Tuesday, January 31, 2006

The Chinese currency question

I just updated the Monitoring the Chinese Currency Regime page. At the outset of this effort, I hadn't appreciated the value of monitoring as much as I do now. The way this works is that data for August, September and October 2005 were closely examined and a model was formed for this `history period', where we know it's a peg to the USD (there is no basket peg, despite claims to the contrary) with miniscule currency flexibility (despite claims to the contrary).

From 1 November 2005 onwards, we have been in the `monitoring period', where new data is compared against the model for the history period. This is a great simplification for the human mind. One doesn't have to rummage in the full set of statistical tests. It is enough to just look at the picture on the web page and you know whether something has changed.

When we had embarked on this work, I thought that the Chinese currency regime was actually evolving, so that in just a few days, the monitoring procedure would reject the null. Even doing updates once a week felt like a long time, because I thought it was a moving target. In the event, remarkably enough, the Chinese have been on one currency regime from 1 August 2005 onwards. When something changes, the monitoring will pick it up!

The world continues to buzz with the consequences of China's pegged exchange rate, and views about what should be done next. Faltering US GDP growth likely hurts deep thinking about the situation: it is likely to encourage short-termism in both the US and in China. In the meantime, the problems with global imbalances seem to only get worse every day.

On 20 January, Morris Goldstein and Nicholas Lardy wrote an important piece in the FT, titled A New Way to Deal with the Renminbi, where they offer a specific proposal to break the existing logjam through four steps:

  1. They ask that in a few months, China does a 10-15% appreciation.
  2. They ask that China substantially increase currency volatility.
  3. They say China should do a fiscal stimulus to partly/full counteract the business-cycle effects of currency appreciation. I.e., they want a bigger G to compensate part or all of the reduction in X-M.
  4. They say China should retain capital controls for now.

What would our monitoring procedure detect? If currency volatility is unchanged, but there is faster appreciation, this will show up as a bigger intercept. If there is a shift away from the USD to a genuine basket peg, which will look like higher CNY/USD volatility, then this will show up as a structural break in the coefficients of other currencies in the regression. Finally, if the peg to the USD continues, but there is greater flexibility around this peg, then it will show up in the monitoring procedure as a higher sigma.

The last point is similar to what Raghuram Rajan has been emphasising: that the open capital account and a flexible exchange rate are distinct policy decisions, and one doesn't have to embrace both at once. Raghu has also emphasised that if abrupt convertibility comes about, there can be a flight of bank deposits from households who don't trust the local currency, and that can bring about a macroeconomic crisis. I disagree in one key respect. China has a massive current account. For all practical purposes, I believe that when a country has a large current account, capital controls are ineffective, because people are able to move 10% to 20% of GDP in or out of the country through overinvoicing/underinvoicing when they have views about currency fluctuations. So Yes, this might not be as big as a flight of the domestic monetary base, but even `retaining existing capital controls' pretty much entails massive capital movements.

Goldstein & Lardy rightly emphasise at the end: Admittedly, this is not an elegant plan. But if it would break the existing logjam in addressing global payments imbalances, it merits consideration. I agree. Atleast it is a plan.

Will the world bite? I believe not. Global imbalances and global warming are similar in that the Nash equilibrium of every-country-for-itself doesn't yield a very good solution. Going beyond the Nash equilibrium requires some mechanisms for negotiations and sharing-of-pain. I worry that the world does not have the political capacity to solve the present situation, that there will be no new Plaza Accord. In the past, large economies were all in the West, where certain shared values, shared goals, and political homogeneity were able to overcome joint problems. This is the first time that a large totalitarian country is deeply embedded in the world economy. (The USSR was a different kettle of fish altogether - it wasn't as big as we think, and it didn't particularly trade).

So I worry that, as Jabba the Hut would say, "there will be no bargain". We'll then just meander on to a painful end, without Plaza Accord II or Kyoto II style solutions. In the meantime, I think we should just tighten our seat-belts by buying out-of-the-money puts on the S&P 500.

Sunday, January 29, 2006

New book from the World bank

The World Bank has produced an important new book putting together current thinking on development : Economic Growth in the 1990s: Learning from a Decade of Reform. All the chapters are available as PDF files at the website. Unfortunately, some critical things, like the bibliography, are not on the website.

