Earnings call transcript: Proact IT Group Q1 2026 revenue surpasses forecast

Proact IT Group reported Q1 2026 revenue of SEK 1,243 million, topping the SEK 1,220 million forecast, while adjusted EBITDA rose 45.8% year over year to SEK 115 million and margin expanded to 9.3%. New cloud contracts increased 24% to SEK 151 million, but managed cloud services declined and a stronger Swedish krona cut organic growth by 3.4%. Shares rose 1.92% on the beat; management also highlighted ongoing cost actions, a Netherlands divestment, and expectations for continued momentum in Q2.

Analysis

The most interesting second-order effect is that this is not just a Proact-specific earnings beat; it is evidence that the AI memory supply shock is leaking from hyperscaler capex into the broader storage/infrastructure channel. That favors distributors and solution providers with inventory access and pricing power in the near term, while pressuring buyers who rely on spot-like procurement and longer lead times. In other words, the winners are the vendors that can reprice faster than their customers can defer, which should keep gross profit ahead of revenue for another 1-2 quarters even if unit volumes normalize.

The market is probably underestimating how uneven the benefit is across the ecosystem. Broadcom and NVIDIA are structurally advantaged if tighter component supply keeps driving accelerated refresh cycles, but TSMC is more exposed to demand timing risk if customers are front-loading orders and then pausing into H2. Apple’s supplier diversification narrative is real, but it should be read as optionality rather than an immediate displacement of TSMC; the more material near-term implication is that every major OEM will push for multi-sourcing and inventory buffering, which is margin-positive for component suppliers and margin-negative for end-device pricing discipline.

Risk-wise, this looks like a 1-2 quarter trade, not a clean multi-year rerating. If memory lead times ease or customer pre-buying rolls over, the current margin tailwind can unwind quickly because the valuation support is tied to cyclical pricing, not durable demand growth. The bigger contrarian read: consensus is too focused on 'AI demand' and not enough on channel restocking; if this is restock-driven, the subsequent digestion phase could be ugly even while headline revenue stays firm.

For Intel, the news flow is directionally supportive but only indirectly: a world where OEMs want second-source flexibility and more supply chain redundancy helps the foundry and custom-silicon narrative, but it does not fix product competitiveness. So the clean trade is not to buy every chip name; it is to own the parts of the stack that capture scarcity rent and service attach, and be selective elsewhere.

Key Decisions for Investors

• Long AVGO vs short TSM over 1-3 months: benefit from supply-chain tightening and pricing power versus TSM’s greater exposure to order normalization if customers have already pre-bought memory and infrastructure components.
• Buy NVDA call spreads 2-4 months out: if AI-driven memory scarcity persists through H2, GPU ecosystem pricing stays supportive; cap upside via spreads because the market is already paying for strong demand.
• Short INTC on rallies, but keep it tactical: use any Apple-diversification headlines as exit liquidity, not a thesis change; the stock may catch sympathy bids, but fundamental benefit is delayed and uncertain.
• Pair long AVGO / short AAPL into any further diversification headlines: Broadcom benefits more immediately from second-sourcing and infrastructure complexity, while Apple’s supplier diversification creates cost and execution friction before any operating benefit.
• If already long cyclical hardware, take partial profits on a 6-8 week horizon: the risk/reward worsens once the market prices in the restock leg, because the next leg is likely digestion rather than acceleration.