Sudhir Mulji: Inflows & exchange fluctuations

The RBI should separate remittance inflows from exchange rate fluctuations

Every Saturday at twelve noon, the Reserve Bank of India (RBI) publishes data of India's foreign exchange reserves as calculated for the previous week.

Commentaries on these figures remind me of the Somerset Maugham story of an Englishman living in an obscure colony, receiving every month by boat a block of six month old London Times which he had disciplined himself to read in chronological order one paper every day. Weekly delays in Reserve statistics may deal with old information but it allows dealers to comment as if it were new.

Certain business newspapers are particularly attentive to these expert views, whether they make sense or not. For example the Financial Express of 4th May reported that "The country's forex reserves rose by a relatively low $287 million... Senior dealers said that one of reasons why the accretion was lower might be because of the decision by the RBI to cap non-resident external deposit rates at Libor."

Naïve readers like me are usually readily influenced by these views of senior dealers but occasionally the analytic part of the mind suddenly begins to wonder how these senior dealers arrive at these admittedly speculative but influential views derived from the data released by the RBI.

It is clear that these authoritative persons assume that an accretion to Reserves is an inflow of remittances, but that is not a valid conclusion drawn from the data presented.

RBI does not show changes in Reserves by separating out remittance inflows from value changes due to exchange rate fluctuations. The dollar values of RBI's reserves alter with changes in exchange rates of the dollar against other foreign currencies.

Assets held by the RBI are denominated in Euro, Sterling and Yen. All these changes are consolidated together to arrive at a net figure published by RBI.

Further RBI simultaneously announces the rupee value of its total Reserves. Thus an accretion of $287 million last week translated into an increase in rupee value of Reserves to Rs 4000 crore or approximately $900 million.

Prima facie this suggests that the inflow through remittance may well have been the usual accretion of about a billion dollars a week, but simultaneously there was a fall in values denominated in currencies other than dollars.

Thus the sort of deduction one may be able to conclude from last week's figures could be that the inflow through remittances was $900 million, but then the total stock of our Reserves fell in value by $ 613 million because the non-dollar currencies devalued against the dollar; after that to arrive at the rupee value of the Reserves we have to make further corrections for changes in the rupee-dollar rate.

Let me summarise for the sake of clarity this confusing argument; every week RBI presents a valuation of India's foreign exchange reserves both in dollars and in rupees. Changes in values occur as a consequence of fresh inflows or outflows arising out of an improvement or deterioration in exports and remittances and also as a consequence of changes in exchange rates.

However there is a tendency on the part of instant commentators to attribute the change entirely to changes in net flows and not to the outcome of RBI's portfolio management of India's foreign assets or indeed to the change in the rupee dollar exchange rate.

The final outcome on a weekly basis can be quite contradictory. Thus for example in the week ending 10th April 2004, the dollar value of our Reserves rose by $1.068 billion while their rupee value declined by Rs 6,426 crore, reflecting an appreciation of the rupee exchange rate.

Further since the change in exchange rates affects the entire accumulated stock of reserves, the affect of current inflows or outflows can be overwhelmed by changes in exchange rates.

Indeed an accretion or depletion of Reserves can occur with no change in inflows or outflows but only through a change in exchange rates and that too of changes in exchange rates between the dollar and foreign currencies like the euro-dollar or the yen-dollar rate.

Any assumption that accretions or depletions of Reserves reflect some significant trend in flows of current remittance could be quite misleading.

Are these accounting explanations of changes in Reserves of any use in economic analysis? It seems that they have some relevance as the recent controversy between Lal, Berry and Pant with Vijay Joshi on policy issues in exchange management depended to some extent on estimates of inflows into Reserves.

To make a simple point: if the entire accretion to Indian Reserves had been a consequence of portfolio management by RBI, there could be very little to argue about the government's exchange policy, although one might still wish to reflect on the use that the Reserves are put to for growth in the Indian economy.

RBI is of course fully aware of the valuation changes in "cross currency movements in exchange rates", (RBI Annual Report 2002-2003). The Report states that "Valuation gains amounted to US dollars $3.8 billion" during the year.

As a percentage of RBI's total foreign asset portfolio in March 2002 of $54 billion, this gain may not appear very significant.

However as a percentage of the annual accretion between March 2002 and March 2003 of $ 21 billion that is from total Reserves going up $54 billion to $75 billion, the percentage gain of $3.8 billion is quite significant. It is to be assumed that for this gain or accretion no rupees had to be expended by the RBI.

Further, if allowance is made for rupee appreciation from 48.5 to the dollar to 46.5, the cost of acquiring foreign exchange declines quite considerably.

And if that be really the case then the argument that RBI in accumulating Reserves loses on the interest difference between Indian and foreign interest rates is quite inaccurate. For at least some losses have been offset by an appreciation in capital gain as a consequence of holding foreign Reserves.

The general point to be made from this analysis is that whilst exchange rate fluctuations may be of little significance over long periods of time, because gains will be neutralised by losses, it is important not to ignore them in shorter periods of weekly changes.

Senior dealers should be rather more careful only to attribute these sharp fluctuations in RBI's overall figures of net accretion to a shift in the trend of inflows. In a short timeframe, changes in Reserves may have been influenced more by fluctuations in cross currency exchange rates than by a decline in the flow of remittances.

The Reserve Bank could assist analysts and foreign exchange dealers considerably by separating out remittance inflows from exchange fluctuations and announcing the value of each of these changes in their weekly notification.

It would be wise to do so because speculative attacks on currencies are encouraged by wild stories and inappropriate assumptions not just by exit polls and political expectations.

Rumours abound when uncertainty prevails, and currencies are a traditional target for ill formed analysis. There can be no better defence to such attacks than providing maximum information.

sjmulji@aol.com