FCNR(B): Has RBI bought forex stability at too high a cost?

The Economic Times September 29, 2026

The RBI’s FCNR(B) [Foreign Currency Non-Resident (Bank)] scheme has been widely lauded for attracting $132 billion. But is the amount attracted the right measure of success? A proper assessment should balance the benefits against the costs. When we do so, it is not clear that the benefits of the scheme outweigh the costs.

Without doubt, the scheme has had two positive outcomes. It has bolstered the RBI’s foreign exchange reserves, which rose from USD 682 billion when the scheme was launched on June 5, to USD 785 billion on September 4. It has also helped strengthen the rupee. In May, the rupee appeared headed towards Rs 100 to the dollar. Since then, it has stabilised in a range of Rs 94–96 to the dollar.

Those are clear results. But the deeper question is: what benefits did these outcomes bring to the economy? The answer to this however is less clear. For a start, it is not obvious why the RBI needed more reserves. USD 682 billion was already a sizeable kitty, more than sufficient to finance any prospective balance of payments deficit for this year, or even the next five years.

As for the exchange rate, there is no obvious reason why Rs 95 to the dollar is better for the economy than Rs 100 to the dollar. A more depreciated rupee does hurt importers by making foreign goods and services such as education, more expensive. But it also gives domestic producers some breathing space as they compete with cheap imports from China. Depreciation also helps exporters by making their products cheaper for foreign buyers and helping them find new markets.

In other words, the economic benefits of higher reserves and a stronger rupee are unclear. What about the costs?

First, consider the financial burden. When Indian banks gave dollars to the RBI, the RBI gave them rupees in return, creating more than Rs 10 lakh crore of liquidity. This surplus liquidity is now pushing interbank interest rates below the RBI’s policy rate. That is a problem because with inflation going up, the RBI may need to raise the policy rate in October. But to do so, it will first have to “mop up” the surplus liquidity to bring interbank interest rates back in line with the policy rate. Otherwise, the rate hike will not be effective.

However, absorbing the excess rupee liquidity would be expensive. For example, if the RBI were to sell the banks, Rs 10 lakh crore of 10-year government securities carrying a market interest rate of roughly 7 per cent, the interest cost would be about Rs 70,000 crore a year. This is a sizeable sum.

There are other options. The RBI could ask the government to bear the cost by issuing new government securities, or share part of the cost with the government. The RBI’s additional foreign exchange reserves would generate some income, for example from US Treasury securities, but even after accounting for this, the net cost to the public sector would remain large.

Alternatively, the RBI could make banks bear the cost by raising the Cash Reserve Ratio (CRR). This has two problems. First, banks that have not received the dollar deposits would also be penalised. Second, banks have to pay 6-7 percent interest rate on the NRI deposits but receive nothing on the surplus liquidity deposited with the RBI as a result of any CRR hike. This would undermine banks’ profitability.

The potential costs do not end there. The second problem is that the RBI has assumed considerable exchange rate risk. For every dollar that banks gave the RBI, they received around Rs 95 in return, assuming this to be the prevailing exchange rate. When the deposits mature in roughly five years, the RBI will have to reverse the transaction at the same exchange rate. But what happens if the rupee depreciates in the meantime? Let's say if the exchange rate is Rs 105 when the deposits mature, the RBI would lose Rs 10 for every dollar swapped.

This is only an illustration, since the future exchange rate is unknown. But even a small depreciation could impose a large cost because the amount involved is enormous. Put simply, returning $127 billion after five years could prove costly for the RBI, over and above the cost of absorbing the surplus liquidity. Future dividends from the RBI to the government would consequently suffer.

Perhaps these costs would have been worth incurring if the scheme had been used to buy time to strengthen India’s balance of payments. But no such measures have been announced either to attract more foreign direct investment or more foreign portfolio investment into Indian equities. Partly for this reason, and also because the war in West Asia has intensified, the rupee has begun to fall again.