And, here's a review of the book by Dani Rodrik, that's forthcoming in JEL.

Cellphones induce greater market efficiency on the fish market

The best Indian economics conference that I know is the NCAER-NBER Neemrana conference, which takes place in Dec/Jan every year. This year, one fascinating paper was by Robert Jensen of Harvard, who has worked on the impact of cellular telephony on pricing efficiency on the market for fish in Kerala. It's a great story, and Swaminathan Aiyar wrote his column in Times of India this morning using it. I haven't been able to locate a PDF of the full paper by Prof. Jensen. It's well worth reading - please tell me if you find the URL.

As Jensen emphasises, what is going on is the power of the `I' in IT. Putting better information in the hands of economic agents transforms their thought process and maximisation. I feel similarly about the work that's been done in India on bringing transparency to previously opaque market, by polling dealer markets. This includes MIBOR (done by NSE) and the NCDEX polled prices for the commodity spot markets (done by CMIE and CRISIL).

(Times of India has a really terrible website, so while I normally like to link to original sources, I have instead linked up to Swami's website. If someone from TOI reads this, please see these guidelines for you are violating most of them.).

Saturday, January 28, 2006

Why did RBI raise rates?

Many observers were surprised by RBI's recent rate hike. It appears to make no sense in terms of where the economy is going. As Ila Patnaik points out in an excellent piece in Indian Express today, it isn't hard to understand what RBI did, but you have to shift your mindset to a modern framework of open economy macroeconomics.

With a pegged exchange rate, and with a de facto open capital account, RBI doesn't actually have much autonomy to run a monetary policy that is crafted to suit India's interests. The weapon of monetary policy is getting used up to deliver low currency volatility, when it could instead be used to deliver low GDP volatility. So even though it just doesn't fit in the immediate context of Indian macro, RBI is responding to the interest rate hikes of the US Fed.

Some people think it's obvious that when the Fed raises rates, every country has to also do so. But as theory and evidence show, strong correlations of monetary policy across countries only happen when you peg or fix. Frankel, Schmukler and Serven have an excellent 2002 NBER paper on the loss of monetary policy autonomy for countries that fix/peg.

Open economy macro is a new field in India, and there isn't much out there. So here are some pointers to work in this field. What is India's exchange rate regime; why do we have such a huge reserves accumulation? Ila solved this in 2003. Does this hurt the implementation of monetary policy? I.e., is there `enough' openness on the capital account for a pegged exchange rate regime to come at the price of monetary policy autonomy? Ila did this in India's experience with a pegged exchange rate, which appeared in India Policy Forum (Brookings Institution & NCAER), Volume 1, 2005 (Here's a WP version). Finally, how can one interpret India's story with capital flows from this framework? Ila and I did a paper India's Experience with Capital Flows: The Elusive Quest for a Sustainable Current Account Deficit which is forthcoming in an NBER book.

Saturday, January 21, 2006

A sharp rise in the VIX

Just last friday (14th January 2005), I wrote a blog entry saying that the VIX was very low, and it made a lot of sense to buy protection using put options on the S&P-500 at an implied volatility of 11%.

The VIX just rose and rose after that! [News story about VIX on Friday] The first opening price after that friday was at 12%. On Friday the 20th, it went up to 14.56%. Here's a table of the relevant data. While it's risen sharply, it's still at very low values by historical standards. I think all eyes will be on the VIX on Monday.

What happened!? Global equity markets have certainly become a lot more nervous. I could discern one potentially important piece of new bad news: that high US consumption based on home equity extraction could be coming to an end.

How free is India, supposedly the biggest democracy in the world?

The self-image of most of us in India is that we have freedom of speech and freedom of press. But in reality, things are much worse than meet the eye.

A simple litmus test is the scoring of `Freedom of Speech' in countries of the world, done by Freedom House. It gives India a score of 38 and a status of "Partly Free". We are a clear notch below the better-run countries of the world. We do better than China, but we're behind South Africa.

What is going wrong? A host of intrusions on freedom can be identified. The Freedom House URL above mentions:

  • The Official Secrets Act as a tool for censorship,
  • Intimidation of journalists including murder and physical violence,
  • State actions to stop advertising in Kashmir Observer,
  • The State monopoly on AM radio transmission,
  • The ban on FM transmission of news,
  • The very existence of Doordarshan,
  • The political slants of a number of TV channels,
  • The periodic bans upon books such as those that point out that Mr. Shivaji was more brigand than hero.

