When I first built correlation-based models for a major European bank, I thought I had cracked the code. The regressions lined up, the backtests looked beautiful, and I believed my work would help steer billions into better investments.
Then came the question that changed everything:
“Why would that relationship hold? Correlation isn’t causation, right?”
I stumbled. I didn’t have a good answer.
That was the moment I realized how easy it is, even for someone with a scientific background, to stop thinking like a scientist once inside finance. We fall in love with prediction. It’s seductive: quick to build, easy to backtest, and impressive on slides. But prediction is fragile. Correlations come and go with the market’s tides.
Science adds something deeper. It asks: Does X cause Y? And under what conditions? That shift flips the script. Prediction shows patterns; science uncovers mechanisms. And mechanisms endure.
Other fields have already made this turn. Development economics embraced randomized trials. Econometrics developed tools to find causal signals in messy data. Tech giants run thousands of experiments a year. Finance is still catching up.
For firms, bringing science back means testing before scaling, rewarding curiosity, and designing structures that learn fast. For markets, it means meeting the growing demand from long-term allocators and regulators for real explanations, not just quick signals.
And for quants, it’s about career survival. Prediction-only careers shine bright — and burn out fast. Scientific quants, by contrast, adapt. They’re harder to automate away, more resilient across cycles, and ultimately more valuable.
The market doesn’t need more predictors. It needs more scientists. And the analysts who rediscover their scientific mindset will be the ones still standing when the next cycle ends.
