Search interesting materials

Monday, May 28, 2012

Evaluating responses to India's macroeconomic crisis

by Shubho Roy and Ajay Shah.

The macroeconomic setting

India's macroeconomic woes consist of high inflation, low GDP growth and a drop in asset prices. The loss of momentum is visible in the seasonally adjusted data:

IndicatorEarly 2009Latest
GDP growth (QoQ, saar) 9.83
Inflation (CPI-IW, pop, saar, 3mma)7.5
Feb 2009
Mar 2012
Jan 2009
May 2012

The picture is not uniformly bleak. The most important asset price of the economy, Nifty, has not dropped across this period. On 1 January 2009, Nifty was at 3033. Today, it is at 4920, which is a good 62% higher. More generally, stock prices have held up rather well so far. The trailing P/E of the broad market index, the CMIE Cospi, stands at 17.3, while the median value across its full history (from 6/1990 to 4/2012) is 17.83. We may think that conditions in India are difficult, but the stock market is saying that they're roughly median conditions in terms of the outlook for earnings growth.

The current account deficit

In recent years, the fiscal condition of the government + PSUs has worsened. This has led to a large gap between savings and investment. The worsening in public finance has diminished savings. There is an accounting identity: The gap between savings and investment is the amount of capital that has to be imported. This is the current account deficit. We have a capital shortfall within India, so we are importing capital.

It is likely that in the coming year, we will have a current account deficit of 4% of GDP, or $80 billion a year, or Rs.1700 crore a day. This means that we have to worry about how foreign capital views India. Under these conditions, if there is even a short hiccup in capital inflows (as appears to have come about after the government proposed to modify the Mauritius route, and more generally with the problems of governance in India), it yields sharp rupee depreciation.

We import a lot of capital; government policy actions interrupt that flow of capital; the rupee depreciates. This is not mis-behaviour of the financial system. The system is not malfunctioning; it is behaving as it should.

What should the responses be?

There are five sensible paths for government to take, in this situation:
  1. We need to see that at heart, this is a problem of macroeconomics. The root cause of the current account deficit is the fiscal deficit. If we want a lower CAD, we need a lower fiscal deficit.
  2. To ensure the smooth flow of Rs.1700 crore a day into the country, we should not spook foreign investors. We should not interfere with the de facto residence-based taxation framework which India is giving foreign investors, as long as they come through Mauritius. This policy framework is, in fact, in India's best interests.
  3. Deeper problems about the loss of confidence of foreign investors, owing to governance problems, need to be solved by strengthening governance. They are important (do not ignore them), but there is no quick fix other than improving governance.
  4. In the face of these difficulties, it would make little sense for RBI to trade in the currency market, to try to block the rupee depreciation. There is good reason for rupee depreciation; the currency market is doing a pretty good job of translating the fundamentals into a price. And, in any case, even if RBI desired to do something about it, its weapons are puny when compared with the size of the currency market and the Indian economy.
  5. It is an opportune time to continue with the liberalisation of the capital account. However, it is useful to think deeply about how to proceed with this. Some kinds of liberalisation can be dangerous. It is important to think about sequencing, and at all times, to worry about unhedged currency exposure. A good deal of expertise has built up on the subject, through the Raghuram Rajan Committee and the UK Sinha Working Group which worked out the medium-term and short-term sequencing of reform.

An evaluation of what has been done

There are three features of recent policy responses which appear to be on track:
  1. By and large, RBI's trading on the currency market appears to be at a low scale, nearing zero in many recent months. This is wise. It increases respect for the brainpower at RBI.
  2. The government raised the price of petrol, so as to cut the fiscal deficit. This increases respect for the brainpower and political capabilities of the government.
  3. The government decided to defer the attack on the Mauritius treaty by a year (though not to shelve it altogether). In the absence of clear policy statements about the importance of residence-based taxation, this shelving does not increase respect for the government.

