TeraWulf CEO Paul Prager Sells 137,500 Shares for $2.3 Million

This filing is mostly a liquidity event, not an information event. A prearranged 10b5-1 sale against a still-large retained stake usually removes a small amount of supply at the margin, but it does not change the core debate around WULF: whether it can convert scarce power and data-center capacity into durable AI-hosting economics fast enough to outrun dilution and financing risk.

The real market driver over the next 1-3 months is not insider activity; it is whether the company can fund expansion on tolerable terms. If the Anthropic-linked narrative proves bankable, WULF can re-rate with other AI-infrastructure optionality names; if not, the market will keep valuing it like a levered bitcoin miner with a costly capex roadmap, which is a very different multiple regime. That also creates relative-value spillover to better-capitalized peers such as IREN and CORZ, which may deserve the premium if the sector gets re-underwritten on balance-sheet strength rather than story.

Contrarian angle: the consensus is probably over-reading the sale as bearish when the bigger signal is that management still has substantial economic exposure. The more important risk is 6-18 months out: if BTC weakens, or if the AI leasing pipeline requires equity issuance or expensive debt, the stock could de-rate sharply despite headline contract wins. The thesis is falsified if WULF posts materially better funding terms, higher-margin AI revenue, and no incremental dilution over the next two quarters.