Chinese ‘Debt Traps’ in Southeast Asia: What the Data Say
This Long Read argues that while the terms of China’s aid and loans to Southeast Asian countries are less favourable compared to those of the World Bank, the weight of the evidence does not lend credence to the existence of a Chinese “debt trap” strategy in the region.
The story of China’s development finance around the world tends to be mired in arguments concerning “debt trap”. Under the “debt trap” narrative, it is said that China has set out with a grand strategy to saddle developing countries with debt by financing the building of big-scale infrastructure projects of dubious economic viability. As these developing-country borrowers struggle to repay their debts and eventually default on their loans, China will be left to seize ownership of their critical infrastructure assets. The poster child of this narrative is the Hambantota port, a project under China’s Belt and Road Initiative (BRI), which fell to Chinese control in a 99-year lease following Sri Lanka’s inability to service debt associated with the project.
On the flipside, Beijing and the host countries reject this characterisation, arguing that Chinese financing is fulfilling a niche for developing countries who have found it difficult to finance sorely needed infrastructure and development projects on their own or with the help of other financial institutions. The ease of borrowing and “no strings attached” approach have made China an attractive lender – leading analysts to describe it as a “financier of first resort”. Encapsulating the essence of this counter-narrative, Cambodia’s Prime Minister Hun Sen, at the 27th Future of Asia Conference in 2022, asked “If we don’t have investment from China, what source of electricity can we have?”.
Given that Southeast Asia has been among the key recipients of Chinese development finance, questions surrounding whether a Chinese debt trap exists in the region loom large. For the most part, scholars analysing China’s development finance programme in Southeast Asia have refrained from commenting directly about the specifics of Chinese-funded projects and instead utilised arguments from the standpoint of individual Southeast Asian countries to provide clarity on the “debt trap” question. Their perspectives shed light on how Southeast Asian countries have wielded a considerable amount of agency and have been able to extract concessions in negotiating financing terms with Beijing; and the “debt trap” argument alone does not match with the fact that numerous Chinese development projects have taken root in the region. Others have pointed out that the bureaucracy behind China’s overseas development finance programme is highly decentralised with different agencies and state-owned enterprises taking ownership of a disjointed agenda. Thus, they conclude it is unlikely that “debt trap” constitutes any form of grand strategy on the part of China.
Building on these analyses, this article utilises the AidData’s Global Chinese Development Finance (GCDF) dataset to examine the structure of China’s development finance in Southeast Asia, making comparisons with the World Bank. The analysis paints a mixed picture. While the terms of China’s aid and loans to Southeast Asian countries are less favourable compared to those of the World Bank, the weight of the evidence does not lend credence to the existence of a Chinese “debt trap” strategy in the region. Taking into account the contextual origins of China’s overseas development finance programme, the less favourable terms of China’s loans to the region stem predominantly from commercial rather than geopolitical interests. Yet, Chinese financial institutions’ risk management against defaults on their loans may entail strategic implications. This complex reality neither fits common Western conceptions of China’s “debt trap” strategy nor China’s self-image as a benevolent development partner.
CHINA’S OVERSEAS DEVELOPMENT FINANCE AND THE GCDF DATASET
China has typically been extremely secretive about details on its overseas development financing and the GCDF dataset represents a pioneering effort in uncovering this shroud of secrecy. Without such data, it is difficult to make concrete assertions about the validity of the “debt trap” argument. This explains why even though the BRI was launched almost a decade ago, there still seems to be continuous debate about a Chinese “debt trap”. In an attempt to uncover the shroud of secrecy over China’s massive development finance portfolio, the researchers behind the GCDF dataset developed a “Tracking Underreported Financial Flows (TUFF)” methodology which systematically synthesises large volumes of open-source data that aim to provide as much detail on China’s overseas development projects as possible. These data come from four main sources: (1) unredacted documentation from various Chinese institutions, (2) information from the aid/debt information management systems from government departments in recipient countries, (3) field research from scholars and NGOs, and (4) English, Mandarin and local-language media reports. Rigorous cross-referencing, verification and data quality control ensure that the data are accurate and standardised across the board. This article utilises various details in the GCDF dataset to shed light on the structure and terms of Chinese development finance in Southeast Asia. Using these insights, this article evaluates arguments about China’s “debt trap” in the region.
