starts with four completed daily sessions. Measure each candle from high to low. If the newest range is smaller than each of the three before it, that candle is the narrowest of the four. The earlier three do not have to shrink in order. Only the last one has to be the smallest. That bar is volatility compression. Mark its high and its low. The next session, a break of the high is a long and a break of the low is a short. Whichever side prints first is the trade. The other order is canceled. The market, not the trader, names the direction of the expansion.
Trader X, as the host reconstructed him, wants the compression to be a sequence, not a single narrow bar. Four consecutive candles, each with a range equal to or smaller than the one before it, until the last bar is the tightest in the whole run. He marks the high and the low of that final candle. A tick through the high is the long. A tick through the low is the short. Once either side triggers, both levels are gone. A reversal is a stop, not a flip. One compression sequence, one signal. Three stages: sustained contraction, a confirmed break of either side, and one-and-done execution.
The idea in plain words
The host coded that rule and looked at where it would have entered. He says many of the hypothetical breaks sat on clean locations, and that no rule is perfect. The coding also produced mediocre breaks that Trader X never took. That is the first real difference from a mechanical breakout. The break is only the trigger. He does not trade every one.
The host could pin down a few filters and had to leave others open. The markets are major futures, ES and NQ. The chart is the hourly most of the time. Other timeframes appeared, with nothing consistent enough to call a rule. He could not find a clock the trader cares about. Major news was sometimes avoided and sometimes not. There is never more than one trade per day in the same market. Several markets can be on at once, up to about five trades in a day, but not two attempts on one symbol in one session.
The exit is variable, which the host says would hurt most prop traders and does not appear to hurt this one. Losers were usually cut in minutes or a few hours. When the break kept going, the position stayed, sometimes for days, and almost always came off in pieces rather than at one fixed multiple. One NQ trade is the example: several exits from the same entry, some pieces closed for a few thousand dollars, the last piece for more than thirty-five thousand. That is not a standing two-to-one or three-to-one. The host's name for the hold is post-liquidity-shock drift: stay after the break itself is finished, as long as price is still moving the position's way. Where the pieces come off is discretion. He could not decode it, and he attributes it to experience.
How to use it
The stop is the part the trader would not explain. The logs did not show a fixed number of points, ticks, or dollars. What the host could see is a trailing stop that widens when volatility expands, so ordinary noise does not tag it, and tightens when volatility contracts, so a fading move does not give the gain back. He tested a basic ATR stop, a standard-deviation stop, and more elaborate volatility models against the fills. None matched exactly. Asked how he measures the volatility, Trader X declined to say it for a public video. The host's advice to anyone who wants the family of idea, not this man's exact stop, is to start with ATR and how a stop is built from it.
Overnight holds are part of the same choice. Futures trade nearly around the clock on weekdays, so an hourly breakout is not forced flat at the cash close.
The distinction the host draws is that Trader X does not decide from price alone. He decides from volatility. Markets are heteroscedastic: variance is not constant. It clusters, contracts, and expands. Retail charts show that as a picture and then trade the picture. This account treats the regime as the state variable. The entry waits for realized volatility to contract through the adapted NR4, then positions for the change from compression to expansion. While the expansion has directional persistence, the stop gives room and pieces stay on. When volatility compresses again and the persistence fades, adverse movement is tolerated less and exposure comes off. Price says where the order is. Volatility says whether there is an order, how far the stop sits, and when the position is no longer the same trade.
What usually goes wrong
Trading every coded break is the first miss. The host's reconstruction produced locations Trader X never took. The break is the trigger, and the filters the logs would support are incomplete: major futures, mostly the hourly, never two attempts on the same symbol in one session, and no clock the host could find. A fixed two-to-one or three-to-one would not match the exits he could see. Pieces came off while the move kept going, and losers were cut in minutes or a few hours. Copying a public ATR stop and calling it this account's stop is the other miss. None of the volatility models the host tested matched the fills exactly, and Trader X would not explain the measurement. The four-to-one payoff, the payoff ratio, and the monthly payouts remain the host's reading of one elite account's recent book.
Questions people ask
How is this NR4 stricter than the textbook bar?
The textbook bar only has to be the narrowest of four completed sessions. The earlier three need not shrink in order. Trader X, as reconstructed, wants four candles in a row, each range equal to or smaller than the one before it. One tick through either side of the last candle is the only signal. The other side is canceled. A reversal is a stop, not a flip.
What did the last twenty-five days show, in the host's reading?
A fifty point one five percent return, a six point eight four payoff ratio, a sixty-one point nine percent win rate, and four point two one reward against one of risk. Monthly payouts ran from about forty-seven thousand three hundred dollars to one hundred two thousand three hundred fifty, in similar amounts across multiple funded accounts. He presents that as a case study, not a forecast.
What part could he not decode?
The stop. No fixed points, ticks, or dollars appeared in the logs. The trail seemed to widen when volatility expanded and tighten when it contracted. ATR, standard deviation, and more elaborate models all failed to match the fills exactly. Where the winning pieces came off was also discretion. One NQ example closed some pieces for a few thousand dollars and the last piece for more than thirty-five thousand.