Prof. Jayanth R. Varma's Financial Markets Blog

Photograph About
Prof. Jayanth R. Varma's Financial Markets Blog, A Blog on Financial Markets and Their Regulation

© Prof. Jayanth R. Varma

Subscribe to a feed
RSS Feed
Atom Feed
RSS Feed (Comments)

Follow on:

Sun Mon Tue Wed Thu Fri Sat

Powered by Blosxom

Wed, 23 Nov 2011

European Banks as America's Shadow Banks

Hyun Song Shin delivered the Mundell-Fleming lecture at the IMF Annual Research Conference earlier this month. This very interesting lecture argues that European banks essentially constitute the US shadow banking system.

While there has been much discussion of how the US has been relying on capital flows from Asia, there is little mention of Europe as a financing source. This is because Europe is not a significant source of net capital flow for the US – after all, Europe has a roughly balanced current account, and is not therefore a source of capital. The picture changes when one looks at gross capital flows instead of net capital flows. This is because European banks borrow dollars in the US and lend the dollars back in the US. This too is well known because it was a major source of distortions in the dollar Libor market and in the currency swap market (see for example, my blog post from April 2008 on this issue).

What makes Shin’s paper important is his demonstration that the sheer scale of this gross flow is much bigger than most people imagined. At least, it is an order of magnitude larger than what I thought it was. In fact, he shows that for a brief period in 2007 and early 2008, the total dollar assets of non US (largely European) banks exceeded the total assets of US commercial banks.

As a result, European banks while not being important sources of net capital, were hugely important sources of liquidity and credit transformation in the US financial system. They created liquid and apparently safe assets out of illiquid and risky loans to US borrowers, and they did this mostly through the shadow banking system (securitization and repos). As Shin points out, this is hugely important in the context of the European crisis. The ongoing deleveraging by European banks could be painful for the US financial system even if none of the big European banks fail.

I am tempted to think of the US as a giant CDO (collateralized debt obligation). China (and the rest of Asia) own the super-senior and senior pieces (Treasury and Agency paper) while Europe holds the equity piece. Much of the complacency about the US financial position is based on the idea that Asia cannot find another home for its money and so the super-senior piece will continue to find buyers. When all else fails, the US Federal Reserve has also provided buying support for this piece through its QE (quantitative easing) programmes. However, the real challenge in selling the CDO is in selling the equity piece because this piece has no natural buyer, and the only buyer in town might be delevering itself out of existence.

Posted at 16:18 on Wed, 23 Nov 2011     View/Post Comments (0)     permanent link

Tue, 15 Nov 2011

Revisiting CME and LCH handling of Lehman default

A year and a half ago, I had a blog post comparing how CME and LCH.Clearnet coped with the Lehman default. I raised a number of questions and concluded by saying that:

In the context of the ongoing debate about better counterparty risk management (including clearing) of OTC derivatives, I think the regulators should release much more detailed information about what happened. Unfortunately, in the aftermath of the crisis, it is only the courts that have been inclined to release information – regulators and governments like to regard all information as state secrets.

While regulators have still not been too forthcoming, considerable new information has become public since then. Somehow I did not get around to revisiting this issue until I received a comment on my blog post a few days ago from Risk Dude saying:

LCH utilized excess margins from other products to auction the IRS book under margin. So it’s a bad comparison.

This comment appears to be quite correct. The best material that I have read on the subject is the book by Peter Norman entitled The Risk Controllers: Central Counterparty Clearing in Globalised Financial Markets, (Wiley, 2011). Chapter 2 of the book deals exclusively with the Lehman bankruptcy, and Norman quotes a personal conversation with the LCH.Clearnet Chief Executive, Liddell in which Liddell says:

We always thought that a common default fund would be the main benefit from being a multi-asset CCP. ... In fact, the big and far more valuable discovery during Lehman was that the initial margin in each market was completely fungible. ... There were inverse correlations with prices moving one way in some markets, another way in others. As we managed to liquidate some of the portfolios more quickly than others, it meant that the margin that was left after some had been liquidated was available to cover risk somewhere else. That was a massive, massive benefit. ... it meant we had a much bigger cushion all the time. ... We didn’t have the same sort of urgent need to get rid of everything straight away.

Norman also explains that over the same weekend that Lehman failed, the energy futures exchange, ICE, was to move all its positions from LCH.Clearnet to ICE Clear Europe. On Sunday evening around 7 pm, the FSA, LCH.Clearnet and ICE Clear agreed to defer this move. This meant that during the liquidation of Lehman position, the ICE positions (and more importantly, the associated margins) were also available to LCH.Clearnet and this was a big benefit.

