The most expensive debt in the stack exists to avoid selling the cheapest equity.

What is mezzanine financing?

It is the layer of a capital structure that sits between senior debt and equity, and it behaves like its position: paid after the senior lender, ahead of the shareholders, priced for a risk closer to ownership than to lending. The name describes the floor it stands on, not a product. In practice the tranche is a subordinated loan, usually unsecured or secured on a second-ranking basis, with a bullet maturity rather than amortization.

Three features define it. First, subordination: in a downside the mezzanine lender recovers only after the senior facility is repaid in full, which is the fact from which every other term follows. Second, the coupon usually splits into a cash portion paid currently and a payment-in-kind portion that accrues to principal, so part of the cost defers to maturity instead of draining liquidity each quarter. Third, an equity component, most often a warrant, that gives the lender a small share of the upside if the business performs.

The purpose is narrower than the structure suggests. A mezzanine tranche exists to fill a specific gap: the distance between the leverage a senior lender will provide and the total funding a transaction requires, in situations where the owner would rather pay a high coupon on a small tranche than sell a larger share of the equity. Whether that trade makes sense is arithmetic, not doctrine, and the arithmetic depends on what the equity would have been worth.

What does mezzanine actually cost?

Counting cash and payment-in-kind interest together, junior capital and mezzanine for a borrower below $10m of EBITDA runs 13.00% to 16.00% (SPP Capital Partners, Market At A Glance, July 2026). Further up the market the junior layer separates into steps: second lien at S+725 to 825 for borrowers with $15m to $40m of EBITDA, subordinated debt at an all-in 11.0% to 13.5%, and preferred equity at 13.5% to 16.5% for borrowers with $40m to $100m of EBITDA (Lincoln International, Private Credit Snapshot, as of 1 May 2026).

For scale, the layers beneath it: at a one-month Term SOFR of 3.65% (CME, 4 August 2026), senior commercial bank cash flow debt for the sub-$10m borrower runs 7.15% to 7.90% all-in and senior non-bank or unitranche 9.15% to 11.15% (SPP, July 2026). The mezzanine tranche is roughly twice the price of the bank money it sits on top of. That is not a market inefficiency. It is the price of standing second in the queue.

The stated yield is also not the whole cost. The payment-in-kind portion compounds: interest that accrues to principal is interest that itself bears interest, so the balance at maturity is larger than the balance drawn, and the refinancing that repays it has to clear that larger number. And the warrant is a cost that appears nowhere in the coupon, because it is paid only in the outcome where the business has succeeded, which is precisely the outcome the owner is working toward.

Deferral is common enough to measure at market level. Payment-in-kind interest was being taken on 10.6% of direct lending loans, amounting to 8.9% of total interest income, the highest share since the fourth quarter of 2020 (Lincoln International, published 7 May 2026, as at 31 March 2026). A structure that defers part of its coupon is a normal structure in this market. A structure that defers it because the cash portion could not be paid is a different thing, and lenders read the difference closely.

When does a mezzanine tranche beat the alternatives?

The honest starting point is that below roughly $10m of EBITDA there is often no room for one. Senior debt for that cohort clears at 2.00x to 2.50x of EBITDA and total debt at 2.50x to 3.25x (SPP, July 2026), so the entire junior layer is about half a turn to three quarters of a turn wide. At that width the practical decision is usually a bank facility versus a unitranche facility, not a stacked structure. The mezzanine question becomes real as the business grows into the size bands where the junior layer widens.

Against a unitranche, the trade is precision against simplicity. A unitranche collapses senior and junior into one document at a blended 9.15% to 11.15%, one lender, one negotiation, one intercreditor problem that does not exist. A senior-plus-mezzanine structure prices each layer separately, which can produce a cheaper blend when the senior band is wide, at the cost of two negotiations and an intercreditor agreement whose drafting matters most in the year nobody expects trouble.

Against equity, the comparison the tranche exists for, the question is what the incremental capital would cost in dilution. A coupon of 13% to 16% on a small tranche is expensive debt. The same capital raised by selling equity in a business the owner expects to grow can be more expensive still, permanently, and without a maturity date on which it leaves. Mezzanine is the instrument for the owner whose answer to that comparison is deliberate rather than assumed.

The cases where it fits recur: an acquisition where senior capacity stops short of the purchase price, a shareholder buyout where one branch of a family exits, a growth investment too large for retained cash flow and too small to justify a sponsor. What the cases share is a defined use, a defined exit, and cash flow that can carry the cash portion of the coupon through a downside year. Where any of those three is missing, the tranche does not fix the problem. It postpones it to maturity, with compounding.

What does a mezzanine lender actually underwrite?

Not the collateral. A subordinated lender knows that in an enforcement it stands behind the senior facility, so its recovery analysis starts from the assumption that the security is spoken for. What it underwrites instead is the stream: whether the business generates enough cash to pay the cash portion of the coupon through a full cycle, and whether a realistic refinancing or sale repays the bullet, including the interest that has quietly accrued to it, at maturity.

That is why the diligence feels closer to an equity process than a bank one. The lender is exposed to the same downside as the shareholders for most realistic bad outcomes while capped at a coupon plus a warrant on the upside, and it prices and diligences that asymmetry. Expect the questions to run to customer concentration, management depth and the credibility of the exit, not just to coverage ratios.

The intercreditor agreement is where the tranche's real behaviour is written down. It sets what the mezzanine lender can do when something goes wrong: whether and for how long its remedies are stood still while the senior lender acts, what payments to it can be blocked and for how long, and who controls an enforcement. Two structures with identical coupons and different intercreditor terms are different instruments. The drafting is a workstream, not a formality, and it is negotiated at close, which is the only time it is cheap.

What should a borrower negotiate in the tranche itself?

The cash and payment-in-kind split, first. It determines how much of the cost hits liquidity now against how much compounds to maturity, and it is the term with the most room, because different providers have genuinely different appetites for accrual. A structure that can toggle a portion of the coupon between cash and accrual in a defined stress, on defined terms agreed in advance, is worth more than a lower headline rate without one.

The equity component, second. Warrant coverage is a real cost paid in the good outcome, and its size, its strike and what happens to it on prepayment or sale are all negotiable. So is the form: a warrant, an option at nominal value, or a conversion right each behave differently on exit, and the differences surface at precisely the moment the business is worth arguing over.

Call protection, third. Mezzanine is bullet debt the borrower frequently intends to refinance early, once the business has grown into cheaper senior capacity. Prepayment premiums price that option, and a borrower who expects to refinance is buying an option against itself. The premium schedule, and whether a sale of the business is carved out of it, belongs on the shortlist of terms argued as hard as the coupon.

None of this is exotic. It is the same rule that governs the rest of the structure: the coupon is the price that is easiest to compare and therefore the one most efficiently set by the market. The terms that carry the dispersion, and therefore the value of preparation and of running more than one provider, are the ones without a number attached at the first meeting.