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
jrvarma@iima.ac.in

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

Follow on:
twitter
Facebook
Wordpress

August
Sun Mon Tue Wed Thu Fri Sat
       
2010
Months
Aug
2009
Months

Powered by Blosxom

Fri, 27 Aug 2010

Loss absorbency of regulatory capital

The Basel committee on banking supervision has put out a proposal to ensure the loss absorbency of regulatory capital at the point of non-viability.

The proposal points out that all regulatory capital instruments are loss absorbent in insolvency and liquidation – “they will only receive any repayment in liquidation if all depositors and senior creditors are first repaid in full.” However, the financial crisis has revealed that many regulatory capital instruments do not always absorb losses in situations in which the public sector provides support to distressed banks that would otherwise have failed.

The solution that is proposed is as follows:

All non-common Tier 1 instruments and Tier 2 instruments at internationally active banks must have a clause in their terms and conditions that requires them to be written-off on the occurrence of ... the earlier of: (1) the decision to make a public sector injection of capital, or equivalent support, without which the firm would have become non-viable, as determined by the relevant authority; and (2) a decision that a write-off, without which the firm would become non-viable, is necessary, as determined by the relevant authority.

I am unable to understand how such tweaking of contractual terms will get around the fundamental problem that governments are unwilling to impose losses on banks and their stakeholders. In the United States when the government injected TARP capital into the banks, it forced the healthy banks also to take the capital to eliminate any stigma associated with TARP capital. Even if the proposed clause were present in the regulatory capital instruments issued by the insolvent banks of that time, clearly the clause would not have been triggered by the injection of TARP capital in this gentle form.

Posted at 15:51 on Fri, 27 Aug 2010     View/Post Comments (2)     permanent link


Fri, 20 Aug 2010

Two curve models

Updated: corrected quote attibution of the second quote.

I am posting for comments an introductory note on two curve models. This note which I wrote largely to improve my own understanding of this subject, represents my interpretation of the results presented in various papers listed in the bibliography and does not claim to contain any original content.

The following two quotations give a sense of what two curve models are all about:

LCH.Clearnet Ltd (LCH.Clearnet), which operates the world’s leading interest rate swap (IRS) clearing service, SwapClear, is to begin using the overnight index swap (OIS) rate curves to discount its $218 trillion IRS portfolio. Previously, in line with market practice, the portfolio was discounted using LIBOR. ... After extensive consultation with market participants, LCH.Clearnet has decided to move to OIS to ensure the most accurate valuation of its portfolio for risk management purposes. (LCH.Clearnet, June 17, 2010)

Ten years ago if you had suggested that a sophisticated investment bank did not know how to value a plain vanilla interest rate swap, people would have laughed at you. But that isn’t too far from the case today. (Deus Ex Machiatto, June 23, 2010)

The topic is quite complex even by the standards of this blog and it takes me twelve pages of occasionaly dense mathematics to explain what I have understood of its basic ideas. To fully understand two curve models you would need to read a lot more of even denser mathematics. If you do not have the patience for that, you should just ignore this post.

What I would really love is for some of my readers to provide comments and suggestions to improve this note and correct any errors that might be there.

Posted at 10:19 on Fri, 20 Aug 2010     View/Post Comments (3)     permanent link


Fri, 13 Aug 2010

RBI on new bank licences in India

I have a column in today’s Financial Express about the RBI’s discussion paper on new bank licences

Reserve Bank of India (RBI) has begun a very open and transparent process of thinking about new bank licences with a discussion paper that outlines the key issues, presents the pros and cons of the alternatives and also highlights the international experience.

While discussing these important issues, it is necessary to keep two things in mind. First, we must learn the right lessons from the experience of the new bank licences given out in the post-reforms period. Second, the global financial crisis has changed the way we think about bank regulation and competition.

Of the 10 new banks licensed in the first phase, the majority were failures in the broadest sense, but a few became outstanding successes. The viewpoint in the RBI discussion paper is that we must identify the causes of the failures and avoid making the same mistakes when granting new licences.

I am of the completely opposite persuasion. The success of the 1993 experiment was that enough licences were granted to permit a few success stories to emerge, despite a low success rate. Capitalism to me is about liberal experimentation and ruthless selection. What is wonderful about the 1993 experiment is that the failures (although many) were on a small scale and were (with one exception) quite painless, while the successes were outstanding. This makes for an extremely favourable risk-reward ratio.

While handing out new licences, the goal should not be to avoid failures; it should be to maintain the same attractive risk-reward ratio. I believe that this again requires the same approach—granting many licences, allowing the market to weed out failures at an early stage, and giving enough freedom to allow the successes to bloom.

