The Inevitable Challenge of Financial Leverage in Modern Banking 

VK Sharma , Former Executive Director , Reserve Bank of India

 vksvs2009@gmail.com

The inevitability  of the challenge of public policy having to countenance higher, but not excessive, financial leverage in banks relative to that in non-bank corporates, arises from the public policy imperative of efficient and effective transmission of monetary policy in the larger interest , in turn, of the public policy imperative of a globally competitive and efficient real economy, in which modern banks play a critical intermediation role . 

( The author’s paper in the Colloquium in April- June 2015 issue of Vikalpa, The Journal for Decision Makers , published by Indian Institute of Management, Ahmedabad http://vik.sagepub.com/  ) 

To see why higher financial leverage in modern banks  is inevitable for efficient and effective transmission of monetary policy is to consider the extreme case of Equity Multiplier of 1 which means the entire assets of a bank are funded by common equity only . To deliver market competitive  equilibrium RoE ( Return on Equity)of ,say 14%, a bank’s RoA ( Return on Assets) will also have to be 14% . So , even if policy rate be 1% , banks will, no matter what , lend at 14% plus only , and no lower ! Modern, competitive, safe, sound and efficient banks have typically operated on an RoA of 1% and EM of 14 times delivering RoE of 14% . Thus both modern banks and non bank corporates have generally delivered market competitive equilibrium RoE of 14% with their financial leverage, as measured by Equity Multiplier, poles apart ! But those who swear by the Modigliani-Miller Theorem aver that such higher financial leverage in banks relative to that in non-bank corporates will not work because higher leverage, or so much lower nominal common equity, will make  cost of bank debt/deposits higher. But in what he believes a first, the author provides a very original insight that lower nominal common equity is complemented by what the author calls “quasi equity” comprising 1) effective and credible bank supervision/regulation, 2) deposit insurance 3) implicit taxpayer guarantee and 4) a central bank’s lender of last resort function ! Of course, another name for all these fouris the so-called  “moral hazard”! And, in fact, the author argues that this “quasi equity” brings banks on par with non-bank corporates in terms of equity and, thus, makes the business model of modern  banks consistent with the Modigliani-Miller Theorem, one for one! Thus, the author shows that while , in terms of nominal common equity, banks and non-bank corporates are poles apart, in terms of being compliant with the Modigliani-Miller Theorem, they both become identical as both deliver similar market competitive equilibrium Return on Equity (RoE), in spite of way too diverging financial leverage , because they both have necessarily to compete in the same capital markets for raising equity capital ! 

4.             If only to complete the story of the inevitability of the challenge of financial leverage in modern banking and central banking , it would only be appropriate to put it in historical perspective . Specifically , in the mid 1800s,  Danish banks had leverage ratio( inverse of EM) of 75% , Americans had 55% and European ones had 25% in early 1900s cf  The Economist of 12 November 2012   " How much capital banks had when they had choice"  ! In other words, banks then had roughly the same business model as non bank corporates have today !  And significantly, monetary economics, monetary policy and modern central banking with lender of last resort function and bank regulation and supervision and deposit insurance , as we know them today , were nonexistent then and followed only later  in the late 1800s and early 1900s, and ,so did with them, the challenge of the inevitability of higher financial leverage ! And from there , it took about a century to reach Basel 2 which prescribed capital adequacy , not  in terms of common equity as a percentage of total assets( leverage ratio) , but risk- weighted assets ,quite apart from introducing the so-called Tier 2 debt capital , thus sowing the seeds of the worst global financial crisis what with financial leverage in global banks , reachingfrom less than 2 times ( leverage ratio of more than 50% ) in the early twentieth century to more than 50 times ( leverage ratio of less than 2% ) in the early twenty first century ! This happened because pre crisis global banks ,in spite of being compliant with Basel 2 on a risk weighted basis ( capital adequacy of 8%) , had risk weighted assets of only 20% of their actual total assets which had the effect of increasing their financial leverage, measured by Equity Multiplier , by 5 times ( leverage ratio of less than 2% as noted above ) ! 

But then, there is no free lunch; there is a price to be paid by economic agents and stakeholders in the form of effective and credible bank supervision/regulation which, significantly, depending upon how effective and credible it is, reduces potential recourse to the other three moral hazards of deposit insurance , taxpayer-funded bailout and central bank as lender of last resort !  The challenge of higher financial leverage in modern banking is inevitable given the imperative of efficient and effective monetary policy transmission in the real economy so that borrowing costs in the real economy  are higher not because of lower financial leverage/ higher leverage ratio but largely, if not only , because of higher monetary policy rates and vice versa  ! This is precisely here that the synergies  between monetary policy , credible and effective banking supervision and , no less ,  lender of last resort function , come in and which is why banking supervision was taken away from the now defunct Financial Services Authority and given back to the Bank of England after the crisis . In decisively and deftly managing the inevitable challenge of leverage in modern banking , effective and credible supervision makes  for efficient and effective monetary policy transmission and , in the process , makes for a globally competitive and efficient real economy ! In other words, there is a trade off between leverage and effective, credible, proactive, preemptive  , and even intrusive, supervision of banks. In other words, the more effective and credible the supervision, the higher the leverage threshold can be and vice versa !  The higher regulatory capital,or lower leverage, is the cost of supervisory failure,, inertia and inaction imposed on banks but borne by the real economy !  To make this seemingly heretical statement realistic , one can make leverage subject to a ceiling as indeed, as we have seen above , Basel 3 has done ; only this ceiling or limit is limited only by the fallibility of those in charge of banking supervision ! So to conclude , to prevent a repeat of the worst global financial crisis , what we need more than, and beyond, Basel 1,2,3,4.......is supervisory temper,culture and attitude !

