## Friday, December 08, 2006

### Regulatory anomalies: RBI edition

Everyone connected with finance in India has their own favourite stories about bizarre things being done in Indian finance as far as the system of policy-making, regulation and supervision is concerned. I recently wrote about the recent SEBI disgorgement order and about the separation between commodity futures and finance.

Today, in Business Standard, A. V. Rajwade has an excellent piece titled Regulatory anomalies which should really be called Regulatory anomalies: RBI edition. His checklist is:

1. The death of the interest rate futures contract:

Take, for example, the question of exchange-traded interest rate derivatives. In theory, a few contracts were introduced with much fanfare in the middle of 2003, but failed to take off. The principal reason was that commercial banks, which are by far the largest players in the bond market, were allowed to use exchange-traded derivatives only for hedging. Inasmuch as all the banks have long positions in government securities, they could have only sold the derivative contracts. With nobody on the opposite side, the market never took off. The irony of the regulatory regime is that, while banks cannot take trading positions in exchange-traded derivatives, they are free to do so in the OTC derivative market. This seems perverse, all the more so when exchange-traded derivatives are far safer than their OTC counterparts!

I agree. In addition to the above, there was one more catch: RBI wanted banks to hold greater capital backing an exposure on the exchange-traded contract when compared with the OTC contract. This was wrong because the exchange-traded contract is backed by the clearing corporation (a zero-leverage, specialised risk management shop holding only transparent assets) while the OTC contract is typically against another bank (a 20:1 leveraged financial firm holding opaque assets). This did not make sense either.

A lot has been made about the difficulties with cash-settled futures on a notional 10-year zero coupon bond. I agree with Rajwade in his emphhasis that the regulatory problems were the real difficulty. When you look at liquid bonds with a roughly 10-year maturity, there are correlations between the market price and the model price in excess of 0.8 - i.e. the contract tracked the (messy) spot market pretty well. The most important reason for the death of the contract was the regulatory treatment.

2. Mistakes in measuring the market risk of bond portfolios:

Consider another example. For measuring the market risk on bond portfolios, the central bank has prescribed the use of modified duration. The modified duration is used to estimate the price change in the value of a bond, arising from higher yields. Regulations require the price change to be calculated based on a 1 per cent yield change at the short end, but 0.6 per cent for longer maturities. This seems to be borrowed from western markets, where the yield volatility is more at the short end, and less at the long end. On the other hand, empirical evidence in India is exactly the opposite: yield volatility at the long end is more than at the short end. Surely we should not be borrowing western models and numbers without testing them empirically for conditions in the Indian market? (The same criticism is applicable to the decay factor of .94 used for calculating EWMA-based value at risk.)

(Aside: the ML estimate of the Risk Metrics lambda for Nifty works out to 0.96, so the difference there isn't economically significant). On the subject of risk measurement for interest rate risk, Jayanth Varma and I had done a fair bit of empirical work with Indian data on the time-series of the spot yield curve, leading up to the SEBI regs for interest rate futures. More generally, on the subject of the interest rate exposure of Indian banks, Ila Patnaik and I have done empirical work which finds that after complying with all RBI regulations, the banks have held huge interest rate anyway. This work was known from early 2002 onwards, but after that, it didn't seem to affect thinking on regulation. To say this in a technical way, after controlling for bank characteristics, the year-fixed-effect dummies are all insignificant. To say this in a non-technical way, even after RBI knew that the banks were carrying huge interest rate risk, nothing was done to solve the problem. [Links to material on these issues]

3. Mistakes in treatment of securitisation:

The securitisation market, which is so necessary for the growth of certain sectors like housing and infrastructure finance, has witnessed significantly reduced activity after the RBI came out with its guidelines on securitisation. While part of the reason could be a change in the interest rate scenario, surely the regulatory prescription that banks cannot account for profit on securitisation upfront, has also contributed to the death of the market. The restriction seems less than logical: on the one hand, the guidelines prescribe that there should be no recourse to the originator in the case of securitised assets; if so, why should the profit not be accounted upfront?

4. Capital controls which impede currency risk management:

Turning now to regulations about the foreign exchange market, recently the Reserve Bank allowed import duty payment (economic) exposures to be hedged in the forward market. On the other hand, it refuses to permit other economic exposures to be hedged. This too seems perverse: after all, import duty exposures are small, while every company in India, producing or consuming commodity kind of goods whose prices in the domestic market are governed by the import parity principle, is facing large economic exposures to the dollar: rupee exchange rate. These need to be managed, particularly when the rupee continues to manifest two-way movement. So what is not a pressing problem (exchange risk on import duty payment) is mitigated, but what is genuinely needed is disallowed! And this is despite both the Committees on capital account making recommendations on the issue.

5. Effectively killing foreign currency accounts:

Recently, the RBI allowed greater access to Indian companies to the maintenance of foreign currency accounts. If the idea is to facilitate such access, surely the stipulation that banks cannot pay any interest on such accounts, should simultaneously have been abolished? It is perverse that we continue with the ban on paying interest, which was introduced in an entirely different set of circumstances (i.e. when the rupee was under pressure), even when the environment has taken a U-turn for the last four years.

6. Complexity:

I also find some of the regulations unnecessarily complex, for the present stage of the market and the kind of business banks do. A few examples: capital adequacy (horizontal and vertical disallowances); derivatives accounting (the separation between below and above 90 days; the rigidity about portfolio hedges, etc.); the plethora of micro-level percentages on cancellation and rebookings of forward foreign exchange contracts.

And finally, his summing up:

Many other examples could be cited, but I suppose that the above list gives an idea of the kind of regulatory quirks, contradictions and irrationalities that exist. Is it that the central bank does not have adequate specialised staff for some of the newer areas of banking? Is it that the decision makers have drowned themselves so much in paper work that they have little time to think? Is it that, in the process, they miss the woods for the trees? Is it that there are too many steps in the decision-making ladder and that, if somebody expresses a negative opinion on a point, the system is reluctant to override it? Perhaps some introspection and review of the systems, quite apart from the specific points I have mentioned above, would be useful, and strengthen the central bank's reputation.

I think RBI staff have IQs which are as good as those found in other government agencies. It is hard to explain analytical and technical blunders such as those described above as deriving from ignorance or incompetence. I think the staff quality and their commitment to their job is just fine, though the RBI HR process leaves much to be desired. The biggest problems lie in the RBI Act and in the complex mandate that RBI is asked to discharge, one that is riddled with contradictions. I think the same staff would do much better in avoiding such mistakes if placed into more focused government agencies where such conflicts of interest were absent, and with improvements in HR procedures.

Please note: LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.