I was recently made aware of a fascinating speech2 on this topic given by Mr. V. K. Sharma (ED, RBI) in Mumbai on November 8th, 2012. In this speech he takes the example of three fixed income derivative contracts and examines why all three of them have failed to live up to their original expectations. He concludes his speech by making some important recommendations on what might be done to redress these failures.
He opens his remarks by commenting on the Interest Rate Swaps (IRS) market in which, “currently the 5 year IRS yield is trading at a negative spread of 120 basis points to 5 year G-Sec [Government of India Security]!” and as a consequence, “… while the G-Sec yield curve is almost flat, the IRS yield curve is steeply inverted…”. For fixed rate receivers, i.e., holders of government securities, he argues that this represents a very profitable and risk-less arbitrage opportunity “…involving buying corresponding maturity G-Sec in the cash market [therefore receiving fixed]3 by financing it in the overnight repo market4 [therefore paying the overnight repo rate], and paying fixed, and receiving overnight, in the IRS market!”. While it is not clear from the text of Mr. Sharma’s speech why, despite the existence of this enormous mispricing, the IRS market is so large (over Rs.50 trillion when the total G-Secs issued are only at Rs.30 trillion), he does express the concern that it represents a “veritable IRS ‘Super-Bubble’” and signifies a “massive ‘financial sector-real sector imbalance’”, created largely on account of, as I understood it, insufficient activity by arbitragers, in particular the public sector banks who have largely stayed out of the market5. This is certainly an interesting situation because it is the presence of arbitragers that is generally associated with existence of asset price bubbles and not the other way around.
In examining the Credit Default Swap6 (CDS) market, Mr. Sharma makes a similar point and argues that the massive mispricing in the IRS market made the CDS far too expensive to buyers and therefore the CDS market remained still-born and simply no arbitrage activity was even possible. He goes on to suggest that this is one of the reasons that the corporate bond market has itself not taken off because buyers of corporate bonds cannot purchase reasonably priced default protection. The Interest Rate Futures7 (IRF) in his view also suffered from insufficient arbitrage activity between the physicals market and the IRF market, allowing large levels of mispricing to persist.
He expresses the concern that in the absence of sufficient arbitrage activity not only have the derivative markets failed to evolve but their inability to transmit liquidity has led to continuing market fragmentation which in turn has been “…the biggest undoing of an efficient, deep, liquid, organically connected and seamlessly integrated financial market which is also a ‘sine qua non’ for effective, efficient, and instantaneous monetary transmission.”. In order to increase the level of arbitrage (and hedging) activity and to anchor the instruments and their prices to the real-sector in a manner that enhances liquidity and efficient price discovery in the physicals market, the following are a few of the recommendations that he makes8:
- Do not launch cash-settled IRF contracts because they will further increase the level of disconnection between the physicals and the futures markets.
- Do not permit IRF contracts that are just settled in the most liquid (and cheapest to deliver) G-Sec because it represents less than 10% of the physicals volume and will not therefore serve the liquidity transmission function.
- Removal of ‘hedge effectiveness’ criterion of 80% to 125% so that there is a great incentive to hedge even if the hedge is less than perfect9.
- Roll back the Held to Maturity (HTM) protection10 offered to banks. Given the very large size of the IRS market banks should easily be able to substitute this with hedges, particularly if they are allowed some leeway in computing effectiveness in accordance with the previous recommendation. This could particularly improve the participation of public sector banks in the fixed income derivative markets.
- Permitting delivery-based short-selling in the cash market and the introduction of term repo, and reverse repo, markets, both co-terminus with the tenure of futures contracts.
I found that the speech made very fascinating reading and offered important insights into why these markets have failed to take off and what might be done about it so that they can grow in an orderly manner.
—
- 1 – This blog post reflects my personal understanding of the speech given by Mr. Sharma and may not have accurately captured what he wanted to convey.
- 2 – Link to the full text of the speech: http://rbidocs.rbi.org.in/rdocs/Speeches/PDFs/SABF121112_F.pdf?cat_id=46&event_id=3
- 3 – Square brackets are mine.
- 4 – Overnight repo stands for Overnight “Repurchase Obligation” and represents the sale by the buyer of the G-Sec (or any other security) today for the agreement (or obligation) to repurchase it at an agreed upon price. The difference in the sale price and the agreed repurchase price represents the “repo rate”. For more detail see: http://en.wikipedia.org/wiki/Repurchase_agreement
- 5 – He reports that private banks constitute 18% of the market, foreign banks 80%, while public sector banks account only for 2% of the market despite controlling about 74% of the total assets of the banking system.