 Significantly, and hearteningly, to the credit of the RBI and the Indian banking sector, Indian banks collectively have an average leverage ratio ( inverse of EM) of 7%+ which,  at about 2.5 times, is way higher than 3% mandated under Basel 3 capital rules to be complied with only in 2018 ! Incidentally, but significantly, Indian banks being already 2.5 times Basel 3 compliant with leverage ratio of 7% + will need to increase equity capital only to maintain their existing leverage ratio i.e to remain compliant with themselves and not at all to comply with Basel 3 as is widely , but erroneously, made out in many quarters ! This conclusion will very much be valid even if the denominator of the leverage ratio is inflated to include all off balance sheet liabilities which ,in the case of Indian banks,  are about 100% of the aggregate assets because this will only reduce the leverage ratio from 7%+ to 3.5%+ which is still higher than Basel 3 requirement of 3% ! I

*********

Getting Credit Default Swap Market  In India Up and Running 

I propose to discuss in my column today the long overdue imperative of developing a vibrant Credit Default Swap market in India . 

In the inflating of the credit bubble in the run up to the Global Financial Crisis , and its inevitable concomitant, the Great Recession from which the global economy is yet to recover , inflated ratings to sub-prime mortgage backed Collateralised Debt Obligations by global rating agencies played a significant role. Given,therefore, the question mark over the infallibility of rating agencies, the Basel Committee needs to revisit the primacy of role assigned to ratings of such agencies for assigning capital charge for credit risk by banks . In fact, credit appraisal and measurement are the most basic functions of intermediation performed by banks traditionally. In the light of this, ratings may be meant for, and be relied upon by, the unsophisticated and uninitiated retail and small investors, but certainly not banks. Besides, given the fact that pre-crisis rating agencies generated almost 40 per cent of their revenues from assigning the so-called inflated ratings to CDOs (Collateralised Debt Obligations) tranches, backed by sub-prime mortgages and the obvious inherent conflict of interest involved, the US Congress and regulators investigated the role and function of rating agencies in the aftermath of the Global Financial Crisis. In view of this, Basel Committee needs to de-emphasise ratings assigned by Rating Agencies for assigning capital charge for credit risk by banks. Indeed, if anything, given the tremendous volumes and liquidity of Credit Default Swaps (CDSs), both single-names and indices-based, it would be far more market-price discovery-driven for banks and regulators alike to rely on prices backed out from these credit derivatives . Indeed, CDSs price credit risks almost on real-time basis as much as Government bonds , foreign exchange, stock and commodities, markets do. Credit rating agencies, in comparison, are inertial and lagged. In the way of example, in the USA, the traditionally very healthy AAA rated mono-line municipal bond insurers MBIA and Ambac changed their business model from insuring only their staple municipal bonds to insuring CDOs and ABS (Asset Backed Securities) . While this went unnoticed by insurance regulators, Pershing Square, a hedge fund, spotted trouble and started shorting both equity and credit risk of these two companies by buying their CDSs. But even after sharp increase in real time CDS spreads of these two insurers , regulators failed to take notice of these early warning signals and any timely preemptive corrective action with the two companies being eventually downgraded several notches by Credit Rating Agencies but only much afterwards ! Significantly, as if to redeem their lost infallibility and reputation, these Rating Agencies , almost immediately after the financial crisis, started a new service which provided implied credit ratings backed out/derived from CDS spreads ! There is thus a very strong case for kick-starting a full-fledged CDS market in India. 

The popular refrain that the last global financial crisis was caused, or exacerbated,  by CDSs is again a myth in that CDSs which are simple plain-vanilla off-balance-sheet/non-fund based credit derivatives, were confused with the CDOs (collateralised debt obligations) which are on-balance-sheet and funded securitised structured credit products. It was securitisation/re-senuritization, involving CDOs that played a seminal role in the crisis and no way the CDSs ! In fact, it is also a myth that securitisation through CDOs was an originate-to-distribute model; rather, really speaking, it was an originate-to-distribute-back-to-originators model! This is because almost all CDOs originated came back to sit on the SIVs (Structured Investment Vehicles)/ conduits sponsored by originating banks themselves. Besides, for all the overdone and totally uninformed fears about systemic risks from the so-called unregulated Over the Counter (OTC) CDS markets, remarkably orderly and non-disruptive auction-based settlement of CDS claims in respect of CDSs written on Lehman Bros., Icelandic Banks, Fannie Mae, and Freddie Mac, incontrovertibly attested to the resilience of CDS markets. Indeed, if anything, CDSs can be an effective and neat answer ,and substitute,for lagged and inertial ratings of credit rating agencies ! And indeed precisely for this reason , RBI should not insist on CDSs being allowed only on listed and rated corporate bonds as what listing and rating purport to deliver is actually delivered far more efficiently on real time basis by CDSs , as typically, CDS markets sniff out financial mess much faster than even equity markets ! 

