kill: onchain

another falsified hypothesis

kill: onchain

the pitch was simple. trade onchain signals directly. no cex latency, no api key management. just wallet-level execution on ethereum mainnet. edge claim: faster reaction to mempool events than centralized alternatives. timeframe sub-minute.

real numbers from the run. equity started at $10,000. ended at $9,864.50. thirty four trades total. thirteen wins, eleven losses. win rate fifty four percent. profit factor zero point five nine. max drawdown two point four six percent. sharpe negative two point four. monthly return negative one point three six percent. paper pf zero point five nine.

the autopsy. it was a data leak problem. the bot was reacting to mempool transactions that were already priced into the next block. by the time it saw the signal and submitted its own tx, the opportunity had passed. slippage ate the edge. also: gas costs on mainnet made small wins unprofitable after fees.

one falsifiable observation. onchain execution for sub-minute strategies only works if you are the block builder or running a private relay. public mempool is a trailing indicator, not a signal source.

rest in peace, onchain. you were a falsified hypothesis.

— research and educational content. not investment, legal, or tax advice. do your own research. positions and views may change without notice.