- 6 – A Credit Default Swap (CDS) is an agreement which refers to a specific Reference Bond and in which the buyer of the CDS makes to the seller a series of payments (comparable to a credit spread) and in return the seller agrees to compensate the buyer of the CDS, if the Reference Bond experiences a default. For more detail see: http://en.wikipedia.org/wiki/Credit_default_swap
- 7 – An Interest Rate Futures (IRF) contract is a futures contract in which the underlying asset is an interest bearing security or a basket of these securities chosen by the exchange on which the contract is listed in a manner consistent with regulatory guidelines. See the link for the RBI guideline on this: http://www.rbi.org.in/scripts/NotificationUser.aspx?Id=6912&Mode=0
- 8 – Full list and their exact wording are in the text of the original speech.
- 9 – Most accounting systems require all derivatives to be marked-to-market but do not generally require this for the underlying financial assets (“physicals”). This could result in substantial volatility in the profit-and-loss accounts of firms and financial institutions. In order to help mitigate this and better align accounting reality with the underlying economic reality, firms and financial institutions are permitted to net off “effective hedges” against the underlying.
- 10 – The HTM protection allows the banks to classify a host of securities as being intended to be held until they mature so that they then need not be marked to market. This protection often severely inhibits trading and hedging activity by banks.
16 Responses
Dr Mor,
Thanks for sharing this speech.
Mr Sharma makes this rather extraordinary observation on pg 8:
“For any ‘cash-settled’ derivative, where physical settlement is possible, tends to become a “non-derivative”, violating the cardinal principle of arbitrage-free pricing/valuation and, therefore, as I said before, comes to exist almost entirely for its own sake and to almost complete exclusion of the larger public policy purpose of sub-serving the hedging needs of the real sector, creating a massive “financial sector-real sector imbalance” and, thus, in turn, become the very antithesis of responsible financial innovation.”
Any cash-settled derivative becomes a non-derivative? Exchange-traded currency futures are cash-settled, and are among the great successes of Indian financial market design. I left my ringside view of the markets in 2010, but even then, the volume of USDINR futures traded daily was significantly more than the spot (and that is an extraordinary statistic, given that futures began trading only in 2008 or so). The fact remains that IRF contracts have been designed very poorly: cash-settlement, for instance, is an absolute no-brainer. How is one to deliver an illiquid G-sec on expiry? Further, the reluctance of the RBI to permit short-dated rates futures is utterly ridiculous. Are we to believe that this market is going to really – shudder – affect the monetary policy transmission mechanism?! I am very glad, however, to note that the HTM permissions are to be rolled back, but I fear that this is too little, too late. Exchange traded contracts need good starts, otherwise one ends up on the wrong side of a chicken-and-egg liquidity problem.
On the issue of the (non)-arbitrage of the IRS vis-a-vis G-secs, if I do lock in an arbitrage over a 5 year term, then I only see my profits if I hold till maturity, correct? Is it possible that traders would rather not lock up capital for up-to 5 years for returns that are risk-less but rather feeble?
Rahul: Thank you for your comments. I have requested Mr. Sharma to respond to all comments on this blog. I hope he will do so.
I have a fundamental problem with the concept of credit rating which probably is a key ingredient the CDS/corp bond market. I cannot fathom how the probability of default for say a AA rated infrastructure company can be same as a AA pharma company and the market prices both the bonds similarly. I would be glad to understand different perspectives on this.
Amit: I am not an expert on this but the way I understand it the ratings represents the “ignorance boundary” for a company. For example for a set of bonds / loans that have the same rating (such as “A”), based on available information, they all believed to be drawn from the same default distribution. We already know that most of these bonds / loans will not go into default but based on the data we have with us we cannnot ex-ante separate them out. The CDS market then hopes to offer protection to holders of a specific pool of bonds by using a benchmark bond with a similar rating. The holders of the pool hope that the idiosyncratic effects (such as those that you mention) would get diversified away and what remains will be “effectively” hedgeable by the CDS.
Thanks Nachiket, my concern is also not just limited to company specific risk but also industry risk, the variables/factors that can impact the infra sector are quite different from those that can impact the pharma sector (although there will be an significant overlap too if one generalises the risks), based on rating rationales that I have read, the ratings have definitely not straddled all the industry linked factor induced credit sensitivities and any factor ignored could have a significant impact if it actually manifests. The idiosyncratic diversification would in my view would largely diversify away the company specific risk factors but may not be able to diversify away the industry factors which are enormous in no. and vary in magnitude of credit impact ( I can say with relatively higher certainity in the infrastructure business).