THE STRUCTURE OF CHINA’S OVERSEAS DEVELOPMENT FINANCE
The unwieldy nature of China’s overseas development finance has made it difficult to provide a framework for understanding its structure. Past studies have tried to provide a more systematic way of breaking down the structure of China’s overseas development finance by drawing comparisons with international development finance lending standards. According to OECD definitions, development finance flows are typically categorised as “Official Development Assistance” (ODA) or “Other Official Flows” (OOF). ODA flows are meant to be concessional in nature and the OECD uses several strict criteria to establish whether development finance flows meet standards of concessionality. For example, ODA loans tend to have much lower interest rates than the market rate. Development finance flows that do not meet the ODA criteria are deemed to be OOF; most OOF flows tend to be priced at or near market rates. This distinction allows analysts to disaggregate between what can be considered “aid” (ODA) or “debt” (OOF). For a more intuitive interpretation, the rest of the article will refer to ODA as “aid” and OOF as “debt”.
As China does not classify its projects along these definitions, the GCDF dataset classifies China’s development finance flows using the “ODA-like” and “OOF-like” categories that strictly follow the OECD’s definitions. These categories allow us to determine the share of aid and debt in China’s development finance portfolio in Southeast Asia. Figure 1 provides a simple visualisation of the ratio of debt-to-aid in China’s development finance flows to Southeast Asia for projects approved between 2000 and 2017 – the period captured by the GCDF dataset.
Within this time period, the overall ratio of aid-to-debt is 1:5, meaning that for every dollar of aid that flowed to Southeast Asia, 5 dollars of debt accompanied it. From these figures, it is evident that China tends to provide loans more than aid in its development programme in Southeast Asia. However, it should also be noted that the 1:5 aid-to-debt ratio in Southeast Asia is lower than the same ratio for China’s global development finance portfolio, which stands at 1:9. While it may be tempting to use these headline figures to support the “debt trap” narrative, further examination of the terms of China development flows is needed. The next section explores in detail three key indicators – interest rates, grace periods and maturity periods – to see how China’s development lending compares to those if international lenders such as the World Bank. For reasons laid out below, the World Bank’s lending programmes serve as an excellent benchmark for assessing the favourability of China’s lending terms.
THE LENDING TERMS OF CHINA’S DEVELOPMENT FINANCE IN SOUTHEAST ASIA
Researchers have previously identified the World Bank’s dual-track development funding windows, which disburses both aid-financed (ODA) and debt-financed (OOF) loans, to be a benchmark to which China’s development loans can be compared. Aid-financed loans are given out under the World Bank’s International Development Association (IDA). Eligibility for IDA loans depends on the country’s relative poverty, which is defined as gross national income below a threshold that is updated yearly. This figure stands at US$1,255 per capita for the 2023 fiscal year. The bulk of the IDA’s funds comes from middle to higher-income donor countries, which meet every three years to replenish the IDA’s resources. This source of funding allows the IDA to price its aid-financed loans on highly concessional terms. On the other hand, the World Bank’s International Bank for Reconstruction and Development (IBRD) provides debt-financed loans to middle-income countries that do not qualify for IDA funding. The IBRD raises its funds through the global financial market, resulting in the terms being more in line with market rates.
The World Bank’s bifurcated structure which separates aid-financed loans and debt-financed loans closely resembles China’s development finance programme, which similarly disburses loans on both aid and debt terms to developing countries. Hence, a like-to-like comparison between the terms of the World Bank’s IDA and IBRD loans can be made with China’s aid-financed and debt-financed loans.
To facilitate this comparison, the first step to take has been to separately calculate the average lending terms of China’s aid-financed and debt-financed loans to Southeast Asia, which are measured along three indicators – interest rates, grace periods, and maturity periods. The average lending terms of the World Bank’s IDA and IBRD loans are derived from a previous study. As this study utilises weighted averages that take into account the value of loans, the calculations for China’s loans in this article will similarly use averages that are weighted by loan value. As data for lending terms in the GCDF dataset are subject to availability, not all projects listed in the dataset are represented in this calculation. Data from 47 aid-financed projects (ODA) and 169 debt-financed (projects) in Southeast Asia are utilised in the calculation of China’s lending terms. Figure 2 provides a summary of the comparison between the lending terms of the World Bank’s IDA and IBRD loans and China’s aid-financed and debt-financed loans.