It adds up to quite a picture of infractions of freedom of speech. Is that a comprehensive list, or are there other violations? I can think of some others. The Indian government covered itself with shame in the episode where Yahoo Groups access within India was shut down because of one mailing list which disagreed with GOI. The State has a big role in the satellites through which TV broadcasting is done in India. The State has barriers against foreign newspapers.

Freedom is built out of a million small battles. One of these was acted out recently, when the State tried to place restrictions on the content that an Indian newspaper could carry, which was sourced from outside the country. This was challenged in the Delhi High Court, which struck down these restrictions as being incompatible with the freedom of speech enshrined in the Constitution.

That's great news! Does this mean someone can legally challenge the ban on FM stations broadcasting news?

Normally, we think that the proliferation of new technology makes it harder and harder to supress knowledge. But as China has demonstrated, electronic media are vulnerable to control, perhaps more so than physical paper was. Yahoo and Microsoft have both betrayed individuals in China, to curry favour with the State. Ila Patnaik had a nice article in Indian Express, on the freedom of speech issues associated with `Direct To Home' TV. As she says, in the case of newspapers, we know and rejoice in the fact that the State does not run newspapers, nor does it run printing presses. But in the electronic media, we accept Doordarshan, and the State has a near-monopoly on printing presses (satellites).

I find the widespread acceptance of Doordarshan and AIR -- government owned media -- to be really disappointing. Compare and contrast against the experience with the US `information' agencies, Voice of America and Worldnet Television. They are explicitly prohibited from broadcasting on US soil. As far back as 1948, lawmakers in the US clearly understood that control of a radio station in the hands of the existing administration distorts the next elections.

Things are bad in India on the related question of civil liberties also. The recent din and fury about phone tapping could hopefully take us to a state where phone tapping is made much harder. At present, violations of privacy are endemic. A friend of mine told me a story where there was a woman who suspected her husband was having an affair. She called the phone company and -- purely by talking with them on telephone -- got them to give all kinds of information about his phone records, and (worse) about his physical location on various dates. It could be that anyone armed with minor skills in social engineering, could call your telco on phone and discuss your movements and phone calls. Scary!! This links up to the civil liberties concerns abut Mapin. On a related note, Robert Cringely has an excellent article today about phone tapping in the US.

Update: Tapping phones might become slightly harder. And, my respect for The Times of India just went up a few notches when it appears that their Patna edition carried the famous cartoons. But this blogger claims that in India, this is illegal.

Friday, January 20, 2006

Another tax on transactions

Stamp duty has been on the burner of late. The `Telgi Scam' involved embezzlement of government tax revenues in the context of the "stamp duty". And the Maharashtra government tried to pull off some disastrous taxation of financial transactions, since most of the securities exchange infrastructure is located in Maharashtra.

Today there is news that the Maharashtra government has backed away from taxing bond market transactions. I don't know if that puts an end to this chapter, or if they will persist in trying to tax other markets such as equity, corporate bonds and commodity futures. There is also news about amending the relevant legislation.

Stamp duty is a tax on transactions. From first principles of public finance, we know that taxing transactions is wrong. Ila Patnaik has an excellent piece in Indian Express, where she says that it is the task of the State to maintain databases of land title, but this is a public good which should be delivered with atmost a user charge such as Rs.5 per transaction, which is roughly NSDL's price for one transaction. To tax this, or to have an ad-valorem charge, is wrong. Imagine how you would feel if when you went up to NSDL, asking him to change the ownership of some shares from Mr. X to Mr. Y, he asked for a fee -- not reflecting the cost of his work of maintaining the database of title -- but a 5% tax.

In the big picture, I see a role for bringing the entire real estate sector into the Goods and Services Tax, as had been proposed by the Kelkar FRBM Task Force, simultaneously with abolishing all taxes on real estate transactions.

Thursday, January 19, 2006

A great new atlas of India is now available

Maps in India are generally terrible. There are two problems: Bad underlying survey data, and bad presentation in converting the map database into maps.

