Academic

To deal with the market failure problems I've discussed last week, I have heard a variety of solutions to the question of what to do to prevent a future bank run, and to prevent banks from holding risky assets (as a colleague noted, another way to think about this was bad risk management by risk managers, who thought they had managed to sell off the riskiest bits of the assets they kept, or who permitted traders to hold assets that they thought were not risky). 

• Buffers between banks, like a Tobin tax on trading, for example, which would encourage a bit more siloing in the system, are a possibility.  
• Another option is capital insurance, one variant of which might be a two-trigger convertible bond that banks could be required to buy.  Such bonds would come due if, say, TED spreads blew up and the individual bank's capital declined precipitously, pouring capital into banks when they need it the most (the triggers thus have an economy-in-crisis test and a bank-in-crisis test).  On the other hand, insured institutions are incentivized to take risks, so capital insurance could mimic the situation for those people worried about the implicit bailout behind too-big-to-fail institutions.  
• Leverage caps - which would shrink the banks - could make them less systemically risky, as long as every bank did not fail at once (the old saw is that in a crisis, all correlations go to 1).  This is what the G20 is talking about, and it would be real reform.
• And, though I haven't heard a lot of support for reforming compensation practices among the finance commentariat (the law commentariat takes this more seriously, of course), I've seen one paper that notes that the intermediaries that compensated their top officials the most, relative to size, appear to have done the worst.  
• The Congressional Oversight Panel suggested creating a list of systemically significant institutions, which get a government guarantee, and accordingly could borrow cheaply - but the idea is that the banks on the list would find regulators camping in their offices, would have to adhere to whatever compensation practices Congress dreamed up for them, would have to sell the jets, stay in Motel 6es, move their HQ to the north slope of Alaska, and so on and so on.  The hope is that banks wouldn't want to be on the list (which would be extremely hard to draw up), that the Goldmans of the world would stay away, even though the government guarantee would make capital for these institutions cheap.

All of this has led people to take a close look at the system.  I've seen one interesting paper that noted, in a stylized form, that funding through securitization involves, through mortgage origination, tranching, SIVs, commercial paper, holders of this paper, and investors in the holders, etc, a seven step supply chain between borrowers and lenders.  Companies with those sorts of long supply chains worry about the efficiencies there, and boring old banking had a three step chain, involving a bank, a depositor, and a borrower.  Is funding through securitization a good model?

All of which will be the basis for much more debate.  We're just sketching out the insights of others here, which means that you, the readers, can view this blog as your own intermediary between the producers of finance scholarship and the consumers of it.

In the meantime, I'm going to ask Gordon, since we?re an intermediary, about purchasing some capital insurance.  The Conglomerate will not fall victim to the next panic with a strong supply of two-trigger convertible bonds backing us up.