Interestingly, the New York Fed-led initiative to improve the OTC CDS markets sought to replicate India's CCIL ( Clearing Corporation of India Ltd) -model, where although OTC foreign exchange transactions are bilaterally negotiated, they are cleared and settled through RBI -sponsored Clearing Corporation of India Ltd. (CCIL). Today CDS prices/spreads are by far the most closely tracked early warning signals for real time changes in credit risk profile of an entity, whether private or sovereign. This is because CDSs make it possible to back out an implied credit price even when one is not being discovered in the underlying cash market instruments like bonds or loans as indeed in the latest instance of Deutsche Bank where its CDS spread widened real time from 100 basis points to 245 basis points in a matter of a week !   Thus, CDS market has tremendous practical application as a reliable diagnostic tool in stress-testing for supervisors and regulators. Besides, a CDS market will also enable efficient trading and hedging of credit risk and synergise development of active and liquid corporate bond and Repo markets. Like equity, credit risk subsumes all other risks as it is a function of forex risk, interest rate risk, leverage risk, liquidity risk, human resources and governance risks and that is why CDSs and equity prices have been , in equilibrium, almost perfectly negatively correlated, that is, as CDSs spreads widen, equity prices fall almost one for one !


Credit Default Swap, like Interest Rate Swap (IRS), or for that matter any other derivative, is no exception to the so-called law-of-one-price/ no-arbitrage-argument based cash market replication principle of derivatives pricing. Without going into mathematical gymnastic proper, price of a CDS, in spread terms, is reasonably approximated by the difference between the spread of a reference bond/loan to corresponding maturity G-Sec ( Government Security) yield and the spread of IRS to the same maturity G-Sec yield. Thus, if Sc be corporate bond spread and Ss be IRS spread to risk-free G-Sec yield of corresponding maturity, then the fair/theoretical/model value/price of a CDS is approximately equal to Sc minus Ss. Tautologically, since G-Sec yield is common to both spreads, another way to approximate CDS price is simply to take the difference between the yield of the reference bond/loan and the same maturity IRS yield. As is well known , finally when the product was launched in India on 7th December, 2011, it was a stillborn and remains so even after RBI’s revised Guidelines issued on 7 January 2013 . In fact, its epitaph was written in the warped, anomalous, quirky and preposterous feature of hugely negative IRS yield spreads to corresponding maturity G-Sec yields and which,alas, exist even today !  For, as one will readily see from the above formula, because of the hugely negative IRS spreads, fair price of a CDS would be so high as to make it both pointless, and useless, to buy a reference bond and also hedge it with a CDS! In other words, one is much better off straightaway buying a corresponding maturity risk-free G-Sec itself because hedged reference bond would have CDS-cost adjusted yield of G-Sec yield minus the IRS spread rather than the normal G-Sec yield plus the IRS spread ! Significantly, if actual CDS premium/price/spread is higher than the above theoretical/model price, then an arbitrageur will sell a CDS (which is equivalent to going long the reference corporate bond) and receive this actual spread and short the reference bond and invest the proceeds of short sale at the going corporate bond repo rate and receive fixed, and pay overnight, in an IRS, and do the opposite arbitrage if the actual CDS spread is lower than the theoretical/model spread/price until the arbitrage opportunity disappears and theoretical/model and actual market prices align again. But sadly, like in a classical catch-22, this arbitrage is just not possible simply because of its complete absence in the IRS market and, therefore, alas, much as we would all wish, a happening corporate bond market cannot happen, inter alia, to supplement huge infrastructure financing needs of the Indian economy!        

“Why even an NDF Dog cannot wag the on-shore tail, much less an NDF tail wagging the on- shore Dog !”

By VK Sharma,

Former Executive Director, 

Reserve Bank of India                                       

       There is quite some compulsive feel, rather than any grounding in logic and reason, to the current hype and hoopla over Non -Deliverable Forward (NDF) market influencing the on-shore USD-INR market ! And,no less, a high degree of interlinkage / correlation between the two is confused with causality. This is because the whole debate lacks in conceptual clarity on how arbitrage actually happens both, in theory, and practice. Only through the organic connect of arbitrage is it possible for information in one market to be seamlessly transmitted to, and influence,the other market. Arbitrage between discrepant prices of an asset in two different markets requires taking simultaneous long and short positions to benefit from, and eventually align, the discrepant prices in two markets. Tautologically, this arbitrage is risk free in that loss on one position is offset by a matching gain on the other and vice versa .  But restrictions on who, and how one, can take positions in the on-shore and NDF USD-INR market come in the way of agents engaging in risk-free arbitrage for NDF market to influence exchange rate in the on-shore market. More specifically, domestic entities, whether banks or businesses / individuals, are not permitted to engage in any transactions in NDF market. But banks are permitted to take open positions in the on-shore OTC (Over the counter) forward market subject to RBI monitored, and supervised, limits and businesses/ individuals are permitted to engage in on-shore OTC forward market subject strictly to actual underlying - exposure requirement and not otherwise . In other words, a forward seller (exporter) has to have actual underlying export and a forward buyer (importer/ foreign currency borrower) has to have actual underlying import/ foreign currency loan. Therefore, if FEMA regulations are not breached / violated, arbitrage between the two markets is impossible because the net position will always be open and not zero which is the hallmark of any arbitrage. Specifically, if forward dollar is trading higher in the NDF market relative to the one in the on-shore forward market, an arbitrageur will typically sell the dollar forward in the NDF market and buy it forward in the on-shore market. But a domestic entity cannot do this because of FEMA regulations which require an actual underlying exposure i.e. either an import order or foreign currency loan. And if there indeed is such an actual underlying exposure, then the net position is not zero- a ' sine qua non ' for a typical arbitrage - but actually net short because the actual underlying short position of import/ foreign currency loan is offset by the on-shore long forward dollar position, leaving the NDF short dollar position open and thus exposing the entity to the risk that the US dollar may appreciate which may potentially more than wipe out the entire arbitrage profit ! Of course, if only there is no underlying actual exposure in the way of import, or foreign currency loan, will a typical arbitrage be possible with the on-shore long forward dollar position exactly offsetting the NDF short forward dollar position, locking in the benefit of higher NDF forward dollar price ! But this will not be possible if FEMA regulations are strictly monitored and actually enforced . In other words,  only in breach and violation alone of FEMA regulations will the much-hyped arbitrage be possible and in no other way ! The same holds when the opposite is the case viz; the forward dollar price is higher in the domestic on-shore market relative to the one in the NDF market in which case an arbitrageur will typically sell dollar forward in the domestic on-shore market and buy dollar forward in the NDF market. But, again, a domestic entity cannot do this because of FEMA regulations which require an actual underlying exposure i.e. export order. And, therefore, as before, if there indeed is such an actual underlying exposure then the net position is not zero - a ' sine qua non ' for a typical arbitrage - but actually net long because the actual long position of export is offset by the on-shore short forward position, leaving the NDF long dollar position open and thus exposing the entity to the risk that the US dollar may depreciate which, as stated before, may potentially more than wipe out the entire arbitrage profit ! Of course, if only there is no actual underlying exposure in the way of export will a typical arbitrage be possible with the on-shore short forward dollar position exactly offsetting the NDF long forwarddollar position, locking in the benefit of higher onshore forward dollar price ! But again, as before, this will not be possible if FEMA regulations are strictly monitored and actually enforced. In other words, again , as before, only in breach and violation alone of FEMA regulations will the much - hyped arbitrage be possible and in no other way ! The foregoing thus conclusively proves and establishes that the NDF USD INR market influencing the on-shore USD INR market cannot just happen, except ,of course, only in breach and violation of FEMA regulations, and to that extent, RBI and Government must agonize only over effectively,decisively, credibly and inviolably enforcing extant FEMA regulations ! 

While still on the subject, as regards statistically establishing causality, this can be credibly and effectively done by recourse to the so-called Granger Causality Test. But it is very important that the USD-INR data points must be taken as frequently as feasible for the Granger causality conclusions to be robust and reliable. Ideally, such data points can be taken at 5 to 1 minute intervals when both the markets are active . Since  NDF market is almost a 24- hour market, the data points can easilyoverlap the market timings of the on-shore market. Significantly, under no circumstances should  Granger Causality Test   be applied to closing data points !

A Cash-Settled Derivative ,Where Physical Settlement Is Possible, Is A Non Derivative

by

V.K. Sharma

Former Executive Director

Reserve Bank of India

1.   In India, we have  a host  of exchange-traded cash-settled derivatives contracts like currency futures and options , Govt bond ( Interest Rate ) futures and single stock futures and options which  all settle in cash rather than by physical delivery on their expiration .  In contrast , in the US , these very same derivatives contracts are required to be settled by physical delivery on their expiration  . And the reason why derivatives , where physical settlement is possible, must settle not in cash , but by physical delivery, on expiration of the contract is that this discipline ensures almost real-time "organic connect" , coupling ,alignment and lockstep convergence between derivatives price and the underlying cash market asset price much like the " organic connect " of umbilical cord between the foetus and the mother until physical delivery .

2.  Significantly, regardless of whether derivatives are over-the-counter (OTC),  or exchange-traded,  the underlying  theory and practice has been the so-called "law of one price" or, what is the  same thing as ,the " no-arbitrage argument", involving replication of derivatives cash flows in the underlying cash markets.  In other words, a derivative of an underlying cash market asset which derives its value from that of the underlying cash market asset, will be so priced/valued that it is not possible to arbitrage between the cash market price  and the derivative price because the derivative is correctly priced relative to the underlying asset in the cash market. For, if a derivative were priced expensive relative to the underlying asset in the cash market , an arbitrageur will engage in risk less arbitrage by selling the expensively priced derivative and simultaneously buying the asset in the cash market by financing it at the going lower ( than the repo rate implied by the expensively priced derivative contract ) repo rate and physically delivering  the asset on expiration and thus realizing the risk free arbitrage profit ! In the opposite case, where the derivative is priced cheap relative to the underlying asset , an arbitrageur will engage in risk less arbitrage by shorting the asset in the cash market, investing the proceeds of short sale at the higher actual repo rate and buying the relatively cheap derivative ( lower implied repo rate) and taking physical delivery of the asset on expiration of the contract and thus again earning risk free arbitrage profit ! This seamless and frictionless risk free arbitrage, made possible by the discipline of physical settlement on expiration, will continue apace until, in equilibrium, the derivative was correctly priced/valued relative to the underlying asset in the cash market , " so much so that this very near perfect arbitrage-free pricing of derivatives itself will result in negligible actual physical delivery on expiration with participants closing out their open derivatives positions with offsetting trades before expiration !" In a physically-settled derivative contract , the only open positions that go into actual delivery at expiration are only the arbitrage positions and not even those of hedgers who close out their positions just before expiration  ! Anecdotal evidence from derivatives exchanges around the world is that in the physically-settled derivatives contracts , actual delivery is almost absent if not exactly zero ! And it is precisely this feature of a physically- settled derivative contract of almost "no actual delivery " on expiration ,due to the discipline of cash-derivative arbitrage, that is the incontrovertible touchstone and hallmark of an efficient, deep and liquid derivatives market which fully delivers on its public policy purpose of real time efficient lockstep price discovery and basis-risk-free hedging in the underlying cash market assets ! So , paradoxically, in the case of physically-settled derivatives , actual physical delivery on expiration "doesn't happen precisely  because it can happen" were the prices of the derivatives and the underlying assets in the cash market to diverge, decouple and disconnect and not move in lockstep  !

3. But in the case of cash-settled derivatives contracts, where physical settlement is possible, because of the absence of risk free cash-derivatives arbitrage and the leverage typical of derivatives , there is considerable scope for a derivative to derive its  price from itself , rather from its underlying asset in the cash market, and for speculators to make derivative price go way too out of sync and  lockstep alignment with those of the underlying asset , which , in turn, has the effect of making the derivative a " non-derivative ", exposing hedgers to not inconsiderable "basis-risk" , and defeating , in the process, the very public policy purpose of a derivative acting as an effective true hedge  !

4.  The foregoing cannot happen in the physical settlement version because , if a speculator manipulates the derivative price to be higher relative to that of the underlying , he will be punished by arbitrageurs who will buy the underlying asset cheap in the cash market and deliver it to the price manipulator at his higher price , making , as we have seen above, a risk free arbitrage profit entirely at the expense of the manipulator ! In the opposite case , risk free arbitrageurs will punish the manipulator by making him deliver an otherwise expensive asset at a lower derivatives price , again , as we have seen above, earning a risk free arbitrage profit entirely at the expense of the manipulator !

5.  Significantly, and interestingly, such seamless and frictionless arbitrage also applies, just as much, to derivatives themselves !  Illustratively, a long position in a forward/ futures can also be replicated by buying a call option and selling a put option with the strike prices for both at the current forward/ futures price.  If the actual forward/futures  price is expensive relative to the ‘synthetic’ forward/futures (call + put options), an arbitrageur will engage in risk-less arbitrage by selling the expensively priced forward/futures  and buying the relatively cheap ‘synthetic’ forward/futures (call + put options) and vice versa !

6.  While the underlying assets in the case of single stock futures/options and currency futures / options are homogeneous , this is not the case with Government bonds underlying the Interest Rate Futures. This is because a standard 10 year Government bond futures contract will cover a deliverable basket of comparable maturity bonds with widely differing coupons. They are made comparable through what is known as conversion factor which is unique to each deliverable bond . To arrive at the conversion factors of individual deliverable bonds , they are all priced to give the yield of the notional/ hypothetical par bond underlying the IRF . These prices are then divided by the par price to give respective conversion factors .

7.     It will be instructive to review the evolution of the physically-settled IRF since its launch in August 2009. After their second launch in August 2009, Interest Rate Futures on 10-year notional government bond had seen two settlements, viz. the December 2009 contract and March 2010 contract.  Significantly, both traded volumes and Open Interest (OI), witnessed decline over the two settlements, eventually decaying very quickly to zero permanently.  In particular, the December 2009 contract, which had a peak Open Interest of Rs. 980 million declined to a pre-settlement Open Interest of Rs. 610 million and settled “entirely” by physical delivery, representing physical settlement of 62% of the peak Open Interest.  In contrast, the March 2010 contract, which witnessed a peak Open Interest of Rs. 570 million declined to a pre-settlement Open Interest of Rs. 420 million and also settled entirely by physical delivery, representing physical settlement of 72% !  Both these settlements were a far cry from the hall-mark and touch-stone of an efficient, frictionless, seamlessly coupled, and organically connected, physically-settled futures market even where physical delivery  typically  does not exceed 1%  of the peak Open Interest !  In fact, on Chicago Mercantile Exchange and New York Mercantile Exchange, settlement by physical delivery is as low as 0.4% even though the contracts are physically settled types . This happened because of the inefficient ‘disconnect’ and ‘friction’ in the IRF market due to only one way arbitrage viz. buying the cheapest-to-deliver (CTD), with the highest implied repo rate (IRR), by financing the same at the lower actual repo rate and simultaneously selling futures.  In fact, as ascertained from one market participant, who accounted for almost the entire Rs. 600 million worth of physical delivery into the December 2009 contract, the implied repo rate of the CTD was 6.75% as against the actual repo rate of 3.4%,  representing  a risk-free arbitrage profit of 3.35% !!  But unlike this, on the other side, for the so-called benchmark, and most expensive-to-deliver, Government bond , the IRR was almost zero to negative, suggesting an arbitrage opportunity of short-selling this bond and investing the proceeds of short sale at much higher actual repo rate and buying the futures contract !  " But ,alas, this arbitrage could not be engaged in for want of short selling for a period co-terminus with that of the futures contract. "  It is the possibility of this two-way arbitrage, working in the opposite directions, that, like a “good conductor” of ‘heat’ and ‘electricity’ in physics, will seamlessly conduct/transmit/permeate/impart liquidity and homogeneity, as it were, from the relatively more liquid (the most-expensive-to-deliver) benchmark government bond to the so-called illiquid (the cheapest-to-deliver) and heterogeneous bonds in the deliverable basket !

8.   As we can easily see above , the physically-settled IRF failed not because of any design defect but because of the absence of the necessary and sufficient conditions for efficient , seamless and frictionless cash -derivatives arbitrage between the deliverable bonds and the futures contract . So , it might just as well be to list out the necessary and sufficient conditions for physically-settled IRF to succeed as indeed it has in all the developed and mature markets across the world. And the necessary and sufficient conditions are as under :

i)   Symmetrical and uniform accounting treatment of both cash and derivatives markets.

ii).  Removal of the ‘hedge effectiveness’ criterion of 80% to 125% which militates against use of derivatives for hedging purposes for it is better to have ‘ineffective’ hedge than to have no hedge at all !

iii)  .  Roll-back of the Held to Maturity (HTM) protection i.e. substituting the current “accounting hedge” with “derivative hedge”.  This is because with HTM, there is no incentive/compulsion whatsoever for use of market-based solutions like IRF which also require constant monitoring, infrastructure, transaction costs like brokerage and margins etc.  

iv).  Delivery-based short-selling in the cash market for a term co-terminus with that of the futures contract and introduction of term repo, and reverse repo, markets, co-terminus again with the tenure of futures contract for borrowing and lending of cash and G-Secs.

Both for IRS and IRF, actual notional/nominal amount of IRS/IRF must be allowed on duration-weighted basis unlike the current regulation which restricts the maximum notional/nominal amount of hedging instrument to no more than the notional/principal amount of the exposure being hedged resulting in under-hedging of risk .

9. But rather than establish and continuously ensure the above necessary and sufficient conditions on the ground for asuccessful physically-settled IRF , a cash-settled  IRF on government bond was launched , instead , which will , to all intents and purposes, exist and trade as “ an end in itself” , rather than “ as a means” to the larger public policy purpose of real time efficient lockstep price discovery , liquidity, depthand basis-risk free hedging in the stock of the heterogeneous government bonds . 

Gold and the Union Budget 2015 -16 : How We Can , and Cannot , Reduce Gold Imports
                                       
                                                        By VK Sharma
                                              Former Executive Director
                                                 Reserve Bank of India
                                 vksvs2009@gmail.com    +919820267474

1.       Hon’ble Finance Minister deserves unreserved compliments for explicitly and unequivocally acknowledging in his Budget speech the imperative of reducing gold imports which were the single cause of the unprecedented 30+ year-high current account deficit of 4.7% of GDP in FY 2013 with gold imports accounting for 3%+ !

2.            The main reason for this was that under the then current import regulations(cf RBI’s FEMA Master Circular), there was preferential, and more favourable , treatment for gold imports as compared to import of any other item,including essential imports, in that import of gold was permitted 1) on consignment basis 2) on unfixed price basis and 3) metal loan basis. This preferential treatment made gold imports free from both price and currency risks for overseas consignors of gold borrowed overseas as effectively it amounted to their short selling such borrowed gold in India and simultaneously covering their short position abroad, with both the risks being borne by end buyers of jewellery and gold ETFs ! This was not the case though with other items,say,coal,edible oils etc , imported on direct import basis. I had accordingly , in December 2012, suggested to the then Governor, Dr Subbarao and the present Governor Dr Raghuram Rajan , then Chief Economic Advisor, Ministry of Finance , to create a level playing field between gold and non gold imports by aligning gold import regulations under FEMA with those for non gold but essential imports . And indeed, aligning gold import regulations with the rest of imports by RBI had the desired effect of taking away this significant, unwarranted and perverse incentive and delivered the desired outcome of dramatically reducing gold imports in double quick time, merely by creating a level playing field between the two kinds of imports so much so that ,as its direct consequence , the current account deficit narrowed dramatically to 1.7 % of GDP in FY 2014 ! And this was achieved , as I said, not by imposing curbs and restrictions on gold imports , as was widely made out in media campaign by some quarters , but merely by creating a level playing field between gold and non gold imports , by aligning gold import regulations under FEMA 1999 with those for non gold , but essential, imports like , as I mentioned above , edible oil and coal !


3.            But surprisingly , RBI only recently rolled back these entirely wholesome and fair measures and no wonder , according to media reports , gold imports have again surged and the current account deficit for the latest fiscal quarter has increased to 2.1% from 1.2% of GDP ! Against the background as somber as above, to reduce gold imports, Hon’ble Finance Minister has proposed in his Budget three schemes and his verbatim statement on which in the Budget Document is as follows:
” India is one of the largest consumers of gold in the world and imports as much as 800-1000 tonnes of gold each year. Though stocks of gold in India are estimated to be over 20,000 tonnes, mostly this gold is neither traded, nor monetized. I propose to:

(i)     Introduce a Gold Monetisation Scheme,which will replace both the present Gold Deposit and Gold metal Loan Schemes. The new scheme will allow the depositors of gold to
earn interest in their metal accounts and the jewelers to obtain loans in their metal account. Banks/other dealers would also be able to monetize this gold.

