Ideas to Action:

Independent research for global prosperity

This note is co-published by CGD and AidData.

Abstract

We examine the behavior of Chinese government lenders in two debt rescheduling episodes: a “low stakes” case involving Seychelles and a “high stakes” case involving the Republic of Congo. For each loan that was rescheduled, we measure the change in its grant element and the net present value (NPV) of the haircut taken by the creditor. In Seychelles, where China’s exposures were small, we find that Beijing offered deeper debt relief than Paris Club creditors, reflected in loan haircuts averaging 78% compared to 61% for Paris Club creditors. In the Republic of Congo, where Beijing had greater exposure and the borrower had limited leverage, China Eximbank actually increased the value of its portfolio in net present value terms by 23%. The fact that the Republic of Congo was worse off after rescheduling its debts with Beijing underscores the importance of exposing these deals to public scrutiny before they are finalized and building borrower country capacity to negotiate more favorable deals.

1. Background/policy context

It is increasingly clear that low income governments and their external creditors will face a debt reckoning triggered by the COVID crisis. Of the 73 countries eligible for the G20 Debt Service Suspension Initiative (DSSI), only 11 present a low risk of debt distress. The purpose of the DSSI is to create fiscal space for low income governments to respond to the health and economic effects of the pandemic in their countries. But the initiative will not address the longer-term outlook for countries whose external debt is unsustainable given that the debt payment standstill does not entail any reductions in payments (i.e., the suspensions are “neutral” in net present value terms).

No doubt, for a small number of DSSI countries, long-term debt sustainability is not a major concern and the payment suspensions will serve their short-term purpose. But for the majority of DSSI countries, some degree of debt forgiveness from creditors will likely need to be on the table. And as much as the relatively modest DSSI has already revealed fault lines across categories of creditors, major cleavages are likely to emerge when it comes to creditor responses to calls for widespread debt forgiveness. Certainly for commercial creditors, who have become increasingly active in low income country markets over the past decade, debt forgiveness in the face of defaults is likely to follow the messy script that has come to define sovereign defaults in the absence of a strong coordination and enforcement mechanism.

But even among official creditors, there are grounds for concern about the prospects for a well-coordinated and timely response to debtor country needs. Historically, coordination issues were fine-tuned through an arrangement in which the largest bilateral creditors worked on a consensus basis through the Paris Club and it in turn coordinated with the IMF and other major multilateral creditors. The Paris Club remains active, and in fact has been prominent in the DSSI. But a club-led approach has collapsed as a result of China’s emergence as the dominant bilateral creditor.

According to new estimates released by the World Bank, the Chinese government accounts for 57% of the debt owed to bilateral creditors by low income country governments (Figure 1), and debt service to China will account for two-thirds of all debt service to bilateral creditors over the next four years (Figure 2).[1]

From this dominant position, China’s behavior as a creditor in the face of widescale debt distress will substantially influence the effectiveness of any international response. In this note, we consider China’s approach to coordination with other creditors through the DSSI and its pre-DSSI approach to debt distress among its borrowers. We also undertake a detailed assessment of two country cases where official creditors from China have already completed debt restructurings. In order to anticipate how Chinese state-owned banks may engage with different types of borrowers during the DSSI era, we analyze the experiences of two countries that sit on different ends of the distribution: a “high stakes” case where Chinese creditors stood to lose a lot of money if the final terms of the rescheduling were particularly favorable to the borrower, and a “low stakes” case where Beijing had relatively limited financial exposure. We also introduce a consistent set of metrics that can be used to judge whether Chinese debt reschedulings are more favorable to lenders or borrowers.

2. China and the DSSI

The DSSI term sheet announced by the G20 indicates that eligible countries will be afforded a standstill in their debt service by “all official bilateral creditors” through December 2020. China endorsed this measure and Chinese president Xi Jinping offered encouraging words for the initiative in his speech to African leaders. Yet, in relation to the G20 term sheet, China’s participation in the DSSI remains uncertain at best and appears to be deficient in several key respects. First, while the Paris Club has communicated DSSI country actions on behalf of its members, China has not participated in these announcements. Nor has Beijing made any parallel announcements. Instead, it has reiterated its pre-DSSI approach of addressing debt problems through bilateral channels, with little or no communication with other creditor countries or borrowing countries. Second, statements from Chinese officials and other official actors suggest that Beijing has taken a more circumscribed approach by limiting the “eligible” lenders and loans. China Development Bank does not appear to be participating, despite the fact that it is an official lending arm of the Chinese government.[2]

Outside of China Eximbank, which now appears to be fully—if slowly—participating in DSSI, Beijing appears to reject the classification of “official” lending that undergirds the G20 initiative (namely, that official lending derives from government ownership of the lender), in favor of a distinction between “commercial” and “non-commercial” lending.[3] Commercial loans, whether offered by a truly private lender, an SOE, or one of China’s major policy banks (CDB, China Eximbank), reflect commercial terms, while non-commercial loans, typically offered through aid channels, are on concessional terms. China’s approach to DSSI seems to classify the non-commercial loans as “official”, that is, automatically eligible for the standstill, while commercial loan standstills will be at the discretion of the lender, consistent with DSSI’s approach to private sector lenders. In his speech to African leaders, President Xi encourages “Chinese financial institutions to respond to the G20’s DSSI”, in much the same way the DSSI itself exhorts private lenders to participate.

Five months after the G20’s DSSI announcement, China’s endorsement of the initiative seems like less of a departure from past practice than may have been implied. The scope of Chinese participation is narrower than a plain reading of the G20 term sheet suggests, and the extent of Chinese coordination across creditors and borrowers, falls far short of a multilateral approach. In fact, President Xi’s public statement seems to be a flat rejection of a multilateral approach.

If China is signaling clearly that it intends to approach debt problems as it always has, it is important to have a clear understanding what this may look like.

3. Stylized facts about the debt forgiveness and debt rescheduling practices of Chinese creditors

There are several things that we know about how official Chinese creditors deal with sovereign borrowers who cannot reliably service their debts. First, we know that Beijing has a decades-long track record of forgiving outstanding debt obligations that result from participation in its interest-free loan program, which accounts for less than 5% of China’s official lending abroad.[4] Since the 1950s, the Chinese Government has extended RMB-denominated, interest-free loans with generous repayment schedules—typically 20 year maturities and 10 year grace periods—to government borrowers. These loans do not require a counterpart funding commitment from the borrower and they typically support projects that do not generate sufficient revenue to ensure loan repayment.[5] The Chinese Government considers these loans to be part of its “foreign aid” program, and when sovereign borrowers have difficulty repaying their debts, these are typically the first loans that Beijing will forgive.[6] Since the Chinese Government offers these loans without paying much attention to the repayment capacity of the borrower or the revenue-generating capacity of the project, protracted negotiations are unnecessary. When Beijing decides to take action on these loans, it can move quickly and decisively.

