The Annual Statement of monetary policy for the current financial year presented by the Reserve Bank Governor, Dr Reddy, at the end of April was preceded by the RBI staff's customary, and invaluable, review of macroeconomic and monetary developments in 2007-08.

The latter document reveals that, for 2007-08, foreign exchange reserves increased by a cool $110.8 billion, or around 10% of our approximately trillion dollar GDP. Since the stock at end-year stood at $310 billion, these numbers also imply that the reserves stock grew by over 50% in the course of the last year alone.

The publication of these numbers provides a useful moment to reflect on a range of interlocking issues connected with reserve accumulation. Much of the recent debate has involved flow issues arising from the rapid increase in reserves over the past few years. These issues include the impact on saving, investment and resource allocation in the economy; the much-discussed links with the exchange regime, competitiveness, inflation and the monetary policy regime; and the distributional consequences of reserves accumulation as versus some assumed counterfactual. With respect to the stock of reserves, the issue rapidly gaining salience, partly given developments in other economies, is the periodic suggestion that India set up a sovereign wealth fund (SWF), now apparently under active review by the Reserve Bank.

Unfortunately the tendency is for these various debates to occur in isolation, whereas from a public policy perspective their welfare and distributional consequences should be assessed as a whole. It is particularly valuable to do so at the present time, when large adjustments are under way in the global economy, and India's international financial prospects are more uncertain than they have been for a while. At the same time the necessary research to understand these channels remains in its infancy.

Our ability to look at these issues holistically has been facilitated by a recent working ICRIER Working Paper*, and by ongoing work by Deepak Lal and associates at NCAER. As readers of these pages are well aware, these issues are also being debated, albeit indirectly, in the context of the Rajan Committee's draft report, available on the Planning Commission website. There it has been suggested that the RBI's own mandate be clarified to make it accountable primarily for keeping long-term inflation low, rather than the present blurred and informal mandate which assumes that the "impossible trinity" (of capital account opening, nominal exchange rate management and autonomy in setting domestic interest rates) can somehow be wished away in a financial version of the Indian rope-trick.

Finally, my own thinking on these issues has greatly benefited from a lively panel discussion at NCAER in early May. The main topic being discussed there was the current inflation and the role of monetary policy in fighting it, an issue I discussed in my last column. But much of the discussion bore also on the topic now at hand, symbolised in the view "if Singapore has done it for so long, why shouldn't we", a question that I attempt to address below.

I will address the flow issues first, because for India today those are the more important for macro-management. Although, as the Sen Gupta paper indicates, most of the relevant theory would suggest that reserve accumulation should be guided by the goal of reaching an optimal target stock, few would argue that this is a convincing explanation of the behaviour of either the RBI or of other Asian central banks since 2001. Instead, reserve accumulation has been one of the responses, perhaps the most visible, to the rebalancing of international portfolios towards emerging markets since 2001.

Sen Gupta shows (in his Figure 5) that India and China have been among the most aggressive in the use of reserve accumulation as a policy response, such reserve accumulation representing active intervention by the RBI in the foreign exchange markets.

Among Indian analysts there is some division on how to characterise the behaviour of the RBI since 2004, and especially since early 2007. I trust I represent their views correctly when I associate Ila Patnaik and Ajay Shah with empirical analysis that suggests that the RBI's intervention has for long periods been designed to stabilise the nominal rupee-dollar exchange rate. By contrast Lal, Shankar Acharya and Surjit Bhalla take the view that the RBI's exchange market intervention has been designed to stabilise the real effective exchange rate. This is a policy of which the latter two approve; Lal, on the other hand, considers such intervention to be "exchange rate protection". He finds this no different in intellectual foundation from the infant industry argument for selective protection, and involving similar opaque, non-debated distributional judgments between those favoured by intervention and those hurt by it.

Even where analysts may differ on behaviour and policy, they do at least (so far!) agree on the logic of the macroeconomic accounts. These clearly indicate that an accumulation of reserves is a gross official capital outflow, which must find its counterpart in the gross saving behaviour of some set of agents in the economy. Since these are unlikely to be in the public sector, they must lie in the household or corporate sector.

Policies that underlie reserve accumulation are also policies that indirectly induce private agents to treat the RBI as a giant mutual fund investing on their behalf in overseas Treasury bills, but with the RBI taking on the exchange, credit and income risk. We also know that the net capital inflow into the economy is represented by the current account deficit (which is projected at 1.3% of GDP by the Economic Advisory Council).

In flow terms therefore, the debate is between those like Sen Gupta and Lal, who would wish policies that would allow larger and more efficient absorption of the capital currently being made available to the economy, and those like Bhalla and Acharya, who seem to believe that mild financial repression to maintain an undervalued exchange rate, in combination with capital controls and sterilisation as necessary, remains a workable strategy.

What then about Singapore, sovereign wealth funds, the yield on assets and the like? The answer lies in the above analysis. Unlike Norway, Abu Dhabi, or Chile, our excess savings are neither based on a mineral rent, nor do they accrue directly to the public sector. Nor could anyone argue that India's investment opportunities return less at the margin than returns globally. It is not even true that our non-tradables sector is more capital-intensive than our tradables sector. How to design a financial system that is capable of bearing and assessing risk so that the private sector will invest wisely is the debate we need to be having. Returns on reserves are, at the end of the day, a sideshow.

*"Cost of Holding Excess Reserves: The Indian Experience" by Abhijit Sen Gupta. ICRIER Working Paper No. 206; www.icrier.org. March 2008

The author is Director-General, National Council of Applied Economic Research. Views expressed are personal