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 article describes a Chinese-equity factor that looks for unusually large intraday increases in trading amount, then aggregates information around those moments into daily observations. Its rationale is that a volume surge accompanying modest, orderly…
The strategy turns conventional stock-selection rules into Boolean features for a CNN classifier. Its inputs include returns and return ranks, turnover, valuation and profitability measures, money-flow signals, volatility, technical indicators, and recent…
This strategy combines machine-learning stock ranking with trend-based entry filters for a manually selected pool of China Securities 150 constituents. The pool emphasizes larger, liquid companies, often described as core or sector leaders. A model scores…
This report tests thirteen momentum and modified-momentum signals in China’s A-share market. The signals include returns over several lookback periods, turnover-weighted returns, exponentially weighted turnover-adjusted returns, and a long-horizon regression…
This account describes extending a backtest of a Chinese small-cap strategy and finding that returns before 2024 were negative. It attributes the later high returns largely to growth in the Beijing Stock Exchange, whose smaller companies strongly affected a…
The document introduces multi-task learning (MTL), in which one model addresses several related tasks and uses shared patterns and task-specific differences. It suggests that this can improve learning efficiency and predictive accuracy compared with training…
This research summary explores whether statistics computed from intraday stock data can add new signals to established multi-factor selection approaches in China’s A-share market. It evaluates realized volatility, realized skewness, and realized kurtosis as…
The article outlines a Monte Carlo approach to pricing European calls on the China 50 ETF under the Black-Scholes-Merton framework. It divides the life of the option into time steps, simulates the underlying price with normally distributed shocks, calculates…
The document explains alpha as the return remaining after comparing a portfolio’s actual return with the return predicted by the Capital Asset Pricing Model. It gives the formula using portfolio return, the risk-free rate, portfolio beta, and benchmark…
This document outlines a forecast of constituent changes for four major mainland Chinese equity indices: the CSI 300, CSI 100, SSE 180, and SSE 50. It explains the potential trading relevance of index rebalancing: when funds tracking an index must adjust…
This short assignment defines quantitative investing as making trading decisions with data-supported methods instead of relying on intuition. It frames the approach as a way to turn an investment idea into explicit rules and encode those rules in a strategy,…
This meetup listing raises two quantitative research questions: why factors with strong information coefficients and information ratios may produce weaker Sharpe ratios, and why combining individually strong Sharpe-ratio factors can reduce portfolio…
This research summary examines whether macroeconomic and market indicators can predict returns to equity factors, with particular attention to the size and reversal factors in Chinese stocks. It uses IVX regression, presented as a response to endogeneity and…
This summary of a Chinese securities research presentation describes several equity research approaches. It outlines an event-driven system intended to assess whether market events have an effective price impact, alongside a multi-factor risk framework and…
The article contrasts three approaches to Chinese A-share trading. The first focuses on leading, popular stocks in the strongest sectors, holding briefly and using the five-day moving average as a reference for exiting after consecutive closes below it. The…
The document presents factor analysis as a core research step for multi-factor, index-enhancement, and market-neutral strategies. It frames the analyst’s work as evaluating candidate factors before strategy researchers use them, with the quality of that…
The report proposes refining the book-to-price (BP) valuation factor by splitting book value into retained earnings and invested capital. It forms two market-value ratios: REP for retained earnings and CCP for invested capital. The stated motivation is that…
This research summary introduces price-impact bias, derived from five-minute stock order-flow and market data. It compares how similar proportions of aggressive buying and selling move prices. Stocks with stronger upward impact are described as easier to…
This Chinese-language document summarizes the development and adoption of algorithmic trading, describing early automated portfolio trading in the United States, later growth alongside computing, and broader access through commercial trading platforms. It…
The document answers a strategy implementation question: how to prevent a stock from being bought again for a specified period after a prior trade. Its proposed approach is to record the relevant date when a position is sold, then compare that stored date…
This forum post describes a BigQuant backtest that raises a pandas KeyError for the column named buy_signal. The strategy code queries daily stock data and creates a factor field, then the trading handler filters context data using buy_signal equal to one.…
This exchange asks how to make a deep-learning model produce two predictions for futures trading, one for a long view and one for a short view. The response explains that producing multiple outputs for a regression or classification model may require a…
This forum post documents a BigQuant workflow in which LightGBM finishes running, but the subsequent score-to-position step fails because its query orders predictions by a column named score that is absent from the input dataset. The error identifies the…
This forum post describes a historical data extraction discrepancy between BigQuant versions 1.0 and 3.0. The author reports calculating a 60-period regression slope relating stock returns to a benchmark return series over January 2010. In the stated…