Both books have an edge in the backtest, and both still had long, ugly stretches. That is the counterintuitive part of an edge: it shows up over hundreds of trades, and inside any few weeks it can look like it isn't there at all. Here are the worst stretches, book by book, so none of them is a surprise.
BACKTESTED AND HYPOTHETICAL
One SPXW unit of each book, Jan 3, 2022 to May 13, 2026, entries at the ask. These are results of a simulation, not of real trading, and they don't guarantee anything about the future. Nothing here is a recommendation. Read the full disclaimer.
It wins 54.5% of its trades, so long losing runs are part of the design. An average loss (−$1,240) is smaller than an average win ($2,364), which is what carries it through them.
It wins 80.3% of the time, so a week of small wins can feel like a sure thing. Then one loss erases most of it. The edge already prices those losses in: they are why it needs to win so often.
History is one path. Phoenix's sessions reshuffled into 10,000 other orders (2021-01-04 to 2026-07-15) hit a drawdown worse than $15,000 in 15.4% of them, so the worst stretch ahead can be worse than the worst one behind.
A card counter's edge is about one percent, and plenty of nights still end down. Quitting after one of those nights means never collecting the edge. The same arithmetic runs here. See what an edge is and how size changes the risk of ruin.
A bad stretch and a broken edge look alike at first. An alarm compares Peregrine's live results with its backtest and flags it when they drift apart, so the difference is measured, not guessed.
Not financial advice. Synapse shows what its engine sees and does. It does not know your account or your goals, and nothing here is a recommendation to buy or sell anything. Backtested results are hypothetical and don't guarantee future results. Read the full disclaimer or the terms.