Angana Jacob - Data as the True Competitive Moat (S7E26)

Bullish:

• $ALTDATA - Data has become the true competitive moat in quant investing as models have been commoditized. The ability to properly source, clean, and connect data is where alpha now lies.

  • “I think the next phase of alpha capture is probably going to come from getting the data properly, getting the data right. It's no longer models alone. It's the quality of the data, the pipelines, and the platform that determines if opportunities can be exploited.”

    Angana·5:14

  • “I believe a meaningful amount of alpha remains uncaptured in data, in data foundations and infrastructure.”

    Angana·5:29

  • “Today that advantage is gone. I think models are getting democratized. TensorFlow, Pytor, standardized risk models, optimizers. Everyone has access to roughly the same toolkit. Academic ideas get published and show up on GitHub almost immediately.”

    Angana·11:01

  • “So where does the real competitive edge come from now? I think from data.”

    Angana·11:26
• $BBG - Bloomberg has a massive competitive advantage in data infrastructure with their ability to provide clean, standardized, point-in-time correct data across all asset classes and systems.

  • “One advantage we have at Bloomberg is that the scope question is largely solved for us upfront. We already sit on an enormous breadth of market data, fundamentals, alternative, all the things we spoke about across asset classes.”

    Angana·16:45

  • “And I think this is where Bloomberg's strength of data and having the same data really matters. The same underlying market reference fundamentals data flows through research environments, the terminal which portfolio managers and traders rely on, and the real time B pipe.”

    Angana·28:47
• $MLINFRA - Machine learning infrastructure and AI-ready data platforms are becoming critical as quant strategies become more ensemble-driven with hundreds of thousands of signals.

  • “What good looks like today for especially at the top end sophisticated platforms is very ensemble driven research. Hundreds of thousands of signals each tied to specific datasets, each with known assumptions and failure modes, each evaluated independently. Machine learning is common to size signals, combine factors, set weights.”

    Angana·31:01

  • “AI obviously requires data, but it isn't simply about more data. It's context aware domain specific data. So the focus for all of this is not just on compute, storage, and speed. It's also on building platforms so that the data can work together and can be connected to surface deeper signals, better optimal portfolios.”

    Angana·33:12

Bearish:

• $AIMODELS - AI models themselves are becoming commoditized as the real value shifts to the data infrastructure beneath them. Models without proper data foundations will fail.

  • “Fifteen years ago, models were definitely the edge. If you had a clever factor model, a proprietary portfolio optimizer, some custom risk framework, that alone could differentiate you. You could hire a few quants, build something bespoke, and you actually genuinely had alpha. Today that advantage is gone.”

    Angana·10:45

  • “I think two firms can run the same model, but the one with the cleaner data, better transformations, and tighter feedback loops should outperform. And that's why the magic isn't in the neural net or the XGBoost. It's also in the data pipelines. The competitive edge has shifted upstream.”

    Angana·12:39
• $HFT - High frequency trading strategies are being forced to move to longer time horizons as their traditional edge has been arbitraged away, requiring completely different data infrastructure they're not equipped for.

  • “I think it's quite well publicized, but again, we saw it quite early that the high frequency firms are moving into medium frequency horizons.”

    Angana·44:53

  • “When HFT firms move into medium frequency, now suddenly they need longer histories, macro contexts. They need quarterly fundamentals, very slow moving data. They need a supply chain. They need all of these things that they did not need before.”

    Angana·45:18