J Bhagwati: Should government borrow in forex?

The Government of India (GOI) currently borrows in foreign exchange (forex) only from multilateral and bilateral sources and not from commercial sources. Given the stretched nature of GOI finances and the massive volume of resources required for social sector and infrastructure development it may sound naïve to ask whether the government should borrow in forex.

Some may advocate that it should borrow whatever it can if the maturities are long and the interest rates are "concessional". This article is about the financial viability and growth impact of GOI's forex borrowings as distinct from grants.

The government's external commercial borrowing (ECB) policies, which are formulated in consultation with the Reserve Bank of India (RBI), determine the maturities and volumes of private sector borrowings in forex. Due to space constraints there is no discussion about private sector forex borrowings in this article.

A simple yardstick which is used to measure the sustainability of forex borrowings is that the nominal rate of interest on forex debt inflows (adjusted for exchange rate changes) should be lower, on average, than the nominal rate of GDP growth.

An implicit assumption in this benchmark is that the forex inflows are invested in the domestic economy and contribute towards raising GDP growth rates. Is this assumption that the government's forex borrowings are fully invested in the domestic economy valid?

Expenditure on social sector or infrastructure projects would be largely rupee denominated. It follows, therefore, that GOI's fresh forex borrowings from multilateral and bilateral sources are unlikely to be fully used.

The unutilised borrowings would end up increasing our forex reserves, which are comfortably around $143 billion, or over a year's worth of imports.

Further, if the government receives Indian rupees from the RBI in exchange for hard currency borrowings there is no immediate additionality of funds in rupees.

Over time domestic money supply would tend to rise but in the past when there has been concern about excess liquidity, the RBI has acted in concert with the government to mop up the additional rupees through stabilisation funds.

It could be argued that if it takes time to use GOI's forex borrowings, in the interim these funds could be warehoused. Let us examine the cost of inflows from the International Bank of Reconstruction and Development (IBRD) versus the rate of return while such funds are waiting to be deployed.

IBRD's amortising loans have average maturities of about seven years. In contrast, to maintain liquidity (at low interest rate risk) the RBI invests a portion of our forex reserves in instruments which have short-term maturities of up to one year.

Interest rates on IBRD's long maturity loans and the rates of return on RBI's short-term investments can be compared only if both rates are expressed on floating rate terms.

Interest rates on IBRD loans are indexed to six-month dollar LIBOR and the lending spread, commitment and front-end fees add up to about 0.7 per cent, that is, the cost is LIBOR plus 0.7 per cent. A sizeable fraction of our hard currency reserves is denominated in dollars and is invested in short-term securities such as the US Government Treasury Bills (T-Bills).

The return on six-month US T-Bills is currently around six-month dollar LIBOR minus 0.3 per cent. That is, the rate of return for warehoused funds less the cost of borrowings is negative 1 per cent per annum. Consequently, the annual cost of warehousing, $10 billion would be $100 million.

The cost of warehousing forex funds received from the Asian Development Bank (ADB) would be of a similar order of magnitude. Loans from the International Development Association (IDA), the soft lending window of the World Bank, have maturities up to 40 years and carry a service charge of 0.75 per cent. These loans are denominated in special drawing rights (SDRs).

The SDR is made up of a basket of currencies that includes the dollar, yen and the euro. The significant appreciation of the yen and other major currencies in the SDR basket against the dollar has resulted in effective IDA interest rates that are comparable to IBRD rates.

A substantial proportion of bilateral loans has been received from Japanese sources and is denominated in yen. The consistent appreciation of the yen against the rupee has meant that ex-post some of these loans have turned out to be costlier than IBRD loans.

In the last several decades, besides high rates of capital investment, the many solutions that have been proposed to accelerate growth in developing countries include, for instance, reduction in income-inequality, improvement in human capital, foreign direct investment (FDI), sound legal system, open trade policy "" the list is very long.

In this wider context, some may view the financial cost associated with warehousing loans from IBRD or other multilateral/bilateral sources as a small price to pay for the long-term growth enhancing impact of aid inflows.

However, cross-country studies have not found any conclusive evidence of a stable positive correlation between aid flows and growth, even adjusting for the bias that subsidised aid tends to go to countries that are in economic distress.

Aid advocates continue to maintain though that forex inflows promote growth if an enabling economic/political environment is in place. Others are of the view that aid enthusiasts overstate the causality relationship between forex inflows and reforms/growth in client countries.

Most observers are more likely to agree that the overwhelmingly important factor is the role of domestically accountable leadership in getting the policy framework right.

Another reason cited in favour of developing countries maintaining relationships with multilateral and bilateral aid providers is that they may be unexpectedly confronted with fiscal/balance of payments problems with the accompanying need to access aid. In addition, it could be argued that loans from multilateral sources come with technical assistance and exposure to international best practises in project preparation and implementation.

To summarise, the financial cost of warehousing GOI's unutilised forex borrowings could be significant over time and the impact of aid inflows on growth is debatable. In the last few weeks, the World Bank and ADB presidents have visited Delhi and there have been reports that the two institutions are expected to increase their lending to India.

As such, this is an opportune time to review the need for the government to borrow from multilateral or bilateral sources let alone increase such inflows.

The views expressed are personal. Comments can be sent at j.bhagwati@gmail.com