Second, we know that China’s policy banks, state-owned commercial banks, and state-owned enterprises (SOEs), which account for most of the country’s official lending, take a different approach with sovereign borrowers who cannot reliably service their debts. When these institutions consider issuing a loan, they pay substantially more attention to the repayment capacity of their would-be borrowers and the revenue-generating capacity of candidate projects.[7] Unlike the administrator of China’s interest-free loan program (the Ministry of Commerce), these institutions expect to be repaid—with interest. Therefore, when borrowers fall behind on scheduled repayments, they usually seek to reschedule rather than cancel outstanding obligations. Even so, debt rescheduling deals with China Eximbank, China Development Bank, and the country’s state-owned commercial banks and enterprises are rare and they take a long time to negotiate.

Third, we know that China’s policy banks, state-owned commercial banks, and SOEs prefer that the terms of these deals be handled discreetly, if not confidentially. They are reluctant to disclose the terms of debt rescheduling deals because they fear that doing so could set off a wave of new rescheduling requests from other borrowers. The possibility of “free-riding” by other creditors provides an additional incentive to shield these deals from public scrutiny.[8] In one recent case, the IMF traveled to Beijing to meet with China Eximbank and discuss a potential debt rescheduling deal with an African borrower. China Eximbank conveyed during that meeting that one of its biggest concerns was the fact that the borrower had major outstanding obligations to a private oil company, and unless this company was willing to also reschedule its loans to the same borrower, an easing of repayment terms by China Eximbank would simply increase the seniority of the borrower’s outstanding obligations to the oil company. So, while it is true that China has spurned multiple invitations to join the Paris Club, its policy banks, state-owned commercial banks, and SOEs understand the value of coordinating with creditors outside of China and try to avoid taking actions that will encourage other creditors—in particular, commercial creditors—to position themselves as “preferred creditors” whose loans should be treated as “priority debts.”

The secrecy surrounding Chinese debt rescheduling deals has also made it difficult for policymakers to understand whether—and when—these deals are more favorable to the lender or the borrower. On the rare occasion when a borrower announces that it has rescheduled its outstanding obligations to a particular Chinese lender, it will often provide an aggregate estimate of the amount of “debt relief” that it secured. But these black box estimates of debt relief are not very useful because they fail to address a key question: whether the lifetime repayments from the borrower to the lender would have been higher or lower in the absence of a rescheduling. A creditor can ease a borrower’s repayment terms and still end up securing a larger lifetime repayment of principal and interest (in net present value terms) than it would have secured if it had left the original terms of the loan intact. So, if policy analysts and policy makers want to understand if a Chinese debt rescheduling is more favorable to the creditor or the borrower, they need objective and comparable metrics that address this counterfactual question.

In the remainder of this note, we demonstrate how this can be done when the terms of a Chinese loan before and after rescheduling are accessible. We also demonstrate how such metrics enable benchmarking of China’s loan rescheduling practices vis-à-vis the practices of other creditors.

4. A tale of two Chinese debt reschedulings

We focus on two African countries—Seychelles and the Republic of Congo (ROC)—that have recently rescheduled their debts with Chinese creditors. They represent sharply contrasting cases. Seychelles’ outstanding obligations to China were very small (approximately $20 million), so even if China agreed to a debt rescheduling deal that favored the borrower, it wouldn’t have had to absorb major financial losses. By contrast, the ROC’s outstanding obligations to China were large (nearly $2.5 billion), so a debt rescheduling deal with preferential terms for the borrower would’ve required that Chinese creditors take on significant financial losses. The Seychelles-ROC comparison is therefore useful because it allows us to see how official creditors from China behave in a low-stakes setting and a high-stakes setting. We also choose to focus on these two country cases because of data availability. Given the general lack of transparency associated with official Chinese lending, there is rarely sufficient data to assess Chinese loan restructurings in a manner that enables estimation of gains and losses to creditors and borrowers in net present value terms.[9]

The data that we have obtained from official sources in Seychelles and ROC also have two “bonus” features. In the Seychelles case, we have obtained data on the pre-restructuring and post-restructuring terms of official sector loans from China and a wider group of official creditors (including Germany, France, Kuwait, the UK, the UAE, and Japan). So, in this case, we can directly compare the rescheduling practices followed by China and other official creditors. The ROC is also a special case, but for a different reason. In 2018 and 2019, it successfully negotiated two separate rescheduling deals with official creditors from China—one deal with a policy bank (China Eximbank) and another deal with a Chinese state-owned enterprise (China Machinery Engineering Corporation, or CMEC)—and we’ve obtained detailed information about the pre-restructuring and post-restructuring terms of the loans that were subject to both deals. Since the outstanding obligations of low-income and middle-income countries to Chinese state-owned companies are particularly opaque and poorly understood, the rescheduling of ROC’s debt to CMEC provides a rare glimpse of how Chinese state-owned enterprises engage with sovereign borrowers who cannot reliably service their debts.[10]

5. Our methodological approach

We use two metrics to estimate the amount of “debt relief” secured by a debtor from Chinese creditors. The first metric is the net present value (NPV) of the haircut taken by the creditor. [11] The second metric is the change in the loan’s grant element, or the change in the amount of grant funding nested within a loan after its restructuring.

Our measurement of the size of the haircut taken by the creditor involves a comparison of the net present value (NPV) of the repayments made under the terms of a loan before restructuring with the NPV of the repayments made under the terms of the loan after restructuring. To do so, we obtained detailed data on pre-restructuring and post-restructuring loan amounts and lending terms, principal arrears and accrued interest before and after restructuring, and loan repayments from the Ministry of Finance of the Republic of Seychelles for all of the official sector loans included in our analysis.[12] The ROC’s Ministry of Finance was unwilling to provide such data. However, with information that AidData has collected on pre-restructuring loan amounts and lending terms and information that the ROC recently disclosed to its legislature on post-restructuring loan amounts and lending terms, we were able to generate amortization tables by employing some simplifying assumptions.[13]

Calculating the net present value of repayments made under post-restructuring loan terms involved several additional steps. First, we obtained the outstanding balance of each loan immediately prior to the restructuring, taking into account both arrears and creditor write-offs. For the ROC, we relied on the original amortization tables (referenced above) to estimate the outstanding balance of each loan immediately prior to its restructuring. In the case of the Republic of Seychelles, the Ministry of Finance gave us detailed information on the outstanding balance of each restructured loan immediately prior to restructuring, accounting for repayments by the borrower and principal arrears and accrued interest prior to restructuring. We then generated amortization tables to measure the total repayments made by the borrower over the restructured maturity period at the new interest rate, based on the outstanding balance at the time of restructuring. We then applied a discount factor to measure the net present value of the sum of these two repayments (i.e. pre-restructuring and post-restructuring repayments), using the OECD’s historical Commercial Interest Reference Rates (CIRRs) that are tethered to the loan’s currency of denomination and the year of the original loan agreement.