In my opinion, for many years, the best alternative was the Lonely Planet Atlas. For some mysterious reason, this was not available in India. But if you could get it, it was high quality cartographic display, on top of low quality underlying maps data. Which is a polite way of saying that the map is remarkably often flat wrong. The LP Atlas seems to have been done in 1995 and nothing has come up after that. The LP website doesn't have it.

A great new alternative has recently sprung up: the Eicher Atlas of India. It is just Rs.370 or so and worth getting. It combines top quality cartographic display with the same flawed underlying data. On the front cover is proudly flaunted a map of the Bombay region - which is a top quality map - but I could spot atleast two mistakes in it. But it's a huge step forward compared with everything else out there.

Maps are a nice problem which help in thinking about the separation between public goods and private goods. The public good is the creation of a map database by running around the country with theodolites and GPS handsets. Once this database is created and released into the public domain, it is a perfect public good (non-rival, non-excludable). After that, the State has no business to be in the map business. The State shouldn't print maps or interfere in what citizens do with maps.

In India, we do wrong on both counts:

  • The Survey of India produces terrible maps data - riddled with mistakes, and hopelessly out of touch with the fast-changing bridges and roads. To add insult to injury, the tax-funded databases produced by the Survey of India are not released with no strings attached into the public domain.
  • The State tries to produce maps, and tries to prevent citizens from having all kinds of maps.

It's a classic Indian public policy mess of not doing the public goods that matter properly, and adding insult to injury by meddling with what free agents in the country do.

I faintly recall a Central Asian city - in Azerbaijan? - where I have heard that there are no street signs, since the Russians wanted to make life difficult for an invasion (when it came). The invaders would be doing fine with GPS, and all the Russians acheived was terrible inconvenience. Update: Naveen Mandava has a pointer to a Rand Corporation study about the security issues of release of geospatial data.

We are like that. For a long time, the State prevented those nifty GPS-based route finding computers from being embedded in cars. I believe some of these legal impediments have been solved, so that these things are now available (Rs.60,000 was the price I heard for an in-car one-city setup). Does someone know more about this?

But the proscriptions against citizens accessing 1:250,000 or 1:50,000 topo sheets remain. Every terrorist wannabe or military type can buy these maps outside India; the only benefit of the license-permit raj is to prevent citizens from leading better lives based on maps.

Wednesday, January 18, 2006

Two big-picture lectures about India

I helped Vijay Kelkar in drafting of two lectures, which actually add up to an interesting integrated view of India:

  1. The first was the Narayanan Oration (April 2004) at the Australian National University: India: On the growth turnpike.
  2. The second was the Gadgil Memorial Lecture in October 2005: India's economic future: Moving beyond State Capitalism.

They add up to interesting reading. The first looks back and thinks of deeper forces, the second looks forward at what is to be done. A slightly related version of the key ideas of the first appeared in Indian Express on Diwali day of 2003. The latter has been written up a bit. Today Andy Mukherjee linked up to it in his column.

Fixing the IPO process

The financial establishment in India is in a tizzy about the "IPO Scam". This involves multiple applications in order to capture the shares reserved for "individual investors". As Jayanth Varma has pointed out, this problem is ultimately caused by mistakes in the design of the IPO.

I wrote a column for Business Standard titled Why track down Roopalben?, where I argue that the problem in the IPO process can best be solved by shifting the IPO market to a pure uniform-price auction, where there is no special treatment of individual investors, there is no `Syndicate', and only a trivial role for the merchant banker.

Oddly enough, I had written essentially the same idea in November 1999 in the context of the IPO of Hughes Software Systems. The basic issue is the same: Should India use some kind of messy computerised IPO process, or should there be a clean pure-auction that is run by computers?

The litmus test of the soundness of the IPO mechanism is the size of IPO underpricing, i.e. the returns from IPO date to listing date. As documented in my 1995 paper, in the bad old days, we had enormous underpricing. The screen-based bookbuilding has helped reduce the size of underpricing, but it is still very large. This suggests that there is a need to improve the IPO mechanism so as to acheive pricing efficiency. The tight "price band" that is set by the investment banker today violates the notion of price discovery by the auction and not by the investment banker, as was the case in the fixed-price offerings which used to take place in the bad old days.

