Tuesday, March 24, 2009

Why the Government is taking the long road of PPIP instead of the nationalization shortcut

Krugman’s Attack

I have to respect Paul Krugman. He is attacking the Obama administration which is politically more aligned with his worldview with the same ferocity and insight as he did the Bush administration which for him was probably like shooting fish in a barrel, and fish he did not particularly like.

Yesterday morning he published a scathing critique of the new Geithner plan in his regular column in the New York Times. It is a well written and well thought out, if gloomy, piece. It seems to me that he thinks the plan is simply another subsidy for Wall Street, and it is that. He also says that it won’t work on account of the valuations involved and he laments the fact that the government is not nationalizing the banks and getting on with things.

This is a cogent argument and while I think he should spend more time explaining why he thinks it won’t work than he does explaining why it is unjust, I think he is missing something with his alternative. There has to be some reason why the government is not taking what Krugman thinks is the shortest path to a full recovery, that is nationalization of the troubled banks. What might that be.

The Shortest Distance Between Two Points: Nationalization?

So what precisely is Krugman recommending? He is recommending that the government guarantee some, but not necessarily all, bank debts. Take over the bank and temporarily takes control of insolvent institutions and cleans up their balance sheets. This is largely what was done at the end of the S&L crisis in the early 90s and indeed it did work then. The government liquidated a lot of banks, paid off their depositors through the FSLIC and then worked off the assets through the Resolution Trust Corp. Investors in the failed S&Ls were wiped out and some of those that could be salvaged were then resold to the public. Krugman is recommending this be done for the largest US banks that are insolvent and it is very likely that this would include Citibank, Bank of America, Wells Fargo and perhaps even JP Morgan.

So why not?

The trouble with this plan is the phrase in Mr. Krugman’s article referring to how some but not all debts of the banking system would be guaranteed. This is an important point. There are basically three groups of creditors to the banking system: the depositors, the bondholders, and trading counterparties. In a nationalization where the government seized an insolvent institution the depositors would be made whole by the FDIC. This would leave the bondholders and the trading counterparties. Presumably Mr. Krugman would want to wipe out the bondholders as they were people who were not using banking services like the depositors but were people who knowingly took a risk lending the bank money they should be wiped out too. This is where we run into trouble.

The trouble is this, you cannot wipe out the bondholders without wiping out the trading counterparties because they are at the same point in the capital structure, that is they have the same rights to any proceeds, if any of the firm. So you can’t say to the bondholders, sorry you get nothing but then make all the counterparties whole.

Who are these trading counterparties anyway and why should I care?

In talking about the plan Geithner made the point that “we are not Sweden” our financial system is “complicated.” What he means by this is that the problem with seizing a bank, making the depositors whole and wiping out the equity and bondholders is that you would wipe out the trading counterparties at the same time. As it turns out, these trading counterparties are the rest of the financial system both the solvent and the insolvent parts of it. Thus by liquidating a few insolvent firms and wiping out their investors who had made bad decisions the government may wind up pushing other firms into insolvency themselves. And why is this? It is because of the nature of the market structure for Credit Default Swaps.

What are Credit Default Swaps?

Credit Default Swaps (CDSs) are ingenious financial products that enable lenders to insure themselves against the default by someone who to whom they have lent money. If I have a loan to IBM but I am worried about IBM defaulting to me I can trade a CDS with another bank. A CDS is a a security that is traded between two firms and its cash flows look a lot like those of the bonds themselves. Let’s say I sell protection to someone. We choose a reference company, in this case IBM, and a term, the length of time I will provide the protection and a notional amount, the amount of principal I am insuring. The way a CDS is structured I would pay them their funding cost, say LIBOR, and they would pay me some premium which corresponded to how likely the chance of default was. The greater the chance of default the greater the premium I will receive and that premium will look a lot like the rate that the reference company would have to pay if they were to issue new loans. If the reference company defaults I have to pay the notional amount less anything that can be recovered from the reference company. As you can see it looks a lot like a loan.

In fact it looks so much like a loan an investor could actually take on the risk profile of a bank without actually setting up a bank just by selling CDSs to banks. Banks raise money from depositors and then lend it out and earn the spread. During the life of the loan they pay interest to their depositors (LIBOR) and take in interest from the company they lent to (the premium.) In the end if the company pays the loan back all is well and you give the depositors their money back and stop paying LIBOR and if they don’t you lose the money and have to pay them the depositors back out of equity or in the case of the CDS you pay the protection buyer. So you can see how useful these things are and in fact many firms essentially became banks by selling these products. This is sometimes called the “shadow banking system.” AIG was simply the largest and not most notorious part of it but hedge funds were also involved as are the banks themselves.

CDS are very useful because they allow non-banks to take on bank risk and allow banks to manage their risks. The market for them has grown into a multi-trillion dollar affair that trades around the clock in all markets all over the world though primarily centred in London, New York and Hong Kong.

If these things are so useful why are they at the center of so much trouble?

As with so many things both God and the devil are in the details. The main problem is in the market structure itself. By this I mean that nuances of the product interact with the way they are traded to preserve the usefulness of the individual instruments while putting the market at large at risk. The problem is the combination of the fact that they are traded OTC and that they look like loans with a stream of cash flows with a potential payout at the end.

The fact that they are OTC means that they are traded against individual firms. The opposite of OTC derivatives are exchange traded derivatives like futures and options traded on the Options and futures exchanges in the US. On exchanges the person who is the other side of your trade is the clearing house of the exchange. The clearinghouse checks the creditworthiness of all the counterparties and takes a fee for clearing the trades and in the event that any given market participant folds the clearinghouse uses its fees (which act like insurance) to make the other side whole. This effectively eliminates the risk of counterparty default. In an OTC market you take the risk that the person you are trading against defaults and doesn’t pay you if the reference company actually does default.

Another problem with OTC trading is that the contracts are not fungible. If buy a future from the Chicago Mercantile Exchange clearing corporation and then sell it back I have no net position, that is I have no risk. With a CDS this is only the case if I trade EXACLTLY the same terms with the original counterparty. This hardly ever happens in the CDS market because there are so many different players its unlikely that you will find the best price twice in the same place. Imagine a situation in which I buy a CDS on IBM from JP Morgan and then the next day I sell if for the same price to Bank of America. With respect to IBM and interest rates I have no more risk. If IBM defaults JPM pays me and I turn around and pay BofA. And if no default occurs I just take the LIBOR payments from JPM and send them to BofA and I take the premiums from BofA and send them top JPM until 5 years from now the trade comes off. But of course what this means is that despite having to exposure to IBM, I do have exposure TO BOTH JPMORGAN AND BANKOFAMERICA FOR THE LIFE OF THE TRADE. Because if either one of them defaults I still have to make the payments to the other and I have my exposure one way or the other to IBM back on my book.

The heart of the matter.

And this is where we find ourselves. There are about $40 trillion of CDS outstanding. 80% of them are issued by the top 10 players. Many of the top 10 players are candidates for potential nationalization. If they were nationalized as Krugman suggests with the depositors made whole but the stock and bondholders wiped out you would also create a massive default in the CDS market which would ripple throughout the entire banking system. Tactically it would have the effect of forcing a lot of people to take back onto their balance sheets risks they thought they had hedged and would likely force many banks to seek to raise capital simultaneously. Strategically it would call into question the whole idea of the CDS market and thereby the riskiness of the banking system altogether. What would happen if all of a sudden life insurance could not be relied on? People would save more. A similar thing would happen if you wiped out the default insurance mechanism.


As a result we are stuck with the Geithner plan.

No comments: