SDE vs EBITDA: which number do lenders use, and how do the two reconcile?

Two measures, built for different questions

EBITDA answers an investor's or lender's question: what does this business earn as a business, with every job in it, including the top one, paid at a market rate? It adds back interest, taxes, depreciation and amortization, and lenders then adjust it for documented one-time items. It does not add back the cost of running the company. At this size it is the measure everything else is built on: senior cash-flow lenders to lower-middle-market companies commonly lend 2x to 3.5x EBITDA, and covenants, pricing grids and valuations all start from it.

SDE answers a different buyer's question: if I own this business and run it myself, how much could I take out of it before paying any debt? It adds back everything one owner receives, including salary, payroll taxes on that salary, health insurance, retirement contributions and personal costs run through the company, along with interest, depreciation and one-time items. It is built for owner-operated businesses sold to individuals who will work in them.

Where SDE still shows up at this size

A company with $10M to $100M+ in revenue is rarely marketed on SDE, but the measure has a way of reaching the lender anyway.

• Add-on acquisitions. A platform buying a smaller competitor often receives a broker's summary quoting the target's SDE. The platform lender will count only the target's adjusted EBITDA in pro forma EBITDA, after a market cost for whoever will run it. See add-on acquisition financing.
• Founder pay set for tax or habit, not the market. Many founder-run companies pay the owner well above or well below what a hired chief executive would cost. Normalizing that pay is an SDE-style adjustment inside an EBITDA calculation, and lenders test it the same way.
• Family on the payroll. Relatives paid above market, below market or not at all distort EBITDA in the same direction as owner pay.
• Management buyouts and successions. When a founder's role is split among two or three successors, the replacement cost is several salaries, not one.
• Price expectations. A seller who has heard what owner-run businesses sell for as a multiple of SDE may carry that multiple into a negotiation for a company that will be financed on EBITDA.

SDE is not a wrong number; it is the wrong number for sizing debt. The error is carrying it from a listing or a founder's view of the business into a loan request.

Reconciling the two from one P&L

The clearest way to see the difference is to walk up from net income. Take a service company, the size of a typical add-on, whose owner pays himself a salary of 220 plus 40 of benefits and payroll taxes, and runs some personal costs through the business.

The broker's summary will say SDE of 780. A lender will not start there. It will ask what a market salary for someone running this company would cost, including payroll taxes and benefits. Suppose the answer is 180. The owner's pay of 260 was more than that, so the lender adds back the excess of 80 to the adjusted EBITDA and arrives at 600. That is the number the lender actually works from, and the reconciling line between the two measures is the 180 the lender deducted.

The replacement cost cuts both ways. An owner who is overpaid makes EBITDA look too low; an owner who draws little or nothing makes it look too high, and the lender will deduct a market cost the seller never paid.

Why the replacement cost is the whole argument

Lenders underwrite a business as if its owner could step away, because one day the owner might, through illness, a sale or a bad year. Someone has to be paid to do the job. That holds even when the buyer or a platform's existing managers plan to run the business: the cost of that time is real, and the lender puts it in the numbers.

The adjustments that most often change the number:

• Two working owners. SDE conventionally adds back one owner's pay. If a second owner or a spouse works in the business, their role still has to be paid at market after the sale.
• An owner doing two jobs, such as running the company and doing the estimating or managing the largest accounts. The replacement may be two salaries, not one.
• Rent paid to the owner. If the seller owns the building and charges below-market rent, the lender uses the rent the company will actually pay after closing.
• Add-backs without evidence. Personal costs and one-time items count only when the records show them. A quality of earnings review is where they are tested. See EBITDA add-backs.

Same business, two debt capacities

Continue the example. The lender's EBITDA is 600. Many lenders also subtract maintenance capital spending, here 70, the cost of replacing vehicles and equipment that wear out, leaving 530 of cash flow available for debt service. Conventional bank lenders commonly look for debt service coverage of at least 1.25x.

On either test, the lender's version supports roughly a quarter to a third less debt. For a platform, the gap shows up as pro forma EBITDA the lender will not count: an add-on bought on the strength of 780 contributes 600 to the combined company's leverage calculation, and the debt raised against it has to fit that figure. The difference comes from more equity, a lower price, a seller note or an earnout.

Applying a leverage multiple to SDE is a category error. The multiple was built for EBITDA, and applied to a figure that still contains the manager's salary it produces a loan no lender offers.

Who uses which number

For how lenders turn either figure into a value, see how lenders value a business.

Get to the lender's number before the letter of intent

The time to rebuild the earnings figure is before agreeing a price, not after a lender has. In practice:

• Reconcile the summary to the tax returns. Lenders start from the returns and the statements; add-backs that do not appear there need evidence. See seller financials vs tax returns.
• Price the replacement role honestly, including payroll taxes and benefits, and every other role the seller or family filled.
• Pull the capex history from the depreciation schedule and fixed-asset records to see what keeping the business running actually costs.
• Use the target's latest full year of figures, never an older year, and a year-to-date P&L through last month-end.
• Plan the diligence. A buyer financing with acquisition debt should expect the lender to rely on a quality of earnings review of the adjustments. See quality of earnings for acquisition loans.

Midas Partners's financing model includes the bridge from reported earnings to the lender's EBITDA, with the market cost of management and each add-back on its own line, so the buyer sees the lender's number before a lender does. Once the documents are in, the full lender package is built in a day, and a senior banker checks every page before the client approves it. For the wider view, see how much debt can my business carry.