Our second metric—the percentage change in the grant element of a loan after rescheduling—estimates the net change in the effective size of the grant nested within a loan by comparing the grant element of the loan under its initial terms to the grant element of the loan under the restructured terms. The IMF defines the grant element of a loan as “the difference between its nominal value (face value) and the sum of the discounted future debt-service payments (net present value) to be made by the borrower, expressed as a percentage of the face value of the loan.” In principle, this measure varies from 0% to 100%, so loans provided on market terms have a grant element of zero, and pure grants have a grant element of 100%.[14] To calculate the grant element of a loan that is provided on below-market (concessional) terms, one needs to calculate the net present value of the future debt service payments that will be made by the borrower. The grant element calculation takes the following form:

An equation calculating the grant element of a loan.

In this equation, r represents the interest rates (or weighted mean of interest rates for a loan portfolio); m represents the maturity length in years (or weighted mean of maturities); g represents the grace period in years (or weighted mean of grace periods); n represents the number of repayments per annum, assumed to be twice a year; and D represents the discount rate of 5.00%. Equal principal repayment is assumed. We have further modified the grant element calculation to account for any arrears write-offs from the creditor (more methodological details can be found in the dataset).

This description provides the general overview of the methodological approach that we used to derive each metric, but depending on the availability of data, we had to make certain assumptions regarding components of individual loans restructurings. Readers should therefore explore the dataset to better understand the assumptions we used and how they are represented in our calculations.

6. The post-2008 restructuring of Seychelles’ debts

Following an acute balance of payment crisis in 2008, the Republic of Seychelles put in place a comprehensive debt restructuring program to reduce sovereign debt as a percentage of GDP to 50% by 2020. Part of this program involved an agreement in April 2009 with the Paris Club to restructure 160 million USD in bilateral debt. Paris Club creditors agreed to a nominal debt stock cancellation of 45%, with the remaining amounts rescheduled over 18 years, including a 5-year grace period. The restructuring agreement with the Paris Club also called upon Seychelles to restructure its outstanding obligations to other bilateral creditors and private creditors (who accounted for the majority of external debt). Then, in February 2010, Seychelles launched an exchange offer for 320 million USD of eligible private sector claims, and in April 2011, it entered into 3 loan restructuring agreements with China Eximbank. In total, the government’s restructuring program reduced the country’s total public debt stock as a percentage of GDP from more than 150% of GDP in 2008 to 60% in 2018.[15]

To compare the debt relief that Seychelles secured from Paris Club creditors and China Eximbank, we obtained data on the pre-restructuring and post-restructuring terms of 8 official sector loans from 4 Paris Club creditors: the UK, France, Germany, and Japan. There were 8 Paris Club creditors in total who participated in the reorganization of Seychelles’ debt, but we were not able to obtain complete data for 4 of these creditors (Belgium, Italy, Russia, and Spain). We also obtained complete data on the pre-restructuring and post-restructuring terms of 3 loans from China Eximbank, which represented all of the Republic of Seychelles’ outstanding debt obligations to Chinese Government institutions at the time of restructuring.[16] The 3 loans from China Eximbank were collectively worth 21.2 million USD and the 8 loans from Paris Club creditors amounted to 55.7 million USD.

Between 1997 and 2005, China Eximbank provided 3 loans to the Republic of Seychelles to fund various housing development projects. The loan balances and lending terms prior to restructuring and after restructuring are provided in Tables 1 and 2, respectively. All 3 loans were initially provided at a fixed interest rate of either 2% or 4%, with maturities ranging between 8 and 14 years, and grace periods ranging between 4 and 8 years. No repayments were made on the 3 loans prior to the April 2011 restructuring. The April 2011 restructuring resulted in substantial maturity and grace period extensions for all 3 loans. The average loan maturity was 10.6 years prior to rescheduling and 29.5 years after restructuring. The average grace period was 5.3 years prior to rescheduling and 24.8 years after restructuring. The interest rates for the 3 loans were standardized at 2% after the restructuring. Therefore, the average interest rate reduction across all three loans was 1.3%.

Table 1. China Eximbank loan details before restructuring

Project Name Face Value of Loan(in RMB) Year of Loan Commitment Maturity (years) Interest Rate Grace Period (years)
East Coast Housing Phase II 50,000,000 1997[17] 14 2% 8
Les Mamelles Housing Development 87,980,000 1998 8 4% 4
Additional - Les Mamelles Housing Project 8,305,000 2005[18] 10 4% 4

Table 2. China Eximbank loan details after restructuring[19]

Creditor Project Name Outstanding Balance at time of restructuring (in RMB) Maturity After Restructuring (years) Interest Rate After Restructuring Grace Period After Restructuring
China Eximbank East Coast Housing Phase II 49,282,473.71 30.5 2% 26.5
China Eximbank Les Mamelles Housing Development 87,891,876.33 35.5 2% 29.5
China Eximbank Additional - Les Mamelles Housing Project 8,213,615 21 2% 17

The 8 loans from Paris Club creditors were initially contracted between 1977 and 2009, and the restructuring agreements for these loans were signed between 2009 to 2015. Repayments were made on only 2 of the 8 loans prior to restructuring. The loan balances and lending terms prior to restructuring and after restructuring are reported in Table 3 and Table 4, respectively.

After restructuring, the maturities of all 8 loans were extended. The largest maturity extension came from Germany (from 7.5 years to 54 years). The average maturity length extension across all 8 loans from Paris Club creditors was 20.7 years. Given that Seychelles had not made any repayments on 6 of the 8 loans prior to restructuring, the borrower effectively received de facto grace period extensions on these loans.[20] The United Kingdom provided the most generous “extension”, from 5 years to 37 years. The average grace period “extension” across all loans from Paris Club creditors was 14.9 years. Interest rates were also reduced for 5 out of 8 loans, with reductions ranging between 2.5% and 7.19%. The average interest rate reduction across all loans from Paris Club creditors was 2.90%.

