My ideas revolve around options trading. I've got experience both as a developer and a quant in the industry. Can tell you that beating the market is hard, but not impossible.
So far I've got the following:
- 20 years of historical end of day data and a running subscription to current end of day data for all optionable instruments in the US (stocks, indices, etfs). Some 8000 underliers and 1,000,000 option quotes per day.
- A software for researching and backtesting options trading strategies. Example: you can test plain Black-Scholes delta-hedging using historical volatility against the market-implied volatility. Spoilers: it doesn't work :) You can test option strategies: spreads, iron condors whatever: problem - you need some "indicator" criteria for entering into such positions, and spoiler alert: nothing seems to work reliably.
- Several option pricing models and associated strategies in tests and research. The ones in tests have passed research phase and now I'm running them in paper trading accounts on real market data. Will see. There are others also in the research pipeline.
- Under development / improvement: a live, fully automated trading system connecting on one hand to an Interactive Brokers account and on the other, to other brokers accounts. Don't wanna give much detail publicly, it's a tough business and competition is rough.
What I'm looking for is one or several other quant developers like me, to associate / collaborate on R&D. Based on my experience there are two directions of making money: either playing under the current market rules and developing better models than the big firms already have (very tough), or establishing a whole new market - think of it similarly to how software industry recycles the same basic constructs by constantly launching "a new language" / "framework" / "platform" every few years. Like new software languages, a new financial market is supposed to solve some problems of the established one. You develop it privately, know the fundamentals, write the software, put all systems in place and then launch it (hoping it catches, like programming languages, there's an element of chance involved). If it does, you've got a headstart.
So message me if you are interested and be realistic and realize that it often takes a very long time (months to years) to form mutually beneficial alliances. Perseverance and patience will be very important.
Make sure that you actually want to do stuff (together), not just talk. Ranting against the world or pulling ideas out of a hat is easy. Doing stuff while involving others is extraordinarily harder.