The issue of multiple applications at the IPO by individual investors is sometimes linked up to the (lack of) the Mapin database which will allow a clarity of identifying each investor. I believe that whether or not Mapin is up and running, the only meaningful notion of finance is where every security has a single price, regardless of the identity of the buyer. Whether I'm an indvidual or a mutual fund, I pay the identical price for a Maruti car or a share on the secondary market. There is no reason why the IPO market should violate such a rule. (I believe we need the Mapin database for other reasons, but not for this one).

Saturday, January 14, 2006

What's the Indian investor to do about global macroeconomic imbalances?

It is the best of times, it is the worst of times. On one hand, the Indian stock market is doing very well. The CMIE Cospi (2540 companies) is at a market capitalisation of Rs.24.7 trillion or roughly $550 billion. What is more, the Cospi P/E is up at 18.

At the same time, the world economy is plagued by important doubts - particularly about the US and the Chinese economies. Business Standard had an excellent editorial, a few days ago, exploring these themes.

What is doubly odd is that while many economists are truly worried about these `global macroeconomic imbalances' (e.g. Ken Rogoff, Martin Feldstein), the financial markets seem to be completely comfortable with what is going on! I use the implied volatility on the S&P 500 as a measure of expectations about future equity volatility. It is down to remarkably low levels like 11%. I can't help thinking it's a good idea to be long volatility.

What is someone invested in India to do? I think that India's exposure to a global business cycle downturn is greater than meets the eye. A lot of the Nifty stocks will take a beating if things go wrong between China and the US. Some Indian companies (e.g. Infosys) will suffer because they export a lot. Others (e.g. Tata Steel) will suffer because distressed Chinese companies will crash product prices.

And given how low the VIX is, it seems like a great idea to buy protection against the global business cycle by: Purchasing out-of-the-money put options on the S&P 500!

There is the small matter of RBI's capital controls, which prevent you or I from doing so. As Jayanth Varma says, the Indian capital account is open to all except the Honest Resident Indian (HRI). Can one use the limit of $20,000 per person per year of outbond capital in order to buy protection using the S&P 500 puts? Has anyone used this limit to do global diversification?

Friday, January 13, 2006


It is budget season in India, and debates about taxation are once again in the air.

Everybody loves to hate the new treatment of `fringe benefits'. Firms are unanimous in being unhappy about the mechanics of the "fringe benefit tax" (FBT). From first principles, what one would like is a system which measures the complete income of all employees accurately, and applies the same tax schedule to all of them. In this cases, whether a person is paid Rs.50,000 per month as a flat salary or whether Rs.20,000 of that is delivered as a zero-rental apartment, the tax rate should be neutral. Critics of the FBT need to fully articulate a non-discretionary mechanism for solving this problem. To the extent that the FBT has pushed firms towards all-cash wage packets, that's a Good thing, and is partly a useful consequence of the FBT. Ila Patnaik has an interesting article on this in yesterday's Indian Express.

In my mind, the picture is very different with the `Securities Transactions Tax' (STT). Business Standard has an excellent editorial on this. This is a tax which is simply wrong. The job of a financial sector is to deliver low transactions costs for trading. Low transactions costs imply higher market efficiency, which is the end-goal of all finance.

The STT stands at 10 bps for cash (delivery) trades, 2 bps for cash (square off) trades and 1.33 bps for derivatives trading. These are huge numbers. Consider a vanilla spot-futures arbitrage for the Nifty futures. Normally, the transactions costs would be driven by the impact cost, which is typically 10 bps on the spot and 2 bps on the future. The STT then implies a huge jump in the frictions of trading - even for this simplest strategy (spot-futures arbitrage). If you tried to run an options book, based on a delta-neutral dynamic strategy, which ran up a lot of trading, the picture would be much worse.

The STT also does something deeply wrong by exempting many parts of finance (e.g. the bond market or the commodity futures market). Given the large size of the STT, this generates huge distortions of activity away from taxed sectors towards untaxed sectors. This is a bit like the customs problem. The best rate for customs duty is 0. But if you have to have a non-zero customs duty, the best system is to have a uniform rate for all markets.

The STT really taps into the deep intuition of a person vis-a-vis finance. Some people think that financial markets are to be mistrusted, that more trading is dangerous, that there is such a thing as markets that work too well. To this instinct, the STT is a good option, for it throws sand in the wheels of finance. To others, like me, the name of the game in financial sector policy is how to obtain more liquidity - not less. If we had wanted less liquidity, why did we bother trying to build modern markets?

