The idea of “alpha capture” is not new. SumZero’s community has created a large data repository that can act as a standalone, crowdsourced alpha capture system. The SumZero Data Feed can also augment and improve existing, internal alpha capture systems. This whitepaper show you how we’ve built various AI model selections that aim to help investors—both quantitative and fundamental—identify outperformance derived from fundamentals-driven investment research.
The industry has reached a period of stasis that siphons wealth to a handful of entrenched individuals and it needs to stop. It's not so difficult to attribute value creation in the investment process, it's just that existing power systems have no reason to actually implement that attribution. Luckily, properly measuring and rewarding contributors of a system are what cryptoeconomic incentives are all about. And if you haven’t realized it yet, cryptoeconomic incentives are a big (real) deal.
The third wave of quantitative investing isn't about chasing milliseconds of noise, but identifying decades of compounding value. The edge now lies in pairing machine learning with fundamental value principles and longer term investment horizons. We're not training algorithms to capture market noise—we're teaching them to recognize business quality that endures.
The field of Natural Language Processing has had a Language Modeling Renaissance as of late and if you haven’t yet heard about it you should pay attention, because it’s about to change the world. At its core, a language model has a very simple objective: learn how to predict the next, most appropriate word in a sentence or document (or other slight variations of that task). At CrowdCent, we use language modeling as one of many tools to enable our investment decision-making pipelines.