It is impossible to figure out right now what will or will not work in the emerging banking environment. It is very unlikely that a successful bank emerging from the new set of licences will be a clone of, say, HDFC Bank. HDFC Bank and its peers succeeded by identifying what the then existing Indian and foreign banks were not doing well or not doing at all, and then setting out to deliver that with high levels of efficiency. But thanks to their very success, that space has now become overcrowded and hypercompetitive. A new bank starting today will have to find a new space in which to make its mark.

No regulator can predict what that new business model will be. As Hayek once wrote: “Competition is valuable only because, and so far as, its results are unpredictable and on the whole different from those which anyone has, or could have, deliberately aimed at.”

What is important is to keep failures small and manageable, and the way to do that is to allow banks to start small. The RBI discussion paper veers towards allowing only large banks in the belief that this will keep out people who are not serious. This is a mistake that financial regulators around the world appear to make.

I still remember that at the height of the Asian Crisis, one of the few healthy and solvent Indonesian banks was one of their smallest banks (Bank NISP). But the Indonesian central bank’s response (probably encouraged by the IMF) was to impose one of the highest minimum capital requirements in the world. It would have been utterly hilarious were it not so tragic.

Equating money with seriousness is a misconception unique to the financial elite. The rest of the world does not think that a student who has got admission by paying a large donation or capitation fee is a more serious student than the one who came in on the merit list. But financial regulators in India and elsewhere have an abiding belief in the ennobling power of money. Sebi has also been talking about increasing capital requirements for its regulatees.

I believe, on the other hand, that the global financial crisis has indicated that in the world of finance, size is evil in itself. Simon Johnson’s brilliant new book 13 Bankers: The Wall Street Takeover and the Next Financial Meltdown argues for a size limit of “no more than 4% of GDP for all banks and 2% of GDP for investment banks.”

By this standard, which I consider quite reasonable, India has seven banks above the size limit, including one that is almost 20% of GDP (I have taken the bank asset data from the RBI’s Statistical Tables Relating to Banks of India, 2008-09). Of these seven banks, only one is in the private sector, but two other large private banks are growing fast enough to cross the size limit in the next few years.

I believe, therefore, that RBI would grant a large number of new bank licences that will rapidly bring down the concentration in the banking system. India needs a lot of banks that are small enough to fail and fewer that are too big to fail.

Somewhere in the long chain from the keyboard to the printed newspaper what should have been a “should” or a “must” in the beginning of the last paragraph became “would.” Who am I to predict what the RBI would or would not do? I do hope however that my unintended prediction turns out right!

Posted at 04:43 on Fri, 13 Aug 2010     View/Post Comments (5)     permanent link


Tue, 10 Aug 2010

Has the greatest financial risk gone away?

I have long argued that the greatest global financial risk is not toxic derivatives or bad loans – it is the unnerving possibility that P=NP. P≠NP is a conjecture about an abstruse problem in mathematics, but too much of computer security depends on it. It is likely that if P=NP, then many financial assets that are recorded as electronic entries could suddenly evaporate because those entries could all be hacked. Since almost all financial assets today are in electronic form, that would be the end of finance as we know it.

During the last couple of days, a purported proof that P≠NP has been circulating on the web (hat tip Bruce Schneier). The hundred page paper by Vinay Deolalikar of HP Research Labs, Palo Alto utilizes and expands “ upon ideas from several fields spanning logic, statistics, graphical models, random ensembles, and statistical physics” to obtain the purported proof. We still do not know whether the proof is correct (see here and here)

It reminds me of the early days of the initial claims of Wiles’s proof of Fermat’s last theorem or Perelman’s proof of the Poincare conjecture. Everybody agrees that it is a serious proof, but nobody knows whether the proof is right. But if Deolikar is right, the biggest financial risk of all has gone away.

Posted at 12:08 on Tue, 10 Aug 2010     View/Post Comments (3)     permanent link


Mon, 09 Aug 2010

RBI proposes CDS market by dealers and for dealers

The Reserve Bank of India has released the report of the Internal Group on Introduction of Credit Default Swaps for Corporate Bonds in India.