Table 3. Paris Club creditors’ loan details before restructuring

Creditor Project Name Face Value of Loan (in original currency) Year of Loan Commitment Maturity (years) Interest Rate Grace Period (years)
Banque de France Paris Club Restructuring 2009 Acc No.1 ODA 21,423,458 EUR 2009 14 2.75% 4.5
Banque de France Paris Club Restructuring 2009 Acc No.2 ODA 3,213,669 EUR 2009 14 2.75% 4.5
Government of Germany 2011 Restructuring 3,422,477 EUR 2007 7.5 2% 3
Government of Japan SMB Prawn Project Restructuring 1,076,543,236 JPY 1990 19 7.86% 6
Government of United Kingdom L249002 Restructuring 2009 164,808 GBP 1990 19 9% 8
Government of United Kingdom L249003 Restructuring 2009 2,806,716 GBP 1993 14 9% 5
Government of United Kingdom Loan 1977 Restructured 2009 6,085,687 GBP 1977 25 6% 5
Government of United Kingdom Loan 1981 Restructured 2009 2,146,031 GBP 1981 26 6% 8

Table 4. Paris Club creditors’ loan details after restructuring[21]

Creditor Project Name Outstanding balance at time of restructuring (in original currency) Year of Restructure Maturity (years) Interest Rate Grace Period (years)
Banque de France Paris Club Restructuring 2009 Acc No.1 ODA 13,355,384 EUR 2015 23.5 2.75% 4.5
Banque de France Paris Club Restructuring 2009 Acc No.2 ODA 2,003,401 EUR 2015 23.5 2.75% 4.5
Government of Germany 2011 Restructuring 3,467,762 EUR 2011 54 2% 20
Government of Japan SMB Prawn Project Restructuring 551,971,641 JPY 2009 37 0.67% 23.5
Government of United Kingdom L249002 Restructuring 2009 99,050 GBP 2009 37 3.50% 24
Government of United Kingdom L249003 Restructuring 2009 1,698,663 GBP 2009 34 3.50% 21
Government of United Kingdom Loan 1977 Restructured 2009 3,307,993 GBP 2009 50 3.50% 37
Government of United Kingdom Loan 1981 Restructured 2009 1,303,712 GBP 2009 46 3.50% 33

In summary, on some dimensions, the loan restructuring agreements that the Republic of Seychelles signed with China Eximbank compare favorably with those that were signed with Paris Club creditors. China Eximbank provided similar maturity extensions and even longer grace period extensions than those offered by Paris Club creditors. However, Paris Club creditors provided significantly more generous interest rate reductions than China Eximbank. In order to account for this complexity, we now use the haircut and grant element methods and metrics that we described in the previous section of this note to determine which loan restructuring agreements were most favorable to the borrower. After calculating the grant elements of the loans and the net present value of repayments before and after restructuring for all 11 loans from China Eximbank and Paris Club creditors, we converted all results to USD to enable comparisons across loans. Our results are reported in Tables 5 and 6.

Table 5. Haircut reductions and grant element changes for the three China Eximbank loans

Creditor Project Name Net Present Value (NPV) of Repayment (in USD) Grant Element
Original Restructured Net Change Net Change as % Original Restructured Net Change
China Eximbank East Coast Housing Phase II $3,732,640.71 $1,022,604.04 ($2,710,037) (72.6%) 25.0% 46.3% 21.3%
China Eximbank Les Mamelles Housing Development $9.972,275.02 $1,610,830.43 ($8,361,444.59) (83.8%) 5.1% 47.5% 42.4%
China Eximbank Additional - Les Mamelles Housing Project $1,083,410.60 $509,128.54 ($574,282.06) (53.0%) 5.6% 37.3% 31.7%
Aggregate $14,788,326.33 $3,142,562.63 ($11,645,763.70) (78.7%) 11.9% 46.5% 34.6%

Table 6. Haircut reductions and grant element changes for the eight loans from Paris Club creditors

Creditor Project Name Net Present Value (NPV) of Repayment (in USD) Grant Element
Original Restructured Net Change Net Change as % Original Restructured Net Change
Banque de France Paris Club Restructuring 2009 Acc No.1 ODA $20,817,930.32 $13,896,356.81 ($6,921,573.51) (33.3%) 16.2% 53.8% 37.6%
Banque de France Paris Club Restructuring 2009 Acc No.2 ODA $3,122,835.60 $2,084,550.73 ($1,038,284.87) (33.3%) 16.2% 53.8% 37.6%
Government of Germany 2011 Restructuring $3,270,908.38 $1,356,362.70 ($1,914,545.68) (58.5%) 13.9% 47.4% 33.5%
Government of Japan SMB Prawn Project Restructuring $10,559,006.92 $790,939.13 ($9,768,067.79) (92.5%) (26.9%) 115.3% 142.1%
Government of United Kingdom L249002 Restructuring 2009 $98,554.70 $5,030.25 ($93,524.45) (94.9%) (39.7%) 62.4% 102.1%
Government of United Kingdom L249003 Restructuring 2009 $2,358,088.38 $124,533.83 ($2,233,554.54) (94.7%) (30.7%) 61.0% 91.7%
Government of United Kingdom Loan 1977 Restructured 2009 $4,071,191.95 $34,277.32 ($4,036,964.62) (99.2%) (10.9%) 71.3% 82.1%
Government of United Kingdom Loan 1981 Restructured 2009 $2,572,892.86 $22,008.12 ($2,550,884.73) (99.1%) (11.9%) 64.1% 76.0%
Total $46,871,409.11 $18,314,008.90 ($28,557,400.21) (60.9%) 0.0%) 67.7% 67.7%

Our findings indicate that, after the restructuring in 2011, the net present value of total repayments to China Eximbank fell from 14.8 million USD (before restructuring) to 3.3 million USD (after restructuring). This reduction represents a 78% haircut in net present value terms. For the 8 loans from Paris Club creditors, the total net present value of repayments fell from 46.9 million USD (before restructuring) to 18.3 million USD (after restructuring). This reduction represents a 60.9% haircut in net present value terms. As such, China Eximbank took a larger haircut than the UK, France, Germany, and Japan did collectively. However, there is significant variation in the haircut size taken by individual Paris Club creditors. Consider the cases of France and the United Kingdom. The total net present value of repayments to the Bank of France fell from 23.9 million USD (before restructuring) to 16.0 million USD (after restructuring), which translates into a 33.3% haircut. Yet, in the case of United Kingdom, the total net present value of repayments fell from 9.1 million USD to 0.2 million USD, which represents a 98% haircut.

We also find that the weighted average grant element of all three China Eximbank loans increased from 11.9% before restructuring to 46.5% after restructuring (a 34.6 percentage point increase). While two of these loans were non-concessional prior to restructuring (based on the OECD-DAC standard of a 25% grant element cutoff for concessionality), all three qualified as concessional loans after restructuring. With respect to the eight loans from Paris Club creditors, the weighted average grant element increased by 67.7 percentage points: from 0% before restructuring to 67.7% after restructuring. All eight loans were considered non-concessional prior to restructuring, but all eight qualified as concessional loans after restructuring.

Our findings therefore suggest that Paris Club creditors did not reschedule the debts of Republic of Seychelles in a way that was substantially more favorable to the borrower than the rescheduling terms provided by China Eximbank. If anything, China Eximbank took slightly larger haircut reductions (in percentage terms).[22] But our other summary metric (the percentage change in the grant element) suggests that, on average, the Paris Club loan reschedulings were more favorable to the borrower than the China Eximbank loan reschedulings.

Also, during our correspondence with the Ministry of Finance in Seychelles, we discovered something important about China Eximbank that, to the best of our knowledge, was not previously known.[23] We learned that China Eximbank will in some cases forgive the outstanding obligations of its borrowers. In the Seychelles’ case, China Eximbank wrote off outstanding obligations related to the East Coast Housing Phase II Project worth RMB 13,963,494.56, leaving RMB 49,282,473.71 as the outstanding balance of the loan at the point of restructuring. It wrote off nearly all of the borrower’s principal arrears as well as the interest that accrued between the point that the loan was contracted and the point that the loan was rescheduled. Therefore, it effectively “reset” the outstanding balance of the loan to its original value (RMB 50,000,000) when it was first contracted in 1997 (see Table 1). It took a very similar approach with the other two loans that it issued to the Republic of Seychelles. In the case of the Les Mamelles Housing Development Project, it wrote off outstanding obligations related worth RMB 23,633,407, thereby resetting the outstanding balance of the loan to RMB 87,891,876.33.[24] In the case of the Additional - Les Mamelles Housing Project, it wrote off outstanding obligations related worth RMB 770,875, thereby resetting the outstanding balance of the loan to RMB 8,213,615. [25] In summary, these acts of debt forgiveness suggest that China Eximbank’s goal was to help the borrower “start over” with the original face values of the loans that it contracted. We also learned, through correspondence with the Ministry of Finance in the Seychelles, that this practice is generally not followed by Paris Club creditors.[26]

So, if anything, our analysis suggests that the Seychelles got a slightly better deal with China Eximbank than Paris Club lenders. But a key caveat is that the Seychelles’ outstanding obligations to China Eximbank were quite small (approximately 21.2 million USD) when the restructuring took place, so the stakes were relatively low for the lender.[27] We now turn our attention to a different case (the Republic of Congo) where the stakes were relatively high for China.

7. The post-2018 restructuring of Congo-Brazzaville’s debts

Between 2006 and 2013, the Republic of Congo (ROC) went on a borrowing spree. It signed a $1.6 billion “strategic partnership” with China Eximbank on June 19, 2006, which allowed it to obtain loans for big-ticket infrastructure projects through a securitization mechanism: Société Nationales des Pétroles Congolais (SNPC)—the country’s state-owned oil company—agreed to deposit a portion of the cash proceeds from its oil exports into an escrow account that is controlled by China Eximbank. A slew of infrastructure projects were approved and implemented through this mechanism, and after the ROC exhausted the first line of credit, China Eximbank extended another $1.6 billion line of credit in 2013 through the same securitization arrangement.

Then, global oil prices declined sharply—from more than $100 a barrel in 2013 to just over $40 a barrel in 2016—and it became significantly more difficult for the ROC to service its debts to the Chinese lender. The Congolese authorities approached the IMF for a bailout, but its initial request was rejected. Public debt had reached an unsustainable level (nearly 90% of the country’s GDP) and China Eximbank was now the country’s largest bilateral creditor, so the IMF signaled that its assistance would be contingent upon a debt rescheduling deal with China Eximbank. Calixte Nganongo, the country’s Finance Minister, initiated negotiations with China Eximbank in late 2017 but did not reached a final debt restructuring agreement until April 2019.

Unlike Seychelles, the ROC did not negotiate a restructuring agreement that encompassed all of its outstanding obligations to official creditors in China.[28] Instead, it rescheduled approximately 65% ($1.6 billion) of its total outstanding obligations ($2.5 billion) to official creditors in China.[29] It focused on 8 loans that would require near-term debt service, rather than recently contracted loans that did not yet require debt service because their grace periods were still in effect. Prior to rescheduling, all of these loans were contracted at an interest rate of 0.25%. Their maturities varied between 13 and 20 years, and their grace periods fell somewhere between 3 and 5 years.

At the point of rescheduling, the ROC’s outstanding obligations to China Eximbank for these 8 loans amounted to $1,612,330,000, and under the terms of the April 2019 rescheduling, the ROC agreed to repay 33% of its outstanding debt obligations under each of the 8 loans ($532,068,900 in total) within 3 years. For the remaining 67% of its outstanding debt obligations under each of the 8 loans, the ROC agreed to meet its obligations according to extended maturities but higher interest rates.[30] More specifically, China Eximbank agreed to extend the maturity of each loan by 15 years, reset the interest rate of a $1 billion loan for Phase 2 of National Route 1 (RN1) Project (375 km Dolisie-Brazzaville Section) to 2%, and reset the interest rates for the other seven loans to 1.5%.

Tables 7 and 8 presents the pre-restructuring terms and post-restructuring terms for each of the eight China Eximbank loans to the ROC. Given that longer maturities typically benefit borrowers but higher interest rates typically benefit creditors, we clearly need to calculate haircut reductions and grant element changes (in net present value terms) to understand if this restructuring agreement is more favorable to the lender or the borrower.

Table 7. China Eximbank loan details before restructuring[31]

Creditor Project Name Face Value of Loan (in USD) Year of Loan Commitment Maturity (years) Interest Rate Grace Period (years)
China Eximbank (1) Djiri Water Treatment Plant Extension Project 84,000,000 2010 16.5 0.25% 4
China Eximbank (2) Djiri Water Treatment Distribution Network Project 88,000,000 2010 16.5 0.25% 4
China Eximbank (3) Phase II of National Route 1 (NR1) Project (375 km Dolisie-Brazzaville Section) 1,000,000,000 2013 20 0.25% 5
China Eximbank (4) Liouesso Hydroelectric Dam Project 96,363,372 2011 19 0.25% 4
China Eximbank (5) Brazzaville Shopping Center Project 56,043,600 2014 14 0.25% 4
China Eximbank (6) Mpila Social Housing Project -- 264 Units 54,800,000 2013 13 0.25% 4
China Eximbank (7) Mpila Commercial Area (Twin Towers) Project 328,100,000 2014 16.5 0.25% 4
China Eximbank (8) Mpila Memorial Project 19,191,458 2014 14 0.25% 4

Table 8. China Eximbank loan details after restructuring[32]

Creditor Project Name Outstanding Balance at time of restructuring (in USD) Maturity After Restructuring (years) Interest Rate After Restructuring Grace Period After Restructuring (no change)
China Eximbank (1) Djiri Water Treatment Plant Extension Project 54,362,240.66 31.5 1.50% 4
China Eximbank (2) Djiri Water Treatment Distribution Network Project 56,950,918.79 31.5 1.50% 4
China Eximbank (3) Phase II of National Route 1 (NR1) Project (375 km Dolisie-Brazzaville Section) 967,859,413.30 35 2.00% 5
China Eximbank (4) Liouesso Hydroelectric Dam Project 74,519,864.41 34 1.50% 4
China Eximbank (5) Brazzaville Shopping Center Project 53,307,390.20 29 1.50% 4
China Eximbank (6) Mpila Social Housing Project -- 264 Units 45,836,844.89 28 1.50% 4
China Eximbank (7) Mpila Commercial Area (Twin Towers) Project 315,365,460.30 31.5 1.50% 4
China Eximbank (8) Mpila Memorial Project 18,254,476.18 29 1.50% 4

