Public research brief
A better risk estimate can improve compounding. It cannot turn a large leveraged loss into a small one.
Bongiorno and colleagues compress a neural portfolio model and train it to minimize realized variance. The economic idea is straightforward: leverage magnifies the compounding penalty from volatility, so reducing variance can be valuable even without forecasting which stocks will rise.
The interesting question is how much of that improvement survives realistic portfolio construction—and whether the remaining loss is acceptable.
Full analysis on Substack
Follow the evidence further.
The complete review examines the findings, implementation limits, and research questions beyond this public brief. Access and subscriptions are handled on Substack.
Continue reading on Substack ↗View subscription optionsAI assists curation and drafting. The research status distinguishes paper reviews from house tests. Read our research approach.