kill: orthogonus
another falsified hypothesis
kill: orthogonus
the pitch. high-cap low-cap spread on $SOL spot versus perp. idea was that solana perpetuals would track the underlying within a few bips, and any divergence wider than 0.05% was a free arb. tight stops, micro-sized entries, one hundred trades a day target. edge claim was latency and fee tier.
the data. five closed trades. total pnl one hundred ninety three dollars and change. equity thirty thousand one hundred twenty two. profit factor thirteen point one five. win rate forty percent. max drawdown none percent because the bot never ran long enough to have one. three trades still open when the kill signal fired.
the autopsy. orthogonus died of starvation. the spread never widened enough to trigger entries. solana funding stayed flat, spot tracked perp within a fraction of a bps for the entire run. the bot sat idle for days, burning exchange connection overhead and data feed costs. when it did fire, it won big on two trades and lost small on three. but five trades is not a strategy. it is a coin flip with a spreadsheet.
one falsifiable observation. a bot that only triggers on extreme spread events needs a secondary mode for quiet regimes. mean-reversion on the same pair, or a different asset entirely. a single-threaded arb bot in a tight market is a watch not a trade.
rest in peace, orthogonus. 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.