The borrower's all-in cost of one unitranche against a split first lien and second lien, with fees, agency costs and market flex priced on one illustrative buyout.
A unitranche costs more than a first lien and second lien split, but by far less than the headline spreads suggest. On an illustrative $240M buyout loan at 6.0x EBITDA, a unitranche at S+550 costs 627.1 bp a year all in, against 514.6 bp for 4.5x of first lien at S+350 and 1.5x of second lien at S+675. The real premium is 112.5 bp, or $2.70M a year, not the 200 bp a comparison with the first lien margin implies.
Worked in full in The Private Credit Investor by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The comparison comes up in every sponsor process where both a direct lender and a syndicate desk are bidding. The borrower sees one number from each side, and the two numbers measure different things: the unitranche spread prices the whole debt stack, the first lien spread prices only its senior part. To compare like with like you have to rebuild the split structure's cost on the same $240M.
| Input | Unitranche | First lien | Second lien |
|---|---|---|---|
| Leverage | 6.0x | 4.5x | 1.5x |
| Amount (EBITDA of 40.0) | 240.0 | 180.0 | 60.0 |
| Spread over SOFR, bp | 550 | 350 | 675 |
| Upfront cost (OID plus fees), points | 2.25 | 2.00 | 3.00 |
| Annual agency and ratings cost, $M | 0.05 | 0.20 for both tranches | |
| Expected life, years | 3 | 3 | 3 |
The upfront line includes the original issue discount and the arrangement or underwriting fees. The syndicated structure also needs public ratings and two agents, which is why its running cost is higher. Both loans are assumed to be refinanced after three years, the life over which a borrower should amortise upfront costs, not the seven-year legal maturity.
Blended spread = (First lien × spread + Second lien × spread) ÷ Total debt
= (180.0 × 350 + 60.0 × 675) ÷ 240.0 = 431.25 bp
In dollars: 6.30 + 4.05 = $10.35M a year of spread, against $13.20M on the unitranche.
In Excel: =SUMPRODUCT(Amounts,Spreads)/SUM(Amounts)
The gap on spread alone is 550 less 431.25, or 118.75 bp. That is already 81.25 bp smaller than the naive 200 bp, and it is the step most often skipped in a board paper.
Upfront, unitranche: 240.0 × 2.25% = 5.40, over 3 years on 240.0 = 75.0 bp a year
Upfront, split: 180.0 × 2.00% + 60.0 × 3.00% = 5.40, also 75.0 bp a year
Agency and ratings: 0.05 ÷ 240.0 = 2.1 bp; 0.20 ÷ 240.0 = 8.3 bp
All in, unitranche: 550 + 75.0 + 2.1 = 627.1 bp
All in, split: 431.25 + 75.0 + 8.3 = 514.6 bp
Premium: 112.5 bp × 240.0 = $2.70M a year, $8.10M over the expected life
With SOFR at an illustrative 4.00 per cent, the coupon is 9.50 per cent on the unitranche and 8.31 per cent blended on the split, or $22.80M of cash interest against $19.95M. The upfront costs are equal here by construction, which is common enough in practice: the higher OID on a unitranche is offset by the underwriting fee and the second lien's deeper discount. When they are equal, the expected life drops out of the comparison and the gap is the same at two years or five.
The split structure's cost is a quote, not a certainty. A syndicated deal launches with flex language that lets the arrangers widen spreads and deepen the OID if the book is thin, and the borrower carries that risk until allocation. A unitranche is agreed with a small group of lenders before signing. The useful question is how much flex closes the gap.
| Flex scenario | First lien | Second lien | Blended spread | All in | Unitranche premium | $M a year |
|---|---|---|---|---|---|---|
| None | +0, +0.00 pt | +0, +0.00 pt | 431.25 | 514.6 | 112.5 | 2.70 |
| Modest | +25, +0.50 pt | +50, +1.00 pt | 462.50 | 566.7 | 60.4 | 1.45 |
| Hard | +50, +1.00 pt | +100, +2.00 pt | 493.75 | 618.8 | 8.3 | 0.20 |
A modest flex removes almost half the premium; a hard one removes nearly all of it. A sponsor who gives the hard case real weight is paying little for certainty. Another way to read the same table: the second lien would need a spread of 1,125 bp, with nothing else moving, before the split cost as much as the unitranche. That number tells you the premium is real in a normal market and disappears only in a dislocated one.
The comparison above assumes both routes deliver the same leverage. Often they do not. If the syndicated market stops at 5.5x, with 4.25x of first lien and 1.25x of second lien, the split raises $220.0M and the sponsor writes a cheque for the missing $20.0M of equity.
Split at 5.5x: 507.2 bp all in over SOFR on 220.0, plus SOFR at 4.00 per cent = $19.96M a year
Unitranche at 6.0x: 627.1 bp over SOFR on 240.0, plus SOFR = $24.65M a year
Extra cost of the unitranche: 4.69 a year to avoid 20.0 of equity, or 23.5 per cent a year before tax
Framed this way, the unitranche premium is the price of the last half turn of debt, and the base rate must be included because the extra $20.0M bears SOFR as well as the spread: leaving it out (3.89 on 20.0, or 19.5 per cent) understates the marginal cost. At 23.5 per cent a year before tax, and less after the tax shield on interest, the last half turn costs about what a sponsor expects to earn on the equity it replaces, so the choice turns on certainty and leverage rather than on price. The same arithmetic is the lender's argument for the spread, and the reason the last half turn is the one a credit committee looks at hardest.
The common mistake is to set the unitranche spread against the first lien margin, 550 against 350, and call the difference the cost of the unitranche. It ignores that the unitranche also replaces a second lien at 675. The second mistake runs the other way: comparing coupons and forgetting that the syndicated route carries flex risk and higher running costs. Both errors go away once every option is expressed as an all-in cost in basis points on the same amount of debt over the same expected life.
Lender's view. The same blend explains a unitranche lender's yield. A unitranche at S+550 is, economically, a first lien and a second lien held together at 431.25 bp plus a premium for the convenience it sells. When a unitranche is later split into first-out and last-out pieces, that premium is shared between them, which is worked in how to calculate the last-out spread in a unitranche.
The book's lender model for a 6.25x unitranche, with its base, downside and stress cases, is in the free workbook for this case. How much a point of OID is worth in spread is worked in how many basis points one point of OID is worth.
It is not comparable with the first lien alone, because it also replaces the second lien. Against a first lien at S+350 the unitranche at S+550 looks 200 bp dearer, but against the blended spread of the split structure, 431.25 bp, the gap is 118.75 bp before fees and 112.5 bp after agency costs. What remains is the price of certainty, speed and a single lender group.
The amount-weighted average spread across tranches that rank together in the borrower's cost. With $180.0M of first lien at 350 bp and $60.0M of second lien at 675 bp, the blended spread is (180.0 x 350 + 60.0 x 675) / 240.0, or 431.25 bp. That is the figure to set against a unitranche spread, not the first lien margin.
When the syndicated market moves against the borrower. In this illustrative case, flex of 50 bp and one point of OID on the first lien and 100 bp and two points on the second lien raises the split structure to 618.8 bp all in, leaving the unitranche only 8.3 bp dearer, about $0.20M a year on $240M.
This article is one calculation from The Private Credit Investor. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: credit, private credit and distressed debt → · All 453 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.