The ForecastOS Super Six: What We Believe

Welcome to Weekly Roundup #14 at ForecastOS! We're glad to have you here, following our journey :)

This week we:

  • Released a landing page for our paper portfolio functionality; you can use it to run a third-party validated backtest out of sample
  • Started building ForecastOS Maverick, our out-of-the-box quantitative investment strategy for asset managers that want to own their own IP
  • Started building ForecastOS FeatureHub, our third-party feature store
  • Began redesigning and rebuilding our marketing / landing pages

We also spent time thinking about why we are building what we are building. The ForecastOS Super Six (not just a fancy road bike) are the 6 core beliefs that drive much of our product development efforts. Our core beliefs, in no specific order, are:

  1. Institutional investors are unwilling to have platform lock-in. They want to own their IP.
  2. Many funds do not have quantitative/systematic investment strategies. They’d like to, but lack the skills, budget (+US$1MM/year), or initiative to develop and maintain an institutional-grade quant strategy.
  3. Forecasts can be improved through hyperparameter tweaking, but gains are limited and mostly one-off. Recurring gains can be made by adding new sources of data/insight/signal. Therefore, to improve forecast performance, focus should be on new dataset/signal evaluation, acquisition, and integration.
  4. Most technical quant investor time is spent doing feature engineering and research. Anything that reduces the time required to be spent on either, or eliminates it, is wildly valuable.
  5. AI is where the puck is right now. It is becoming commoditized. For most use cases, open-source AI is as good or better than proprietary AI algorithms. Data, not AI, is where the puck is going. The value is in the data, and the insights therein, which makes AI performant. The best investors, or forecasters in any context, onboard and understand new sources of predictive data the fastest.
  6. Data providers are bad at estimating what their data is worth. Helping them answer this question in a client-specific context is valuable, as they can use this information to better price their offerings.

Do you agree with our beliefs? Let us know!

Work: What's Coming Next

It's important to keep velocity high. We keep ourselves accountable by sharing what we hope to finish over the next week.

This week we will:

  • Build ForecastOS Maverick, our out-of-the-box quantitative investment strategy for asset managers (ongoing, March 2024 ETA)
  • Building ForecastOS FeatureHub, our third-party feature store (ongoing, March 2024 ETA)
  • Redesign and rebuild our marketing / landing pages (ongoing)

Until next week!


Subscribe to The ForecastOS Blog

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
[email protected]