Apart from these three good moves, a slew of dubious ideas have been afoot.
A. Enlarging the scope for dollar-denominated borrowing by Indian firms
On 20th April, 2012: external Commercial Borrowings regulations were amended to:
  1. Increase the limit on power companies to refinance their borrowings in Rupees with foreign currency loans (also called External Commercial Borrowings or ECBs).
  2. Allow companies to borrow in foreign exchange to make capital expenditure for maintenance and operations of toll systems (See here)
  3. Companies were allowed to refinance their ECBs with subsequent ECBs at higher interest rates (See here).
Evaluation: Do we really want Indian firms to hold dollar denominated debt? In particular, firms in the field of infrastructure who have cashflows in rupees? Sensible firm should see the high ex ante currency volatility and stay away from borrowing in dollars without hedging; so the impact upon capital flows will be small at best. And firms that do borrow in dollars and keep it unhedged are probably not going to fare well.

B. Enlarging the scope for dollar-denominated borrowing by banks
On 4th May, 2012: The maximum interest payable on forex deposits by NRIs in Indian banks was increased (See here):
  1. For deposits between 1 to 3 years the increase was 75 basis points.
  2. For deposits between 3 to 5 years the increase was 175 basis points.
Evaluation: Banks are disaster-prone 19th century institutions. Do we really want them to hold more unhedged foreign currency exposure? Of all places in the economy, this is the worst place to keep unhedged currency exposure. The wise ones will not borrow in this fashion, so the impact upon capital flows will be small at best. And the unwise ones, that borrow in dollars and keep it unhedged, are probably not going to fare well.

C. Reducing the economic freedom of exporters
On 10th May, 2012: the right of exporters to hold foreign exchange was reduced by 50% (See here):
  1. Exporters were allowed to keep their forex earnings in special accounts called EEFC accounts. They were not mandated to convert it into Rupees. This allowed them the ability to fund imports for their business without going through costly conversions.
  2. Now only 50% of their export earning will be allowed to be kept in forex. The rest will be converted into Rupees against their wishes.
Evaluation: In the old India, FERA made ownership of foreign exchange an exotic and rare thing. Many businessmen in India engaged in import/export misinvoicing and tried to hold assets outside the country. In the early 1990s, C. Rangarajan's RBI embarked on a modern arrangement. Exporters were given greater economic freedom. We are now rolling the clock back by 20 years; we are tampering with current account liberalisation.

The number "50%" has not been justified in the RBI notification. Any exporter, with significant raw material import cost will now pay unnecessary transaction charges. In global trade, where every country takes the utmost effort to keep their exports competitive, any small distortion impacts on export competitiveness; this is pushing in the other direction - it is an attempt to reduce India's export competitiveness.

This is a new low in Indian economic policy. Every internationally oriented household in India will now be more keen to hold assets and liquid balances outside India, safe from the clutches of Indian capital controls. This measure will thus exacerbate capital flight and worsen the problems of the rupee. Success in the marketplace will tend to accrue to businessmen who break laws as opposed to the law-abiding ones.

D. Damaging the currency futures market
On 21at May, 2012: restrictions were put on exchange-traded derivatives (See here):
  1. The net overnight open positions that the banks hold shall not include positions in the exchanges.
  2. The positions in exchanges cannot be used to offset positions in the OTC market for
  3. The position of banks in currency exchanges shall be limited to $100 million or 15% of the market (whichever is lower)
Evaluation: The world over, there is a clear understanding that the exchange is a superior way to organise financial trading. When compared with the OTC market, the exchange has superior transparency and risk management. Policy makers need to continually modify policies so as to favour a migration of all standardised products away from the OTC market to the exchange-traded contracts. RBI's moves go in the wrong direction.

How do we ensure that the price on a financial market is driven by fundamentals? The answer : We must have a deep and liquid market, and a broad array of sophisticated speculators. RBI's actions are going in the exact opposite direction. They are trying to make the market illiquid. But it is in an illiquid market that we will get market inefficiencies and weird behaviour of the price. They are increasing the chance that something nutty happens on the rupee.

This circular is also a reminder about poor legal process at RBI. Every action by a regulator must articulate a rationale. Financial regulations are motivated by exactly two possibilities - consumer protection or micro-prudential regulation. The government agency that wields the power of financial regulation must show the clear rationale, describing what is the market failure that this regulation is seeking to address. The government agency must show the cost-benefit analysis, explaining why the costs of this action outweigh the benefits. As is typical of financial regulators in India today, RBI's documents show no rationale.

It is possible to conjure one conspiracy theory. The attempts at damaging the liquidity of the currency futures market should be seen in connection with previous work on damaging the liquidity of the OTC market. Perhaps there is a grand plan here. The scale of RBI's trading on the currency market is implausibly small when faced with the size of the Indian economy, with the size of India's cross-border interactions and the size of the currency market (both onshore and offshore). Perhaps these recent moves are designed to damage the liquidity of the market, so as to get to a point where RBI intervention can make an appreciable dent on the price. Perhaps the game plan is to gnaw away at the capability of the currency market through a series of moves, and then take off doing large scale manipulation of the market. If this is the game plan, it reflects very poorly on the economic policy capability at RBI. It would also generate massive profit opportunities for the speculators of the world, who would short the rupee when the large scale manipulation commences.

Rumours about other bad ideas abound. E.g. it is suggested that RBI will sell dollars to exporters directly. How is this different from selling dollars on the market?? It is suggested that the currency futures and the OTC markets should be completely cutoff by banning the arbitrage. How would this solve the macroeconomic problems which bedevil the rupee?

Microeconomic distortions are not a good way to address macroeconomic problems

What does one make of this spectacle? A simple principle worth reiterating is:
Problems rooted in macroeconomics must be addressed using macroeconomic instruments.

We got into this mess because of inappropriate fiscal and monetary policy. We need to solve these -- monetary policy must get back to the business of delivering low and stable inflation, we have to fight inflation until we see y-o-y headline inflation (i.e. CPI-IW inflation) going to the 4-to-5 per cent range. Alongside this, fiscal policy needs to correct itself. Each of these has a clear direction to move in, and movement on any one is valuable regardless of what the other does.

A big element in the picture is the loss of confidence, in the eyes of the private sector, on an array of issues ranging from ethical standards to the sophistication of fiscal, financial and monetary policy. This is an important problem and it needs to be addressed. The spectacle of a government flailing at the macro problems using micro instruments is worsening matters. Perhaps there is constant pressure to announce `new measures' to solve the problem. Deeper solutions are hard, and there is enthusiasm for `doing something' (large or small) [example].

We've seen this movie before. In the last decade, again and again, RBI tried to wield capital controls as a tool for macroeconomic policy. They failed. It is disappointing to see the lack of learning.

Some of the moves above have come out of the reflexive socialism that lurks within the Indian bureacracy. Perhaps, in a crisis environment, the ordinary immune system within each government agency, which keeps the sub-clinical socialism under check, is not working as well. This hurts from two points of view. It betrays the lack of capability of these government organisations; it reminds us that the Indian State is strewn with people who have a low knowledge of economics and a taste for dirigiste. It also reminds us of the policy risk: Precisely when the best capabilities are required (in a crisis), we seem to be slipping into the lowest quality policy initiatives.

Everyone who sees the government / RBI engaged in one ill thought out measure after another gets worried about India's future. How can a $2 trillion economy flourish while such immense powers are placed with individuals and institutions with such weak capabilities? This further damages confidence, which deepens the macroeconomic crisis.

Acknowledgements: We are grateful to Apoorva Ankur, Sumathi Chandrashekaran, Pratik Datta and Kaushalya Venkataraman for useful suggestions.


  1. Agree with mmost of this except for the EEFC criticism. One can argue whether 50% is the right number. But the macroeconomic rationale for this 'microeconomic distortion' is fairly clear - when the capital account if not fully convertible, you want to prevent mechanisms that make it convertible. Sitting on EEFC funds is a speculative capital account bet on the rupee disguised as a current account facilitator. We have to ask ourselves - (1) how many exporters have significant (upto 50% or more) of their working capital and invoicing costs in dollars (2) Are they not outdone in all comparisons of magnitude by those who don't and are simply using the provision of EEFC funds to bet on the rupee without having to go for an explicit asset purchase or cash management product.

