The situation
Retail investors can see a fund's past returns anywhere. What they cannot easily see is how a particular mix of holdings would have behaved through past downturns, or what the range of outcomes looks like ten years out. That modelling lives in spreadsheets and professional tools.
StockPAL puts it in a browser, for one person's portfolio, with the arithmetic shown.
What we built
- Portfolios of tickers and weights, with live prices.
- Backtesting against historical data: growth, drawdowns, worst periods.
- Monte Carlo forecasting: thousands of simulated paths and the percentile bands around them.
- Risk-profile allocation: suggested mixes for conservative, balanced and aggressive profiles, and the gap between them and the portfolio as held.
- Throttle-safe market data: the data provider's limits are respected by design, so a busy day cannot lock the product out.
How it is built
Price history is pulled once per ticker and cached; the simulations run on the server and the charts render in the browser. Nothing is advice, and the product says so.
For the engineer reading along: Python and FastAPI, MySQL, numeric work in NumPy and pandas, a token-bucket limiter under the market data provider's cap with cached history so repeat runs cost zero calls, hosted on the shared Axis infrastructure.
What it does today
StockPAL is a small product used by a handful of people, including us. It is here because it shows the numerical and data-provider side of the work: simulations, time series and a third-party API with hard limits.
Running it
Shared infrastructure, its own process, monitored by Watchdog. Not to be confused with similarly named products from other companies; ours lives only at stockpal.axisdigital.ai.
Have something that needs building, or running?
Tell us what you are trying to do. We reply within one business day, and the first conversation is free.