In India, there is a deeply held belief that debt inflows are bad, that all policy levers should be used to block debt inflows. In my column in Business Standard titled Doing everything wrong on debt inflows on Wednesday, I argue that a more nuanced understanding of the problems of debt inflows throws up difficulties with every aspect of existing policies.
I offer a five-part design of a sensible framework on handling debt inflows within a framework of capital controls. We could be moving towards convertibility, which would put wonks like me out of job on these kinds of questions. :-) But this five-part strategy could be put into use immediately, pending full convertibility.
I asked Joydeep Mukherji about what is going on with local currency debt issuance in Latin America. He said: Latin governments are trying seriously to get foreigners to invest in their long-term local currency debt. The Mexican government issued 20 year bonds 2 years ago and is thinking of issuing 30 year bonds now in Pesos, with fixed nominal interest rates. Most of these bonds are bought by foreigners, who have the choice of hedging currency risk if they so desire. Colombia is doing the same, but issuing in Pesos but in the foreign market (due to imperfections in its own markets). Brazil, as you cite, is trying to get FII money into its own debt market to lengthen the yield curve. These sovereigns have been trying to take advantage of the currently liquid state of global capital markets by rapidly reducing their dollar exposure, reducing a major source of vulnerability to their finances.
Update (17/6/2006): in a speech two days ago, Randall Kroszner said:
Since 2000, ten-year nominal fixed-coupon bonds in local currency have been introduced in Brazil, Colombia, Indonesia, Mexico, and Russia, while Korea issued a ten-year fixed coupon bond in 1995. To illustrate in more detail, the governments of Mexico and Korea have been able extend the average maturity of their local-currency debt significantly in just the past few years. The Mexican government issued ten-year maturities in 2001 and then 20-year maturities in 2003. The proportion of local-currency debt in Mexico maturing within one year was nearly 90 percent in 2002 and is now below 75 percent. (I have included floating rate debt in the one-year maturity category.) The Korean government continues to increase the proportion of its domestic currency debt in longer maturities, with the one-year-and-under segment falling from roughly one-half in 1999 to one-quarter by the end of last year.
Two, bond yields in local currencies of emerging-market countries have also declined. It is perhaps not surprising that, given their high rates of saving and generally high level of economic development, the governments of Hong Kong and Korea can borrow at close to industrial-country levels. More notable, however, is that the Mexican government can borrow in pesos at a ten-year maturity at rates that have averaged roughly 9 percent. And Mexico is not unique in this regard. Other middle-income emerging markets with ten-year local-currency fixed rate bond yields in the single digits include Chile, Malaysia, Russia, and Thailand, to name but a few. For countries with longer maturities, implied short-term interest rates five years ahead also have been declining and have reached very low levels, although there have been some increases in the past few months.
I got lots of comments on these issues. Here are some comments, and my responses:
Comment: But the FII limit is not binding and hence it's not really a limit. My response would be in two parts. First, my article was about the structure of capital controls, pointing out that it's just wrong to limit FII investments into INR debt (both GOI and firms) while supporting liability dollarisation by both the State and by firms. So whether the limit binds or not, the policies are wrong. The policies claim to keep India safe but are actually in the opposite direction. Second, I think that when India advertises tiny limits to the world, the biggest institutional investors get turned off. If India said that our FII framework for debt is the same as our FII framework for equity, then the big institutional investors would get going setting up offices in India to learn and trade Indian bonds.
Comment: Do foreign investors really want to buy INR denominated GOI bonds? My personal experience is: Yes. By world standards, India's GDP (~ $800 billion) is a big number and by world standards, we issue a lot of debt. Global investors would generally not ignore a country of this size. A lot of people feel that the 25-year perspective on the INR is quite positive. The USD, the Euro and the JPY have problems. I think it's not hard to make a case to global fixed income portfolios that they should be putting 2% into India for the purpose of diversification and for the purpose of riding the optimistic future of the INR.
Comment from Jayanth Varma: I agree with Ajay Shah that this is truly absurd but it is fully explained by the political economy of the situation. The corporate sector usually gets what it wants through intensive lobbying. This happened in East Asia before the crisis and it is happening in India now. I think a good businessman understands that liability dollarisation isn't pretty, so I'm not sure the corporate sector yearns for the present policy framework. I thought the political economy was about the influence that banks have with RBI. Whether it's ECB or NRI deposits, it looks like policies are being designed to feed fees to banks and shrivel securities markets.
Comment from Jayanth Varma: If foreign investors are allowed to hedge currency risk, then the true national exposure may still be that of foreign currency debt if it is an Indian entity that stands on the other side of the hedge. In fact, the effective position of the nation can be that of short term foreign currency debt. This problem is not insurmountable but some thought needs to be given to it. I think in terms of a world with covered interest parity (CIP). Fairly soon now, India will have a GDP of $1 trillion and gross flows across the border of $1 trillion per year. I think in terms of a world where infinite capital is deployed into CIP arbitrage. Then whether a few billion dollars here and there seek currency hedging doesn't change anything big. If we think that India will have a debt/GDP of roughly 60% and if 10% of the debt market is owned by foreigners, then that's (at the maximum) roughly $60 billion of outstanding hedge positions on the currency derivatives market. This is not a big number compared with flows of $1 trillion per year sloshing through the currency market.
Comment from Jayanth Varma: As Martin Wolf points out, many countries have been able to overcome original sin and borrow in their own currencies once they have persuaded their own citizens to lend to them. We need a proper government debt market instead of the captive market that we have now. I agree that the INR yield curve is distorted owing to financial repression (e.g. banks, insurance companies, pensions) and the lack of speculative price discovery. That should impede some foreign investors. But I don't think it's a show stopper. I have met many foreign fixed income investors who are keen on getting invested in INR debt at existing (distorted) interest rates.