Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
4,613 documents
The article describes practical TradingView workflows for discretionary and early-stage trading research. The author uses watchlists to monitor macro instruments and track a basket of factor ETFs through a synthetic instrument weighted by units held. A…
This article develops intuition for using convex optimisation to turn forecasts into portfolio positions under practical constraints. It begins with a long-only, unlevered return-maximisation example, then adds existing holdings and transaction costs to show…
This short reflection argues that machine-learning and statistical methods are tools for analysis, not trading edges by themselves. It offers questions for assessing whether a discovered effect has a plausible structural, economic, financial, or behavioral…
This article examines how a retail-sized account can trade a broad futures universe when positions must be whole contracts. A diversified portfolio performs well in a fractional-position backtest, but integer rounding prevents the smaller account from…
This outline describes a study of trading an equity curve: reducing a system’s exposure after weak performance and restoring exposure when a simulated account recovers. It frames the approach as an overlay with separate rules for detecting poor performance,…
The post compares a stateless trend-following approach with trade management that changes as a position develops. It describes a test system using a moving average signal, volatility-scaled positions, and stop losses. Dynamic volatility control resizes…
The document describes how to add a risk overlay to a systematic futures strategy and where to place it in a process that uses dynamic position optimization. The overlay scales unrounded target positions by a multiplier, while separate controls address…
The document considers whether volatility targeting improves a trend-following strategy. Its motivating example is a long position that gains as price rises but is reduced because the position’s risk has increased. The author notes that trend following often…
The document outlines a discretionary, spreadsheet-friendly approach to constructing long-only portfolios or allocating among trading strategies. It emphasizes choosing inputs that are easier to estimate and interpret, especially Sharpe ratios and…
The document describes a systematic way to select a fixed subset of futures markets for an account with limited capital. It first filters for liquidity, then estimates each instrument’s expected trading costs and the penalty from contract sizes that prevent…
The document examines how small account size limits diversification when futures positions must be held in whole contracts. This creates abrupt position changes as forecasts, volatility, or account value shift, which can misalign risk targets and raise…
The document describes a systematic overlay for reducing a trading system’s positions when estimated portfolio risk rises above chosen limits. It starts by comparing realised portfolio volatility with expected risk and argues that expected risk can vary…
This document examines maximum drawdown as a way to set capital or risk targets. It frames drawdown in relation to annualised volatility, Sharpe ratio, and the length of the measurement period, and compares drawdown-based sizing with Kelly-style risk…
The article describes the components and operating choices of a long-running automated futures system. Its rules combine trend and momentum signals, breakouts, carry, relative signals, skew, acceleration, and mean reversion. Forecasts are weighted across…
The document presents a static optimization approach for choosing tradable futures positions when a small account cannot hold fractional target weights. It minimizes portfolio tracking error relative to an ideal target, while also accounting for trading…
The document frames stop losses as one part of a broader risk process. It describes a trailing stop that moves upward as a position reaches new highs, with the aim of limiting the amount of accumulated profit that can be given back. The examples are…
This article considers whether a trading system should maintain a fixed expected portfolio risk or allow risk to vary with signal strength. The author's system targets a long run average volatility, while its daily expected risk varies with both aggregate…
The article examines a performance measure based on the highest geometric return achievable at a strategy's optimal leverage. Under Gaussian returns and unrestricted leverage, it explains why Sharpe ratio can determine the preferred strategy, while a nonzero…
This document lists recurring trading errors, including failing to define a system, abandoning stop losses, risking too much capital, setting stops without sound money management, relying on fixed profit targets instead of trailing stops, trading…
This article uses bootstrap resampling to examine uncertainty in portfolio statistics and allocation decisions. Resampling observed returns with replacement creates alternative histories and a distribution of estimates, rather than a single point estimate.…
This article explains how bootstrap resampling can represent uncertainty in estimates used for portfolio decisions. Instead of relying on one calculated mean or on returns generated from an assumed distribution, it repeatedly samples observed returns with…
This article describes a simulation framework for estimating the probability, time, and expense of passing two proprietary trading firm evaluations. It compares challenge rules, including profit targets, daily loss limits, drawdown constraints, minimum…
The document examines whether improving volatility forecasts is worth the effort when volatility estimates are used to scale trading positions inversely. It contrasts basic estimates based on recent realized volatility with more involved approaches,…
The document considers how to draw regular income from a trading account while accounting for uncertainty in the estimated Sharpe ratio. It outlines three ways to represent that uncertainty: a distributional formula under a Gaussian return assumption,…