The World Bank’s lending terms are consistent with the idea that IDA-eligible low-income countries should receive more concessional loans while middle-income countries, who are better able to pay loans, receive harder terms. Progression from IDA to IBRD loans marks a transition to higher interest rates as well as shorter grace and maturity periods. This can be seen in the 1.3% and 2.8% weighted average interest rates for IDA and IBRD loans, respectively. China’s loans to Southeast Asia are somewhat consistent with this logic, where aid-financed loans have marginally lower average interest rates of 3.2% while debt-financed loans are higher at 3.6%.
… both China’s aid-financed and debt-financed loans are more expensive than the World Bank’s. On average, aid-financed loans from China have interest rates that are 1.5% higher, while debt-financed loans have interest rates that are 0.8% higher
Based on the average grace and maturity period of China’s loans, there does not seem to be a consistent logic to which these terms are set. Counterintuitively, the weighted average grace periods for China’s aid-financed loans to Southeast Asia have been shorter than their debt-financed ones, at 4.8 years compared to 6 years. As for maturity periods, there is not much differentiation between China’s aid-financed and debt-financed loans, which have average maturity periods of 17.6 years and 17.8 years respectively.
Based on the interest rate figures, we can see that both China’s aid-financed and debt-financed loans are more expensive than the World Bank’s. On average, aid-financed loans from China have interest rates that are 1.5% higher, while debt-financed loans have interest rates that are 0.8% higher. However, for some Southeast Asian countries, this disparity seems to be larger. Table 1 provides the country-level weighted average terms for China’s debt-financed loans. To facilitate comparison with the World Bank’s rates, IBRD terms have also been included in the table.
Debt-financed loans given to Myanmar, Indonesia and Vietnam seem to have average interest rates that are substantially higher (above 4.0%). On the other hand, Malaysia, the Philippines, Cambodia and Laos seem to secure lower interest rates for debt-financed loans (3.1% to 3.3% on average). While there are suggestions that certain Southeast Asian countries have been able to secure better deals with China through tough negotiations, it remains that these rates are across the board higher than the IBRD’s 2.8% benchmark.
A similar picture can be gleaned from the terms for China’s aid-financed loans, which are provided in Table 2. For comparison, the IDA terms are included in the table. The interest rates on China’s aid-financed loans in these Southeast Asian countries are also higher than what the World Bank offers. That being said, as there is much less data available for China’s aid-financed loans, more caution should be exercised when interpreting these country-level figures.
CONTEXTUALISING THE TERMS OF CHINA’S DEVELOPMENT LOANS IN SOUTHEAST ASIA
The figures presented in the earlier sections might lend credence to the idea that China is saddling Southeast Asian countries with debt on unfavourable terms, possibly prompting a “debt trap” narrative in the region. However, this is an oversimplification of the issue as these figures need to be interpreted within the historical context of China’s overseas development finance. There are several reasons indicating that China’s development loans to Southeast Asia, expensive as they may be, are unlikely to be part of a “debt trap” strategy.
What Chinese lenders see as essentially a risk management strategy is chiefly where fears of a Chinese “debt trap” strategy lie. This is especially so if the development loans are concentrated primarily in sectors of strategic importance.
As analysts point out, when China began its overseas development finance programme through its “Going Out” strategy in the late 1990s, its primary concern was its massive foreign currency reserves. The Chinese government was concerned that channelling these reserves into its domestic economy may increase the macro-economic instability in the form of inflation or currency revaluation. Hence, it quickly accelerated the pace of its overseas investments and lending. Locked mostly in U.S. Treasury bonds, China’s foreign currency reserves generated an estimated 3% in annual returns. At the same time, there was an increasing consensus within China that its foreign reserves could be better invested in higher-yielding assets.
