<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>TradeStack Journal</title><link>https://tradestacks.dev/blog/</link><description>Engineering and evidence-led paper-trading research.</description><language>en-IN</language><item><title>Inside TradeStack: Engineering a Rule-Based Options Research Engine</title><link>https://tradestacks.dev/blog/building-a-rule-based-options-research-engine/</link><guid isPermaLink="true">https://tradestacks.dev/blog/building-a-rule-based-options-research-engine/</guid><pubDate>Thu, 19 Nov 2026 00:00:00 GMT</pubDate><description>How TradeStack moves from Upstox market data to strategy decisions, risk checks, simulated positions and auditable research without making profitability promises.</description></item><item><title>Recording Market Sessions with DuckDB, Parquet and a Simulation Clock</title><link>https://tradestacks.dev/blog/market-data-replay-with-parquet-and-duckdb/</link><guid isPermaLink="true">https://tradestacks.dev/blog/market-data-replay-with-parquet-and-duckdb/</guid><pubDate>Thu, 19 Nov 2026 00:00:00 GMT</pubDate><description>How TradeStack approaches tick recording, reproducible session replay, depth-aware fill simulation and the limitations of backtesting fidelity.</description></item><item><title>A Smaller Replay Loss Is Not Proof of a Profitable Strategy</title><link>https://tradestacks.dev/blog/what-a-smaller-replay-loss-really-means/</link><guid isPermaLink="true">https://tradestacks.dev/blog/what-a-smaller-replay-loss-really-means/</guid><pubDate>Thu, 19 Nov 2026 00:00:00 GMT</pubDate><description>TradeStack's October 16 controlled replay compared two entry-quality filters with a baseline. Here is what the reported numbers mean—and what they cannot establish.</description></item></channel></rss>