principally, you can divide the thing into more than two tranches of claim. (Very safe super-senior claims get paid first, quite safe senior claims get paid next, then somewhat risky mezzanine claims, then quite risky equity claims.)
. . .Much of what happens in finance is some form of this move. And the reason for that is basically that some people want to own safe things, because they have money that they don’t want to lose, and other people want to own risky things, because they have money that they want to turn into more money. If you have something that is moderately risky, someone will buy it, but if you slice it into things that are super-safe and things that are super-risky, more people might buy them. Financial theory suggests that this is impossible but virtually all of financial practice disagrees.
The link goes to the Modigliani-Miller theorem. That theorem says that moves like this do not eliminate risk–they just redistribute it.
I like to say that the nonfinancial sector wants to hold short-term riskless assets and issue risky long-term liabilities; the financial sector accommodates this by doing the opposite. But the debt and equity claims issued by the financial sector have to be owned by someone. Ultimately, households hold those claims, but with government regularly stepping in, particularly to redistribute losses.
Every day, Levine seems to come across a new scheme by which financial intermediaries are getting away with issuing liabilities that are riskier than they are represented as being. These are schemes that will not end well.
Very well said.
If I read Levine’s article correctly he seems to think contra-Kling that this will end well? He is saying in 5 years people will think tether’s scheme is just what banks do? Or did I misunderstand that.
I’m surprised it seems Levine misunderstands M&M thm. M&M assumes perfect markets. E.g., regulation increases demand for super-safe assets, and distress costs (and bailouts) affects risk-return of super-risky assets. So the Value additivity principle doesn’t hold in reality. (V(A+B) = V(A) + V(B))
The Modigliani-Miller theorem says nothing of the sort (i.e., it does not claim that this is impossible). It merely says that if you slice the risk in different ways, and stay well away from bankruptcy risk, the rates people will charge to accept the various risk tranches will such that the overall weighted risk for the project will be unchanged. If you want to tie that to a statement about the number of individuals who will buy the different risk tranches, you’ll need an additional theorem.
Yes — but I think bankruptcy risk is OK under M&M as long as there are no costs to processing bankruptcy. It’s the deadweight loss of bankruptcy costs that’s not allowed by M&M (because it assumes frictionless markets).