Would a $100/month security master be useful for quant research and backtesting?
retained text
We're trying to make a decision about something we've built, and I'd appreciate some feedback from people working with historical market data. We originally set out to build a platform for agentic financial research and backtesting. One of the problems we ran into was accurately identifying securities over time, especially when dealing with ticker changes, reused symbols, mergers, spinoffs, and delistings. We ended up building our own historical security master to solve that problem. It now covers 23+ years (2003–present) of US exchange-listed securities, including: * **35,000+ ticker listings**, with historical start/end dates * **25,000+ security lineages**, tracking ticker changes, symbol reuse, and security histories * **Multiple security types:** Common stocks, ETFs, ADRs, preferred shares, SPACs, trusts, funds, notes, and others * **Corporate actions:** Mergers, acquisitions, spinoffs, splits, dividends, and conversion ratios * **Identifiers:** CIK, CUSIP, ISIN, FIGI, EIN, linked to the appropriate issuers and securities * **Point-in-time history**, so you can determine what existed, what it was called, and how it was related to other securities on a particular date It currently focuses on US exchange-listed securities, so it doesn't cover futures, forex, bonds, or crypto. We've also built several datasets on top of it: * Historical index constituent memberships * Daily historical market capitalizations for thousands of companies * Historical ETF and fund holdings All of this was originally built for our own research platform, but we're considering exposing it independently through REST APIs and MCP servers for use by developers, researchers, and AI agents. **Here's what we're trying to figure out:** If something like this were available for around **$100/month**, with unlimited API/MCP requests under a reasonable fair-use policy: 1. Would this solve a problem you currently have? 2. How are you handling historical security identification and corporate actions today? 3. Would you use something like this directly, or is it functionality you already get from your existing data providers? 4. Is $100/month reasonable for this kind of data, assuming the historical accuracy and coverage hold up? We're not trying to sell anything in this thread. We're genuinely trying to understand whether there's demand for this as a standalone data service or whether it makes more sense to keep it as part of our research platform. Any feedback, especially from people who have dealt with these problems firsthand, would be appreciated.
document fetched 2026-11-04 00:44 · canonical https://reddit.com/r/algotrading/comments/1xn37wh/would_a_100month_security_master_be_useful_for