Applied Digital vs. TeraWulf: Which Neocloud Stock Is the Better Buy?

TeraWulf signed a 20-year, $19B deal with Anthropic covering 401MW of critical IT load (to come online in waves, fully by early 2028). Morgan Stanley raised its price target to $72, implying the stock could more than triple within 12 months, and the article argues Applied Digital may benefit from similar AI data-center compute demand. Applied Digital has an edge on nearer-term scale (3GW active pipeline; 1.41GW contracted IT load) while TeraWulf’s key differentiator is owning power generation assets (923MW contracted, ~half with Anthropic) which could reduce grid-price risk.

Analysis

The market is likely underpricing that in this niche, the scarce asset is not GPUs but durable power access with financing attached. That favors the operator whose cost base is structurally more controllable over a 6-18 month horizon, because once contracts scale, small differences in power economics can swing equity value far more than headline backlog. The headline move may be over-indexing on contract value while ignoring how much of that value is pre-committed to capex, interconnect, and financing dilution.

Near term, the easier-build model can still win on multiple expansion because public markets reward visible site launches and faster revenue conversion. But that also creates a hidden liability: if utility pricing tightens or grid access becomes constrained, the model with leased power has less protection and more margin volatility when contracts roll. In contrast, owned generation should earn a higher EV per MW over time, even if the stock is slower to re-rate today.

The contrarian miss is that “more contracted load” is not the same as better equity quality. For these names, the right question is not backlog size but who can convert backlog into free cash flow without repeated equity raises; that is where many investors will be disappointed over the next 1-3 quarters. A meaningful reversal would come if the supposedly easier builder secures unusually cheap long-dated power renewals and the market sees no dilution risk, or if the power-owner stumbles on execution and financing.

Key Decisions for Investors

• Long WULF vs. short APLD on rallies as a 3-6 month relative-value trade: thesis is that owned power deserves a higher mid-cycle multiple once the market stops treating backlog as fungible. Falsify if APLD locks in attractive long-duration power economics and posts cleaner site-to-site conversion than expected.
• If looking for outright exposure, prefer WULF on 6-18 month horizon; use any post-news pullback to build. Risk/reward is better if the stock is still discounted to future power optionality rather than current backlog.
• Do not chase both names simultaneously at current sentiment; instead wait for the next financing update or construction milestone. The key catalyst is whether either company can fund expansion without punitive dilution.
• Set an alert for WULF/ APLD earnings and any new power procurement disclosures: a step-up in utility costs or lease renegotiation would be the fastest way to break the APLD bull case.
• If you want a tactical momentum trade, use short-dated call spreads in the name with the next likely contract announcement, but size small; these names can gap on deal headlines, yet the gains are often diluted once capex reality is priced in.