(ii)      Also develop an alternate financial asset, a Sovereign Gold Bond, as an alternative to purchasing metal gold. The Bonds will carry a fixed rate of interest, and also be redeemable in cash in terms of the face value of the gold, at the time of redemption by the holder of the Bond.


(iii)      Commence work on developing an Indian Gold Coin, which will carry the Ashok Chakra on its face. Such an Indian Gold Coin would help reduce the demand for coins minted outside India and also help to recycle the gold available in the country. ”


 4 .         The proposed Sovereign Gold Bond Scheme is ,in its design and logic, exactly like a Gold ETF ( Exchange Traded Fund) with the difference that the latter doesn’t pay any interest but delivers gold returns to investors by investing the entire proceeds of subscription in metal gold which is held in demat form ! Significantly, all the 14 Gold ETFs in India ,between them, hold no more than 40 to 50 tonnes of gold, which is a mere 5% of annual gold imports of 800 to 1000 tonnes mentioned by Hon’ble Finance Minister! So a gold ETF does not at all reduce metal gold demand and all it does is substitute gold demand from individual metal gold investors with like demand from a professionally and expertly managed gold ETF instead  ! Exactly so will also be the case if the Government does the same to deliver gold returns to Sovereign Gold Bond holders on redemption ! In the highly unlikely event of Government even remotely contemplating using the subscription proceeds for a purpose other than investing only in metal gold will expose it to extreme price risk as it will be committed to delivering gold returns to bond investors on redemption at the ruling gold price ! This is because this will amount to Government incurring  huge short position in gold potentially  exposing the exchequer to gold price increases !   A sense can be had of the enormity of this proposition by considering the fact that gold price moved all over from a then all time high of $ 850 per oz in late seventies to a life time low of $ 270 in the late nineties and then back up to an all time high of $ 1920 in September 2011 !! Assuming the worst case , and hence the most prudent and conservative scenario , of gold price rising from its recent low of $ 1180 per oz to its all time high of $ 1920 ( 65 % increase) due possibly to potential monetary stimulus in Japan and Eurozone , the resulting shock could be fiscally overwhelming and way too disruptive . The rationale for this is that if we again assume conservatively that Gold Bonds would attract the equivalent of 1000 tonnes of annual gold import demand , it will translate into the rupee equivalent of ₹ 1.5 lakh  crores ( $ 25 billion ) of net loss on redemption not counting the interest paid which will ,all else being equal , increase fiscal deficit by like amount because of subscription proceeds not being invested fully in metal gold ! Any multiple of 1000 tonnes will have multiplier effect and in the extreme scenario of the theoretically maximum subscription equivalent to the entire metal gold stock in India of 20000 tonnes , will translate into a whopping net loss of around ₹ 30 lakh crores ($ 485 billion) to the exchequer on redemption at the assumed ruling price of gold of $ 1920 per oz and ,as stated before , will , all else being equal, increase fiscal deficit by like amount, representing 20% of the current GDP ! Not only this , large investors in Sovereign  Gold Bonds could also cause  the so called short squeeze , close to the due date of redemption , to profit at the expense of Government by speculatively pushing gold price higher fully aware that the Government is hugely short in gold , making the above mentioned whopping net loss to the exchequer even worse !


 5.            Contextually, in February 2013 , someone from Commodity Exchange space , who should have known better, had proposed a variant on the theme of the above Gold Bond Scheme . He suggested that the subscription proceeds of the sovereign gold bonds be invested in infrastructure projects and to deliver gold returns to bond holders without price risk and loss , Government may hedge its price risk by buying gold call options ! But this hedging proposition , involving buying gold call options covering ,as stated in the preceding paragraphs, 1000 tonnes worth of annual gold import demand , while tenable in theory of options , will be untenable in practice as this will inundate call option writers/sellers and result in call options getting deep-in-the-money because of resultant price increase way above the starting strike price at which Government will buy these call options , pushing the delta hedge ratio to almost 1 which would mean call options writers will have to physically buy almost 1000 tonnes of the metal , leaving the physical metal gold demand in the domestic market pretty much the same, if not more, and which anyway, as now, will only be imported ! This will also apply just as much to Government seeking to hedge against gold price increase by buying gold futures because given the sheer size of the the hedging demand , call options and futures will both have the hedge ratio of 1 meaning that the demand for the metal gold will be about 1000 tonnes; only that it will shift from Government to the metal gold market and which , if not supplied in the domestic market will , as now, have to be met through imports ! This is the very basic theory and practice of futures and options hedging and pricing which is governed by the so-called no-arbitrage argument or,what is the same thing as,the law of one price. In other words, there can be no free lunch or the case of eating one’s cake( investing cash proceeds of gold bond sales in projects and not in gold) and having it too( replicating and delivering gold returns without investing in physical gold)  !! In a lighter vein , if only to dramatise to best effect , the analogy of futures and options hedging with a dialogue of the character Gabbar Singh in film Sholay is most apt and I quote : ” Oh village folks if there is anyone who can save you from Gabbar, he is none other than Gabbar himself !” So also ” If there is any asset which can deliver gold returns with gold price risk, it is none other than gold itself !” Of course, the same holds for any other asset like equity stocks, bonds, foreign currency, real estate etc ! So to conclude, the proposed Sovereign Gold Bond Scheme will not deliver in practice !