    The other way to look at the EEFC conversion is that to the extent exporters are different from importers, it reduces the lag between the flow of dollars from exporters to importers and thereby reduces the pressure on the reserve financing of imports, bringing the situation closer to the equilibrium ideal of reserve financing of the CA/balance of payments deficit.

    1. Dear Ritwik,

      We're discussing a number. Does a bureaucrat know whether the number should be 40% or 50% or 60%?

      For the firms where the correct number is above 50%, this is a current account restriction which hampers their international trade.

      There are a few big ideas here that are worth recounting. Current account and capital account liberalisation are tightly interlinked.

      1. More than half of global trade flows take place within MNCs; that is, for a country to participate in the current account, the country needs to be open to FDI.

      2. Similarly, international trade drives the demand for currency trading, currency hedging, cross-border payment services, cross-border credit management, etc.: crossborder finance follows trade.

      3. Countries that open the current account tend to open the capital account and vice versa. This was worked out by Aizenman and Noy in a pair of clean papers.

      4. Shang Jin Wei and co authors have a nice paper which shows that incremental capital controls are akin to customs duties in terms of their impact upon current account activity.

      5. Arvind Subramaniam and coauthors have nice work on trade misinvoicing in order to beat customs duties.

      6. Finally, Ila Patnaik and Abhijit Sen Gupta and I have a paper in Open Economies Review which is a multi-country analysis of trade misinvoicing; one element that we see is that more capital controls yield more trade misinvoicing: people are using the current account to circumvent capital account restrictions.

      In short, some people like to think that current account opening is good cholesterol and capital account opening is bad cholesterol, so let's do one without the other. This is not how the world works. Both are tightly intertwined.

    2. Dear Prof Shah.

      Absolutely agree with the importance of being more open to FDI and the general points being made. Agree also with the broad thrust of making the capital account more convertible, at current levels of openness. But note that this does not apply at all levels of openness- a large body of Raghuram Rajan's research, for example, is about the superiority of FDI over FII and what that means for capital account opening. But I digress.

      My main point with the EEFC defence, perhaps not articulated well, is that it should be seen as akin to a TDS. Dollars are going to flow, ultimately, from exporters to importers. The time lag of this flow is assumed away in equilibrium analysis of the BoP & exchange rate but just like liquidity constraints on economic actors, it matters for the latitude of exchange rate interventions. In a country faced with a dual BoP/ inflation problem like India, the central bank might want to reduce all these 'frictional' pressures on the BoP to enable it to maintain its interplay between interest rates and exchange rates in the pursuit of monetary policy objectives. EEFC caps give this latitude.

      Your objection of this raising transaction costs for exporters applies in an important quantitative sense only to those exporters who are also big importers simultaneously, of the order of more than 50% of their export earnings. I take your point about most global trade flow being within MNCs (is it within or is it between) but given that big MNCs are more likely to be accounting and distributiong their earnings in Dollars/euros than rupees, this logic applies more to currencies of those countries where these MNCs are based/ accounted for. I am not sure the 'right figure' point has much merit as long as one is convinced that a number is roughly right (50% being a high enough cutoff, on the face of it) - it's a bit like asking what happens at age 18 that makes you eligible to vote when you weren't at 17.5.

      But even more fundamentally, let's go back to the point about Fx reserves being constant and actually at zero in a true BoP equilibrium with perfectly flexible exchange rates. RBI intervenes not just when it sells reserves to defend the rupee, it also intervenes when it accumulates reserves to not let the rupee appreciate too much. The fact that our reserves went up from $10 billion to $300 billion over the last 20 years suggests that if we lived in a world of no liquidity constraints and the RBI were to truly un-intervene, the rupee would appreciate dramatically. But clearly, we don't. So to the extent that liquidity constraints matter, an attempt by the central bank to ease the current account pressure on its reserves at a time of macroeconomic stress should be seen as proper macroeconomic management, not a microeconomic distortion.

