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.
1,033 documents
This paper documents RED-2400 family v2, a collection of five public benchmark datasets intended to support research on Solana-native and decentralized-finance market microstructure. Collected during a concurrent 57-day period using public and free data…
This study addresses a variance double-counting problem in synthetic multi-asset equity data. When full asset-return generators are combined with a market factor, the simulated returns can include market variance twice. The proposed correction centers and…
The document develops a trading strategy with trailing stops under the assumption that currency exchange rates behave in a mean-reverting way similar to Ornstein–Uhlenbeck processes. It derives an explicit expression for the running maximum of an…
This paper proposes a two-factor asset-pricing model for economies where firm capitalizations have a sufficiently heavy-tailed distribution. Alongside conventional market risk, the model includes diversification risk, proxied by an equally weighted…
This work examines how a centralized trading desk should handle incoming orders when future order flow is uncertain. The desk can warehouse orders in hopes of offsetting them against later opposite-side flow, or send them to the market and pay costs…
The paper studies order flow imbalance (OFI), an indicator based on the asymmetry between bid-side and offer-side events that is associated with short-term price direction. It uses Hawkes processes to estimate OFI while modeling lagged dependence between bid…
This paper develops a method for timing both entry into and liquidation of positions in mean-reverting price spreads. It frames the decisions as sequential optimal stopping problems and explicitly includes transaction costs. A refined signature-based optimal…
The document presents momentum risk-taking as real-time management of short-term equity volatility. Its proposed system turns a theory of how news affects share prices into discrete tables of expected returns across investment horizons, which then guide…
This study tests for a day-of-the-week pattern in daily returns of the NASDAQ Biotechnology Index over a 16-year period, from January 2002 through December 2015. It reports that returns were lowest on Mondays and significantly higher on Wednesdays,…
The paper develops a model of multiple assets traded by investor groups with different strategies: some follow trends while others base decisions on fundamental value. It extends behavioral finance frameworks by representing finite cash and shares,…
The document presents a decade of evidence on applying trend-following strategies to cryptocurrency markets, beginning with Bitcoin’s early market history. It argues that crypto markets share characteristics with historical commodity markets and reports that…
This work studies execution and statistical arbitrage in constant product markets, a common automated market maker design used by decentralised exchanges. In these markets, a trading function determines exchange rates. The analysis treats exchange-rate…
This study presents SimStock, a framework for identifying similar stocks from time-series behavior rather than relying only on conventional sector or regional labels. It combines self-supervised learning with temporal domain generalization to form…
This note introduces Eigenrisk Parity, a portfolio construction approach that allocates risk evenly across the principal components of a covariance matrix. The broader idea is to use symmetry rather than rely on precise forecasts of returns and correlations.…
This study examines whether publicly displayed trading intentions reduce execution costs, using protocol-native TWAP orders on Hyperliquid as an example of “sunshine trading.” It compares visible TWAP executions with hidden metaorders reconstructed from…
This paper uses European put options to construct a p-index risk measure for evaluating stock investment strategies. The index represents the insurance cost per insured dollar needed to ensure a minimum return on a specified future date. The study applies it…
This paper examines how investors’ beliefs and asset prices may be distorted when their learning model understates structural breaks in factor risk premia. It proposes a minimal Bayesian framework in which this misspecification creates persistent forecast…
This study applies deep multi-task learning to diversified time-series momentum portfolios in continuous futures. Its central idea is to learn portfolio construction and volatility-related tasks together, rather than treating the momentum signal and…
This study develops a day-ahead bidding and operating policy for a hybrid plant that combines wind generation with an electrolyzer. The policy sets power bids and schedules hydrogen production using contextual information, with linear decision rules learned…
The paper proposes an adaptive Temporal Fusion Transformer approach for short-term cryptocurrency price forecasting. It addresses non-stationarity and sharp price variation by segmenting the series into variable-length subseries that end at relative maxima:…
The paper proposes a Structural Matrix Autoregressive model to study asset returns, realized volatility, and trading volume together in a large-dimensional setting. It is designed to capture both dynamic spillovers among the variables and dependence across…
The document presents an ecological alternative to equilibrium-based accounts of financial markets. It treats the wealth allocated to a trading strategy like the abundance of a species, then studies a toy market populated by value investors, trend followers,…
The document develops a volatility forecasting framework for binary prediction markets, where prices represent bounded probabilities and contracts resolve at known deadlines. It combines a Wright–Fisher component to model how uncertainty is forced to resolve…
The study examines whether better probabilistic day-ahead electricity price forecasts translate into better battery trading outcomes. It identifies two limitations of quantile-based trading strategies: they do not reward honest probabilistic forecasts, and…