Monday, January 09, 2006

Capital controls in operation: Tales from the front

We all very well remember the 1960s and 1970s, when we had a complex system of quantity controls and price controls. Three things were in place: There were many quantity controls, there were many price controls, and it added up to a complex system with thousands of levers.

We know what happened in that period. Firms got focused on how to earn profits by beating the system, rather than being efficient. People who had a permit earned a rent on that permit. It was easy money. They got addicted to it, and then started political lobbying to ensure that nobody else got the scarce permits. The functioning of companies, and the technology of production, got enormously distorted in ways which emphasised earning profits by learning and exploiting the system of controls rather than being smart or being efficient.

But we learn from history that we learn nothing from history. Hence, this movie is now being played in finance, where we have the same three pieces: there are many quantity controls, there are many price controls, and it adds up to a complex system with thousands of levers.

Today, the Hindu Business Line has an article by Rajesh Abraham telling a story about a `banking stock price index ETF' product by Benchmark which has been phenomenonally successful. Benchmark are good guys, and they are among the better purveyors of index funds in India, and I like index funds in general, but Rs.2,668.39 crore invested in a bank index ETF??

I read the piece, and it's the old 1960s and 1970s story being played again. FIIs are prohibited from buying bank stocks directly, so they flock to the bank ETF as a way of getting that exposure.

On a related note, U. R. Bhat has an opinion piece in Economic Times talking about participatory notes. PNs are OTC derivatives written outside India by a finance company. The article is not as blunt as it ought to be, in one respect: No regulator in India can know what PNs are written by what firm outside India to whom.

PNs can be described as OTC derivatives written outside India. Most of the time, the PN seller would want to hedge himself by taking an offsetting position in India. Ah, I see! This is just the old 1960s / 1970s story of earning a rent on a permit. Some FIIs in India have the permit, and they're getting a fee by renting it out to the poor folk who don't.

In addition, PNs are an elegant and desirable mechanism for avoiding the procedural frictions and transactions costs of doing business in India. The buyer of the PN is avoiding these hassles, and the seller of the PN is getting a fee for taking the trouble. As U. R. Bhat says: Quite often investors who are eligible to get a FII or FII sub-account registration do participate in the Indian market through PNs. This is the case because of the procedural issues relating to registration, establishing broker and custodian relationships, undertaking forex transactions and more importantly, dealing with tax certifications, filing of income-tax returns and getting tax assessments completed. In addition, there is the added uncertainty about the tax status of foreign investors with the revenue authorities interpreting the provisions of tax treaties differently on different occasions with issues like permanent establishment, classifying investment income as business income etc, being subject matters of frequent disputes. For an investor, PNs offer an elegant solution to procedural hassles and tax uncertainties, by transferring these risks to the main FII, albeit at a price, to enable the investor to focus attention on scouting for good investment opportunities.

The price of the PN reflects the size of these frictions, and the scarcity of FII permits.

What would the 1960s / 1970s response to these things be? More control and more policemen! FIIs are beating limits on bank stocks by buying the Bank ETF? Let's Ban FII investment in ETFs! FIIs are coming in through PNs? Let's Ban PNs! We already have RBI pushing the latter (couched in the vocabulary of `unclean capital flows', which would make a 1970s bureaucrat proud). I haven't seen anyone proposing the former, yet. But who can tell? Those who fail to learn from history are doomed to repeat it.

To read more on these issues in the context of capital controls, see Kristin Forbes' excellent survey article which is part of this forthcoming NBER book.

The pension reforms story goes on

Things are looking grim on pension reform. It's been 1.5 years of the UPA government, and they haven't yet got the PFRDA Bill to vote in Parliament. The left trade unions - who seem to have veto powers on this one - seem to have taken a tough stance of opposing reforms. There don't seem to be enough sane voices within the CPI or the CPI(M) to overcome the vested interests of the left trade unions. Things are just stuck.

My most-recent Business Standard column is titled What next on pensions. I argue that you don't need the PFRDA Act to build the New Pension System. We can proceed on building NPS right now, using alternative mechanisms for regulation, and do the PFRDA Act after the next elections, when the CPI(M) no longer has veto power over all legislation.

