When managing partners at boutiques tell me what their junior analysts cost, they usually quote the base salary. “We pay around $110K for a first-year.” That number is not wrong, but it is so far from the actual loaded cost of running an analyst that it functionally distorts how partners think about the economics of their firm.
The real number for a US-based first-year M&A analyst in 2026 is closer to $300,000 per year. And once you understand what that $300,000 is actually buying you in output, the question of whether to adopt AI tools stops being a philosophical one. It becomes a P&L question with an obvious answer.
Building up the true loaded cost
Start with base salary. A first-year analyst at a bulge bracket in 2026 earns $100,000-$125,000 base, with the elite boutiques in the same range, per Mergers & Inquisitions’ 2026 compensation report. Call it $110K mid-range.
Add the year-end bonus. Bonuses at the analyst level run 50-100% of base, with most firms paying in the 65-85% range. Top performers at firms like Centerview and Moelis can exceed 100%. Call it $80K mid-range.
That gets you to $190K in cash compensation, which is roughly the all-in number that gets quoted in industry surveys. But you are nowhere near the actual cost yet.
Add benefits and payroll taxes. In the US, the standard load is 25-30% on top of cash compensation: health insurance, retirement contributions, payroll taxes, life and disability insurance. Call it 27% on $190K, which adds $51K. You are at $241K.
Add the allocated cost of office space. Manhattan-class office in 2026 runs $90-$120 per square foot fully loaded. An analyst at a small boutique gets maybe 150 square feet of allocated space: desk, share of conference rooms, share of common areas. That is $13K-$18K per year. Call it $15K.
Add tools and data subscriptions. Bloomberg terminal: $30K per user. PitchBook or Capital IQ: another $15K-$25K depending on tier. Document management, CRM, deal database: another $5K-$10K. The analyst doesn’t get a full Bloomberg seat, usually a shared one, but the allocated cost is real. Call it $15K.
Add management overhead. A senior analyst or associate is supervising and reviewing the junior analyst’s work. Industry studies of professional services firms typically allocate 15-20% of a junior’s time as senior management overhead. At $400K+ all-in cost for the senior, 15% of their time supervising one junior is $60K of senior cost loaded onto the junior.
Add recruiting and training. Boutiques typically spend $20K-$40K to recruit a first-year analyst from a top program, plus another $10K-$15K in the first six months on training, ramp-up, and lost productivity. Amortized over a typical two-year analyst tenure, that is roughly $20K per year. Probably the most under-counted line item in the entire calculation.
Add the cost of turnover risk. The standard M&A analyst tenure is two to three years. About 30-40% of analysts leave for buy-side roles or MBA programs before they become genuinely productive. The probability-weighted cost of replacement is another $15K-$25K per year of expected loss. Call it $15K.
Total loaded cost: $190K cash compensation + $51K benefits + $15K office + $15K tools + $60K management + $20K recruiting/training + $15K turnover risk = roughly $366K per first-year analyst.
At elite boutiques where analyst comp can hit $250K, the loaded cost crosses $400K.
What that $300K-400K is actually doing
Here is where the math gets uncomfortable. If you observe how an analyst’s time is actually spent across the year, the breakdown looks roughly like this. About 40-50% of hours go to document mechanics: data extraction, Excel rebuilds, IM drafting, pitch material updates, formatting. About 20-25% go to research and analysis: market sizing, competitor benchmarking, precedent transactions. About 15-20% go to deal process: data room admin, NDA tracking, buyer correspondence. The remaining 10-15% goes to actual judgment work: calls, internal strategy, deal-team discussions.
I have seen these percentages reported with minor variation across half a dozen industry surveys and they tend to converge. The headline is consistent: roughly half of analyst hours go to document mechanics that have nothing to do with the judgment-requiring core of the role.
So your $366K analyst is producing maybe $40K-$60K worth of judgment work per year, and the rest is mechanics work that an AI tool can now do at a fraction of the cost.
What this changes in practice
I am not arguing that boutiques should fire their analysts. The analysts are essential to the firm. They are also the next generation of associates, VPs, and MDs, and the apprenticeship model has real value that AI does not replace.
What I am arguing is that the right way to think about AI tooling is not “what does it cost.” It is “what does my analyst’s time cost when I have them doing mechanics work instead of learning judgment.” When you compress the 40-50% of analyst hours that go to mechanics, you don’t get rid of the analyst. You get an analyst whose effective output doubles, whose career development accelerates, and whose firm can run more deals per partner without scaling headcount linearly.
The cost of AI tools for M&A in 2026 is typically 5-15% of the loaded cost of a single analyst. The return is structural, compounding, and increasingly visible at the firms that have already deployed it. The question isn’t whether the math works. The question is how long you can run on the older math before the boutique down the street eats your mandate flow.