Much toil and time is given to the study of markets and businesses. New insights can be interesting; the characters we encounter entertaining; and the bargains rock bottom. For the rainmakers out there, the latter are likely most useful, as every field needs a periodic downpour.

Yesterday it was the bondholders of Nova Chemicals Corporation that got the rain, as their $250 million in outstanding 7.4% medium-term notes were paid in full (cusip: 669936AA4). On the surface, the event may not strike you as noteworthy, but the story gets more interesting when you realize that these very bonds were trading at 58 cents on the dollar during the week of February 16th. Worried that Nova would not be able to pay these notes off or obtain the requisite financing by April 1st, buyers left Nova’s market deserted. Despite a book value of $750 million at December 31, 2008, the equity sellers on February 20th valued the company at less than $150 million.

Nova’s savior arrived though on February 23rd, as International Petroleum Investment Co. of Abu Dhabi came bearing the gift of $6 per common share, and a $250 million credit backstop. Nova’s debt and equity holders danced in the rain.

Reflecting on this scenario, let’s speculate about the relative attractiveness of Nova’s equity and bonds during the week of Feb. 16th. Imagine that we could go back in time; what would the expected value of each investment to be?

Looking at the equity then trading at $1.50 per share, let’s say you estimated the likelihood of bankruptcy (and an equity value of zero) at 40%. Then put a 40% likelihood on a successful debt refinancing, while positing that—in that scenario—the equity would trade marginally higher to $2 per share. And last, take the likelihood of a $6 per share buyout at 20%. All told, the expected value for the equity buyer would be (0*.4+2*.4+6*.2), or $2 per share—not at all a bad return (33%) in six weeks’ time.

How about the debt though, trading at 58 cents on the dollar? In the worst case scenario—bankruptcy—estimated at 40%, the court would have to order a restructuring, and the return to bondholders could vary widely. Authors Van de Castle, Keisman, and Yang have found the median recovery rates for defaulted senior notes to be near 43%. Since Nova has relatively abundant debt at this time (with 1.36 billion in long-term debt as of 12/31/08), let’s estimate that the recovery rates in bankruptcy would be lower than average, say 30%. Again, we’ll put a 40% likelihood on a successful refinancing, which would return 100 cents on the dollar, and the same return would follow from the 20% chance that Nova is acquired. Here the expected value for the debt holder would be (30*.4+100*.4+100*.2), or 72 cents on the dollar. On a purchase price of 58 cents, this would represent a 24% return—again, quite impressive for six weeks’ work.

In sum, either security looks favorable, given these assumptions. However, it is clear that a few small changes in our assumptions will much more negatively impact the equity’s expected value. For example, if the chance of a buyout were only 10% and bankruptcy instead 50%, the equity purchase would have a negative expected value, while the bond would still provide positive returns (30*.5+100*.4+100*.1). Therefore, in similar situations, I suspect that the prudent buyer of the debt will bear far fewer costs for any mistaken assumptions. Better to make rain with the debt holders, unless one’s assumptions are firm.

Disclosure: No position at the time of this writing.

Interesting analysis.

Do you know of any data on company turnarounds vs bankruptcy?

I remember reading that in any 10 year period around 30% of companies will go through a major crisis.

From memory of the 30% that stumble around 1/3 will fail and 2/3 will recover.

Do you know of any similar figures more relaible than my memory?

David