The CIM lands in your inbox. Revenue is $4.2M, broker-adjusted EBITDA is $910K, asking price is $3.6M. At a 3.96× multiple and a $3.08M SBA loan, DSCR looks like it works. You book the intro call.
Forty-five minutes later you've rebuilt the P&L from scratch and the real EBITDA is $587K. DSCR fails at the asking price by a margin that no creative structure fixes. You just wasted a morning.
I do the P&L rebuild before the intro call now. Here's exactly how.
Most CIMs include an adjusted EBITDA schedule. It will show owner compensation add-back, one-time expenses, personal expenses run through the business, depreciation, sometimes a rent normalization. The number at the bottom is the number the broker wants you to use as your base.
Do not use it as your base.
The broker's job is to present the business at the highest defensible valuation. That is not a criticism — it's just what the role requires. Your job is to model what the business actually earns under your ownership. Those two numbers are almost always different.
Start from the raw P&L. Build your own schedule.
Owner compensation is the biggest one. A broker will add back the full owner salary and benefits — sometimes $250K — and replace it with a placeholder "market rate manager" salary of $80–90K. On a $4M MSP with 18 employees, a competent GM costs $150–165K in New Jersey. The difference between $90K and $160K is $70K of EBITDA. At a 4× multiple, that's $280K of enterprise value the broker built in with a number you cannot replicate.
T&E is the second. Personal travel, meals, vehicle allowances. Brokers add it all back. I keep 55–60% of it as a real operating expense, because whoever runs the business after close will have real travel and client entertainment costs. On a $4M services business, T&E might be $45–60K. Keeping $30K of it reduces your add-back by $30K.
Related-party rent is the third. Owner owns the building, charges the business below-market rent of $4K/month, and the CIM helpfully notes the market rate is $7K. The add-back normalizes to market rate rent — but sometimes the seller is not willing to sell the building, so you'll actually be paying market rate from day one. The add-back is correct, but you need to verify the real lease terms before you rely on it.
Lenders will sometimes use your personal draw requirements rather than a straight market-rate salary for DSCR purposes. That helps your underwriting number, but it does not help the business.
If you're buying a $4M MSP and running it yourself, you need to model what it would cost to replace yourself — because the lender who follows you will underwrite it that way, and so will any buyer if you ever sell. Build it into the model at market rate. If underwriting needs a lower draw assumption to clear 1.25×, that tells you something about the deal, not about the salary.
Every deal I've rebuilt has come in below the broker's number. The gap varies, but on IT services and B2B deals in the $2M to $5M EV range, the difference between broker-adjusted EBITDA and a properly rebuilt figure is consistently in the $80K to $200K range. At a 3.5 to 4× multiple, that's a lot of enterprise value built on a salary assumption you cannot replicate.
That does not mean every deal is overpriced. It means the first conversation should happen with your number, not theirs.
Before I book any intro call, I take the asking price, estimate a TPC-based SBA loan at 80% of TPC, calculate annual P&I at current rates, and check whether my rebuilt EBITDA — less a real GM salary and taxes — clears 1.25× DSCR.
If it does not clear at the asking price, I pass or I come in with a price that makes the math work. I do not spend three months on a deal that was never going to fund at the broker's number. The rebuild takes 35–40 minutes. It is the most valuable 40 minutes in the process.