Friday, February 26, 2010

The indomitability of risk

Since my last post leads in to a point that I made in November 2008, I am recycling it here:

The 3 Laws of Risk

1. For every potential return created, we create an opposite potential risk. If banks wish to earn investment income, they need to engage in a financial transaction. The more transactions they enter, the greater the potential income and risk.

2. Risk, once created, cannot be destroyed or absolutely decreased (unless potential return is also decreased or given up). It can only be transferred, shared, or dispersed.

3. In a system where risk has been shared and dispersed among interconnected institutions, any random adverse change in risk in any locality will adversely change the risk in other localities of the system. This is true of all kinds of risk, whether credit default risk, market risk, volatility risk, interest risk, currency risk, sovereign risk, or capitulation risk. They are all inter-connected.

These three laws then lead us into extrapolating five major implications of risk on a financial system.

5 implications of risk on a financial system

1. Any firm trying to earn additional return by taking on the risk of another party, either by entering into a swap, selling an option, guarantee, or insurance to the other party, incurs a potential risk of loss that runs up to a maximum of the loss of the counterparty. AIG and Citibank, more than any others, have learned this the hard way.

2. Any firm that tries to hedge its risk by entering into a swap with another party only succeeds in transferring the original risk to another party. That’s what Goldman was successful in doing to AIG.

3. Worse, if each party in the chain that offers to be a counterparty in this risk management process tries to earn a minimal return on the transaction, it creates additional counterparty risk that was not there before. Hence the more firms involved in a chain of counterparty swaps and derivatives transactions, the greater the resulting risk created in the system. Any and all returns earned from entering into these types of transactions should not paid out to employees or shareholders of the firms, but kept as capital buffer in the event of such systemic meltdown.

4. The more interconnected a system, and the more of these types of transactions it has, the greater the likelihood of a systemic meltdown. Due to the need to bailout insolvent but crucial entities (to the continued functioning of the system) in the event of a meltdown, these firms should not be allowed to enter into derivative transactions.

5. There is no such thing as a benign or risk-free environment. You can never create an environment where risk has been controlled or made benign. It only means risk has been put at rest. But the more risk is put at rest, the greater its potential blowup energy. You may only succeed in causing it to implode more fiercely, and in unexpected ways. This makes the case for having a centralized clearing house that more important. With a clearing house, we can better track where in the system risk is building up, and we can correspondingly require greater collateral protection before something unexpected blows up.


investment catcher said...

This is a wonderful website!! 
勉強になります。ありがとう! Thank you!!

I'd be pleased if you exchange reciprocal link with me.

Easy investment : Balance sheet

The Arthurian said...

I am drawn to organized expressions of thought such as this from Ritwik Priya, and your list here.

If you are up for disorganized critique, I offer one: Though you have done well with it, I think you are too entangled in details. Depends on one's objective, to be sure. But I can take your First Law of Risk and use it to solve the economic problem. You write: "The more transactions they enter, the greater the potential income and risk."

Solution: Don't let them enter so many financial transactions. As Simon Johnson has it: "But is it really healthy – or even sustainable – to have a finance sector as large as what we face today?" Or as I have it, the problem, simply, is excessive reliance on credit.

I see that you get there eventually, in your Implication #4. But then your idea there sounds like micromanagement. And you move on to your #5, which seems to be your conclusion: that a Centralized Clearing House is needed so that we may continue to have too many financial transactions.

Offered as something to kick around.

Rogue Economist said...

Thanks Arthurian, for the comments and thoughts. My aim here really is to just put forth some principles (the phrases in bold), though I did get to some operational implications (the phrases following the bold).

The desirability of a specific operationality may actually vary from person to person. Some people may want to ban derivatives altogether, some may just want to regulate it, and some may just want participants to have sufficient capital and/or collateral. And then some people want to put them in a clearinghouse, some onto exchanges.

Before we agree on the action, we should at least acknowledge the principles - and the most basic of which is that risk once created does not disappear (without undoing a transaction), it is only reallocated to other parts of the system.

The Arthurian said...

Hello Rogue. Sir, your last paragraph above is exquisite. And (for the record) I thought your post was very good. (Otherwise I would not have been drawn to comment.) And you seem to have understood that it was my intent, at least, to be interesting (as opposed to, say, rude). For this I am grateful.

You got invited to be a contributor to Seeking Alpha?? Nice.

RE: Your "desirability of a specific operationality" paragraph... You list a variety of possible policy preferences. I think this is what Alan Blinder and Robert Heilbroner have called "disarray." I think we are past the time for such niceties.

Surely you have a preference, no? Or are you satisfied to fiddle while... well, you know.

Rogue Economist said...

Arthurian, to answer your question as to my preference, I wouldn't mind it if we did away with CDS altogether. But that is coming from someone who never had any use for them, I'm sure that US capital marketmakers would say otherwise.

But personally, if you think a particular investment is sufficiently risky that you need to buy insurance on it VIA CDS, then you are better off not taking the risk at all. The presence of CDS merely encourages you to buy something which you know to be junk because in your mind, with the CDS, you can collect on it somehow.