Direct lending model: project a unitranche loan to a hotel group for five years with 1% amortisation and a 50% excess cash sweep, then work out the lender's yield if it buys the loan at 98.

Leverage falls from 4.0x to about 1.65x by year 5 and the debt from Rs 600 crore to about Rs 390 crore; bought at 98, the lender earns about 12.11%. The sweep repays about Rs 180 crore early as EBITDA grows. The discount does most of the work on yield; the sweep does most of the work on risk, leaving Rs 180 crore less debt at refinancing.

Imagine a landlord who agrees that, whenever a tenant's business has a good year, half of the spare cash goes to paying off the tenant's arrears. A bad year costs the tenant nothing extra; a good year pays the debt down faster. A cash sweepA loan term requiring the borrower to use a set share of its excess cash flow each year to prepay the loan. ties repayment to performance: the better Quilona does, the faster the loan shrinks. Each year, start from EBITDA, take off capex, interest on the opening balance and tax, then the Rs 6 crore of scheduled amortisation. Half of what is left prepays the loan; the other half stays with the owners.

Lay out the lender's cash flows: Rs 588 crore paid out at the start, then each year's interest, amortisation and sweep, and the remaining balance at par at the end of year 5. The rate that sets their present value equal to Rs 588 crore is 12.11%. Bought at par, the same loan would yield exactly the 11.50% coupon, so the 2 point original issue discountA loan or bond sold below par; the gap is earned back as the principal is repaid at 100. adds about 0.6% a year. Without the sweep the yield would be 12.07%: the sweep returns principal sooner, so the discount is earned back a little faster, but here the difference is only a few hundredths of a percent because the loan is refinanced in year 5 either way.

Its value is in risk, not yield. With the sweep, Quilona reaches year 5 at 1.65x instead of 2.41x, so the refinancing the lender depends on is far easier and a bad year hurts far less. Hotels are cyclical: a slow tourist season or a delay in opening the new rooms would stall the 12% growth. The limit of the model is that it assumes growth arrives on time. Run the downside before signing: if EBITDA stays flat at Rs 150 crore, excess cash flow is only about Rs 36 to 45 crore a year, the sweep repays only about Rs 18 to 22 crore a year, and leverage would still be about 3.1x in year 5, roughly double the base case. A lender comfortable with that case can lend; one that needs the growth to be safe should not.