Case 039Credit, distressed and capital structureHard

Ovrin Chemicals' bonds trade at a 600 basis point spread, which with 40% recovery implies about a 10% annual chance of default. Its equity trades at 8x EBITDA, which looks healthy. One market is wrong. Set out the assumptions, build the capital structure trade and say how you size the two legs.

The situation

Ovrin Chemicals has EBITDA of Rs 500 crore and net debt of Rs 2,000 crore, 4x EBITDA, almost all of it in five-year bonds paying a 9% coupon. The five-year risk-free rate is 7%. The bonds trade at a spread of 600 basis points, a 13% yield, which puts their price at about Rs 85.9 per Rs 100 of face value.

The shares trade at an enterprise value of 8x EBITDA, Rs 4,000 crore, so the equity is worth about Rs 2,000 crore. Your credit team uses a 40% recovery for Ovrin's bonds. The fund can borrow Ovrin shares at about 1% a year and finance bonds at the 7% risk-free rate.

Your task

Which default probability does each market imply, what assumptions sit behind each number, how would you set up a trade between the bonds and the shares, and how do you size the two legs?

Quick check

What annual default probability does the bond spread imply?

Worked solution

Try it on paper, then open one step at a time.

30-second answerThe answer to give first

The bonds price about a 10% annual default rate, some 41% over five years; the shares, at 8x EBITDA with debt at 4x, price nearer 11% over five years. For the shares to agree with the bonds, the equity would be worth about Rs 273 crore, not Rs 2,000 crore. Buy the bonds and short the shares, at least Rs 46 crore of short per Rs 100 crore of bonds so a default still pays, and size the whole trade on the scenario where both legs lose.

Step 1What does each market say about default?

Two people size up the same shop: one will lend its owner money only at a punishing rate, while the other would happily buy the shop at a healthy price. They cannot both be right about how likely the owner is to go bust. Bonds and shares are two claims on the same business, so each price carries a view of default, and a large gap between the two views is the raw material of a capital structure tradeA position in two or more securities of the same company, such as long bonds and short shares, that profits when their prices move back into line.. The bonds are the easy half: spread over loss in default, 6% over 60%, is 10% a year, 41% over five years.

The shares need a model. Treat them as a claim on enterprise value above the debt, as the structural model credit analysts use does: value of Rs 4,000 crore against debt of Rs 2,000 crore, with the business's value moving about 25% a year. With twice as much value as debt, the chance of value falling below the debt within five years comes out near 11%, about 2.2% a year, a quarter of what the bonds imply. Run it backwards and the gap is stark: for the shares to price a 41% five-year default chance, enterprise value would be about Rs 2,273 crore, 4.5x EBITDA, leaving the equity worth about Rs 273 crore.

The relationship

p_{\text{bond}} = \frac{s}{1-R} = \frac{6\%}{60\%} = 10\% \qquad DD = \frac{\ln(V/D)}{\sigma\sqrt{T}} = \frac{\ln 2}{0.25\sqrt{5}} \approx 1.24

s	the credit spread, 600 basis points
R	the recovery assumption, 40%
V/D	enterprise value over debt, 4,000 / 2,000
sigma	the assumed volatility of enterprise value, 25% a year
DD	distance to default: how many standard deviations value can fall before it reaches the debt

What it says in wordsThe bonds give a default rate from spread and recovery; the shares give one from how far value sits above the debt, measured in units of volatility.

The bonds' 600 basis point spread with 40% recovery implies 10% a year, 41% over five years, while enterprise value at twice the debt implies about 11% over five years; for the shares to agree with the bonds the equity would be worth about Rs 273 crore, not Rs 2,000 crore.

Step 2Which assumptions could close the gap without either market being wrong?

Set them out before you trade, because the interviewer will test each one. Recovery, the liquidity premium, the EBITDA figure and the volatility are the numbers that decide whether the gap is real. A lower recovery raises the bond-implied default rate, so it widens the gap rather than closing it. Part of any spread pays for illiquidity rather than default, so the bonds' true default view is below 10%. The shares may be valuing EBITDA of Rs 500 crore that is about to fall: if the chemical cycle is turning, 8x on today's EBITDA hides a riskier balance sheet, and the credit market may simply be earlier. And volatility of 25% is a guess; at 40% the equity-implied five-year figure rises to about 22%.

Step 3How do you build the trade?

If you think the bonds are too pessimistic, buy them and use the shares as the hedge. Size the short so that a default still makes money: the bonds fall from 85.9 to 40, a loss of about Rs 46 crore per Rs 100 crore of face, so short at least that much equity, which would go to zero. This example uses Rs 50 crore. If the bonds are right and Ovrin defaults, the short pays for the bond loss. If the shares are right and the spread tightens to 250 basis points, the bonds gain about Rs 12 crore while a 10% rise in the shares costs Rs 5 crore. If nothing happens for a year, the coupon and the pull towards par, less funding and borrow costs, earn about Rs 4.7 crore.

Scenario	What moves	Net, Rs crore
Default within a year	bonds to 40% recovery, shares to zero	+4.1
Credit was wrong	spread 600 to 250 bp, shares +10%	+7.1
Nothing changes for a year	coupon and pull to par, less funding and borrow	+4.7
Debt-funded buyback	spread 600 to 900 bp, shares +20%	-18.9

Long Rs 100 crore face of bonds at 85.9 and short Rs 50 crore of shares makes money if Ovrin defaults, if the credit market was wrong, or if nothing happens, but loses about Rs 19 crore if the company borrows to buy back shares.

In a default the Rs 50 crore short more than covers the bonds' Rs 45.9 crore loss, and if credit was wrong the bonds gain Rs 12.1 crore against a Rs 5 crore short loss, but a debt-funded buyback makes both legs lose, about Rs 18.9 crore in total.

Step 4How do you size the whole trade?

Size it on the scenario where both legs lose. A debt-funded buyback, or a leveraged buyout, pushes the shares up and the bonds down at the same time, and on this position that costs about Rs 19 crore per Rs 100 crore of bonds. If the fund allows this trade to lose 1% of a Rs 2,000 crore book, Rs 20 crore, then about Rs 100 crore of face is the limit. Read the bond documents for covenants that restrict new debt or payouts to shareholders: strong covenants make the bad scenario less likely and justify a larger position. Say which market you think is wrong and why, then show that the trade does not need you to be right about the timing.

Where candidates lose it

The usual loss is picking a side on instinct, credit is always smarter or equity is always smarter, without setting out the recovery, liquidity, EBITDA and volatility assumptions that decide it. The question asks for the assumptions first because the answer depends on them.

The second is getting the hedge wrong. Too little short and a default loses money; the costliest case, though, is the one people forget, a buyback that makes both legs lose at once, and it is the case the position must be sized on.

What the interviewer asks next

Recovery is 20% instead of 40%. What default rate do the bonds imply, and does the trade change?
How would you use credit default swaps instead of the bonds?
What in the bond documents would make you larger in this trade?
Ovrin announces an asset sale that will repay a third of the debt. Which leg reacts more?

Company names and figures are illustrative.