Our findings indicate that, after the restructuring in 2019, the net present value of total repayments to China Eximbank rose from 1.3 billion USD (before restructuring) to 1.6 billion USD (after restructuring).[33] This increase represents a 23% increase in net present value terms. To be clear, China Eximbank did not take a haircut from its ROC restructuring. The deal that was struck benefited the creditor more than the borrower.[34] This outcome stands in sharp contrast to the outcome of the 2008 rescheduling of Seychelles’ outstanding obligations to China Eximbank. We also find that the weighted average grant element of all eight China Eximbank loans to the ROC decreased from 40.3% before restructuring to 36.6% after restructuring (a 3.7% decrease).[35]

Table 9. Haircut reductions and grant element changes for the eight China Eximbank loans

Creditor Project Name Net Present Value (NPV) of Repayment Grant Element
Original (in USD) Restructured (in USD) Net Change Net Change as % Original Restructured Net Change
China Eximbank (1) Djiri Water Treatment Plant Extension Project $57,430,996.55 $56,894,348.34 ($536,648.21) (0.9%) 37.2% 38.5% 1.3%
China Eximbank (2) Djiri Water Treatment Distribution Network Project $60,165,805.91 $59,603,603.03 ($562,202.89) (0.9%) 37.2% 38.5% 1.3%
China Eximbank (3) Phase II of National Route 1 (NR1) Project (375 km Dolisie-Brazzaville Section) $796,545,795.74 $1,047,162,225.85 $250,616,430.11 31.5% 42.8% 35.3% (7.5%)
China Eximbank (4) Liouesso Hydroelectric Dam Project $66,202,635.44 $69,418,902.63 $3,216,267.19 4.9% 40.2% 39.9% (0.3%)
China Eximbank (5) Brazzaville Shopping Center Project $42,726,376.41 $48,604,043.47 $5,877,667.06 13.8% 33.9% 36.9% 3.1%
China Eximbank (6) Mpila Social Housing Project -- 264 Units $46,686,125,78 $52,972,176.33 $6,286,050.55 13.5% 32.5% 36.3% 3.8%
China Eximbank (7) Mpila Commercial Area (Twin Towers) Project $242,230,172.35 $279,852,367.50 $37,622,195.16 15.5% 37.2% 38.5% 1.3%
China Eximbank (8) Mpila Memorial Project $14,631,134.96 $16,643,871.53 $2,012,736.56 13.8% 33.9% 36.9% 3.1%
Total $1,326,619,043.15 $1,631,151,538.68 $304,532,495.53 23.0% 40.3% 36.6% -3.7%

Turning now to Table 9, one can see that these aggregate changes disguise some loan-by-loan differences. Although the maturity extensions were identical across all eight loans, the outcomes of these adjustments were different for each loan. In 2 of the 8 loans, China Eximbank took a haircut; in 6 of the 8 loans, the grant element of the loan increased—meaning that the new terms were considered more favorable to ROC than the original terms. One can also see in Table 9 that the haircut and grant element “benefits” which accrued to China Eximbank were largely driven by a single loan. Loan #3, which funded Phase 2 of the National Route 1 Project (375 km Dolisie-Brazzaville Section), was unique in that it was the largest loan (with a face value of $1 billion) and its restructured interest rate was the higher (2%) than the restructured interest rate (1.5%) that was applied to the 7 other loans.[36]

Our analysis of the ROC restructuring therefore suggests that, when the stakes are high for China Eximbank, it will negotiate aggressively to ensure that its balance sheet is protected. Another likely reason why the ROC secured less favorable terms than the Seychelles is that it was negotiating with its largest official creditor and could not afford to default on its obligations.[37] Debtors generally repay their creditors when they fear being excluded from international capital markets after a default. As such, the ROC had substantially more to lose than Seychelles by walking away from the negotiating table with China Eximbank.

Another key distinction is collateralization. None of the loans that Seychelles has contracted with China Eximbank are collateralized.[38] However, according to the IMF, the “Congolese authorities are required [by China Eximbank] to keep a minimum deposit balance equivalent to about twenty percent of total outstanding loans in an escrow account” that is controlled by the creditor. Consequently, if the ROC defaulted on its repayment obligations, China Eximbank would have had no need to take the ROC to court to recover its claims. It was in a position to seize a fully liquid asset within its direct control. The “grab and go” nature of the collateral that the Congolese authorities granted to China Eximbank almost certainly limited their leverage during the 2018/2019 rescheduling negotiations.

Apart from the eight loans that were restructured through the 2019 agreement with China Eximbank, we obtained data on the 2018 restructuring of a $551.5 million loan from China Machinery Engineering Corporation (CMEC) that the ROC contracted in 2005.[39] The pre-restructuring and post-restructuring terms of this loan are summarized in Table 10. Our calculations of the haircut reduction and grant element increase are reported in Table 11.

Table 10. CMEC loan details before and after restructuring

Creditor Project Name Face Value of Loan (in USD) Year of Loan Commitment Maturity (years) Interest Rate Grace Period (years)
China Machinery Engineering Corporation Energy Transport Network Associated with the Hydroelectric Power Station of Imboulou Project 551,507,000 2005 17 0.20% 5
Creditor Project Name Outstanding Balance at time of restructuring (in USD) Year of Restructuring Maturity After Restructuring (years) Interest Rate After Restructuring Grace Period After Restructuring (no change)
China Machinery Engineering Corporation Energy Transport Network Associated with the Hydroelectric Power Station of Imboulou Project 343,630,496.46 2018 33 0.20% 8.5

Table 11. Haircut reduction and grant element change for CMEC loan

Creditor Project Name Net Present Value (NPV) of Repayment Grant Element
Original (in USD) Restructured (in USD) Net Change (in USD) Net Change as % Original Restructured Net Change
China Machinery Engineering Corporation Energy Transport Network Associated with the Hydroelectric Power Station of Imboulou Project $331,557,351.05[40] $246,670,965.62 ($84,886,385.43) (25.6%) 39.7% 59.4% 19.7%

With respect to the CMEC loan rescheduling, we can see that the lender agreed to a far more generous restructuring than China Eximbank. It took a 25.6% haircut (equivalent to $84.9 million) and the grant element of its loan to the ROC increased by 19.7 percentage points: from 39.7% before restructuring to 59.4% after restructuring. Why would CMEC—a Chinese state-owned enterprise—agree to a rescheduling agreement that is substantially more favorable to the borrower than China Eximbank’s rescheduling agreement with the same borrower? We think the most likely explanation is that CMEC and China Eximbank had different levels of bargaining power vis-à-vis the ROC. CMEC is a company that only occasionally extends vendor financing (i.e. supplier credits) to overseas borrowers.[41] It is not a bank that repeatedly lends to the same borrowers. In fact, the Congolese authorities have not contracted any additional loans with CMEC over the last 15 years and there are no indications that they plan to borrow from it in the future. Therefore, CMEC probably had limited leverage vis-à-vis its borrower during debt rescheduling negotiations.

