Posts

The CFTC will publish the Commitments of Traders report twice per week from the end of 2026. The series keeps its name, but the day it describes, the delay before it's public, and the number of rows in a year all change. A backtest built on the weekly report has been testing that structure as well as the rule.

Lou, Yan, and Zhang found that Treasuries cheapen in the days before an auction and recover after it. In Treasury futures from 2010 to 2026, the daily gaps around the late-month auctions line up with the month-end rally, and a trade on the month-end calendar alone did better. Intraday, the futures still move by about half a basis point of yield around the close of bidding, but no intraday trade we tested shows a reliable profit after a tick of cost per round trip.

We took ten things traders take for granted, from gold as a safe haven to copper as a guide to stocks, and tested them on 16 years of data in 19 futures markets.

Traders treat round prices as support and resistance. In 16 years of one-second data in six CME futures, price that reaches a round level goes through it more often than price that reaches an ordinary one, and least often turns back one tick past it. Approaches to round levels also seem to stall short of them a little more often, though that result is less firm. We checked the findings tick by tick, against placebo levels, and a dozen other ways.

A backtest that runs back to 1926 only helps if what it measures existed for the whole period. We tested which edges lasted through changes in the market and which didn't, using a century of daily US returns, 16 years of ES intraday data, the closing-auction seconds that moved in 2018 and 2019, and the fee that widened ETF spreads in April 2026.

A continuous futures series is a chain of expiring contracts joined by somebody's rules. Here's what those rules are, what goes wrong when you take them on trust, and how to build and check a series of your own.

QQQ and SPY jump at 15:50:00 and 15:55:00 just as the index futures do, SPY included, although its own exchange's auction runs to a different timetable. Eight years of Nasdaq trades show where the ETFs match the futures and where they part company: at the close, after the bell, and since April 2026 in what it costs to trade them.

Index futures jump on a timetable. Peering into eight years of NQ, ES, YM, and RTY bars shows that the closing-auction clock makes 15:50:00 and 15:55:00 the most violent seconds of the regular session, and other time-related observations.

Development is an architecture problem before it is a model problem. Most traders are wondering about the best LLM to use. That's the wrong question to ask.

Many short-term trading rules need facts a daily bar throws away. We coded 16 of them and five of Jim Dalton's, ran them on 16 years of one-minute data in ES, crude oil, Treasury notes, the grains, and SPY, and report what held up and what a daily-bar backtest can't see.

About 4,000 articles from Technical Analysis of Stocks & Commodities, 1982 to 2026, read as a single dataset: what the field keeps rediscovering, what it has never tested, and where the uncrowded ideas are.

Equity data is reported, under rules written in 1975 and rewritten in 2005, and is changing again on 6 December 2026.

The hidden decisions in your market data can change your signals, your position sizes, and what your backtest appears to prove.

A blind panel of four frontier language models each read 468 academic trading papers and independently converged on the same short list of futures-tradable ideas. This sets out what they agreed on, where they split, and what survived an adversarial fact-check.

Many aspiring traders are told that using daily bars is good enough for all but short-term intraday strategies. We tested the guess many backtesting engines make about the path inside a daily bar against 16 years of one-second data in six futures markets. It gets the order of the high and low right four days in five, as it would on a random walk, but for a stop and a target near the entry it's close to a coin toss, and it flatters the tight-stop, wide-target brackets an optimiser is drawn to.

Every serious quant trader is likely using AI now to research and build models. Mastering how your coding agent propagates knowledge you build up is essential.

One of the most fundamental decisions that aspiring algorithmic developers make is that of which platform to choose. This decision is more consequential than most developers realise.

1. Almgren, R., & Chriss, N. (1999). Optimal execution of portfolio transactions. Journal of Risk, 3(2), 5--39.

The current state of retail backtesting practice is characterised by a significant gap between the apparent sophistication of the tools and the actual rigour of the analysis they produce.

Based on the foregoing analysis, I propose the following minimum standards for a backtest to be considered credible evidence of a potentially viable trading strategy.

Novice and retail trading system developers suffer from what might be termed the signal isolation blindspot: an overwhelming focus on a single dimension of the problem, generating a profitable entry…

Off-the-shelf backtesting frameworks are not neutral tools.

Python's dominance in retail algorithmic trading is a sociological phenomenon, not the result of a deliberate technical evaluation.

Traders who are aware of the overfitting problem described in the preceding section often turn to a family of validation techniques (Monte Carlo simulation, synthetic data generation, and…

Harvey, Liu, and Zhu (2016) demonstrated that the threshold for statistical significance in backtested strategies must be adjusted for the number of strategies tested.

Every OHLC bar, regardless of its timeframe, is a lossy compression of the true price path. A bar records the open, the high, the low, and the close, but nothing about the trajectory between them.

The most fundamental question in any backtest is deceptively simple: would this trade have been filled, and at what price?

A backtest is only as reliable as the data on which it is built, and the quality of freely available and even commercially distributed historical market data is far worse than most traders realise.

The barriers to entry for quantitative trading research have fallen dramatically over the past decade.