Notes
Reference, not a feed.
Technical notes on execution, research method and derivatives. Each is written to stay correct rather than to be current, so there is no chronology here and nothing goes stale.
They share a common subject: the gap between what a backtest reports and what a market does, and the specific mechanisms that open it.
01 / Execution
Execution
What it costs to turn a decision into a position, and why the components behave differently.
- What a backtest gets wrong about fills A backtest reports the price you asked for. A market reports the price you got. Almost every way that gap opens up is systematic, and almost every one of them flatters the result.
- Slippage and market impact are not the same cost One is a measurement against a benchmark you chose. The other is a price change you caused. Collapsing them into a single basis-point assumption hides the only part of the cost that responds to how the order is worked.
- Latency and slippage are different problems Latency is a delay. Slippage is a price difference. Delay produces a price difference through adverse selection, which is why the two get conflated and why the wrong one usually gets optimised.
- Partial fills are a state machine, not a cost adjustment Most backtesters model an order as filled or not filled. Live, an order is a small state machine, and the branches a backtest never visits are the ones that fail first in production.
02 / Research method
Research method
Data discipline and evaluation. Mostly about the ways a result can be true in sample and meaningless out of it.
- Point-in-time data and the shapes lookahead takes Lookahead is rarely a line of code that reads tomorrow's price. It is usually a dataset that quietly knows how the story ended, and the correction is a property of the data rather than of the model.
- Walk-forward validation for intraday strategies Intraday data offers many observations and few independent ones. Walk-forward is the right frame, but the details that make it honest are the ones most often skipped.
- A high win rate is not an edge Win rate describes how often a strategy is right. Expectancy describes whether being right pays. They are close to independent, and strategies are routinely selected on the first while being evaluated on the second.
03 / Derivatives
Derivatives
Second-order risk, positioning estimates, and the difference between describing a market and forecasting it.
- Gamma exposure: signal or description? A dealer gamma estimate is a restatement of positioning, computed from a chain everyone can see. Whether it predicts anything is a separate question from whether it describes something, and the two are routinely merged.
- Why gamma walls do not produce edges A large open interest strike is a real feature of the option chain. The step from there to a price level that attracts, repels or holds is where the reasoning stops being supported.
- Modelling 0DTE options execution An option with hours to live has gamma that dominates every other exposure, a spread that is wide relative to its premium, and a decay profile that makes holding period a first-order decision rather than a detail.