However, when China Eximbank entered into rescheduling negotiations with the ROC, it found itself in a fundamentally different position. It had extended dozens of loans to the ROC worth several billion dollars since 2000, and as the country’s largest bilateral creditor, it knew that the Congolese authorities would want to preserve their ability to borrow from it in the future.[42] Also, given that the ROC had a large number of outstanding obligations to China Eximbank at the time of the rescheduling negotiations, the lender had more tools at its disposal to impose discipline on the borrower. A case in point is its use of cross-default clauses, which specify that a default on one loan automatically triggers a default on other loans extended by the same creditor. In a forthcoming study with Sebastian Horn, Christoph Trebesch, and Anna Gelpern, we undertake a detailed review of the terms and conditions in Chinese loan contracts and find that 100% of China Eximbank loan contracts include cross-default clauses (in a sample from 23 low-income and middle-income countries, including the ROC). Therefore, as a “one-off” creditor to the ROC, CMEC had substantially less bargaining power than China Eximbank. [43]

These bargaining power dynamics almost certainly influence borrower calculations and behavior. If the ROC understood that the costs of defaulting or falling into arrears on China Eximbank loans would be significantly higher than the costs of defaulting or falling into arrears on a loan with CMEC, then it probably also understood that it had more leverage over CMEC and could afford to bargain for a more favorable deal than the one it would likely obtain with China Eximbank.

8. Conclusion

The Chinese government’s commitment to a multilateral approach to managing the debt crisis in developing countries remains in doubt. On the one hand, Beijing joined other G20 countries in putting forward the DSSI. Yet, almost immediately, divergent interpretations of the DSSI emerged between China and the IMF, World Bank, and Paris Club. Specifically, there is a fundamental disagreement over what constitutes an official credit, including what constitutes an official lender. The DSSI itself is a modest initiative. It does not entail loan forgiveness and absent further action by G20 creditors, the 2020 payments will come due in 2021. It would seem then that China’s disagreement with other key actors on basic terms does not bode well for more difficult discussions ahead pertaining to deep debt restructurings.

Our analysis of the Seychelles and ROC cases suggests that Chinese creditors behave very differently in low-stakes and high-stakes settings. In the Seychelles, China Eximbank pursued a debt restructuring within the same general parameters that other official lenders followed, even though it did not formally participate in Paris Club rescheduling process. In other words, in a low-stakes setting where other official lenders took significant haircuts, China Eximbank followed suit and it did so in spite of the fact that the loans in need of rescheduling were originally issued on close-to-market terms.

However, the ROC case is a cautionary tale. In this rescheduling, which poses a substantially higher level of financial risk for Beijing, the country’s largest bilateral creditor (China Eximbank) refused to take a haircut. In fact, it negotiated a debt rescheduling agreement that actually made the borrower worse off than it would have been in the absence of a debt rescheduling agreement.[44] China Eximbank will receive $300 million in additional loan repayments from the ROC (in NPV terms) under the terms of the deal that it struck with the Congolese authorities in 2019.

From a borrower country perspective, these widely divergent outcomes underscore the importance of carefully analyzing the proposed terms of a debt rescheduling before it is finalized. The metrics and methods that we have used in this note call attention to the kinds of tools that borrowers will need during the DSSI era to measure whether and to what extent the proposed terms of a Chinese debt rescheduling favor the creditor or the borrower.


[1] In June 2020, the World Bank published data on public and publicly guaranteed debt stocks and debt service by creditor country. The publicly available debt stock data from the World Bank’s Debtor Reporting System (DRS) indicate that China is the single largest bilateral lender to 46 of the 63 poorest countries in the world (that participate in the DRS). The publicly available debt service data from the DRS also indicate that China is the single largest bilateral lender to 45 of these 63 countries. See https://datatopics.worldbank.org/debt/ids/

[2] CDB describes itself as “a development financial institution which is wholly owned by the PRC government and reports directly to the State Council.” It also acknowledges that its objective is to carry out the policy goals of the Chinese government.

[3] According to the World Bank, official bilateral credit refers to “lending by sovereign governments and all public institutions in which the government share is 50 percent or above and encompassed by the SNA definition of general government, central government; state and local government; central bank; and public enterprise.”

[4] The Chinese Government refers to these loans as 无息贷款 or 免息贷款. They should not be confused with concessional loans (优惠贷款) from China Eximbank, which were previously called foreign aid loans (援外优惠贷款).

[5] For example, interest-free loans from the Chinese Government have in many cases supported the construction of presidential palaces, parliamentary buildings, convention centers, theaters, opera houses, hospitals, and schools.

[6] Finance ministries in low-income and middle-income countries often treat these loans as though they are “as good as grants”—that is to say, they fall into arrears with the expectation that their debts will eventually be forgiven. These loans can be rescheduled rather than cancelled, but in many cases they are initially rescheduled (through a grace period and/or maturity extension) and then forgiven at a later date.

[7] They generally favor projects that create or maintain revenue-generating assets, such as railways that generate ticket sales, highways that collect toll revenues, and power plants that charge residential and industrial consumers based on their levels of electricity use.

[8] China’s policy banks and state-owned commercial banks, which in many ways behave like commercial creditors, are especially concerned about the possibility of “free-riding” by other commercial lenders.

[9] To undertake such calculations, one must be able to access the terms of lending before and after restructuring.

[10] The IMF and the World Bank generally exclude loans from Chinese state-owned enterprises from their calculations of official sector debt.

[11] Haircuts refer to creditor losses in debt restructurings.

[12] We owe a debt of gratitude to Dick Labonte, Ryan Ahtave, and Stephana Hoareau for generously sharing these data.

[13] We estimate the total repayments made over the initial maturity period of a loan at its initial interest rate, assuming 2 repayments per year. We then apply a 5% discount factor to measure the net present value of repayments made by the Republic of the Congo to its creditors.

[14] While most World Bank loans fall within the parameters of 0% and 100% for grant element calculation purposes, it is technically possible to use the same equation and obtain values that are less than 0% or greater than 100%. Negative values simply indicate that a loan is priced near market terms and values in excess of 100% indicate that the pricing of a loan is highly concessional.

[15] Total debt is estimated to be 85.1% of GDP at the end of 2020. The government hopes to reduce it to 52% of GDP by 2021.

[16] We thank the Ministry of Finance of the Republic of Seychelles for confirming this point.

[17] According to the April 18, 2011 debt rescheduling agreement between the Republic of Seychelles and China Eximbank, the loan agreement for this project was signed by the lender and the borrower on August 4, 1997. We thank the Ministry of Finance of the Republic of Seychelles for sharing a copy of this agreement. SAIS-CARI identifies the loan commitment year as 2007 in its recently published database of Chinese loan commitments to African governments from 2000-2018.