  6.               Turning to the other Budget proposal on the so called monetisation of 20000 tonnes of domestic stock of gold which ,as Hon’ble Finance Minister very rightly observed in his speech, is mostly neither traded nor monetised, I will begin with his own statement that the domestic metal gold stock is mostly not traded . Of course, globally, deep and liquid metal gold deposit and lending markets exist but only for two reasons; the first being active trading and the second gold miners borrowing metal gold to raise money at ultra low interest rates to fund their capital investment in either new mines or in expansion of their existing mines . And there there is no third reason for this !


 7.               The dynamic of the global gold deposit and lending markets in New York, London, Singapore and Hong Kong involves gold borrowing demand coming from short sellers and large gold miners . In particular, short selling arises from speculators, hedge funds and other participants betting on declines in gold price so that they can make profit from their bet coming right by buying back gold at a price lower than the price at which they originally short sold and repaying the metal gold loan  ! But since there is the market discipline of delivery into a short sale, short sellers have necessarily to borrow metal gold to deliver into the short sale which gets adjusted after some time when short sellers either buy metal gold back to cut their losses due to stop loss limits or to book their profits and using the gold so bought back for repaying the borrowed gold !


8.                  Another reason for short selling metal gold is engagement by market participants in cash futures arbitrage when gold futures are cheaper relative to the spot market i.e. when no-arbitrage argument /law of one price of derivatives is violated . What then they actually do is buy cheaper gold futures and short sell expensive gold in the spot market and carrying the arbitrage trades into maturity and earning totally risk free profits due to such mis pricing. Such arbitrage trades continue until, as a result of these trades , such price discrepancy eventually disappears ; the faster this happens ,the more efficient the markets ! Since globally there is large scale demand for such borrowed metal gold , supply comes from gold deposits and , like in any bank deposit and loan market , there are deposits and lending/borrowing rates known as gold lease rates and these are way too low compared to currency rates because they are in theory and practice nothing but the difference between the uncollateralised currency interest rates and and those collateralised by metal gold . Specifically, when gold is short sold , the sale proceeds of short sale are used by the short sellers to collateralise their borrowing of metal gold borrowing which amounts to lending by short sellers of cash from short sale to metal gold lenders at an interest rate lower than the interest rate they will lend at otherwise ; the more desperate the metal gold borrowers are to borrow, the less interest rates they will charge to the lender of the metal gold and so much higher will be the metal gold lease rates ! On the other hand in an aggressive bull market in gold when prices shoot up , the demand for borrowed gold will relatively be far lower and ,therefore, gold lease rates will be lower and can be, and have indeed been, negative !! These gold lease rates are in both theory and practice a fraction of a currency ‘s interest rates ! Illustratively, currently these are 0.09 % , 0.11% , 0.15 % , 0. 25% and 0.40% for 1 month , 2 months, 3 months, 6 months and 1 year , respectively! Incidentally, in the Indian context , there has been frequent demand from some quarters that interest rates paid on gold deposits be higher than currently paid without appreciating the fact that each asset has its own yield curve like for example just as you cannot pay Indian Rupee rates on yen and dollar deposits , so also you cannot pay Indian rupee interest rates on gold deposits ! Coming back to gold lease rates , they have also been negative during the episodes of aggressive bull market in gold when lenders of metal gold far exceed the borrowers of gold to a point that lease rates become zero but there are still storage and insurance costs which make effective lease rates negative!


9.             As I mentioned above , other than short sellers and arbitrageurs there is only one more kind of metal gold loan borrowers , namely, large gold miners who borrow metal gold for longer periods to sell the metal in the spot market and use the sale proceeds to fund capital investment in new mines or in capacity expansion of their existing gold mines and have a natural hedge against gold price rise as they use the metal gold mined to repay their metal gold loans without any price risk and thus have the natural advantage of raising capital at a fraction of cost of debt and equity capital unlike other non gold businesses !


10.           Significantly, as regards lending gold to jewellers , to call it a metal gold loan is an oxymoronic misnomer because, effectively and substantively, it is simply nothing but a sale and purchase transaction as , both in finance theory and practice, a deposit and lending transaction is one where whatever is borrowed has necessarily to be repaid like ,for example, businesses in India borrow in rupees   and generate rupees and thus repay in rupees . So also, jewellers borrowing metal gold will only generate rupees by selling jewellery in the domestic market , and certainly not gold , and so will have necessarily to buy metal gold back  from the domestic market and which, if not supplied locally , will again only be imported  to repay gold borrowed , and to that extent it is no brainer to see that the proposed monetisation scheme will not deliver in terms of reducing gold imports   ! In other words, one will be back to a square one !!


11.             So to conclude whether the proposed gold deposit and lending scheme will deliver in practice will depend critically on whether we have in India both active gold trading involving ,as I said before, short selling and arbitraging and gold mines of global scale ! And as we already know the above necessary and sufficient conditions are not satisfied in the Indian context . In view of the foregoing reasons , both the Sovereign Bond Scheme and Gold Monetisation Scheme will not deliver ! What , however , will is what actually and effectively did only as recently as in 2013-14 , that is, restoring by RBI of the status quo ante by again creating level playing field between gold and non gold imports by stopping gold imports on 1) consignment basis 2) unfixed price basis and 3 ) metal loan basis !