26th March, 2024Philippe Belot

LOCAL CURRENCY LENDING: The challenge faced by Multilateral Development Banks

Multilateral Developments Banks (MDBs) have been very successful since the first of them, the International Bank for Reconstruction and Development (aka the “World Bank”), was established in 1944. It has become the model for many other multilateral, regional development institutions. In 2022 the seven largest MDBs[1] had assets in excess of USD 1,500 billion. However, the world has changed significantly since those banks were established and MDBs now face several challenges[2]. One of them is local currency lending.

MDBs have traditionally been lending in hard currencies (HC) and particularly in US dollars. Several reasons explain this: historical reasons (the economic and financial dominance of the United States after the second World War), practical reasons (hard currencies benefit from liquid markets making funding easy, especially considering the high credit ratings enjoyed by MDBs[3]) and accounting reasons (except the AfDB[4], the MDBs mentioned above, have the US dollar or the euro as their accounting currency). Borrowing in hard currency can also sometimes be favored by clients, as it presents several advantages too (see box below).

The advantages of hard currency borrowing for clients:

  • In many developing countries where the domestic currency is not freely convertible, hard currencies can be difficult to obtain. Still, they are necessary to procure capital goods and trade internationally.
  • Hard currencies are universally accepted, convertible and a store of value.
  • They tend to be more stable and predictable.
  • Funding costs and Interests rates are usually lower.
  • Long-term maturities are available.
  • The liquidity of HC markets means that hedging is easy and cheap.

In many countries however, the time when MDBs were “the only game in town” is over. The financial sector has become stronger and plays an increasing role in the financing of emerging markets economies. In the most advanced of those markets, MDBs are looking for ways to continue bringing development impact while remaining additional. Their areas of focus are changing to encompass (i) private sector development (ii) financial inclusion (for instance SMEs or gender-smart lending) (iii) municipal infrastructure; and of course (iv) climate change.

Some of this new focus requires lending in local currency (LC).  Most private or SME clients and municipalities do not have HC revenues or assets which would warrant bearing liabilities in HC. Lending in HC to smaller borrowers did take place, sometimes at the instigation of clients themselves, attracted by lower interest rates, but in many instances it has ended badly.  HC appreciation versus LC and financial crisis have caused havoc among borrowers, bringing accusations of MDBs acting irresponsibly by transferring Forex risk to their clients. Some borrowers can absorb HC loans because they have sufficient export revenues in matching currency, but for those who don’t the risk can be too high. Even when borrowers understand the risk they are taking, they may not have the capability to manage it efficiently. Hedging contracts tend to be short term and may not cover the full term of the borrowing. And in many emerging markets, the hedging instruments may not be readily available, or affordable.

Hence, MDBs are looking for ways to lend in LC to widen their client base and offer them a more fitting product. This looks like a simple question. Don’t most commercial banks offer loans in multiple currencies?

Not so simple. MDBs do not take deposits and thus cannot engage in the classic bank activity called “transformation” (transforming client’s deposits into loans). They do not have subsidiaries or branches in countries where they could raise money locally[5].

The money that cannot be raised through deposits has to be raised through inter-bank or capital markets. MDBs are indeed large borrowers, aided by their high credit rating making their funding cost very low. If they can rely on the markets to raise hard currencies for their clients, why can’t they do the same for the currencies of their countries of operation? The reason is that those domestic currency markets are often shallow, without reliable interbank benchmarks, and limited possibilities to hedge. Those characteristics are precisely what makes emerging markets’ currencies softer.

MDBs have risen to the challenge and embarked on several strategies to provide the LC required by their projects. One of them consists in operating through local financial intermediaries (FI), banks or microfinance institutions. In this two-tier lending model, MDBs fund the FIs in HC. In turn, the FIs deploy their capital to their clients in LC. This transfers the FX risk to the FI, who have broader shoulders than their customers. They may also have HC savings and deposit accounts to fund the MDB interest and principal repayments. When MDBs target specific classes of clients in this way, it becomes “directed lending”[6]. This very successful model is increasingly used to finance SMEs and green projects, for example.

