China's high and rising corporate debt

China’s corporate debt to GDP ratio, a measure of corporate leverage, is now among the very highest globally. It has risen nearly 65 percentage points within a decade, the fastest increase among the major economies.

A decade after the 2008 crisis, the total debt of the non-financial sector (government, corporate and household debt) worldwide has surged to a new high of nearly 242 percent as a share of GDP in 2017. In China, this ratio has almost doubled to the same level, making the country the most indebted emerging economy.

In contrast to many standard narratives, our research identified four driving forces behind China’s corporate debt dynamics. Taken together, they suggest that declining investment efficiency may have been one contributing factor to the marked rise in Chinese corporate leverage during the decade following the 2008 global financial crisis:

1. A lower share of corporate capital investment financed by internally generated funds of the sector contributes significantly and consistently to faster build-up of corporate debt, for both China and elsewhere. China’s share of internally funded capital expenditure (capex) is the lowest of all 41 economies in our data sample – possibly due to poor corporate earnings combined with aggressive capital investment.
2. Lower levels of government debt tend to lift corporate leverage as measured by corporate debt to GDP ratio. China has a light government debt load and a heavy corporate debt burden, so its corporate leverage is high. This is partially due to Chinese local governments financing some of their hastily chosen and questionable investment projects in the wake of the 2008 global financial crisis with corporate debt rather than through conventional fiscal borrowing.
3. The effects of these two driving factors behind corporate leverage (i.e. internally funded capital expenditure share and government debt) are equally persistent elsewhere but appear stronger in China’s case. This may be due to the bigger presence of so-called zombie state enterprises and government-designated investment projects. Neither bodes well for investment efficiency.
4. Finally, China’s investment rate is the highest among all the economies we examine, and a higher investment rate is found to raise corporate leverage significantly. Possibly, declining efficiency requires more investment to achieve a given economic growth, in turn contributing to China’s high corporate leverage.

This China Monitor summarizes some key findings of our forthcoming academic paper “The Role of Internally Financed Capex in Rising Chinese Corporate Debts”.¹ Our research was prompted by concerns about China’s high and rising corporate debt and questions over whether the domestic invest- ments funded were worthwhile. We will also shed light on the China’s latest deleveraging campaign and assess its sustainability.

We use a broad international dataset of 41 countries to examine the roles of two determinants of corporate leverage – the share of internally funded corporate capital expenditure (capex) and government debt level. Our findings challenge and filter out over-reliance on heavily China-centric explanations that automatically emphasize special “Chinese characteristics” of the country’s financial market and government policy.

A major force driving rising overall indebtedness, for the world and for China in particular, has been the steep increase in corporate debt. China’s corporate debt to GDP ratio, a measure of corporate leverage, is now among the very highest globally. It has risen nearly 65 percentage points within a decade, the fastest increase among the major economies (see exhibit 2).

1.1. Most narratives to explain high levels of corporate leverage are useful but flawed

Many narratives have been proposed to explain why Chinese corporate leverage is high. One explanation, put forward by the former Chinese central bank Governor Zhou Xiaochuan (2002 – 2018), is the exceptionally high domestic saving rate.³ The idea is that a high level of savings means there is an abundance of what economists call “loanable funds”, which one way or another need to be lent to companies. One problem of this narrative is that high savings also means strong cash flows and hence less need to borrow. In theory, therefore, high savings can lead to higher or lower corporate debt. Another problem is that China’s savings rate had already declined from a peak of 50 percent of GDP in 2006 to around 45 percent by 2016, precisely when Chinese corporate leverage climbed rapidly.

Another narrative centers on zombie state firms, companies that need bailout in order to operate. According to this story, many Chinese zombie state companies could not service their debt but have been propped up by state bank loans. Hence bad corporate debts borrowed and non-performing bank loans lent both keep on piling up, leading to high and rising corporate leverage. However, the importance of state banks and state enterprises within China’s overall trended lower over the period 2000 to 2016, when its corporate leverage surged. By contrast, there was notable Chinese corporate deleveraging in the decade before the 2008 crisis, at a time when state companies held a much bigger slice of the economy.⁴

Moreover, it is important to remember that some of China’s top private property developers are also among the biggest Chinese corporate bond issuers nowadays. Thus, the Chinese private sector was also a major borrower over this period as well, and the zombie story can only explain so much.

