We ran a symmetric version of the strategy on bitcoin alongside the one on ETH. Same signal, same structure, a second market. It looked reasonable for a long time, which is the whole problem with this story.
Then we put it through the same audit as the ETH book: the full set of configurations our parameters can produce, corrected for the size of that search. ETH cleared the bar. Bitcoin did not, by a wide margin. The two numbers are in Letter 001.
The leg closed on 1 August. Not paused, not reduced. Closed.
The bitcoin book had an attractive optimum. There was a configuration that looked better than anything on the ETH side, and for a while we treated it as the more promising of the two.
It turned out to be an artefact of the lens. Measured in dollars it looked excellent; measured in the coin, which is the denomination the account actually lives in, it was ordinary. Nothing about the trading had changed between those two views. Only the unit of account had.
We had built a view of a market on a number that existed because of how we were looking at it.
Nothing was visibly broken. The book was not losing money at the moment we shut it. There was no incident, no drawdown, no bad week to point at. The only thing that had happened was that we measured it properly and it did not pass.
Shutting down something that is not currently failing requires the decision to rest entirely on the measurement, with nothing in the present to corroborate it. That is a different act from cutting a loser, and considerably less satisfying.
One account, one asset, one strategy. Bitcoin still has a role, but as a thermometer rather than an instrument: it tells us something about the weather without our having to hold a position in it.
A fund that adds an instrument every quarter grows wider. We chose to grow deeper, and we admit that is the less photogenic option.