What is proposed is a market by dealers and for dealers. Users can only buy CDS protection, and they have to buy them from dealers (banks and other regulated entities) who are the only people allowed to sell CDS. But the most diabolical recommendation is the following:

The users can, however, unwind their bought protection by terminating the position with the original counterparty. ... Users are not permitted to unwind the protection by entering into an offsetting contract. [Paragraph 2.7.6(ii) on page 19]

This leaves the unwinding users at the complete mercy of the original dealer from whom they bought CDS protection – that dealer can fleece the users knowing fully well that they cannot go elsewhere. Under these terms, it would be utterly imprudent for a company to use CDS at all. Well designed corporate risk management policies should demand the availability of competitive quotes both at inception and at unwind, and should therefore completely prohibit the use of the proposed CDS market. Of course, India has a number of imprudent companies with poor risk management policies; perhaps, the RBI proposed market is suitable only for them.

The other frightening part of the proposals is that at a time when the entire world is worried about the dangers of an opaque CDS market, the report envisages the creation of a CDS market without a trade repository let alone a clearing mandate. The report envisages a trade reporting platform at some unspecified future date, but the establishment of this platform is not a precondition for CDS trading to begin. As far as clearing is concerned, the report makes the right noises, but it is clear that the RBI is not very keen on this.

Even when the trade repository starts functioning, it is unclear what transparency it would bring. First of all, the recommendation uses the word “may” which deprives it of operational significance:

The reporting platform may collect and make available data to the regulators for surveillance and regulatory purposes and also publish, for market information, relevant price and volume data on CDS activities such as notional and gross market values for CDS reference entities broken down by maturity, ratings etc., gross and net market values of CDS contracts and concentration level for major counterparties. [Paragraph 4.2.1 page 40]

Second, the report provides a trade reporting format (Form I in Annex IV) and this format does not include any data on prices at all. This means that even when the reporting platform starts working, it would not provide price transparency even on a post trade basis. What more could the dealer wish for when it comes to fleecing the customer?

One relatively minor issue which I am not able to figure out is whether RBI intends CDS to be used to hedge loans and not only bonds. The report clearly states that only bonds can be reference obligations for CDS, but it is silent on whether loans can be deliverable obligations. Some parts of the report appeared to be deliberately written vaguely to allow loans to be hedged. For example, “The users can buy CDS for amounts not higher than the face value of credit risk held by them” (Paragraph 2.7.6(i) page 19). That would allow loans to be hedged, and what is deliverable would presumably be decided by the Determination Committee which can be counted on to go with the banks on this issue. Whether loans can be hedged is not terribly important, but if the intention is to permit it, why not say so explicitly?

Coming back to the important prudential issue, I believe that India needs a CDS market, but I am concerned that a CDS market as proposed by the RBI would create more systemic risks than it would eliminate. If these are the only terms on which a CDS market can be had, it would be better for the country that we do not create such a market at all.

Posted at 15:41 on Mon, 09 Aug 2010     View/Post Comments (2)     permanent link


Criticism of monetary policy

There has been a lively debate in India on senior central bank officials criticizing monetary policy decisions in which they may have participated. This debate has tended to focus on the harm that such alleged indiscreetness can do, while I think the important question is how to design the conduct of monetary policy in a manner where open debate does not cause harm.

Consider this passage in a paper last month by a member of the US FOMC that decides monetary policy in that country:

The U.S. is closer to a Japanese-style outcome today than at any time in recent history. In part, this uncomfortably close circumstance is due to the interest rate policy being pursued by the FOMC.

Or this from a UK MPC member who last month titled his speech provocatively as “How long should the song remain the same?”:

The normal monetary policy reaction to a sustained period of above target inflation would be to tighten policy, to create demand conditions which are more conducive to restraining price increases and bringing inflation back to target. But so far, the Committee has not supported that course of action – and is keeping monetary policy extremely loose. ...

Last month, however, I dissented from this approach and voted for a small rise in interest rates. And in today’s speech I want to set out the thinking behind my view of current economic prospects and the implications for UK monetary policy that led me to that decision. ...

The MPC has a clear remit, which is to keep inflation on target at 2% over the medium term. ... we need to adjust the policy settings we put in place to head off the downside risks to inflation identified in the immediate aftermath of the big financial shocks in late 2008 and early 2009.

I think such open debate and criticism strengthens the conduct of monetary policy by allowing divergent points of view to be heard and considered. Alternative analytical frameworks can thus be developed and are available to the policy makers if and when they choose to change their mind. My knowledge of either the theory or practice of monetary policy is very limited, but I like to believe that monetary policy is closer to a science that progresses through informed debate, rather than a dark art that derives its mystique and efficacy from a veil of secrecy.

Unfortunately, criticism of monetary policy decision by senior officials themselves can be prevented from doing harm only in a culture of transparency where minutes of monetary policy deliberations are published openly so that dissenting voices are not misinterpreted by the markets.