Once again, I hope that regulators will disclose more details about what happened during those dark days.

Posted at 14:20 on Tue, 15 Nov 2011     View/Post Comments (0)     permanent link

Mon, 07 Nov 2011

Unifying initial and maintenance margins

The bankruptcy of MF Global prompted CME Clearing to temporarily unify maintenance and initial margins (h/t Kid Dynamite). I would argue for a permanent abolition of the distinction between these two margins.

Initial margin is the margin that market participants must pay when they initiate a position while the maintenance margin is the level at which the margin must be maintained subsequently. I believe that this distinction is quite silly. The whole purpose of daily mark to market is to ensure that every day begins on a clean slate. There is absolutely no difference between a position initiated yesterday and a position initiated today because yesterday’s losses and gains have already been settled in cash. To pretend otherwise is a delusion. CME Clearing stated in its notice that:

Maintenance margins are set to provide appropriate risk management coverage. Initial margins are set to provide an additional buffer against future losses in the account.

In these troubled times, no one can argue against additional buffers, but the idea of applying buffers selectively to some positions is absurd. The situation becomes even more ludicrous when we consider a large portfolio of positions where new positions do not necessarily correspond to new risks.

In the days when fund transfer was slow and painful, it made sense for customers to deposit extra margins to avoid the hassle of managing daily cash inflows and outflows. This is no longer relevant with today’s electronic payment systems. In any case, that should be a matter for the individual customer to decide. There is little merit in the clearing corporation micro-managing the cash management policies of its customers.

CME Clearing is absolutely right in saying that “Maintenance margins are set to provide appropriate risk management coverage.” It should stop with risk management and leave cash management to market forces.

I am quite disturbed by the statement from CME Clearing that “This is a short term accommodation to maintain market integrity and provide temporary relief to customers whose accounts have been disrupted by this event.” One expects risk managers at clearing corporations to be ruthless and uncompromising on protecting the clearing corporation from risks. Words like accommodation and relief should not be part of their vocabulary at all. Instead CME Clearing should have said that since it is illogical for customers to pay a higher margin merely because their positions have been transferred from MF Global to another clearing member, they are permanently unifying their margins.

Posted at 19:28 on Mon, 07 Nov 2011     View/Post Comments (0)     permanent link

Sun, 06 Nov 2011

Pricing of cheques and electronic payments

India seems to be now moving to a system where it costs the originator more to make an electronic payment than to issue a cheque. This is dysfunctional because the cheque imposes costs both on the receiver and on the payment system (the paying bank, the collecting bank and the clearing house). Of course, in a free market, banks are free to levy charges as they deem fit and charges do vary significantly across banks in India. What is interesting is (a) that the resulting market equilibrium is so perverse, and (b) that this perversity is a recent phenomenon.

In the past, many banks were not charging for electronic payments through the National Electronic Fund Transfer (NEFT) system operated by the Reserve Bank of India. Recently, more and more banks seem to be introducing charges at the maximum permitted rate of Rs 5 per outbound electronic transfer for small transactions. By contrast, most banks charge only about Rs 2 per cheque leaf, and the first 20-30 cheques per year are typically free.

This means that it may now be advantageous for many retail consumers to issue cheques instead of using NEFT for inter-bank transfers within the same city. This privately optimal decision does however have a huge negative externality. The cost to the paying and collecting bank of processing this cheque might be about Rs 50 each, and there are costs elsewhere in the system (for the payee who has to deposit the cheque and for the clearing house which has to process it).

What I would like to know is whether this pricing decision is optimal for the individual banks. Perverse as the end result is, the pricing can be rational for individual banks if customers sophisticated enough to use NEFT are assumed to be relatively price insensitive. In that case, the customers continue to use NEFT and the bank simply pockets another source of revenue. The alternative hypothesis is that the costing and transfer pricing system in many banks is badly broken. In that case, the costs of processing paper cheques is not fully reflected in the pricing decision, while the cost of the electronic transfer is a transparent out of pocket cost (the fee charged by NEFT). A naive cost-plus-pricing system does the rest of the mischief. It would be interesting to figure out which hypothesis is closer to the truth.

In either case, the problem can be solved by a Pigovian tax on cheques.

Posted at 20:29 on Sun, 06 Nov 2011     View/Post Comments (6)     permanent link