Running a hedge fund desk means one thing above all else: identifying trades before the move or getting in early enough to ride signals with statistical conviction. Our approach was systematic: pattern analysis across hundreds of instruments, isolating setups that had made new 1-month or 3-month highs and in which forward returns were positive 70%, 80%, or 90% of the time, often with double-digit magnitude. High win rates, high reward-to-risk.

The challenge was doing this at scale: loading hundreds of charts, screening for qualifying setups, and statistically stress-testing each. That process demanded automation, especially as we added more models. That's what our Trading Signals model does.

We built these models in TradFi more than a decade ago. Six years ago, we ported them to crypto, adding cryptocurrencies and crypto equities. We also kept running them with large-cap stocks, ETFs, and, more recently, AI-related names. Every day, the models run the same scan: find the pattern, test the setup, flag the signal.

Cryptocurrencies have generated fewer signals recently; crypto equities and AI stocks have been more active. The universe evolves as markets do, and if a signal is crypto- or crypto-equity-related, it also appears in the dashboard (Parts 3 and 4).

Today, the models identified four new trades.

One of the new trades, would you buy this new high? See previous signals