      Plus, I buy the RBI point about EEFC funds being used as a capital account bet. We should enable direct punts on the rupee at some point of time, sure. But so long as we haven't, why should exporters benefit? Plus, if an exporter wants to punt on the rupee shouldn't they invest their free cash in a fund which does that? Let exporters acquire dollar assets explicitly, rather than under the guise of transaction costs.

  2. RBI faced with currency fluctuations that they don't like is behaving just like FMC faced with edible oil price fluctuations that they don't like. The intellectual capacity of both agencies is far from market economics.

  3. "There are three features of recent policy responses which appear to be on track".
    Here's my US$0.02 (Ok, Rs0.02, if that's too expensive these days):

    "By and large, RBI's trading on the currency market appears to be at a low scale, nearing zero in many recent months. This is wise. It increases respect for the brainpower at RBI."

    The last six months (since ~Oct'11) have seen the RBI spend ~US$20bn in the FX markets. The sharp intervention (and the resultant drag on M0) has sucked out nearly Rs1trn from domestic liquidity, creating acute repo, partially addressed with OMOs (Rs1.3trn in FY12). Even now, borrowing at the repo window has tended to rise above the RBI's stated comfort levels of 1% of NDTL. Unless therefore you specify what you mean by 'Many recent months", one does not really know the implication of that comment.

    Agreed, 'policy measures' have to some extent replaced USD sales in April, and IMHO the optimal tool, given extant market conditions. Coming after burning up of US$20bn with dubious results, one can only imagine the impact of a sudden, zero intervention. A far faster slide, than the rather orderly decline we are seeing now.

    Was the intervention right in the first place? No two opinions on that. But when presented with a _fait accompli_, as the RBI found itself in April, alternatives (to perceptions of capitulation) have been limited. All of these measures could be justifiably castigated, in better market conditions, and as they improve, one is certain to see gradual rescinding. There's precedence. cf. the all-in-cost ceiling for ECBs and the relative ease of accessing them over the last four years. It makes perfect sense to decry the tinkering with FX market microstructure. One's left with the feeling however that this article should have made better sense earlier, perhaps in February.

    "The government raised the price of petrol, so as to cut the fiscal deficit. This increases respect for the brainpower and political capabilities of the government."

    For the record, Petrol losses are NOT subsidized by the Govt--details available from the PPAC website. One could potentially argue about _combined_ OMC losses, but even then, MS losses are puny compared to that of HSD/LPG/SKO. Given the Govt.'s stated position on decontrolled Petrol prices, one should ostensibly see no merit on its part when OMCs raise prices, right?

    Reality, as we well know however, is slightly different. 'Allowing' IOC/BPCL/HPCL to raise MS prices is indeed laudable, given the political implications. Look at this another way, though. If we are happy and congratulating the raising of decontrolled Petrol prices after six months, then we tacitly accept the positioning of the ruling dispensation. With that understanding, the proposal of reducing the fiscal deficit to curtail the CAD--while completely the right thing to do--strains credulity, as does allowing capital convertibility. Today we're in a position where measures to reduce the fisc. would perhaps find partial acceptance, but that comes after a 20% fall in the currency.

    "The government decided to defer the attack on the Mauritius treaty by a year (though not to shelve it altogether). In the absence of clear policy statements about the importance of residence-based taxation, this shelving does not increase respect for the government.
    Apart from these three good moves, a slew of dubious ideas have been afoot."

    How can something that doesn't increase respect (and rightly so) be a good thing? GAAR differing by a year does not solve the problem one bit, but rather illustrates a degree of capitulation in the face of sharp, adverse globally negative publicity (even if the intentions could perhaps at a stretch be construed as genuine). It's amazing how the Govt. asks for rating agencies for a raise, and then turns the clock back with an unmitigated PR disaster in the form of GAAR.

    1. You say: "The last six months (since ~Oct'11) have seen the RBI spend ~US$20bn in the FX markets and this caused trouble for domestic liquidity".