Thanks Nachiket, my concern on the ratings is largely relating to industry specific risk factors (which are quite large in no. and differ in thier impact on the credit)which in my opinion differ significantly between sectors although if one wants to generalize there could be a significant overlap. In generalization though one tends to overlook some of the very specific sub-factors (which got generalized), also in generalization one sacrifices the ability to determine credit impact of a particular factor and the impact covariances that different factors put together would have on a credit.
In respect of available information which you referred to that would be a non negotiable aspect and if complete relevant info is not available (the domain of which can be determined by the rating agency) the rating agency should decline to rate (rather than qualify the rating).
If bond market/CDS is using credit ratings (if they do not address the above points) to price credit premiums in the bonds there is a significant probability of mispricing a bond and although for an institutional investor this is okay (since he should do his own credit analysis) but for an individual investor it may not be appropriate
“I found that the speech made very fascinating reading and offered important insights into why these markets have failed to take off and what might be done about it so that they can grow in an orderly manner.”
Have had a couple of days to think about the construction of this sentence, and I must say, well played, sir. Who would possibly think a finance blog would call for a Straussian reading?: (http://en.wikipedia.org/wiki/Leo_Strauss#Strauss_on_reading)
I am rather skeptical about the prospects for developing a truly vibrant derivative markets on fixed income securities such as interest rate swaps, G-sec futures contracts, and futures on wholesale funding rates akin to the Eurodollar futures. The reasons are not that difficult to grasp:
1. Government has a fairly cozy issuance and investment relationship for G-securities with the Nationalized Banks, RBI and LIC. Together, these three classes of institutions hold nearly 85% of all G-securities. This is reflective of a “private placement” market rather than a liquid publicly traded market. Unless this is changed and a much broader ownership is brought about, I do not see much of a prospect for derivatives in fixed income markets.
2. RBI provides many dispensations to commercial banks, including HTM accounting. If all the G-securities held by the commercial banks (including those held in HTM) were to be marked to market, what would be the health of our commercial banks? What would it say about the cost of funding for the Government of India? These are counterfacuals for which no one appears to be interested in learning the answers.
3. If Basel III provisions and international accounting standards on bank balance sheets are enforced on Indian Banks today, what would be the recapitalization needs of our banks? How would that play into the G-sec cash market’s opacity?
4. There is no active way to short-sell G-securities to express one’s views about the relative values in different parts of the Government yield curve, not anywhere near what one sees in developed markets. How could then one arbitrage between swap and cash markets, for example?
My point is simple: we have a long way to go in making the G-sec cash market to be transparent. Unless this is accomplished first, it may be too early to bemoan about the absnce of an active derivative markets on fixed income debt.
Suresh Sundaresan
Chase Manhattan Bank Professor of
Economics and Finance
Columbia University
Dear Professor Sundaresan,
Thank you for participating in this conversation. You make a very valid point about the development of a vibrant G-Sec market. This is undoubtedly an essential first step. The point you make about HTM accounting seems also to echo one of the suggestions by Mr. Sharma and from recent news report the RBI seems to be planning to move in this direction as well: http://profit.ndtv.com/news/banking-finance/article-held-to-maturity-ratio-likely-to-be-cut-in-april-rbi-deputy-governor-317271. There was also a notification by the RBI in this regard (http://www.rbi.org.in/scripts/NotificationUser.aspx?Id=7647&Mode=0 — see point number 64).
However since we do have a fixed income derivatives market and the need for credit and market risk transfer remains paramount I feel that it would still be important to understand, while we are waiting for the G-Sec market to evolve, what kinds of design principles we should use in these markets. For example the issue of not having cash-settled futures — is the concern that Mr.Sharma raises a valid one — particularly given the very poor liquidity situation that you point out?
Sincerely,
Nachiket Mor
Dr. Mor, he entire argument of Mr. sharma hinges on the assumption that GsEC yields are fair and something is wrong with swap yields being lower. if the Gsec market is highly illiquid compared to swap market, illiquidity dsicounts in Gsec yields can make them higher than swap yields even though on credit basis swap yields will be higher. Hence the policy implications is not to think there is Swap-bubble, but there illiquidity in Govt bond market,and hence there has to be a policy to develop benchmark securities through regular issuance and trading. when you have a liquid yield curve all other anamolies disappear. i believe his observation that it is an anomoly is correct, but his policy inference is not. the real culprit is illiquid no-trading government bond market. when we solve this, a lot many other problems will start getting resolved. Gangadhar Darbha,
Thank you Gangadhar for participating in this dialogue. The point you make is a very valid one and not immediately apparent to an outsider. What specific recommendations would we make to the Government or to the Reserve Bank of India on this? I have also request Professor Sundaresan to share with us a more detailed version of his ideas for this blog.