There are many other stories around to explain China’s high corporate leverage. They typically share some common weaknesses.

• First, these stories are often China-centric, assuming that the high Chinese corporate leverage is a result of some supposedly unique or special “Chinese characteristics”.
• Second, most of these narratives simply assume that financial considerations do not matter at all in Chinese corporate borrowing decisions.
• Third, these theories typically do not subject themselves to “horse-race tests”, whereby the influences of different factors on corporate leverage are checked using the same criteria to ascertain how relevant they are.

As a consequence, these explanations may miss, or at least under-appreciate, some of the leading drivers behind china’s corporate debt dynamics.

1.2. Two underappreciated factors stand behind corporate leverage

Our approach differs from most other explanations. We highlight two often-ignored determinants or explanatory variables of corporate borrowing, to factor in both financial conditions and government policy.

One is the “internally funded capex share”, which captures the effect of financial conditions when taking on or retiring corporate debt. The other is the level of government debt, which partially captures the influence of politics and policy. For both these two factors, we do allow for some special “Chinese characteristics” effects but reject the notion that only Chinese corporations are inherently debt hungry, while non-Chinese firms are naturally thrifty.

Let’s first start with financial considerations and incentives facing firms. We argue that corporations, Chinese or otherwise, take on new debt or pay off existing debt in response to the conditions they face, such as interest rate, earnings, cash flows and capex needs. At least, this possibility should not be simply swept under the carpet.

What might these financial considerations be when firms decide to raise or retire debt? One obvious candidate, as suggested by well-known corporate finance theories, is the “ratio of internally generated funds over capex” or equivalently “the share of internally funded (financed) capex”. Internally generated funds mostly consist of retained earnings and depreciation. Intuitively, a higher share of internally funded capex suggests stronger cash flows relative to capex needs and hence a less need for new borrowing, and vice versa. This is one central hypothesis to be tested in our paper: the ratio of internally generated funds over capital expenditure negatively affects corporate leverage.

This negative effect can be more pronounced in China’s case. As corporate earnings slow just when government-driven “corporate” investment accelerates, the ratio of internally generated funds to capex of course declines more, necessitating even greater “corporate” borrowing to fund capex (see exhibit 4). Also, as earnings weaken, even more zombie state firms may struggle and need to be bailed by additional loans from state banks. Or both.

Next, we turn to our second proposed determinant of corporate borrowing: government debt. As noted earlier, China is burdened with a heavy corporate debt load but has manageable government debt obligations (see exhibit 3). This interesting profile prompts the question of why the Chinese government is prudent while Chinese corporations are aggressive in leveraging up. To explore this question, it is useful to consider two sets of forces, “shadow borrowing” and “shadow banking”, that may have influenced the dynamics between China’s government and corporate debts.

1.3. Shadow borrowing explains China’s mix of high corporate leverage and modest government debt

“Shadow borrowing” refers to debt obligations issued by corporate entities linked to Chinese local governments. These -linked, -sponsored or -owned legal corporate entities are collectively called “local government financing or funding vehicles” (LGFVs). Before 2015, the Chinese local governments faced the strict legal prohibition of borrowing, as specified by the 1994 Budget Law, but also had powerful incentives to circumvent it via these LGFVs, since local government officials got promoted faster by pursuing higher local GDP growth. In a knee-jerk reaction to the 2008 global financial crisis,⁵the Chinese government in 2009 launched a huge stimulus program funded in part by “shadow borrowing”.⁶Effectively, the Chinese government at that time tacitly turned a blind eye to “shadow borrowing” by the local governments.

