Filtering earnings options signals with Jev
An options system has GPT extract claims from earnings materials, Jev classify the evidence type, and QuantLib reprice the trade, so a candidate that gets direction right but is crushed by a volatility collapse is blocked before execution.
People ask AI which stock will jump after earnings. With options, it can get the direction right and still lose their money. A company beats expectations. The shares rise. The call bought before the announcement falls in value. Nothing has to be broken. The buyer paid for uncertainty. After the announcement, implied volatility can collapse enough to outweigh the gain from the stock's move. A convincing earnings thesis can completely miss that second bet. That's the problem we built our Jev options system around. GPT extracts the actual claims from company filings and earnings material. Jev classifies the evidence: new information, repeated narrative, conflicting evidence, or something that needs deeper review. Python and QuantLib do the numbers: reprice candidate options across different stock moves, volatility changes and holding periods. IBKR supplies current quotes and handles order submission. The loop tests whether the proposed trade survives several plausible outcomes, including being right on direction and wrong on volatility. A candidate that fails the predefined risk checks never reaches execution. We compare the Jev filter with a rules-only version on earnings events kept outside development. The extra model has to earn its place. That changes what the system looks for. A bullish headline starts an investigation. The option's price, exposure and behavior under stress determine whether the idea gets any further. AI can give you a convincing reason a stock should rise. Our system asks what happens to your option even if it's right.
This post quotes another post: x.com/hanakoxbt/status/2101709924828934222