Forecast Friday 📈📉🤷

Putting my forecasting and analysis skills to the test - introducing the 4Lights Matched Correlation Market Model

Happy Veterans Day!

I’ve done a lot of forecasting in my career. Some of it (>50%) turned out correct; some turned out wrong; some of it was useful in decision-making and strategy; some of it might have been frivolous. However, forecasting financial markets is different — market data is beyond dynamic. My experience in forecasting and prediction is more rooted in slower moving processes.

Regardless, this was a fun and challenging mental exercise that I’m excited to share (and keep sharing). Note: This is for informational purposes only and is not intended to be personal financial advice; be cautious — there is always risk involved with financial decisions!

So how does it work?

I’ll elaborate more on that later, but essentially the model runs through 40+ years of market data (~20 years for newer financial products like the CBOE VIX index) and finds the periods in time most highly correlated to the most recent 25 days of data. The model then uses that historic performance to forecast the next 15 days.

I would be shocked to find out that anything I’ve done here hasn’t already been done many times (literally millions of times per second) by hedge-funds, quantitative analysts, and algorithmic trading computers; however, I’m still excited to see how well it performs!