Seasoning, Tenor and the Take-Out Decision on a 42 km Bypass
Satluj Bypass Tollway Pvt Ltd holds a 20-year concession from a state government for a 42 km four-lane bypass, awarded in 2023 on a DBFOT basis with a tolling right and no annuity. Project cost is ₹1,850 crore, funded 30:70 by equity and a consortium of three banks. The ₹1,295 crore loan carries 9.6%, a 3-year construction period, and 12-year repayment with a 12-month moratorium after the commercial operation date (COD), now expected in June 2026. The sponsor's adviser proposes refinancing at COD through a 20-year bond. The lead banker objects that the SPV has no operating record and that the instrument will price above 10.5%. A take-out window from a government-backed infrastructure lender is available: up to 80% of outstanding principal, 15-year residual tenor, 8.9%, subject to the project having completed one full year of tolling at above 90% collection efficiency and meeting a 1.20 DSCR. A partial credit enhancement (PCE) scheme for infrastructure bonds is also available against a fee. The concession agreement permits refinancing after COD without state consent.
1. Recommend a refinancing sequence from financial close to year five, and justify it with numbers rather than with a preference for any one instrument.
Read the worked answer
Test each route against tenor, cost and seasoning.
Cost arithmetic. Refinancing ₹1,295 crore at 10.5% instead of 9.6% adds ₹11.7 crore of interest in year one (1,295 × 0.9%) - value destruction, not refinancing. The take-out at 8.9% on 80% of principal (₹1,036 crore) saves ₹7.25 crore a year against the bank loan, and its 15-year residual tenor matches an asset with roughly 16 years of concession life left at COD+1. A 12-year bank loan against a 20-year concession leaves a refinancing cliff at a point when the sponsor has weaker bargaining power.
Sequence. (i) Bank debt funds construction - the only instrument that will lend pre-COD. (ii) At COD+12 months, exercise the take-out on 80% of principal, provided the 90% collection-efficiency and 1.20 DSCR tests are met. (iii) Leave 20% (₹259 crore) with the banks or repay from surplus; a PCE-wrapped bond becomes viable only after two to three years of tolling history, when the wrap fee is offset by a tighter spread.
Judgement. The bond is the right end-state but the wrong start-state; seasoning is the binding constraint, not price.
2. What conditions must be locked in during construction so that the year-one take-out actually happens, rather than being negotiated under duress?
Read the worked answer
The take-out is only as good as the covenants signed before the first disbursement. Insist on:
- Escrow of all toll receipts with a waterfall that funds O&M, then debt service, then DSRA, then distributions - a take-out lender will not accept a structure where the sponsor controls collections.
- DSRA of two quarters of debt service, funded at COD from equity, not from drawn debt.
- Cash sweep of surplus above a 1.20 DSCR, so that a good traffic year de-risks a bad one.
- Direct agreement with the state giving the lender step-in and the right to substitute the concessionaire, plus assignment of termination payments.
- Independent engineer certification of COD - take-out eligibility runs off a certified date, and a disputed COD shifts the seasoning clock.
- No dividend or subordinated repayment until DSRA is funded and the 90% collection test is met.
- Change-in-law and force majeure relief preserved through refinancing, since the concession permits refinancing without state consent but does not automatically preserve relief.
The principle: every take-out condition should appear as a construction-period covenant, otherwise the sponsor can stall until the bank has no alternative but to extend.
3. Six months into tolling, collections average 78% of the base case. How should the refinancing plan change?
Read the worked answer
The 90% collection-efficiency gate fails, so contractual take-out at year one is unavailable. Do not request a waiver as a first move - a waiver granted on a failed traffic test signals weakness to every subsequent lender and invites a price reset.
Work through the options in order of cost:
- Re-test the DSCR at 78% of base revenue. Revenue falls much faster than DSCR because O&M is largely fixed; if the base DSCR was 1.25, the stressed figure may fall near or below 1.0. If it is below 1.20, the 78% case itself disqualifies the project, not just the seasoning test.
- Partial take-out. Offer 40–50% of principal to the take-out lender, keeping the remainder with the banks. This preserves the relationship and gives the sponsor a rate benefit without pretending the traffic test is met.
- Sponsor equity cure. Bring forward equity to prepay a slice of debt so that the DSCR recovers to 1.20 on the 78% case. This is the cleanest fix because it converts a traffic shortfall into a capital-structure fix.
- Extend seasoning by 12 months with bank consent, at a higher spread, and revisit once the ramp-up curve is visible.
PCE-wrapped bonds should be deferred: a wrap prices credit risk, and a failed traffic test is precisely the information the market will reprice.