It's possible to sign a contract between government and the Central Recordkeeping Agency (CRA), and obtain enforceability through that contract. It's possible to have SEBI regulate pension fund managers. It's possible to have contracts with the banks and post offices who're the front-end of the new pension system. Through this three-pronged approach, it's possible to use contracts, and SEBI's powers on regulating fund management, to get the New Pension System up and running.

We've gone through some interesting flip-flops on this. Under the NDA government, Jaswant Singh and S. Narayan wanted to do NPS first, let it stabilise, and then do the law. When the UPA government came along, P. Chidambaram felt that it was feasible and important to do the legislation first. For a while, it looked like the legislation would go through (we got till an ordinance). But now things look gloomy i.t.o. the ability of the UPA to do economic policy, so I say, why not just go back to the previous strategy?

The very next day, Gautam Bhardwaj wrote an article, also in Business Standard, where he has a different take on the problem. He emphasises that the original goal of Project OASIS (1998-2000) was the great masses of the uncovered sector, and not civil servants. It was only later, in December 2002, that the Government of India chose to utilise the institutional mechanisms designed by Project OASIS for the purpose of solving the civil servants DC pension problem.

Now it looks like the baggage of the civil servants problem is slowing down the core task of a pensions regulator and a DC pension system. Gautam argues that what we ought to do is proceed on building PFRDA and NPS for the uncovered sector, while separately haggling with the left parties about what should be done for civil servants.

The third new piece on pension reforms is the briefing prepared by the new effort Parliamentary Research Service, which is located at the Centre for Policy Research in New Delhi. Their analysis of the PFRDA Bill, and associated briefing materials, seek to bring multiple perspectives to bear on pending legislation, and help Members of Parliament, and the public at large, become more effective in coping with pending legislation.

SEBI as the regulator of (all) securities markets

Business Standard has another editorial on a financial architecture question: on the organisation of commodity futures regulation. At first blush, it's an obvious problem - look around the world, and all the important exchanges of the world trade futures on all kinds of underlyings under a single roof. But in India, we're headed for a messy separation of the exchange-traded equity derivatives ecosystem from the exchange-traded commodity derivatives ecosystem.

If you are an exchange trading equity index futures, you won't be able to launch a gold futures product, and vice versa. If you are a securities firm, you will need to setup two distinct subsidiaries: one trading equity futures and another trading commodity futures. If you are a customer, you will need to have two trading accounts when one would have sufficed. This will induce three kinds of difficulties:

  1. Costs will go up. India will fail to harness economies of scale and economies of scope.
  2. There will be adverse effect on competition owing to walling off subsets of the industry from each other. In the best of times, it's hard to get active competition into the exchange industry, given the network effects associated with market liquidity. We make it worse by walling off exchanges from competing with each other.
  3. It will take many years, and possibly many a disaster, before FMC replicates the learning that has already taken place at SEBI in terms of building an adequate regulatory capacity.

It looks like the UPA Cabinet has decided to support an amendment to the Forward Contracts (Regulation) Act. I haven't yet seen the amendment; it's unlikely to become public until it gets tabled in Parliament. Even if the amendment drafting is perfect, we'd get a bad policy (separation of commodity futures into a separate industry with a separate regulator).

In practice, it could easily get worse than that, if weak human capital is put into the drafting of the law. I have closely watched the evolution of SC(R)A and the SEBI Act over the last 15 years, and I know how hard it is to get the drafting right. There isn't much knowledge of finance in the Department of Consumer Affairs, where the drafting is being done, and I don't know that they are outsourcing the work to people who know. So we could easily get a flawed implementation of a bad idea.

Rational runs on cooperative banks

The weakest part of Indian finance is probably the cooperative banks. EPFO's Employee Pension Scheme (EPS) probably has a funding gap of 1% of GDP, but the cooperative banks are most-likely worse than this. Rediff has a fascinating collection of news items about fraud and crises in cooperative banks.

From a public policy perspective, it is generally thought that customers of cooperative banks are the worst off in being helpless, uninformed depositors - unable to understand the risk of banks and unable to take care of themselves. This motives all sorts of paternalism in terms of putting down government money to rescue failed cooperative banks, even beyond the ordinary processes of deposit insurance.

