Offshore Banking Units (OBUs) of domestic banks have borrowed from banks in the euro zone at 2 percent interest rate to meet their euro liabilities against Usance Payable At Sight (UPAS) issued by them to Bangladeshi importers. Simply stated, the UPAS issuing bank provides financing to the Bangladeshi importer so the foreign seller can receive the payment at sight basis (payment due on demand requiring the party receiving the good or service to pay immediately upon being presented with the bill of exchange) in euros while the importer enjoys a longer credit term from the bank.

Since the issuing bank has to pay euros to the seller, while they will only get the payment from the importer at the end of the credit term, the issuing bank either uses its own euro liquidity or borrows euros if it does not have such liquidity. These interest payments on such borrowings are outflows of foreign exchange from Bangladesh. Bangladesh Bank (BB) on the other hand has liquid assets in Euro currently invested at negative interest rates. This is causing erosion of BB's euro liquid assets.

There is therefore a potential win-win if BB uses these euro liquidity for lending to Bangladeshi banks. As long as the interest rate is positive but less than the rate at which Bangladeshi banks borrow from euro zone institutions, it is a win-win for both parties. So why not go ahead? According to a report published in this paper on September 12, BB has decided to lend to the local banks' OBUs "from the portion of its reserves it would invest in Euribor-linked products."

Some technical details need to be made clear at the outset. If BB lends the euros to Bangladeshi banks, it will acquire a claim in euros on domestic banks while domestic banks incur a liability in euros to BB. Since these are transactions between domestic residents, the BB claim in euros on domestic banks does not count as official foreign exchange reserves. Official reserve assets normally consist of liquid or easily marketable foreign currency assets that are under the effective control of, and readily available to, the central bank. To be liquid and freely useable for settlements of international transactions, the reserves need to be held in the form of convertible foreign currency claims of the authorities on nonresidents. This is why BB's reserve management guidelines do not allow investment of foreign exchange reserves to acquire claims over domestic residents, no matter what the currency denomination of the claim is. BB of course has made some exceptions in the past.

By proposing to invest the negative yielding liquid euros to lend to domestic private banks, BB is seeking to maximise the value of reserves and save outflows of foreign exchange from the country. Such intentions are no doubt laudable. The principles of reserve management require doing this within prudent risk limits so that reserves are always available when needed. This means reserve asset portfolios must be highly risk-averse, with a priority for liquidity and security before profit or carrying cost (such as a small negative interest rate) considerations. Lending in euros to domestic banks against UPAS is equivalent to doing a repo in euro—BB sells euro now and agrees to buy it back at some future date when the importer pays the bank at the maturity of its usance (predetermined credit period) LC.

There is inevitably a trade-off between risk and return in the context of setting reserve management priorities. The BB claims in euros on domestic banks will not be tradable in international markets as their current liquid investments in euros are. By definition, the former is less liquid and vulnerable to macroeconomic shocks to the Bangladesh economy in ways that the liquid euro assets acquired from the international market are not.

An argument for investing the reserves in claims on domestic residents can be made if the current level of reserves is considered excessive. Assessing the adequacy of reserves requires consideration of the multiple roles played by reserves, the external risks and vulnerabilities, and the opportunity cost the country faces. A number of traditional approaches—including import and short-term debt coverage—have been used and remain relevant for particular sets of countries, typically capturing individual risks.

Compared to strong 20.4 and 11.8 percent growth in FY16 and FY17 respectively, the growth in reserves was negative subsequently in FY18 and FY19, reflecting the slack in export growth, sharp increase in imports and drop in remittances before the latter recovered in FY19. Imports (goods and services) coverage of reserves has declined from 6 months in FY18 to 5.5 at the end of FY19, according to BB data. Given that BB intervenes in the foreign exchange market to keep the exchange rate relatively stable against the US dollar, assessment by the IMF's 2018 Article IV report suggests an adequacy ranging from 3 to 8.8 months of imports given the specific characteristics of Bangladesh.

Allowing domestic investment of foreign exchange reserves can open a pandora's box. Banks' lending to domestic borrowers from their OBUs will be encouraged to seek refinancing in foreign exchange from BB by offering interest higher than what BB is getting from its current holdings of reserves abroad. This risk unleashing a process where maximisation of returns may become the dominant consideration at the cost of liquidity.

There is also a moral hazard risk. When commercial banks borrow short or long term from foreign financial institutions, they are subject to the international financial market discipline that penalises excessive risk taking. When banks borrow from BB, they will be subject to the discipline exercised by the BB. If this is perceived to be malleable or less stringent than the market, there may be an inducement to take excessive risks in issuing UPAS without adequate due diligence or, for that matter, other instruments if the scope of the reserve window is expanded.

The response to the problem of liquid euro reserve erosion due to negative returns does not have to be switching to investment in domestic claims. BB can look to invest this liquidity in alternative internationally tradable assets with non-negative yields. Any adverse development on the external front would require recourse to foreign exchange reserves. Thus, a highly liquid portfolio has to be a necessary constraint in the reserve investment strategy.