Hi HN, I’m an independent researcher (former systems thinking background). I wrote a Python script to test a hypothesis: that financial models often undervalue thermodynamic efficiency (Creation/Negentropy vs Bureaucratic Drag). I tested a simple metric: R&D / SG&A on the S&P 100. I expected noise, but found a statistically significant positive correlation (0.376, p=0.0168). It seems to act as a filter for identifying high-volatility 'growth' engines versus stagnating incumbents.

Code (Python/yfinance/scipy) is open source: https://github.com/gmirsky2/mirsky-ratio-analysis Full methodology paper is on Zenodo: https://zenodo.org/records/17956159

Curious if anyone in quant finance has seen a similar pure-ratio metric used before.