[18] According to the April 18, 2011 debt rescheduling agreement between the Republic of Seychelles and China Eximbank, the loan agreement for this project was signed by the lender and the borrower on January 8, 2005. We thank the Ministry of Finance of the Republic of Seychelles for sharing a copy of this agreement. SAIS-CARI identifies the loan commitment year as 2007 in its recently published database of Chinese loan commitments to African governments from 2000-2018.

[19] The maturity after restructuring includes the time period prior to restructuring. Because no payments were made prior to the restructure, the entire pre-restructure period is treated as a de facto grace period for the restructured loans.

[20] The loans issued by Banque de France included repayments prior to the restructuring and also had no grace period extension.

[21] The maturity after restructuring includes the original time period prior to restructuring. For loans where no payments were made prior to restructuring (Germany, Japan, and the United Kingdom), the entire pre-restructure period is treated as a de facto grace period for the restructured loans. The loans with payments made prior to restructuring (France), had no extension from the original grace period.

[22] Two official creditors outside of the Paris Club also rescheduled their loans to the Republic of Seychelles: Abu Dhabi and Kuwait. Therefore, as another point of comparison, we calculated the haircut reductions and grant element changes to these loans (three from Abu Dhabi and one from Kuwait). Our analysis demonstrates that Kuwait took a 25.4% haircut and the grant element of its loan increased by 19.7 percentage points, while Abu Dhabi took an 65% haircut and the grant element of its three loans increased by 52.6 percentage points.

[23] In a recently published policy brief, Kevin Acker, Deborah Brautigam, and Yufan Huang correctly note that interest-free loans from China’s Ministry of Commerce (MOFCOM) “can be cancelled, while loans from Eximbank and Chinese companies can be restructured.” They also write that “[b]orrowers should keep in mind that Chinese loans other than [interest-free loans from MOFCOM] have almost never been, and likely will never be, considered for cancellation […].” However, an important caveat is that savvy borrowers can secure partial debt forgiveness when they are rescheduling their outstanding obligations to China Eximbank.

[24] Recall that the original face value of this loan when it was first contracted in 1998 was RMB 87,980,000.

8,305,000 (see Table 1)

[25] Recall that the original face value of this loan when it was first contracted in 2005 was RMB 8,305,000 (see Table 1).

[26] To be clear, some Paris Club creditors do write off principal arrears and/or accrued interest, but there is no evidence that they strive to help the borrowers “start over” with the original face values of the loans that they contracted.

[27] The fact that Seychelles secured favorable rescheduling terms with Paris Club creditors (in 2009 to 2010) prior to rescheduling its outstanding obligations to China Eximbank (in 2011) may have also given it more bargaining power.

[28] Unlike Seychelles, the ROC had not fallen into significant arrears on any of these China Eximbank loans. It had mostly stayed up-to-date with the original loan repayment schedules. One potential reason for this difference is that the ROC’s loans from China Eximbank were collateralized, but the China Eximbank loans contracted by Republic of Seychelles were not.

[29] There is a high level of uncertainty about the ROC’s total outstanding obligations to Chinese government lenders. In 2019, the Congolese authorities reported that the figure was $2,527,400,000. However, according recently published DRS data, the 2019 figure is $3,751,870,000.

[30] The grace periods for all eight loans were left unchanged.

[31] Several values in this table are imputed. For more information on how we imputed missing values, see our dataset.

[32] Several values in this table are imputed. For more information on how we imputed missing values, see the underlying dataset.

[33] An important caveat is that we rely on discount rates that which are tethered to the OECD’s historical Commercial Interest Reference Rates (CIRRs) rather than exit yields that account for borrower-specific levels of risk (see Schlegl et al. 2019). As such CIRRs underestimate the uncertainty of future cash-flows and our haircut estimates should be treated as lower bound estimates.

[34] The fact that the China Eximbank did not take a haircut reportedly made it more difficult for the ROC to reschedule its debts to private oil traders on favorable terms.

[35] However, both the original loans (with an average grant element of 40%) and the restructured loans (with an average grant element of 37% ) are still considered concessional loans (as opposed to market-rate loans) according to both the OECD standard of a 35% grant element cutoff and the World Bank standard of a 25% grant element cutoff.

[36] When we remove Loan #3 from our analysis, China Eximbank has an even smaller NPV “benefit” of only $54 million (as opposed to $304 million), and the grant element of the new loans is 1% more concessional (as opposed to 3.7% less concessional).

[37] At the time that Seychelles finalized its rescheduling deal with China Eximbank, approximately 90% of its external debt was held by commercial banks, Paris Club creditors, and multilateral institutions. Therefore, China Eximbank was not one of its most important creditors.

[38] Authors’ correspondence with officials from the Ministry of Finance of the Republic of Seychelles, November 2020.

[39] SAIS-CARI recently described this 2018 deal as a rescheduling of a CMEC supplier’s credit (loan) for the Imboulou Hydropower Project. However, after obtaining the rescheduling agreement between CMEC and the Republic of Congo, we learned that it does not pertain to the $238,000,000 supplier’s credit that CMEC extended to the Republic of Congo for the Imboulou Hydropower Project in 2003. It pertains to a $551,507,000 supplier’s credit that CMEC extended to the Republic of Congo in 2005 for a separate electricity transmission line project that evacuated power from the Imboulou Hydropower Project. This 2005 supplier’s credit is not recorded in SAIS-CARI’s 2000-2018 database of Chinese loan commitments to African governments.

[40] Note that this figure ($331,557,351.05) represents the net present value of repayments, which are discounted from the original loan agreement date. This discounting procedure is why total net present value payments do not meet or exceed the original face value of the loan ($551,507,000).

[41] Vendor financing is a generally risky enterprise. For Chinese companies like CMEC that implement large-scale overseas infrastructure projects, it is preferable for the host government to finance such projects with a loan from China Eximbank, China Development Bank, or another Chinese state-owned bank. Then, if the borrower defaults or falls into arrears, the bank is responsible for debt collection. If a Chinese company like CMEC offers a loan to the host government, it must shoulder all of the repayment risk that the bank would have borne.

[42] In fact, recently published DRS data suggest that the ROC was contracting new loans from the Chinese government (most likely with China Eximbank) during the same period of time (2018-2019) when it was negotiating a debt rescheduling deal with China Eximbank. We thank David Mihalyi for calling our attention to this point.

[43] Our forthcoming study with Sebastian Horn, Christoph Trebesch, and Anna Gelpern demonstrates that a cross-default clause was not included in CMEC’s 2005 loan agreement with the ROC. In the absence of a large portfolio of other loans to the same borrower, CMEC would not have gained much leverage by including a cross-default clause in its 2005 loan agreement with the ROC.

[44] On the other hand, this characterization is premised on treating these loans as official credits. From the standpoint of commercial creditors, which estimate the NPV of reschedulings according to country risk premia at the time of rescheduling, one would have expected an even more favorable NPV outcome for the Chinese lenders.

Rights & Permissions

You may use and disseminate CGD’s publications under these conditions.