The M&A rebound is real. It is also five trades in a trench coat.

TL;DR

• Q1 2026 US M&A deal value: megadeals +319% YoY (EY-Parthenon). Total deal count up 18%.
• US venture capital deployed: US$267 billion — the highest quarter on record, breaking Q4 2021's dot-com-adjacent peak.
• 89% of Q1 VC dollars went to AI-related investments.
• Five deals = 73% of total M&A value. AES go-private at US$33.4B led. The others: SpaceX/xAI consolidation, two private-credit take-privates, one energy-infrastructure mega-roll-up.
• The "rebound" narrative is directionally correct. The statistical portrait is a single concentrated AI-infrastructure-and-energy trade wearing M&A clothing.

What happened

On 14 April, EY-Parthenon released its Q1 2026 M&A report. Headline: mega-deal activity up 319% year-over-year, highest quarterly total since Q4 2021. Mid-market activity (US$100M–US$1B deals) grew 16% — solid, not spectacular. Large-deal activity (US$1B–US$10B) grew 42%. Mega-deal activity (>US$10B) grew the 319%.

Separately, Crunchbase reported US venture capital deployment in Q1 of US$267 billion — the highest quarter in history, narrowly ahead of Q4 2021. AI-related investment accounted for 89% of deployed capital. Non-AI venture deployment was US$29 billion, which would be a below-average quarter by post-2018 standards.

Looking inside the M&A data, five deals account for roughly 73% of total disclosed value:

1. AES go-private (US$33.4B) — BlackRock's Global Infrastructure Partners plus ADIA taking the US utility private.
2. SpaceX–xAI consolidation (US$80B effective merger value) — Musk's combination in late 2025/early 2026, finalised pre-IPO.
3. Ares-led private-credit roll-up (US$21B) — a holding-company consolidation of three mid-size private-credit platforms.
4. Williams Companies / Energy Transfer merger (US$47B) — US midstream energy consolidation.
5. CoreWeave / Lambda merger (US$18B) — AI GPU-cloud consolidation.

The common thread is not industry. The common thread is compute, power, or the infrastructure underneath both. Four of the five deals are directly AI-data-centre-infrastructure plays. The fifth (Williams/Energy Transfer) is the fossil-fuel backbone that data centres run on. Even the "diversified" deal-value line isn't diversified.

What it actually means

EY-Parthenon's report is titled "The Great Rebound." That headline is doing a lot of work.

"Rebound" implies a return to broad activity across sectors. That is not what the data shows. What the data shows is a single thesis — that AI needs compute, that compute needs data centres, that data centres need power — expressing itself as a wave of mega-deals that collectively look like a recovery. Strip out the five named transactions and total US M&A value falls back to a level that is up ~22% year-over-year, which is respectable but not historic.¹

The more interesting framing is this: a concentrated thesis that large enough to count as a macroeconomic event is indistinguishable, in the M&A data, from a broad-based recovery. Both produce the same league-table charts. The diagnostic test is to ask what would have to be true for Q2 to look similar — if the same five industries repeat the same deal shape, the thesis is narrowing, not widening. If new sectors show up (consumer, retail, financial services), it is broadening.

The second thing to notice: venture-capital concentration is tighter than M&A concentration. 89% of US VC into AI is, historically, a reading that only shows up when a new category is absorbing all funding — mobile in 2009–2010, SaaS-cloud in 2014–2015, crypto in Q1 2022. Each time, the category's reported share peaked roughly 12–18 months before a valuation correction that took 40–70% off the top tier. Whether that pattern holds for AI is not a historical certainty. It is a base rate.

Hype deconstruction

"M&A is back." Yes, at the mega-deal tier. Not in the mid-market, where deal-count growth of 16% is merely respectable. If you run a mid-market advisory book, Q1 felt more like a recovery than a rebound.

"89% AI concentration is exciting." It should be sobering. Concentration ratios at this level have a poor forward base rate. The category itself is real; the allocation concentration is a risk marker.

"The Great Rebound is a new cycle." Early to call. Historically, M&A cycles broaden in year two — consumer, retail, and services deals pick up as buyers get comfortable. If Q2 and Q3 remain five-deal-dominated, the cycle is a thesis, not a cycle.

"VC funding is democratising." It's the opposite. 89% into one category, with a small number of hyperscalers driving the top of the round-size distribution, is the tightest funding concentration in 15 years. Founders outside the AI hot zone are reporting 30–40% longer fundraising cycles than 2024 baselines.

