Iren vs. Applied Digital: Which Is the Better Long-Term Play?

Iren and Applied Digital have pivoted from crypto mining to AI/high-performance data centers, with Iren securing a $9.7 billion AI cloud contract with Microsoft and reporting net income improvement from a $51.7 million loss to a $384.6 million gain in Q1 fiscal 2026; Iren’s stock is up over 400% year-over-year with a forward P/E around 50 and P/S near 20. Applied Digital, whose stock is up over 500% Y/Y, posted a 250% revenue increase in its latest quarter, holds multibillion-dollar leases (including CoreWeave) and a $16 billion backlog, supporting more predictable long-term cash flows. The developments materially reframe each company’s growth profile—Applied Digital for contract-backed stability and Iren for optionality—while both remain exposed to high valuation risk and AI-cycle volatility.

Analysis

Market structure: Hyperscalers, large colo operators and GPU/equipment suppliers are primary winners — Applied Digital (APLD) benefits from long-term lease cash flows and Microsoft (MSFT) benefits as an anchor cloud buyer; smaller crypto-only miners and undifferentiated colos face margin compression. The $9.7B MSFT deal and APLD’s $16B backlog signal demand for AI racks that likely outstrips near-term GPU/server supply and grid capacity, supporting pricing power for secured-capacity providers over the next 6–24 months.

Risk assessment: Key tail risks are regulatory moves against crypto/AI data exports, a rapid GPU price collapse, or grid/transformer bottlenecks that halt builds — any of which could wipe 30–70% of forward EBIT for leveraged builders. Near-term (days–weeks) expect headline-driven volatility; medium-term (3–12 months) backlog conversion and equipment delivery are critical; long-term (1–3 years) watch interest rates (10y >3.5%) and WACC-driven multiple compression.

trade implications: Favor contract-backed operators: establish a 12–18 month bias long APLD and hedge execution/valuation risk via put protection; use capped downside short tools (put spreads) on richly valued, optionality-heavy IREN where P/E ~50 and P/S ~20 imply vulnerability to multiple reversion. Cross-asset: higher yields hurt REIT-like valuations (EQIX, data-center landlords) and increase funding costs for capex-heavy builders; rising copper/transformer lead-times support select industrial suppliers.

contrarian angles: Consensus underweights conversion risk — backlog ($16B) is not guaranteed revenue within 12 months and MSFT deals can be heavily milestone‑contingent; the market may be overpaying for optionality (IREN) while under-discounting grid and supply-chain execution risk. Historical cloud buildouts show smaller providers get squeezed when hyperscalers internalize capacity; expect winners to be those with locked, indexed power contracts and multi-year take-or-pay clauses.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in APLD with a 12–18 month horizon (or buy 12-month LEAPS calls ~25–35% OTM). Take profits at +50% or if quarterly backlog conversion <50% vs guidance; set a hard stop-loss at -20% on the equity leg.
• Open a tactical 1–2% bearish exposure to IREN via a 6–9 month put spread (buy 20% OTM put, sell 40% OTM) to limit capital at risk; unwind if IREN announces >$5B additional contracted, non-contingent revenue or posts net income beats >20% for two consecutive quarters.
• Execute a dollar-neutral pair trade: long APLD / short IREN equal notional (rebalance quarterly). Exit the pair if APLD’s backlog converts to <30% revenue within 12 months or if IREN secures >$5B in fixed-price, take-or-pay contracts.
• Use options to hedge sector volatility: sell covered calls on existing NVDA or APLD positions to monetize elevated IV and buy 3–6 month protective puts on data-center REITs (e.g., EQIX) if 10y Treasury rises above 3.5% (expect REIT drawdown ≥10%).
• Reduce pure crypto-mining exposure by ~50% vs benchmark weight and reallocate into AI-infra names with secured power contracts; only re-increase mining exposure after 2 consecutive quarters of stable ASIC pricing and grid capacity confirmations.