      Look at intervention as per cent of M0 and we are at small numbers compared with the great days of currency trading by Y V Reddy and Rakesh Mohan. Express it as per cent of the global daily currency turnover (roughly $70 billion a day) and it is a pittance.

      Yes, what little intervention was done was not sterilised and has caused a mess. That's a separate issue. The really important point is that if you think about what impact cost they could have got by virtue of doing this intervention, it's tiny. And I'm giving them the benefit of the doubt and thinking that short-term impact cost is it - that the market does not have resilience to come back to the market price after the order has been digested. From a market microstructure perspective, we pretty much got the market price, not the manipulated price.

      You say: "But when presented with a _fait accompli_, as the RBI found itself in April, alternatives (to perceptions of capitulation) have been limited. All of these measures could be justifiably castigated, in better market conditions, and as they improve, one is certain to see gradual rescinding."

      If you are suggesting that these measures are slightly justified, I disagree. "Capitulation" is a pejorative phrase. A much better term is "market price". Why is GOI manipulation of the petrol price wrong but RBI manipulation of the rupee-dollar rate right? And, I may point out, in the end they have ended up at the same place - the market's price of the rupee and not a government-controlled one. Not a bad outcome - but very poor governance and policy making along the way.

      You say: "For the record, Petrol losses are NOT subsidized by the Govt--details available from the PPAC website. One could potentially argue about _combined_ OMC losses, but even then, MS losses are puny compared to that of HSD/LPG/SKO."

      For NAS calculations, the aggregate public sector saving/dissaving is the sum of GOI and PSUs. So a loss at the PSU drives down S, raises the I-S, drives up the CAD.

      You say: "With that understanding, the proposal of reducing the fiscal deficit to curtail the CAD--while completely the right thing to do--strains credulity"

      It is an accounting identity. CAD = I-S. You want to make a difference to the CAD, cut the fiscal deficit. That's it. Nothing else will do it.

      About GAAR, you say: "How can something that doesn't increase respect (and rightly so) be a good thing?"

      Can you visualise the mayhem that would have resulted if they had gone ahead with GAAR, and as promised in Parliament by the MOS, used GAAR to (effectively) unilaterally abrogate the Mauritius treaty? Out of this episode, one does not gain respect for policy making capabilities, but one does heave a sigh of relief.

    2. But Prof Shah, why do you look at trading, specifically Fx selling activity to arrive at market 'manipulation' as you call it. As Scott Sumner and most others analysis developed world central banks would put it, there is no such thing as doing nothing in monetary policy. Passivity is itself a great lever of monetary policy. When the RBI gained reserves from $10 billion to $300 billion, was it not manipulating the rupee to favour exporters?

      Secondly, why're you treating the exchange rate as a market price that is to be understood in a vastly different context than, say, the price of short term government bonds (or the price of overnight liquidity), both of which are considered legitimate tools of monetary policy? Again, as you undoubtedly know, and most monetary economists from the developed world will tell you, a central bank's power over the nominal economy arises from its control over its balance sheet through the monetary base. Against the monetary base, it chooses to keep certain assets - government bonds are the primary, but Fx reserves and gold are non-trivial, especially for a developing country central bank. It intervenes in the markets for all of these assets, by manipulating (increasing, decreasing, refusing to increase, etc.) its domestic currency base. When the RBI buys gold, do you believe it is unfairly trying to manipulate a market price? When it sets the short rate, thereby setting the base off which g-secs will be priced, do you think it is unfairly/incompetently manipulating a market price? Then why only for exchange rates?

      If the argument is simply that the intervention will not work because the RBI is not a big enough player in the rupee dollar market, then it can be understood and/or empirically challenged. But I don't quite agree with this floating exchange rate fundamentalism.

  4. It would be interesting to know how much of the INR move has been due to short term effects vs fundamentals. Is the rupee cheap or did we just jump to a new fundamental equilibrium. Are we going to head back to 45 and lower if and when Europe issues subside. If we see progressive taxation and FDI measures from the govt, how can we quantify its effects on the INR. Ofcourse, there are models like REER etc, but is there a rough way of guesstimating the size of the expected move, given event X?


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

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