A number of options exist when MDBs wish to finance projects directly in LC, for example:

  • Cross-currency swaps, when available. The MDB borrows the money in HC, swaps it for LC and on-lends the proceeds in LC. This is an expensive solution, though, as the bank pays interest on the HC plus a margin on the swap, a risk premium that integrates the risk of LC depreciation. Also, cross-currency swaps rarely cover long maturities, which is what MDBs are intent on providing, and refinancing risk can become acute. Finally, projects with complex cash flows would require entering into a series of swaps matching the cash flows of the underlying investment transactions, which can be difficult, costly and inflexible.
  • Another solution consists in raising LC through off-shore or on-shore bonds. MDBs are sought-after borrowers and raising money in that way not only satisfies the needs of the MDBs but also participates in the deepening of local capital markets. However, bonds being wholesale instruments, the proceeds need to be invested or hedged until used in individual lending transactions. And then, reflows are unlikely to match the repayments of the bond. In short, it creates a position in local currency that most MDBs are unwilling to run.
  • The MDB can initiate swap transactions with the central bank itself, offering an alternative where the cross-currency swap market in the currency concerned is not sufficiently developed. This benefits both parties, providing the central bank with HC and the MDB with access to LC. But is dependent upon the willingness of the central bank to cooperate and essentially transfer the FX risk to the public purse.
  • Guarantees can be a way to support projects otherwise financed by local banks in LC. In this case, the MDB does not bear Forex risk except on the guarantee fee itself, which is paid in LC, or – but only in case of default – on the principal. One variant of guarantees are unfunded risk participation agreements. This comes handy in syndications where MDBs and local banks finance a project jointly in LC, but the MDBs’ tranches are unfunded.

The lack of affordable hedging[7] in many emerging markets restricts MDBs’ ability to borrow and lend in LC because they would then incur significant Forex risk. Their mandate requires them to take project risk only and avoid speculating on volatile currencies. Their often conservative treasury and funding policies reflect this lack of appetite for currency risk.

Some MDBs have solved the LC issue in a radical way, creating “pools” in LC. This consists in keeping the proceeds from bonds or swap issuances in the pool until the funds are utilised for lending. Reflows are also retained in the pool, creating in time a self-perpetuating source of funds for transactions in that currency. Flexibility is maximum, cash can be managed. This works well in any in large markets where transactions are numerous. The LC pool becomes an internal source of funding like any other. But it is unhedged and thus subject to loss should the value of the LC depreciate against the value of the HC.

Conclusion:

MDBs have several instruments at their disposal to finance their clients in LC. As no one offers a perfect solution, the selection of the instrument will depend on the MDB’s characteristics, its lending volume, degree of sophistication, nature of its portfolio and the issue it is trying to solve by lending in LC. Success in LC lending depends on a careful analysis of the issue that LC is trying to solve. It also requires an appropriate risk appetite of their management and Board, as well as an innovative and sophisticated treasury.

References:

[1] International Bank for Reconstruction and Development aka “World Bank”, European Investment Bank, International Finance Corporation, Inter-American Development Bank, Asian Development Bank, African Development Bank and European Bank for Reconstruction and Development.

[2] GBRW published in 2018 a paper entitled “Ten challenges facing MDBs”.

[3] The seven largest MDBs mentioned above benefit from a AAA credit rating.

[4] The AfDB uses a unit of account or UA equivalent to the IMF Special Drawing Rights, themselves calculated as a basket made of 43% USD and 29% EUR.

[5] MDBs have a large field presence, however only through representative offices. Funding and lending remains centralised.

[6] GBRW has explored the topic of directed lending for BII, contributing to the Discussion Paper “Directed Lending: Current Practices and Challenges” (January 2021).

[7] The creation of TCX Fund is an attempt at solving this issue.