Posted at 07:29 on Mon, 09 Aug 2010     View/Post Comments (0)     permanent link


Wed, 04 Aug 2010

Time stamping by stock exchanges

I just finished reading an interesting study of the flash crash in the US on May 6, 2010 by Nanex which is a data feed company which provides high frequency real time trade and quote data for all US equity, option, and futures exchanges (over one million updates per second). The study was published in mid June and linked by Abnormal Returns a week later, but I got around to reading it only today after several blogs talked about it two days ago.

The study claims:

Beginning at 14:42:46, bids from the NYSE started crossing above the National Best Ask prices in about 100 NYSE listed stocks, expanding to over 250 stocks within 2 minutes (See Part 1, Chart 1-b). Detailed inspection indicates NYSE quote prices started lagging quotes from other markets; their bid prices were not dropping fast enough to keep below the other exchange’s falling offer prices. The time stamp on NYSE quotes matched that of other exchange quotes, indicating they were valid and fresh.

With NYSE’s bid above the offer price at other exchanges, HFT systems would attempt to profit from this difference by sending buy orders to other exchanges and sell orders to the NYSE. Hence the NYSE would bear the brunt of the selling pressure for those stocks that were crossed.

Minutes later, trade executions from the NYSE started coming through in many stocks at prices slightly below the National Best Bid, setting new lows for the day. (See Part 1, Chart 2). This is unexpected, the execution prices from the NYSE should have been higher -- matching NYSE’s higher bid price, unless the time stamps are not reflecting when quotes and trades actually occurred.

If the quotes sent from the NYSE were stuck in a queue for transmission and time stamped ONLY when exiting the queue, then all data inconsistencies disappear and things make sense. In fact, this very situation occurred on 2 separate occasions at October 30, 2009, and again on January 28, 2010. (See Part 2, Previous Occurrences).

If this is really true, then instead of criticizing only high frequency traders, we must also direct some of the blame at the exchanges which behaved irresponsibly. Why cannot the exchange provide time stamps both of the time that a quote entered the queue and when it exited the queue?

Organizations with none of the self regulatory responsibilities of an exchange do this kind of thing routinely. One of the things that I read today was this study about dissemination of press releases at public websites. Yahoo! Finance provides two timestamps when it reports a press release on its web site – first is the timestamp of the press release itself and the second is the time stamp of when it was published on Yahoo! Finance. By comparing these two time stamps, the study concludes that “the average delay was 83 seconds. The fastest was 24 seconds and the slowest was 237 seconds or almost 4 minutes. The median was 80 seconds.”

It did not require a regulator framing rules for a public website to provide two timestamps in its stories. But perhaps exchanges will do the right thing only if they receive a direction from the regulator. And perhaps the regulator will find itself easily persuaded that this simple thing is too costly, complicated or confusing to implement.

Posted at 21:21 on Wed, 04 Aug 2010     View/Post Comments (1)     permanent link


Mon, 02 Aug 2010

Risk Management for Derivative Exchanges

I wrote a chapter on risk management lessons from the global financial crisis for derivative exchanges for a book edited by Robert W. Kolb on Lessons from the Financial Crisis: Causes, Consequences, and Our Economic Future.

Kolb Book Front Cover

During the global financial crisis, no major derivative clearinghouse in the world encountered distress, while many banks were pushed to the brink and beyond. This was despite the exchanges having to deal with more volatile assets—equities are about twice as volatile as real estate, and natural gas is about 10 times more volatile than real estate. Clearly, risk management at the world’s leading exchanges proved to be superior to that of the banks. The global financial crisis has shown that the quality of risk management models does matter.

Three important lessons have emerged from this experience:

  1. The quality of risk management models can be measured along two independent dimensions: crudeness versus sophistication and fragility versus robustness. The crisis of 2007-2009 has shown that of these two dimensions, the second dimension (robustness) is far more important than the first dimension (sophistication).
  2. An apparent structural change in the economy and the financial markets may only be a temporary change in the volatility regime. Risk models that ignore this can be disastrous.
  3. Risk models of the 1990s, based on normal distributions, linear correlations, and value at risk, are obsolete not only in theory but also in practice.

Most of the chapter paper deals with these lessons from the crisis of 2007-2009. In the final section, the paper argues that as derivative exchanges prepare to trade and clear ever more complex products, it is important that they refine and develop their risk models even further so that they can survive the next crisis.

The chapter is based largely on a paper that I wrote in February 2009.

Posted at 13:12 on Mon, 02 Aug 2010     View/Post Comments (2)     permanent link