Original research from our own data: what we tested, what we measured, and what we refused to ship.
Gamma is not the only Greek that generates hedging flow. Charm turns the passage of time into buying and selling, vanna turns changes in implied volatility into the same, and both peak exactly where 0DTE trading lives. A plain-language explanation.
A Weinstein Stage-2 trend sleeve backtested at 616% total return, Sharpe 1.16, profit factor 3.30 and a deflated Sharpe of 0.9997. Every number is real. We shelved it, and the three reasons are more useful than the headline.
Our Wyckoff signal tracker backtests at profit factor 1.37 across 95,319 trades and returns 0.96 live. The obvious conclusion is that the backtest overstated by 43%. We checked, and the two numbers are not measuring the same thing.
Gamma levels are not support and resistance in the usual sense. They are concentrations of hedging obligation. Here is what each level actually represents, how the flip point is computed, and which of these levels we grade as folklore.
Wave counting is irreducibly subjective: the ZigZag threshold determines the count, multiple valid counts coexist, and labels are provisional until hindsight. So our engine never emits one count. It emits ranked candidates with the price that breaks each one.
The at-the-money straddle gives you the market's priced range for an underlying over a defined period, in one number, with no model assumptions beyond the price itself. Here is how to compute it, how to read it, and the four ways it gets misused.
We extracted 201 Gann rules from trading books. None ever reached twenty trades and 116 were skipped as unevaluable. Gann angles need an arbitrary scaling constant, which makes them unfalsifiable. We shipped only the Square-of-9 price ladder, and only as confluence.
Implied volatility is not a prediction the market makes. It is a price, expressed in volatility units, backed out of what options are trading for. Getting that backwards is the single most common error in options analysis.
IV rank measures where volatility sits in its range. IV percentile measures how often it has been lower. One outlier day can push them 70 points apart, and most platforms show only one without saying which.
The bid-ask spread tells you the cost of a small trade. Kyle's lambda tells you what happens when the trade is large: how far price moves per unit of order flow. It is the cleanest single measure of market depth, and it is estimated with a regression anyone can run.
Volume is unsigned. A million shares traded tells you nothing about who was in a hurry. Signing that volume is what turns a volume bar into an order flow reading, and the conventions for doing it are where most of the errors live.
Every options print carries a strike, a size, a price and a timestamp. It does not carry intent. Here is how sweeps, blocks and splits are classified, how aggressor side is inferred, and the four pieces of missing information that make most unusual-activity alerts unreadable.
Wyckoff, Elliott Wave, Gann, Fibonacci, Ichimoku, RSI and MACD. We turned 1,275 book rules from these frameworks into machine-evaluable form and backtested them. Thirteen ever accumulated twenty trades. Gann produced none at all.
We extracted 358 Smart Money Concepts rules and 153 Dow Theory rules from trading books. Two ever reached twenty trades. Meanwhile the deterministic engines built from the same methods raise our buy list from +4.66% to +5.87%. Both facts are real.
Implied volatility exceeds subsequent realized volatility on average, across markets and decades. That gap is real and it is harvestable. It also produces the most misleading equity curves in finance, for a structural reason worth understanding before you read another win rate.
One symbol does not have one implied volatility. It has a surface across strikes and expiries, and the shape of that surface carries information the headline number destroys. Why index puts cost more, and what an inverted term structure means.
A five-minute bar at the open and a five-minute bar at lunch are the same duration and nothing else. Sampling by clock time oversamples quiet periods and undersamples the ones that matter, and it breaks the statistical assumptions most indicators quietly rely on.
VPIN estimates the share of trading that is informed, using volume buckets rather than clock time. It became famous for the 2010 flash crash and it has a serious academic critique. Both are worth knowing, along with the reason a raw VPIN number is meaningless on its own.
Market makers hedge the options they sell, and that hedging buys and sells the underlying. Whether it buys strength or sells it depends on one thing: the sign of dealer gamma. A plain-language explanation of the mechanic, the formula, and the assumption everyone glosses over.
If a trade has a fixed target, stop and time limit, the usual 1/(1+RR) baseline is not the right null. We measured it directly: for our equity setups the true random-entry rate is 27.9%, not 33.3%, and the error grows with barrier width.
We extracted 21,191 mechanical rules from 656 trading books and backtested every one. After correcting for the number of trials, none showed a statistically significant edge. Here is the method, the result, and what it changed.
Across 11,615 resolved futures signals we win 64.9% of the time and average -0.0208 R per trade. Here is the arithmetic that makes those two facts compatible, and why a published win rate tells you almost nothing on its own.
A win rate computed only over closed trades is uninterpretable, and our own book shows why: the two worst positions were still open. This is a worked example of reading a track record honestly, using the only one we can audit line by line, which is ours.