The rupee has finally lost some of the aura of invincibility that surrounded it over the past year. Between the end of January and the second week of February, it has shed about 3 per cent of its value against the dollar.

More importantly, this period has seen a fairly major shift in market perceptions -- traders are reconciling to the prospect of medium-term decline in the exchange rate instead of dismissing any episode of rupee depreciation as a transient deviation from a path of continuous appreciation.

Will the rupee continue to depreciate? Should it? Let me take up the second question first. Clearly the depreciation is a shot in the arm for beleaguered exporters both of merchandise and services like IT. However, from a slightly broader perspective, the economic cost of sharp depreciation in the current global environment might just outweigh the benefits.

An appreciating currency effectively acted as a shield against "imported" inflation stemming from high global commodity prices. These pressures have intensified over the past month with oil prices climbing over the critical $100 a barrel barrier and food prices remaining firm.

A combination of local price impulses and these global pressures has pushed up domestic inflation. Wholesale price inflation for the week ended February 23 printed at a somewhat alarming 5.02 per cent, marginally higher than what at least the markets consider to be the RBI's tolerance limit. In such a situation continuous depreciation could spell trouble.

Is the rupee likely to depreciate much further? That depends on how the global risk environment pans out. One important facet of this is what is happening to "carry trades" -- investors borrowing cheap in low-interest markets like Japan and investing in higher-yielding assets like emerging market (including Indian) stocks. In fact, informal estimates suggest that over 40 per cent of incremental flows into India in 2006 and 2007 came from carry trades.

Carry trades have virtually come to a halt this year. As each layer of the unimaginably complex 'sub-prime' crisis has unpeeled, risk aversion has escalated. Carry traders have unwound their positions and paid back their loans. The rapid appreciation in the favourite 'carry' or funding currencies like the yen and the Swiss dollar bears testimony. This spigot of easy money that chased Indian assets is likely to stay dry for a while. Thus, going forward, the rupee is unlikely to get support from these funds.

In fact, with rising risk aversion and unwinding carry trades, portfolio flows could dwindle further. It is important to recognise that despite the sharp fall in stock-market indices, we have not seen a sharp sell-off by foreign funds yet.

Going by Sebi data, FIIs actually bought a net amount of $773 million between January 1 and February 10. It is not difficult, given the state of global markets, to envisage a situation in which FIIs turn net sellers, dumping Indian stocks and rushing to safe havens like the US treasury market or euro-zone money markets.

Risk aversion has not impacted on portfolio flows alone. Hard data are not easy to come by but I suspect that a number of other flows that drove the rupee's upward momentum last year have slowed down sharply. Anecdotal evidence suggests that there is very little volume in the international market for loans to Indian companies. This is not surprising since the problems in the international financial markets are concentrated in the banking system and have manifested in an acute liquidity squeeze and credit squeeze.

This reduction in the access to international credit comes at a time when the repayment cycle for existing loans is likely to pick up. (External commercial borrowings picked up from 2005 and the most common tenor for these loans was three and five years. That should logically mean that repayments should start rising from 2008.)

Rising repayments, coupled with a smaller quantum of fresh loans, could lead to a rise in capital outflows pressuring the rupee down further. Foreign direct inflows should, theoretically, not be driven by cyclical variation in risk-premia but my sense is that the more "risky" components such as direct investment in real estate and private equity flows are also moderating.

Finally, this phase of depreciation has come at a time (the first quarter of the calendar year) when the current account has historically been surplus or just mildly negative. The next quarter could see a much wider current account gap with obvious implications for the rupee.

Thus, to cut a long story short, the odds of further depreciation in the rupee are fairly high. I have also claimed that more depreciation could be dangerous from an inflation perspective. Can the RBI and the government thwart the depreciation then? The RBI has to turn from a persistent dollar buyer that it has been over the past year to a dollar seller. While it has enough reserves to dip into and offer dollars from, the transaction would tend to reduce local liquidity (the RBI sells dollars and sucks out rupees).

Thus liquidity management will have to be deft to prevent episodes of dollar shortage from translating quickly into rupee liquidity shortage.

Finally, it is perhaps time to revisit some of the policy impediments to inflows that were put in place to stem the flood of dollars last year. Despite the heavy dose of scepticism about the efficacy of capital controls, both the restrictions on commercial borrowings and the curbs on the use of derivative instruments by foreign investors (the P-note restrictions) have played their role in moderating the inflow of dollars. The grapevine has it that the government and Sebi are toying with the idea of lifting these curbs, given the changed circumstances. Now, that seems like a good idea.

The author is chief economist, HDFC Bank [Get Quote]. The views here are personal