This is a summary of links recently featured on Quantocracy as of Saturday, 05/30/2026. To see our most recent links, visit the Quant Mashup. Read on readers!
-
How to Build a Reliable Algo Trading Infrastructure [Concretum Group]More and more traders are using Claude Code, ChatGPT, Cursor, and other LLMs to build and automate their trading systems. It works. You can go from strategy idea to a working bot in a day. The code compiles, the backtest looks good, orders fire on paper trading, and you move to production. Then stuff breaks. Not the strategy logic – the infrastructure around it. Over the years, weve repeatedly
-
Trend following (1/4): Replicating your own program [Beyond Passive]The published literature on trend-following replication treats the program being copied as a black box. When the program is your own, this is the wrong way around and fixing it changes the result more than I expected. The story of trend following as a systematic strategy reaches back to the 1970s, when a handful of futures traders observed that prices in commodity markets tended to persist in
-
When Short Sellers Create Overnight Alpha [Concretum Group]Last week, we shared some findings of an intraday short-selling signal taken from our internal research archives. Today, picking up on the same theme, we present some evidence behind an effect we believe stems from the very presence of short sellers in stocks with the same characteristics highlighted in our previous piece. We recommend you first read our original analysis here. Identifying Stocks
-
Trend-Following Filters Part 10 [Alpha Architect]Two previous articles, Trend-Following Filters Part 7 [1] and Trend-Following Filters Part 9 [2], examined, from a digital signal processing (DSP) time domain perspective, digital filters commonly used by technical analysts to aid in making trading decisions. The filters examined in Part 7 include moving average (MA), linear weighted moving average (LWMA), and exponential
-
The Sharpe stability ratio of trading strategies [Macrosynergy]The Sharpe stability ratio measures the consistency of risk-adjusted PnL value generation. It divides the mean Sharpe ratio over sequential overlapping lookback periods by its estimated standard error. Thereby, it quantifies significance and intertemporal stability. Both are critical for selecting factors and for assessing the commercial viability of a strategy. If two strategies produced the same