MARA Holdings, Inc. (MARA) Presents at H.C. Wainwright 28th Annual Global Investment Conference Transcript

At H.C. Wainwright's conference, panelists highlighted AI infrastructure demand centered on scalable power capacity, stating hyperscalers are expected to spend more than $1 trillion on AI next year. The discussion cited over $160 billion in AI colocation deals secured by Bitcoin miners and more than 14 gigawatts of total power capacity among the participating companies. The remarks underscore a constructive investment case for digital-infrastructure operators repositioning power assets toward AI workloads, though no company-specific financial results or new contracts were disclosed.

Analysis

The investable distinction is not nominal megawatts but deliverable, interconnect-ready capacity with transmission, cooling, fiber and creditworthy counterparties. CORZ and WYFI appear best positioned for a re-rating if contracted AI/HPC revenue converts into disclosed backlog, because recurring colo cash flows can support infrastructure-style valuation rather than Bitcoin-beta multiples. MARA and CLSK retain greater sensitivity to hashprice and Bitcoin treasury volatility; their power optionality deserves little incremental value until management shows that AI workloads generate superior contribution margin net of retrofit capex and foregone mining economics.

Near term, the panel itself is unlikely to change estimates; the catalyst path over 1-3 months is signed contracts with duration, escalators, customer credit support, and financing terms. The critical second-order constraint is construction financing: converting mining sites to high-density compute can require substantial electrical and cooling capex, potentially forcing equity issuance precisely where balance sheets and free cash flow are weakest. BTDR and SLNH have more asymmetric upside from monetizing underutilized power, but also higher dilution and project-execution risk.

Consensus may be overvaluing aggregate power claims while undervaluing the scarcity premium for commissioned capacity. Hyperscaler demand does not automatically translate into miner economics: customers will favor sites with low outage risk, permitted expansion, liquid network connectivity and transparent governance. A sustained decline in Bitcoin hashprice or a weaker-than-expected AI lease rate would expose miners whose valuations already capitalize both mining upside and data-center optionality.

Key Decisions for Investors

• Prefer a 3-6 month long CORZ / short MARA pair, sized market-neutral: CORZ has the clearer path to contracted infrastructure cash flow, while MARA remains more exposed to Bitcoin and balance-sheet volatility. Reassess if MARA discloses a binding AI contract with customer-funded retrofit capex or if CORZ fails to convert announced capacity into backlog by the next earnings cycle.
• Maintain BTDR and SLNH as event-driven watch names rather than core longs. Initiate only after disclosure of contract term, MW placed in service, expected capex per MW and financing source; require projected AI/HPC EBITDA yield to exceed the implied mining return by a meaningful margin.
• For CLSK, avoid paying a data-center multiple absent evidence of non-mining revenue. A 10-15% Bitcoin pullback or deterioration in network hashprice would be a better entry point for its core mining exposure than the current AI-infrastructure narrative.
• Set an alert around quarterly disclosures of contracted versus merely available power across CORZ, WYFI, BTDR and MARA. The key falsifier for the infrastructure thesis is rising capex and interest expense without corresponding contracted revenue, which would signal that the power option is becoming a funding liability.