LessonThis is a lesson restated in new words. Any figure in it is an example from the source, not this desk's record.
Order flow led, and fading extensions finished last, when the firm sorted roughly 10,000 traders from its latest quarter into five families. Nine in ten of that book failed inside a month, and one in twenty ever received a payout, on the host's tally. The rates, clocks, and dollars below are the firm's own figures, and they are not an outside audit.
Why it matters
The usual story credits risk rules and mindset. The cut he published is the method. Almost every account fell into one of five buckets: fading a stretch, trading a volatility break, riding a persistent move, reading price directly, or reading the tape. Each bucket gets a pass rate, days to the pass, a monthly figure among people who were paid, and the share who ever collected.
The idea in plain words
Fading an extension, too far up or too far down, can use an oscillator, a divergence in cumulative volume delta, or an Ornstein-Uhlenbeck arbitrage model. Just 5 percent passed. Those passes averaged 25.6 days. Accounts that did get paid averaged a month of $2,441. Inside that paid slice the best tenth reached $4,917 and the best twentieth $11,086. A pass followed by a first check was 1.8 percent of the fade cohort. He places the style fifth, and he says $2,441 is below a typical monthly wage in most rich countries.
Volatility breaks hunt news, stop cascades, and anything that is not a random wiggle. Fewer winners are accepted for a wider payoff, built from levels, headlines, and liquidity, and he says the entries have to be nearly exact because one expansion is asked to repay several scratches. Passes were 8.3 percent and averaged 13.9 days, about half the fade group's calendar. Paid months averaged $3,268. Among those payouts the best tenth was $6,643 and the best hundredth $15,618. A first check reached 3.7 percent. Fourth place. The direction can be correct while the volatility still clips the stop.
Riding a persistent move, from a moving-average cross through a Kalman filter, dies when a directional-looking chart is only noise. He says most labeled trends lack statistically significant momentum. Passes were 9.8 percent and took 22.8 days. Paid accounts averaged $4,376 a month, about 34 percent over the breakout payouts and 79 percent over the fade payouts. A first check reached 5.4 percent. Third. No top-decile dollar amount is stated for this family, so none is added here.
Direct price reading covers stairs of rising peaks and troughs at the simple end, and stop-runs, demand blocks, breaks in structure, and sharp displacement at the other. Passes were 11.7 percent in 19.1 days. Paid months averaged $4,137, ahead of breaks and fades, behind persistent-trend accounts. The best tenth of those payouts was $7,482. A first check reached 6.8 percent, a 25.9 percent lift versus the trend family and a 278 percent lift versus fades. The top 1 percent dollar amount is missing from the caption. He still ranks the bucket second and wants it read lightly, because the habits inside it differ.
How to use it
Tape reading asks who is pushing, where the push is absorbed, and when control slips. He points to volume delta, footprints, and the resting book, calls the work advanced and hard for beginners. Passes reached as high as 15 percent. The average clock was 17 days. Paid months sat near $4,862, ahead of the other four. The best tenth of those months was $8,437 and the best hundredth $15,021. At least one payout reached 9.5 percent of the users. Weakest pass rate to strongest, his order is fade, break, persistent trend, direct price, then tape.
After spread, commission, and slippage, the average trade still has to stay positive on data the rules were not built on. Futures, spot, and CFDs do not hand that edge over.
What usually goes wrong
Loss limits only slow a losing process. Treating the leaderboard as proof that tape reading confers an edge by itself is the other miss. He notes a selection effect. The work is harder, so the people doing it may already be further along, choosier, or more rule-bound than a typical fade trader. In this set they were likelier to be running something that paid. What changed was how often a real edge was found and then traded. The label alone created nothing.
The monthly dollars belong only to people who were paid. The share who both passed and collected is much smaller: 1.8 percent among fades and 9.5 percent among tape readers. Across the whole book, his figure is that 5 percent ever collected.
Questions people ask
Which family cleared a test soonest?
Volatility breaks, at 13.9 days on average, against 25.6 days for fades. Tape reading, at 17 days, had the highest pass rate.
Does one paid month prove the style?
Temperament and loss limits can protect an advantage that already exists. They cannot invent one from randomness. The result still has to stay net positive after costs, over many trades, beyond a hand-picked sample.
Why is the price-reading bucket hard to trust?
It mixes plain structure with stop-runs and displacement. He still places it second, with an 11.7 percent pass rate and a 6.8 percent first-payout rate, and the absent top 1 percent dollar figure stays absent.