Alpha explainers
Papers, threads and videos turned into plain language, with buildable strategy ideas.
Regime-Based Portfolio Allocation with Hidden Markov Models and Reinforcement Learning
A new paper labels the market as calm, transitional or crisis with a Hidden Markov Model, then holds whatever has worked best in that regime. In its out-of-sample test, a learned allocation policy beat buy-and-hold SPY on Sharpe ratio (0.83 vs 0.65) with a smaller max drawdown (-23.5% vs -35.7%). We explain how it works and how the same two steps (classify the regime, then switch exposure) can gate trading bots.
Repo deep-dives
Commit-pinned reviews of open-source trading and research repos.
hmmlearn deep-dive: a solid HMM baseline, with a look-ahead trap
hmmlearn is a popular, BSD-licensed Python library for Hidden Markov Models, a common way to label market regimes. Our review finds it a solid research baseline, but its default prediction calls use the whole data series, which leaks future information into backtests unless you compute probabilities step by step. It is also in limited maintenance, so we recommend it for research and benchmarking rather than live bot gating.
Verdict · Selective fitRequest a deep-dive
Want a repo reviewed? Requests are batched daily, up to 3 deep-dives per day.