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.
3,481 documents
The study applies machine learning to estimate an optimal rebalancing frequency for pairs trading. It uses minute-level prices for 50 large cryptocurrencies from Binance during 2022 and 2023. For each asset pair, a simulated pairs-trading algorithm…
This Chinese equity screening rule combines a turnover range of 3% to 12%, a positive product of the day’s price change and large-order net flow, and revenue in 2021 at least 10% above revenue in 2018. The post presents the combination as a way to pair…
This Chinese-language report summary examines intraday price, volume, and trading characteristics using minute data to build factors and compare their behavior in stocks and futures. The factor families include return-distribution statistics such as realized…
This Chinese-language article curates research papers and books on reinforcement learning (RL) in finance. Its examples span trading and market applications, Q-learning combined with Black–Scholes, option hedging and dynamic replication, portfolio…
The discussion addresses a practical gap between signals based on closing prices and orders that may be impossible to execute at a daily price limit. A trader may see a buy signal when a security opens at its upper limit, or a sell signal when it opens at…
This Chinese-language research summary assesses whether mainland Chinese equities may be nearing a market bottom using three indicators: northbound investor flows, mutual-fund equity exposure, and a turnover-to-volatility measure. It reports that…
The document describes a proposed workflow for using BigQuant to generate planned orders and a separate brokerage terminal to place trades when the account is held outside the broker supported for direct live trading. It retrieves planned orders through…
This brief technical-indicator note defines Breakout of Bar using separate lookback windows for highs and lows. Its upper component subtracts the highest prior high over the selected number of days from the current high; its lower component subtracts the…
The document introduces covariance as a measure of how two assets move together, then outlines a matrix-based approach to portfolio risk. It describes collecting observations for several stocks, calculating each series’ mean, subtracting that mean, and using…
The document explains how a multi-strategy backtest module combines strategies. It applies user-set weights to the return series from each strategy’s backtest and aggregates those weighted series into a single performance curve. The weight therefore applies…
This study proposes using the roughly 42-month cycle in year-over-year asset-price series to identify broad market turning points. It treats price behavior as a combination of trend, cycle, and random components, and argues that year-over-year log changes…
This report surveys equity factors associated with trading frictions and outlines how academic anomaly research differs from investment-oriented factor research. Academic studies seek a suitable pricing model and portfolios whose returns the model does not…
This debugging note follows an attempt to combine a daily top-ten stock selection strategy with exits when a holding gains about 9% or loses 5%. The author reports that `context.options['data']` is a DataSource and has no `positions` attribute, so it cannot…
This support note investigates why a portfolio log showed a sale on one date while the holding still appeared in positions and no sell signal was visible until the following day. The explanation identifies the test universe’s end-date setting: it extended…
This tutorial explains term frequency–inverse document frequency (TF-IDF), a method for assigning greater weight to words that occur often in one document but less often across a corpus. Term frequency captures within-document prevalence, while inverse…
This study explains combinatorially symmetric cross-validation (CSCV) as a way to estimate the probability that a strategy selected as best in a backtest will rank poorly across alternative partitions of that history. The backtest period is divided into…
This study examines a database of 486 patent-derived factors for Chinese listed companies and tests their usefulness in stock selection. It organizes patent information by type, procedural status, metric, and lookback period. Measures include patent counts,…
This report studies whether gaps between consensus forecasts for macroeconomic indicators and their released values help explain subsequent returns across asset classes. It places the signal in an allocation framework that combines long-term strategic…
This research note examines whether Chinese A-share ownership data can help select stocks. It groups the data into shareholder counts, holdings by the ten largest shareholders, institutional ownership, and major shareholder transactions. Most of these series…
The article defines an earnings surprise factor, ESP, as the difference between actual and expected quarterly net profit divided by the absolute expected amount. It describes deriving quarterly analyst expectations from annual forecasts under an assumption…
This research summary considers when event-driven stock signals should be represented as factors within a broader multi-factor model. It cites analyst recommendations, institutional research visits, and restricted-share unlocks as events that may contain…
This brief Chinese-language forum entry asks whether stock-selection backtests using random forest classification and a StockRanker ranking model can be trusted when features were not standardized or treated for outliers. The accompanying tag states that…
The document explains practical definitions for three functions used in the Guotai Junan Alpha191 factor set. Count is illustrated as summing a binary condition over a rolling window, such as counting days when the current close exceeds the previous close.…
This model seeks to rotate between Chinese equity growth and value styles, represented by high and low price-to-earnings indices. It combines six signals drawn from technical conditions, domestic and overseas liquidity, and periods when equities may be…