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 discussion considers how to allocate capital to an automated crypto strategy whose backtest suggests substantial drawdowns, and when to scale its exposure or stop trading as its edge may fade. The response emphasizes that changing position size is a…
The document asks whether a time-series model that chooses a fully invested long or short position could be trained by maximizing simulated terminal portfolio value instead of minimizing squared prediction error. It proposes simulating daily positions from…
The document considers how to allocate a fixed portfolio among four hedge funds when their returns are assumed independent and each fund has a purchase limit. It explains that zero correlations do not prevent mean-variance optimization: the covariance matrix…
The document considers how to choose portfolio weights that are proportional to risk budgets divided by asset volatilities, while meeting a target portfolio volatility and other constraints. It contrasts this objective with a risk-budgeting formulation that…
The document argues that risk management can sit beside an ML or AI alpha model rather than trying to interpret its internal features. A separate portfolio controller can translate forecasts or directional signals into feasible positions, accounting for…
The document surveys alternatives to Kelly sizing, focusing on how statistical arbitrage portfolios balance positions. Sector neutrality groups assets by industry, while risk neutrality uses a model of shared factor exposures to help an optimizer construct a…
The document applies the Kelly betting formula to historical results from selling option spreads and asks whether a calculated allocation of about 20% of capital is too high. The response explains that Kelly is aggressive and assumes a well-specified bet…
The document considers how to express portfolio performance as a percentage when daily trading PnL and exposure are available but beginning and ending account asset values are not. It highlights that exposure may change because of accumulated PnL, yet cash…
This document explains how to allocate a market-cap-weighted crypto index when no constituent may exceed a chosen maximum weight. It presents a sequential calculation: start with each asset’s share of total market value, scale that share by the weight and…
This note frames trading under an assumed known stochastic price model as a finite-horizon stochastic control problem. The trader’s state includes capital, current position, price, and model parameters; the model evolves price and parameters using noise. The…
This note explains why conversion factors affect hedge ratios during the trading life of a deliverable Treasury futures contract but do not change the physical delivery quantity. Because the futures price does not move one-for-one with a particular bond, the…
The document presents a two-day example of a pairs trade that goes long one stock and short another, using roughly equal dollar amounts in each leg. It calculates the opening and closing cash flows from the share quantities and prices, then adds them to…
The discussion examines how to tell whether a market-timing strategy adds skill beyond exposure to broad market returns. A strategy that is mostly long during a rising market may appear successful while offering little independent evidence of an edge. One…
The document discusses hedging exposure to oil prices with forward contracts and asks whether one contract per asset is automatically the right hedge. The responses emphasize that the appropriate ratio depends on the objective and on how well the forward…
The document presents a long-short portfolio with a small long position, a much larger short position, and cash held alongside the positions. Its reported portfolio contribution is sharply negative even though the portfolio’s net market value is positive.…
The document addresses an attempt to maximize expected portfolio return at a specified volatility using R's optim function. The proposed penalty based on an if statement gives the optimizer a discontinuous objective, and the answer instead describes…
The document addresses how traders might identify market conditions associated with changes in algorithmic strategy performance. It proposes structural-break analysis as one avenue, naming CUSUM tests on recursive residuals or levels and explosiveness tests…
The document asks how to define a spread for pairs trading after estimating a linear relationship between two asset prices. It compares subtracting the fitted intercept and scaling one price by the estimated gradient with omitting the intercept or simply…
The document distinguishes two portfolio problems: choosing investments and sizing a group of trading strategies. For multiple strategies believed to be profitable and reasonably uncorrelated, one answer suggests allocating capital in inverse proportion to…
The document poses a portfolio insurance problem in which an investor trades a stock and holds the remainder in a zero-interest bank account at a finite set of dates. The stock is modeled as geometric Brownian motion, and the objective is to choose stock…
The document compares risk management for retail algorithmic traders with controls at larger quantitative firms. It describes institutional safeguards as layered: risk teams may set trading exposure limits, while compliance may focus on legal and regulatory…
The document explains how to estimate the units of a EUR/JPY spot short when the account balance is denominated in USD. The key step is to convert the available USD balance into the base currency, EUR, using the EUR/USD exchange rate. In the example,…
The discussion addresses a long only trading system whose estimated Kelly fraction becomes negative after losses or changing estimates. Under the simple Kelly setup, a negative fraction indicates that the estimated expected return for the specified bet is…
The document asks whether a portfolio optimization problem can use an equality constraint on the sum of absolute position weights instead of the usual equality constraint on their signed sum. The distinction is that the L1 constraint fixes total absolute…