The latest data release from the Global Emerging Markets, or GEMs, consortium of 21 multilateral development banks and development finance institutions is a big step forward in informing private investors about the historical credit risks of lending to emerging markets and developing economies, or EMDEs.
Earlier versions of the dataset were too aggregated – across years, sectors and countries – to be useful. Now we have default rates and recovery rates available by year for 30 years, by sector and by country.
In addition, we have more information about the data patterns – in addition to averages, we have confidence intervals which provide information about the amount of variation around the average. We are not yet at the point of getting the data disaggregated by year and country and sector at the same time, as private investors have called for.
Given the limited number of data points, this level of disaggregation would in many cases leave you with too few observations to be meaningful – even a single loan default would move the average default rate dramatically, so the latest release strikes a good balance between revealing significant differences by year, country and sector, without being too ‘noisy.’
Only fully disaggregated data will respond to private investor demands for the information they need for their risk models. GEMS should go further and publish this for those countries and sectors where it has enough data to be meaningful.
Five takeaways
What can we learn about the riskiness of lending to emerging markets and developing economies from the latest data release?
1. MDB sovereign lending is incredibly low risk. While default rates look scary at some times and in some places, MDBs are able to recover over 99% of their loans. This lends support to the proposals that MDBs can stretch their capital further, and lend more per dollar of capital without much fear of capital losses.
2. Lending to private entities is also less risky than commonly perceived, especially local currency lending. On average, 3.3% of loans over the past 10 years defaulted, with a median recovery rate of 90% which rises to 97% for local currency lending. This is less than investment grade, which is to be expected, as blue chip firms generally don’t need to borrow from MDBs and DFIs, but comparable to a single B risk rating from S&P.
3. EMDE loans could add diversification benefits to investor portfolios. The correlation of loans to private entities in the GEMS dataset with loans from all countries rated single B by S&P is only 0.46.
4. Private loan performance is surprisingly resilient to macroeconomic shocks. In the past 20 years, average default rates on loans to private entities have risen by more than 100 basis points only in one year – 2020 at the height of the COVID pandemic – when they reached 5.38% compared to the 30 average of 3.56%, before falling back to below average levels in the following years. Indeed, default rates on loans in the GEMS portfolio during the 2008 Global Financial Crisis were lower than for single B rated loans from all countries.
5. Infrastructure, which is a priority for mobilizing private capital to address climate change while meeting rising energy demand, has shown relatively low default rates for loans to private parties. Those rates: 3.27% for energy and 3.01% for utilities, below the 30 year average for all loans to private parties of 3.56%. This supports the idea that EMDE infrastructure assets could find a place in risk-diversified portfolios of pension funds and insurance companies.
Risk mitigation
Keep in mind that these default rates reflect the risk mitigants available specifically to MDBs and to some extent DFIs: preferred creditor status, a long-term relationship with governments and local counterparts, and the backing of powerful shareholders.
It also reflects well on their approach of conducting rigorous due diligence and structuring upfront to reduce risk, and being a patient investor in the event of adverse events, staying the course to resolve non-performing loans rather than taking bigger losses by selling them.
Private investors can share in these risk mitigants by coinvesting with MDBs and DFIs, and there are more ways than ever for them to do so, including IFC’s MCPP co-investment platform and the ILX Fund. They can also learn from the approach of MDBs and DFIs, and see EMDE loans as a long-term asset class which repays patient investing and a thorough approach to asset selection.
It would be useful to see comparable loan default and recovery rates for private investors in the same countries and sectors to compare their performance with MDBs and DFIs – something the consortium should pursue with market data providers. But the unfortunate truth is that for many of the countries and sectors covered by the GEMS database, there have been few private lenders.
So with all its limitations, the GEMS data provides the best source of information on historical default and recovery rates in many countries and sectors, over the longest time.
There is much in the current database which can be helpful to reassure private investors that lending to EMDEs is not a leap into the unknown, but rests on a track record of loan repayment and recovery comparable to other debt assets, with risk diversification benefits.
Hence, this further step towards data transparency can help improve perceptions of EMDE investment risk, lowering one obstacle to mobilizing more much-needed capital into these countries.
Neil Gregory is a senior research associate at ODI.