Dear Dr. Mor
It is nice to have such a fascinating conversation on fixed income derivatives. Obviously the kind of gross illiquidity that the government securities market currently exhibits can’t be for a single reason. But the HTM restriction and its consequent impact on G-sec liquidity is a bit of a red herring. I have tracked securities turnover in AFS as well and they are not particularly great either..hence fair valuing the risk on b/s is possibly not as great an incentive as it is being made out to be.Incidentally, with the exception of German Bunds, the liquidity of sovereign bonds in Europe is not that great either. Actually, like most of Europe, what India suffers from is a relative absence of the hedge fund types, who are not particularly inhibited by accounting fair values and the OIS-Cash basis and therefore can accrue the yield differential.
According to me, what India also lacks is a inter bank term curve unsecured benchmark a.k.a Libor around which a swap curve can develop and once an unsecured swap curve is in place, the valuation of CDS (i.e. the asset swap spread ) becomes a little more easy to deduce. Currently we are blaming the illiquidity of CDS but can anybody specify the curve against which the idiosyncratic risk be measured. It surely cannot be OIS since the underlying curve represents secured expsoure and hence cannot be used to value the unsecured senior bonds of a corporate. Infact , to me without an unsecured term money curve, the CDS cannot be fairly priced and hence the market is being still born.
The basic point is that the unsecured funding cost benchmark has to be first put in place before one can even talk about a cash-carry arbitrage and once that is in- place, possibly liquidity across asset classes will be a matter of time.
Regards
Indranil
Thank you Indranil for participating in this dialogue. You make a very important and a new point. Would you have suggestions on how one might go about implementing your idea? Should RBI do something? Should FIMMDA take the lead?
I concur with views of Prof. Sunderasan, Indranil and Gangadhar darbha. As they’ve articulated there are various issues that are plaguing the Fixed-Income markets in India.
The most pressing issue is liquidity in the cash market. As Prof. Sundaresan and Indranil said, there are too many issues, such as, HTM dispensation, lack of a broader ownership and unavailability of sophisticated investors with varying views and preferences.
There are no easy solutions here. Somehow public-sector banks have to be nudged
to be active participants in the G-sec market. However, as long as Government
ownership of banks continue it is unlikely that neither the SLR would go nor
HTM benefit. So, unless we permit more FII participation and allow them to
participate in all hedging products, I can’t see how the liquidity in the cash
market would improve. But then again that solution comes with it the attendant
problems of financial stability.
Mr. Sharma’s argument about the negative spread between G-sec and OIS and the lack of arbitrage activity has to be viewed from a perspective on market realities. Fundamentally the price of any fixed-income product is determined by the credit risk premium and the liquidity-risk premium. In India, as Mr. Darbha said, the liquidity-risk
premium attached to G-sec more than offsets the credit-risk premium attached to
the OIS. Moreover, due to the persistent supply of G-secs, the movement in G-sec
yields is restricted (for example, in a falling interest-rate scenario, the
bond yield does not fall as much as it should have). Secondly, to undertake a
cash-carry arbitrage, the trader must be convinced that he/she can exit the
position when the spread narrows, since these positions are not held to
maturity. Given the illiquidity in the G-sec market, traders may not be able to
exit the position and that acts as a deterrent. Finally, such a position has to
be funded with overnight repos and the flexibility of switching from market
repo to RBI repo and the movement in overnight rates will be critical. In this
regard and for the development of a CDS market, the development of a term repo
market and an unsecured funding benchmark are essential. In fact, the
development of a term repo market will also aid in short-selling.
One final point is with respect to Mr, Nachiket Mor’s observation as to “what kinds of design principles we should use in these markets”? My view is central bank should stay out of designing any product. Central bank, as the regulator of debt markets, should
create an enabling environment and facilitate the development of derivative
markets. The designing should be left to the exchanges and the market participants.
Thank you Mr. Nair for participating in this dialogue. From your commentary it appears to me that in your view nothing can really be done about this situation other than full capital account convertibility. If so then this is worrying. I would urge you to suggest some concrete ideas for market participants (via FIMMDA), RBI, and MOF that could move the process forward.
Now the Economist Magazine jumps in wtih their comments on Mr. Sharma’s piece: http://www.economist.com/news/finance-and-economics/21571901-retiring-official-raises-alarm-about-derivatives-india-derivatiff?frsc=dg%7Ca. They express a concern about the very large positions that international banks hold: “Foreign banks seem to be involved in a trading game that dwarfs their balance-sheets and bears a minimal relationship to India’s real economy. It may be harmless but regulators should be on their guard.”