In summary, the benefits to the economy of the FCNR(B) scheme are unclear, while its costs are significant. In the end, it has left the system weaker than before. This raises a basic question: was the scheme really needed?

The GDP debate, redux

The Business Standard September 15, 2026

The recent GDP data release has rekindled the debate about the credibility of official statistics. It seems odd that while the global economy was facing a severe energy crisis, one of the world’s largest fuel importers was experiencing an economic boom. But just because something seems odd does not mean it did not happen. Also, MoSPI has made several changes to address known problems in GDP estimation. So why are people still questioning the data?

To answer this, we need to understand the problems with the old series, and how MoSPI has addressed them. Two of the most serious problems were the lack of double deflation and inadequate measurement of informal sector output. Let’s consider how these have been addressed.

In the new GDP series, MoSPI seems to have adopted double deflation. This is welcome because, ideally, firms’ nominal value added should be deflated twice: once for output and once for inputs. In the past, MoSPI, unlike G20 countries, deflated the nominal data only once. It did not deflate the value of inputs. This was a problem because changes in input prices, such as large swings in commodity prices would distort real GDP.

In principle, therefore, the shift to double deflation is a good thing. But its impact depends upon how it is implemented. This is where questions arise, because MoSPI has not released enough detail.

For instance, we do not know how widely the double deflation method has been applied. Does it cover all nominal numbers, only the manufacturing sector, or only part of manufacturing? This matters because partial adoption of double deflation could introduce new distortions.

Consider a rise in steel prices. Steel producers will see their margins improve, while steel users will see their margins squeezed. The overall effect on the economy may be neutral. But if double deflation is applied to one group and not the other, the estimated real output will be distorted. If the squeezed margins of steel users are “restored” through double deflation while the higher margins of steel producers are not adjusted downwards, measured output will be overstated.

The choice of deflator is also critical. Most countries use the Producer Price Index (PPI). But for decades, India’s statistical office did not have the prices that factories charge their wholesalers or retailers or data on the thousands of inputs used in production.

This raises the question: what has changed? MoSPI’s own documents say that the input PPI is still under trial. If so, how reliable is it? On the output side, services account for most of the Indian economy, yet the new services PPI covers only seven sub-sectors. What about the remaining services for which no PPI exists?

Next, consider the measurement of informal sector output. MoSPI says this has been addressed through the new survey of unincorporated enterprises (ASUSE) and the Periodic Labour Force Survey (PLFS). But the quarterly ASUSE does not provide value added. So how was informal sector GDP estimated?

Finally, let's step back from the methodological details and look at the big picture painted by the GDP numbers. The problem is that the different parts do not seem to fit together. According to MoSPI, import volumes fell by 1 per cent in the April-June quarter. How, then, did the economy grow by 8 per cent and the import-dependent secondary sector, including manufacturing, grow by 10 per cent in real terms?

The data also show that the value of imports rose by 31 per cent, implying that import prices increased by a staggering 32 per cent. Some of this increase was absorbed by the government, while GST rates were cut (a year ago) and public capex also increased. But these measures were far too small to offset a shock equivalent to around 1.5 per cent of GDP. Someone still had to pay the higher energy bill. This included industry, restaurants and consumers, some of whom had to buy LPG cylinders on the black market. Yet the data suggest that despite this squeeze on incomes, spending increased sharply.

In summary, despite MoSPI’s efforts, it remains difficult to assess whether the problems with the earlier GDP series have been adequately addressed. Part of the problem is transparency. MoSPI has released some information, but it is not enough. The press releases and Frequently Asked Questions (FAQs) provide little detail. The committee reports are more detailed, but it is unclear which recommendations were finally implemented. To restore confidence, MoSPI should release a comprehensive Sources and Methods document, as has always been the practice, with detailed answers to these open questions.

Most important of all is to rekindle the spirit of cooperation. Given the erosion of trust in official GDP data over the years, the best way forward for both stakeholders and MoSPI is to work together to identify the problems and find solutions.