Stakeholder landscape

Cross-layer implications

• Sector breadth. The real test of the cycle is Q2. If industrials, consumer, and retail show mid-teens deal-value growth, the rebound is broadening. If they remain flat-to-single-digit, the "rebound" is the AI thesis widening its definition.
• Private-credit concentration. Mega-deal debt financing is increasingly sourced from private credit. The Blue Owl redemption stress (prior piece this set) and the Q1 mega-deal concentration are related — one funding channel is under stress while carrying a larger share of large-deal financing.
• Indexation. AES going private removes it from utility indices; SPCX entering public markets adds to tech/infrastructure indices. Net effect is a rotation of index weighting toward the AI-infrastructure thesis.
• Energy demand. Four of the five megadeals are structured around data-centre power consumption forecasts. If power-demand forecasts from the AI thesis disappoint in 2027–28, multiple deal theses reprice simultaneously.
• Regulatory. US antitrust tolerated the energy consolidations because they support AI supply-chain thesis that the administration backs. That tolerance is conditional on the thesis performing.
• Cross-border. Non-US cross-border deal activity is up only 11% in Q1 — the "rebound" is principally a US phenomenon. European M&A remains subdued; Asian activity is focused on Singapore hub consolidation (covered elsewhere in the April set).

What this means for you

If you are an M&A banker. Pipeline quality matters more than pipeline volume. A pipeline concentrated in AI-infrastructure mandates carries single-thesis exposure. Diversify selection effort into consumer, healthcare, and industrials now so that when the category-rotation happens, you are not rebuilding coverage under pressure.

If you are an LP allocating to venture. The 89% concentration reading is a risk signal. Rebalance toward managers with proven non-AI deployment; write smaller AI allocations at higher valuations than you would in a diversified cycle.

If you are a CEO considering a strategic sale. Q2 is the window. If you have non-AI strategic fit, price discovery is still generous; if you are AI-adjacent, multiples are at historic highs and will normalise in 2027. Either way, a closed transaction in 2026 is better than an open one in 2027.

If you are a CFO evaluating a public strategic acquisition. Cost of capital has normalised but is not cheap. Private-credit financing for sub-US$2B deals is still functional but pricing has widened 80–120bps since January. Lock quotes, don't work-in-progress them.

If you are a corporate development leader. The five-deal concentration is the signal to buy into categories adjacent to the AI thesis — data-centre supply chain, HVAC, industrial automation — where valuations have not yet compressed around the same theme. Window probably 12–18 months.

If you run economic analysis at a bank or institution. Q1 M&A as a macroeconomic signal is less informative than Q1 M&A in previous cycles because of the concentration. Strip out the top five deals and the broader trend picture is "normal," not "rebound."

Uncertainty ledger

• Q2 breadth. Whether the thesis broadens or narrows in Q2 is the single most important unknown in the cycle. Watch consumer, financial services, and healthcare deal-count growth rates specifically.
• Regulatory action. At least one of the five megadeals faces antitrust remedies. Which one, and how painful, will reshape the deal-case playbook for H2.
• AI-infrastructure power demand. The underlying demand forecasts for compute and grid load are the assumption that justifies four of the five deal valuations. A 20% downward revision to 2027 forecasts would reprice multiple deals simultaneously.
• Rate cuts. The Fed path is disputed. A 50bp cut by year-end changes mega-deal IRRs materially; a hold changes the deal mix toward strategic rather than financial buyers.
• Cross-border. European Q2 data will show whether this is a US-only thesis or broadening.

Bottom Line

The Q1 M&A rebound is real, and it is also five trades in a trench coat. Strip out AES, Williams/Energy Transfer, the private-credit roll-up, the AI-compute merger, and the SpaceX-xAI consolidation, and total deal value falls to "respectable 2024-style activity." The 319% megadeal growth is the AI-infrastructure thesis expressing itself as an M&A cycle. The 89% VC concentration in AI is the same thesis expressing itself as a venture cycle. Both are the same statement. Whether it is a cycle or a thesis will be resolved by whether Q2 broadens or narrows. Until then, every league-table headline that describes this as a "broad-based recovery" is rhetoric a chart would not support.

Written in the tradition of — M.

Sources

• Tier 1: EY-Parthenon Q1 2026 M&A Outlook; Crunchbase Q1 2026 Venture Report; Bloomberg; Refinitiv; S&P Capital IQ deal data
• Tier 2: Axios; Pitchbook; The Wall Street Journal; FT Lex