Carlyle Q1 2026 slides: AUM hits $475B despite earnings miss

Carlyle reported Q1 2026 distributable earnings of $327 million, or $0.89 per share, missing the $1.00 consensus as unrealized performance allocation reversals drove a $132 million GAAP net loss. Offsetting the earnings miss, AUM rose 5% year over year to $475 billion, fee-earning AUM increased 6% to $333 billion, and the firm posted record-like activity with $13.0 billion of inflows and $12.2 billion of realized proceeds. Carlyle also maintained its $0.35 quarterly dividend and repurchased $205 million of stock, while management reiterated confidence in long-term growth and fundraising momentum.

Analysis

The clean read-through is that the headline “tax scare” is a transitory tape event, but it lands on top of a real vulnerability: semiconductor multiples are still priced for uninterrupted AI capex, so any policy shock that changes after-tax economics can compress the entire supply chain faster than end-demand fundamentals would justify. NVDA is the obvious beta outlet, but the more interesting second-order effect is on the infrastructure cohort that depends on hyperscaler capital budgets staying aggressive; if CFOs start defending free cash flow, the first cuts usually show up in networking, memory, and lower-priority acceleration projects before they hit flagship AI GPU orders.

For Carlyle, the market should look through the earnings miss and focus on the mix shift toward fee-earning and perpetual capital. The real implication is that private-markets managers with scaling credit and secondaries platforms are becoming more bond-like in revenue durability, even if GAAP earnings remain noisy quarter to quarter. That makes CG less a pure “earnings beat/miss” trade and more a self-funding compounding story, with buybacks and dividends providing a floor while AUM growth expands forward fee base over the next 2-4 quarters.

The contrarian view on NVDA is that the move may be partially overdone if the policy change is framed as a one-time adjustment rather than a structural tax regime. If the market is effectively pricing a permanent margin haircut, that is probably too aggressive given NVDA’s pricing power and mix toward software/system-level value capture. The better tell over the next 1-3 months is not the headline stock reaction but whether peers with less differentiated economics underperform more persistently; that would confirm a broader capex de-risking rather than a single-name scare.

Key Decisions for Investors

• Reduce NVDA beta tactically for 1-2 weeks; if the policy headline persists, expect the stock to trade on margin compression fears rather than fundamentals. Use a 2-4% risk trim rather than a full exit because downside is sentiment-driven, not thesis-breaking.
• Express the broader AI capex risk with a short basket of lower-quality semiconductor adjacencies vs long NVDA into the next 30-60 days; the trade works if the market rotates from growth scarcity to after-tax cash flow discipline.
• Add CG on weakness over the next 1-3 months: the combination of buybacks, dividend yield, and growing fee-earning AUM offers a cleaner cash-return profile than the market is implying. Risk/reward is favorable if you buy below recent trading levels and target a rerating toward stable asset-manager multiples.
• Pair long CG / short a more earnings-sensitive private-equity peer for 1 quarter: the trade isolates recurring fee growth and perpetual-capital mix against firms more exposed to realization volatility.
• For NVDA, consider downside hedges via put spreads rather than outright shorts; the event risk is binary but time-limited, and put spreads monetize a 5-10% drawdown if the tax narrative broadens without overpaying for convexity.