Calderathe social layer for ObsidianGet your own page
Mike Jones.mdGardenFees decide everything
@mike / Garden

Fees decide everything

🌿 buddingPlanted Oct 12, 2026Tended Oct 12, 2026169 words

"Fees always killed this" is the most expensive lesson in short-timeframe trading.

  • Risk to reward beats win rate. At 1:1 risk to reward you need well over 50% wins after costs. At 1:3 you break even at about 26%. Several of my early strategies were roughly 1:1 at under 40% wins: structurally losing, however clever the entry.
  • Measure net, always. Every trade carries the spread at the time it was taken. The dashboard computes net results at query time with adjustable assumptions (spread, commission, slippage, borrow cost for shorts), so changing an assumption never needs a re-run.
  • Filter the universe by cost. On US equities, dropping the widest-spread half of the candidate list cut the average spread by about two thirds and kept most of the volatile names worth trading.
  • Maker vs. taker matters. Posting liquidity instead of taking it changes which strategies are viable at all.

Part of Trading with an agent fleet. See also Scouts and winners and Crypto strategies on US equities.

Links to this note

Trading with an agent fleetI run systematic trading as an experiment in what a fleet of AI agents can do on its own. The agents run on Cerberus…Scouts and winnersThe method behind my Trading with an agent fleet setup, in one line: seed everything as a scout, promote only what…Data sources I use and wantThis is the note I most want people to link to.