A big crisis in a cooperative bank took place in 2001, with Madhavpura Mercantile Cooperative Bank (MMCB). In a fascinating recent paper, Rajkamal Iyer and Jose-Luis Peydro have examined this episode. They seem to have obtained some unique data from RBI, which isn't easily available normally.

The special feature of MMCB that they focus on is a unique mechanism for interbank contagion: many other cooperative banks had held deposits with MMCB (!). Hence, when MMCB went down, there were fears about other cooperative banks also failing. As is well known, cooperative banks have negligible equity capital, and are hence unable to absorb even the smallest adverse shocks. So if a cooperative bank had even 1% of it's assets with MMCB, this would be enough to induce insolvency.

At the time, there was a run on many a cooperative bank in Gujarat. What Iyer & Peydro find is that there was more rationality in these runs than meets the eye. They find that the contagion was significantly related to exposure-to-MMCB. Banks which had a greater exposure to MMCB experienced higher depositor flight. Banks which were more sound experienced smaller withdrawals. Depositors didn't panic irrationally.

I see this as some kind of `strong form efficiency': depositors seem to have been able to obtain information about cooperative banks which is not in the public domain, and then make judgments about which banks were sound and which weren't.

This evidence contradicts the popularly held notion in India that depositors - particularly depositors in cooperative banks - are helpless, are ignorant and require paternalistic support from the government in the form of various kinds of risk-management subsidies for their transactions with banks.

The paper is here in pdf format. An easily-read version of the same ideas appeared in the H. T. Parekh column in EPW.

India's story with cooperative banks is far from finished - as the Rediff index page above clearly suggests. A recent, and excellent piece on this subject, by Tamal Bandopadhyay appeared in Business Standard.

MAPIN is back; so let's get the privacy right

MAPIN is a database about individuals, which uses fingerprints to ensure that one individual cannot have multiple accounts. This makes possible interesting applications in the financial sector. For example, if a person is debarred from working in the securities industry for X years by the regulator, the database makes it possible to prevent him from resurfacing under a new identity. SEBI has begun by making MAPIN mandatory for a few people (board of directors), but the long-term goal is to have a large number of people in the system.

Many pillars of society have complained about being treated like common criminals, and having to supply fingerprints. There are also concerns about the safety and privacy of the data.

For a while, SEBI seemed to want to kill the MAPIN system, but they have changed their mind now and seem to be headed to restart MAPIN. Business Standard has an interesting editorial on the MAPIN database. They say that the system is required, in dealing with problems like insider trading and market manipulation. They show a host of questions that the citizen should worry about on questions of safety and privacy of data, and argue that while MAPIN is needed a concerted policy focus on privacy.

The questions they pose are:

  • If a private investigator could tap Amar Singh's Reliance telephone, what is to ensure privacy of the information with MAPIN?
  • Can computer-scanned thumbprints, obtained from MAPIN, be used to frame a person at a crime scene?
  • The Amar Singh case involved attack by a private individual. But very often, in India, the worst perpetrators in terms of violation of privacy are employees of the government. Is all MAPIN data available to the IT department?
  • Can the police query one record? Can the police run a search on the full database?

My understanding of the treatment of thumbprints is that scanned images of fingerprints are put through a feature-detection algorithm, and a compact vector of characteristics is stored in the system. Testing that two fingerprints are identical is then synonymous with testing whether the two vectors are close to each other. The actual scanned fingerprint is not stored. (And, it a person can be tricked to hold a glass of water, his fingerprints can be extracted from it).

The attacks on privacy can occur in two ways: based on policy and based on violations of policy.

Violation of policy would involve attacking NSDL computer systems, unethical employees at NSDL, etc. NSDL needs to comprehensively persuade the country that it is doing a good job of blocking attacks which are based on violations of policy.

Policy-based attacks can be where a policeman walks up to NSDL and asks for information. NSDL is only the agent of SEBI. So when a policeman walks up to NSDL and asks for information from MAPIN, NSDL needs an instruction in writing from SEBI to release the information. So the question of policy-based attacks goes one deeper: SEBI needs to write down a privacy policy for the securities settlement database and MAPIN; the Income Tax Department needs to write down a privacy policy for TIN, and so on.