When Not to Trade

When Not to Trade

The discipline of sitting out

About this tag

Portfolio risk is what you carry in aggregate, not what you carry on any single trade. A stop loss caps the downside of one position. It says nothing about how that position behaves alongside the other eight when everything moves at once. The gap between perceived exposure and actual exposure is where portfolios quietly come apart - individually sized trades that add up to one undiversified bet.

The accounting is where it goes wrong. An allocation sized at a small fraction of the book at entry can drift silently to a much larger share as surrounding positions bleed out. Duration compounds the same way: every hour a trade stays open, the book has more committed to that one outcome. Neither drift shows up in per-trade stops. Both show up in the total when it is too late to rebalance cleanly. The quiet positions - stablecoin farms, hedged structures, low-leverage corners of the book - are often sized small precisely because nobody modeled what they look like under stress.

This tag collects observations on risk measured at the book level rather than the trade level. Exposure mismatch between what the position log says and what the whole book is actually doing. Diversification counted by tickers rather than by behavioral independence. The difference between a portfolio that stress-tests each line item and one that stress-tests the book as a single entity - and why those two exercises produce different answers.

The framing is structural, not directional. The point is not which position will fail but what the whole book does when the one assumption nobody questioned gets answered for them. Notes here document where aggregate exposure accumulates, how book-level heat builds unnoticed, and why the forgotten position is usually the one that decides the outcome.