Given the minimal costs and storage requirements thus far, I think I'll begin to load in fundamentals as well. Something like computing 50-day moving averages for five years of a security's closing prices just took 00:00:00 to execute (less than a second). That function runs in parallel for each symbol (I'm throttling at 4 instances currently), orchestrated by Logic Apps. For $8 of Azure SQL, I can run a Python function I wrote to load and effectively index 7GB of daily OHLC history (about 5 years of data for 10,000 securities). I can't believe how low-code and low-cost my industry has become! For $4/mo, I have SMS notifications from a dedicated 800 number, letting me know when trading data has been updated in my system. This week, I started from scratch, this time powered by Azure SQL, Communication Services, Function Apps and Logic Apps.

Really fun stuff, but I was never brave enough to put my money where my bollinger bands were.

Ten years ago, I wrote quite a bit of C# and SQL to download stock history and then develop (and backtest) screens against it. Reignited my obsession with #quantitativetrading this past week on #Azure.

Steam box raspberry pi