One of the vehicles for China’s overseas investment was through its development finance portfolio which predominantly comprises debt rather than aid and is typically priced above the reference 3% rate in Southeast Asia. As an additional point of reference, the domestic rate of return of infrastructure in China stood at 3.1% in 2012.
These higher interest loans carry a heavier risk of default that Chinese lenders manage through the utilisation of repayment safeguards. In a report by the creators of the GCDF dataset, they found that collateralisation is the risk mitigation strategy of choice for Chinese lenders. In the context of a typical Chinese development loan, this means that borrowers have pledged some form of asset, future revenue or commodity export which can be seized in the event of a default. What Chinese lenders see as essentially a risk management strategy is chiefly where fears of a Chinese “debt trap” strategy lie. This is especially so if the development loans are concentrated primarily in sectors of strategic importance.
In Southeast Asia, the GCDF dataset recorded 54 instances of collateralised development loans totalling US$23.9 billion (constant 2017). Table 3 provides a country-level breakdown of these loans. Approximately 95% of these collateralised loans went into sectors of strategic importance – energy, mining, transport and communications. However, it should be noted that this sample of 54 projects is based on what the GCDF authors could collect based on publicly available information. For the vast majority of projects, details on collateralisation are not available. That being said, even if we expect that many more Chinese development loans in the region are likely to be collateralised, it is still a stretch to claim that a “debt trap” strategy is in play.
Dr Bradley Parks, Director of the AidData research lab, which created the GCDF dataset, said in an interview that China heavily prefers to collateralise on cash or liquid assets as well as commodity exports. We see this in multiple examples in the sample of 54 collateralised projects. The most notable case is a collateralised debt-financed loan for the China-Laos Railway Project, valued at US$3.63 billion (constant 2017). The Laotian government did not pledge any assets or provide a sovereign guarantee for the loan but provided instead collateral in the form of revenue from a bauxite mine and three potash mines. In another example, a US$724 million (constant 2017) debt-financed loan for the Myanmar-China Gas Pipeline Project was collateralised using future revenues from the project.
In the instances where illiquid physical infrastructure has been pledged as collateral, it is worth remembering that these development loans are meant to generate a financial return for China’s lenders. If seizure of assets is the purported goal of “debt trap”, it would mean that the loans of Chinese lenders ultimately became non-performing. It makes little sense, if any, for any financial institution to hope that their loans fail. Previous research has also indicated China’s strong preference for restructuring debt before the possibility of defaults. China’s approach to debt restructuring has been dominated by debt maturity extensions with outright debt cancellations being limited to the very few zero-interest loans. It is thus likely that the seizure of pledged collateral would only be a final resort for Chinese financial institutions dealing with non-performing loans.
China’s development finance portfolio in Southeast Asia heavily relies on debt-financing with interest rates that are high compared to the World Bank. There are also questions surrounding the debt sustainability of Southeast Asian countries that have taken such loans from China, as well as doubts about the due diligence processes in which Chinese lenders finance “high-risk” projects that overwhelmingly require collateralisation as a risk management strategy. However, the evidence, on the balance, does not lend credence to a purposive Chinese “debt trap” where China sacrifices its genuine commercial interests in the hope of securing assets of strategic importance, and purposefully saddles the host countries with unsustainable debts.
The data explored in this article present a more nuanced picture in which China assumes a role akin to that of a banker who has devised sure ways to guarantee repayment on its loans, which allows it to take on big-scale, high-risk projects. The seizure of strategic assets to repay Chinese loans is more an outgrowth of this process than a premeditated geopolitical goal. China’s development finance is largely driven by commercial interests although its risk management against defaults may entail strategic implications. These strategic implications, which are amplified by the state-led nature of China’s overseas development finance, have underpinned the anxiety behind the “debt-trap” narrative. However, the complex reality neither fits Western conceptions of China’s “debt trap” strategy nor Beijing’s self-image as a benevolent development partner.
This is an adapted version of ISEAS Perspective 2022/88 published on 8 September 2022. The paper and its references can be accessed at this link.
Darren Cheong was Research Associate with the Regional Strategic and Political Studies Programme, ISEAS - Yusof Ishak Institute.