Since the publication of our paper on the IFC’s project portfolio last week, we have received several helpful comments from readers. They plausibly suggest that the portfolio may be (even) less risky than we suggested, with even more space to pivot towards the low income countries where the IFC can make the most difference. But until the IFC publishes more information, we won’t really know.
Disbursements a better measure of risk than commitments?
Many people, including IFC staff, have commented that disbursements and outstanding amounts are a more important indicator of risk than commitments. Many private sector deals are committed to but never actually disburse or disburse and get quickly repaid. This might be because the borrower found other sources of finance which were less burdensome than the IFC or the economics of the investment changed for the worse and the loan ultimately did not go forward, for example. Those outcomes are probably more likely when the IFC is in competition with a number of private financiers, which is more likely in less risky environments. In turn that suggests if we used disbursement data, our point about the declining riskiness of the IFC’s portfolio would be further reinforced. But with available information we can’t be sure that is right.
Looking beyond country risk
A repeated (and accurate) point we heard is that there are other issues to consider beyond country risk. If IFC is making subordinated loans in highly levered corporate investments launching new businesses with new technologies, it is taking on a lot of risk. If it is making senior secured loans in standard investments carrying little debt that generate foreign exchange, it is not taking on much risk at all. How all of this balances out is hard to tell with the information that the IFC makes public. We know that the overall share of debt versus equity in IFC’s portfolio has remained fairly constant, but this is only part of the story. We were also told that it is often the case that multilaterals invest in senior secured debt in investment grade countries—and it’s hard to justify when there is so much private institutional capital willing to make such investments. But then again we were informed that the IFC may be investing more in the more challenging sectors and geographies of richer developing countries.
Is IFC’s lending to SMEs really indicative of greater risk?
Our argument that IFC’s lending to small and medium-sized enterprises (SMEs) might be an indicator of taking on more risk was also challenged. We learned that the IFC (and other multilateral lenders to private sector banks) are generally making senior corporate loans in to bank and non-bank financial institutions that on-lend. As a condition for those loans, they are asking their bank/non-bank clients to generate certain types of assets, such as SMEs or microfinance. But they are not taking on SME or microfinance exposures themselves; rather they are taking credit exposure to the lender as a senior lender. This is relatively low risk, especially if the exposure is to banks that are too big to fail in their domestic markets. We were told it is hard to find a development rationale for senior corporate lending to Too-Big-To-Fail banks in many emerging markets—but that kind of lending is a very large share of what development finance institutions do to “get money out the door.”
Another way to look at the issue of risk is by examining data on defaults, in particular “loss given defaults” (LGDs). Typically, we were told, the board of IFC and its sister institutions are extremely intolerant of default/headline risk so management acts accordingly. With regard to non-payment, we can’t find public data on IFC defaults, but we can quote this Moody’s report on IFC’s non-performing loans:
At the end of FY2016 total reserves against loan losses ($1.8 billion) represented 7.4% of the disbursed loan portfolio. That figure includes $965 million of specific reserves, which cover 74.6% of currently non-performing loans, indicating that outstanding NPLs pose little risk to IFC's profitability.
We’re grateful for these comments. The ones we can better evaluate suggest, if anything, that we have over-estimated the risk that the IFC is willing to take on. But other concerns are very hard to quantify given what we know about the IFC’s investments even after a lot of effort to compile data. We don’t know about final commitments or the covenants about on-lending, for example.
And that brings us back to why we started the scraping exercise to begin with: there is far too little information in the public domain about what the IFC and its sister organizations are doing with money that ultimately belongs to the taxpayers of shareholder countries.