Devesh Kapur & Arjun Raychaudhuri: NRI deposits for core sector

Montek Singh Ahluwalia's suggestion should be slightly modified for this.

For nearly half a century Indian policy makers operated under conditions of foreign exchange scarcity. Today they face the opposite problem""how best to manage the country's growing reserves (now nearly $175 billion). India is hardly alone in facing this embarrassment of riches from burgeoning reserves. Many Asian countries are grappling with this problem and at least one central bank of a major Asian country is reportedly technically bankrupt. Jaimini Bhagwati, in his excellent analysis on India's foreign currency assets ("Return on FCA," Business Standard, December 14, 2006), argues using data from the RBI's 2005-06 Annual Report that while the nominal rupee denominated rates of return on India's foreign currency assets (FCA) for 2005-06 was 3.9%, inflation in India was about 5%, leading to a real rate of return on India's FCA to around minus 1.1%. His solution is to divide the reserves into several portfolios with differing risk-return characteristics.

Bhagwati's focus is on the asset side; however, India also needs to creatively rethink the liability side of its foreign exchange reserves. India is unique in that a significant fraction of its reserves ($35 billion) are in the form of NRI deposits. The source of this inflow goes back to 1970, when the first scheme to attract NRI inflow in the capital account was introduced. Although a decade later deposits barely exceeded one billion dollars, they grew rapidly in the 1980s (to $12.4 billion by March 1990). Following the onset of the 1991 reforms, and especially since the late 1990s, they have grown very rapidly and are now around $35 billion.

The reasons for this growth are manifold and are discussed in the RBI Bulletin (November 2006). More interesting, however, are the costs of these deposits. The Table gives a basic summary of costs of NRI deposits, assuming that that the size of RBI investments and reserves are roughly comparable. By this calculation these schemes have notionally cost the RBI $2.7 billion over the last four years. Of course the costs depend on which benchmark interest rates are used. In the past the inflow has allowed the GoI to retire costlier debt prematurely. Today, these could be seen as an insurance payment that gives assurances to markets, thereby reducing the international cost of borrowings by Indian companies.

Nonetheless, there can be no doubt that while NRI deposits were important for India during the many decades of foreign exchange scarcity, they are increasingly costly as reserves mount. The schemes are seen as politically important for India's engagement with its diaspora even though the ability to invest in relatively high-yielding risk-free assets may be one reason why NRIs (unlike their Chinese counterparts) invest so little in FDI, whose long-term contributions to India might be more considerable. Furthermore, NRI depositors have grown accustomed to the artificially high deposit rates and this will make it even more difficult for the government to introduce a more diverse set of investment vehicles. Rather than just do away with these schemes is it possible to more creatively harness these NRI deposits""in particular the proposal raised by the Deputy Chairman of the Planning Commission, to use reserves for infrastructure financing? That proposal rightly raised concerns from many quarters, but is it possible to resurrect it but use only one specific source of foreign exchange reserves, namely NRI deposits, by transforming the term structure of these deposits while giving NRIs guaranteed access to recurring cash flows plus the flexibility of unexpected withdrawal of funds?

The first step in answering this question is to get a clear picture of what an investor really wants. Well-diversified investors typically want to allocate investments in a mix of low-risk (and lower return) assets that provide a stable cash flow and higher risk (and higher return) opportunities. Currently, NRIs looking to partake of the Indian growth story can only invest in the high risk-return equity markets. Those looking for a stable cash flow can invest in attractive NRI banking deposit schemes. However, given the burgeoning repertoire of financial instruments, this binary choice available to the retail depositor makes little sense. Is it possible to create a financial product that will give NRIs returns that are directly linked to a definable aspect of India's growth (e.g. infrastructure), a low-risk profile (a minimum level of assured return), cheap exit options (no higher than exiting a fixed deposit), and the ability to generate adequate cash flows? Unlike virtually any other country that receives substantial remittances, more than half of India's remittances ($12.5 billion of $24 billion) are local withdrawals from NRI deposits""nearly a third of NRI deposits ($35 billion) are withdrawn for local expenditures.

One possibility from the array of innovative financial products is an annuity-type financial instrument that meets the cash-flow requirements of NRI depositors while at the same creates a long-term instrument key to funding infrastructure. A potential structure would be to use a single premium variable annuity product which is similar to that of conventional pension annuities. However, it should also allow both for a minimum guarantee on the annuity payment, and flexibility to withdraw funds early at a pre-specified cost. A hypothetical example of such a product would be that it guarantees a 5% rate of return to the customer, with 50% of the upswing in investment return, over 30 years, payable every ten years. Assuming that infrastructure projects can meet a target return of 13%, this implies that the consumer gets 9%. The balance 4% is required to pay for the costs of administration, screening investments and more importantly for underwriting additional benefits for the investor. Typically, these benefits include guaranteed minimum monthly payments and/or a guaranteed minimum withdrawal rate where the customer can withdraw up to 7% of his capital every year. Studies show that the total cost of adding such flexibility ranges from 1% to 2.5% for the insurer. Such a structure would also have the advantage that customers would tend to lock in their funds, as the scheme would only distribute the gains from growth periodically.

There are undoubtedly many caveats, not least the need to establish market-based benchmarks that underlie conventional annuities. Creating several competing SPVs for the specific purpose of investing NRI funds in various infrastructure projects may be one possibility. What is critical, however, is that investments from such vehicles be entirely driven by commercial and market principles. The government will always need to support projects which are socially beneficial yet commercially unviable""but those should be funded from budgetary resources. Whatever the precise characteristics of these financial instruments, a more innovative approach to NRI capital inflow is no longer just desirable but is becoming essential.

Devesh Kapur holds the Madan Lal Sobti Professorship for the Study of Contemporary India at the University of Pennsylvania. Arjun Raychaudhuri is a consultant with Mercer Oliver Wyman in London. Mr Raychaudhuri's views above are his own and do not represent those of his organisation