A developer needs bondholder consent to sell a subsidiary that a covenant protects, and offers a 0.5% fee for approval. What does it cost the issuer, how should a bondholder decide, and what if the sale weakens the credit?

The fee costs Nirvani at most Rs 5 crore, and holders should approve only if the proceeds are tied to cutting debt. If the Rs 700 crore repays debt, leverage falls to 3.1x and the bond gains; if it stays in the business, leverage rises to 5.0x and the bond loses about 2.7 points, far more than the 0.5 fee. The fee pays for the risk only if the bad outcome is less than about 41% likely.

The fee is paid on bonds that vote yes. If every holder consents it costs Rs 5 crore; at the bare two-thirds threshold, Rs 3.33 crore, plus legal and agency costs. Against a Rs 700 crore sale, the fee is small change, which is exactly why holders have leverage here. A consent solicitationA formal request by an issuer asking bondholders to vote to change or waive a term of the bond, usually with a fee for those who agree. is the cheapest way for an issuer to change a covenant; refinancing the whole bond would cost far more.

It depends entirely on where the money goes. Selling a business that earns Rs 90 crore of EBITDA removes 20% of Nirvani's earnings. If the Rs 700 crore repays debt, leverage falls from 4.0x to 3.06x; if it does not, the same debt sits on smaller earnings and leverage rises to 5.0x. A family that sells a rented flat and clears its home loan is safer; one that sells the flat and spends the money has lost its rental income and kept its loan. At a spread duration of 3.4, 30 basis points tighter is about +1.02 points of price and 80 basis points wider is about -2.72.

Split it into two questions. First, if the change is going to pass, you should consent, because the fee is the only difference between the branches. That is the pressure the fee structure is designed to create: each holder is paid to say yes whether or not yes is good for holders as a group. Second, as a group, is the fee enough? Weigh the branches: 0.50 plus 1.02 with probability one minus p, and 0.50 minus 2.72 with probability p. That breaks even at p of about 41%. With the proceeds promised only to general corporate purposes, a holder cannot rule out the bad branch.

Change the terms so the bad branch goes away rather than haggling over the fee. Ask that the proceeds go to repaying debt, for example an offer to buy back bonds at par or above from the sale money, and keep the covenant in force for other subsidiaries. Holders of a third plus one rupee can block the change, so an organised group has real bargaining power. If Nirvani will not commit the proceeds, a higher fee is the fallback, but even 1% would not cover a 2.7 point loss. The limitation: spread moves are estimates, and a rating action could move the price more than the desk expects.