As a result, much of such “shadow borrowing” undertaken by local governments was often recorded as corporate debt. This in part helps explain China’s interesting mix of exceptionally high corporate leverage and a relatively modest government debt load.

It raises a big question of how much of these LGFV debts should be regarded as “corporate” or “government” debt (see exhibit 3). Furthermore, their actual demarcation can evolve over time, in light of the recent sizable swaps of some LGFV debts into official local government bonds starting in 2015.

1.4. ​​​​​​​Shadow banking inflates corporate debt

At the same time, “shadow banking” also expanded. China’s financial system experienced broad-based deregulation over the decade ending 2015, becoming less tightly regulated.⁷Many LGFVs that had initially borrowed from banks during 2009 soon came under rollover and refinancing pressure. To accommodate the mess, “shadow banking” expanded considerably. Shadow banking also catered to the rise of real estate companies and less-privileged private firms as holders of corporate debt. The outcome was that China’s corporate debt might have been inflated, while official government debt data could be understated.

One implication is that smaller government borrowing relates to bigger corporate borrowing, and vice versa. This negative relationship is not China-specific but also applies to other economies. Japan is a case in point. As the Japanese corporate sector deleveraged in the 1990s and 2000s, its government stepped in to undertake fiscal borrowing, in an attempt to counter the adverse impact on the Japanese economy. Hence one observes a negative relation between ballooning government debt and shrinking corporate debt in Japan’s “lost decades”. Again, in some European economies, local banking crises hurt corporate lending, prompting fresh bank capital injections funded by larger government deficits. There too, government and corporate borrowings negatively relate to each other.

In any case, such a negative relation between government and corporate debts should not be taken as a unique “Chinese characteristic”. However, we allow for the possibility that this negative effect may be stronger in China’s case than elsewhere, as Chinese local governments may have strong incentives to tap the LGFVs.

In summary, our expectation is that the share of internally funded capex share and the government debt level will both negatively affect corporate leverage, probably more so in China’s case. These expected negative effects are two central hypotheses we test in our statistical regressions, using a broad international panel dataset that covers 41 economies. We take into consideration many other possible factors, such as the impacts of the global financial crisis, income level, growth, investment rates and saving rates. The appendix offers more details about the methodological and data issues.

In China’s case, some “shadow borrowing” was used to fund government-designated and hastily chosen investment projects as an emergency response to the 2008 global financial crisis. There are legitimate concerns about the efficiency of and expected returns on some of these projects.

Third, the above two conclusions hold simultaneously. This suggests that both determinants may have influenced corporate leverage at the same time. Even after excluding China from our international data sample, these negative effects still persist, suggesting that our narrative is not just another special “Chinese characteristics” story.

However, the effects of both determinants on corporate borrowing appear stronger in China’s case, indicating some relevance of the “Chinese characteristics” narratives. Relative to their international peers, Chinese companies borrow more in response to either weaker corporate earnings or possibly understated government borrowing.

There could be many possible causes behind the stronger Chinese responses, including government stimulus and the prevalence of zombie firms. Government investment projects funded by the LGFV debt were rushed through to stimulate China’s economy in the face of adverse shocks. Also, China has more zombie firms than the other economies covered; its corporate leverage therefore rises even faster when corporate earnings slow. Both possibilities do not bode well for investment efficiency.

Finally, a higher investment rate strongly adds to corporate debt load, while a higher saving rate actually slows corporate leverage (see exhibit 7).Hence, a high saving rate neither explains nor justifies China’s high corporate leverage, casting doubt onthe China-centric story that China’shigh corporate leverage is mainly attributable to its high saving rate, while highlighting the prominent role of the investment rate.

One implication is that a decline in investment efficiency may contribute to a higher corporate debt/GDP ratio, since sustaining a given economic growth rate would require a higher investment rate. For instance, sustained high investment could lead to diminishing marginal returns, in turn depressing the internally funded capex share.

3.1. The short term: the latest deleveraging campaign has mixed blessings

In the short term, what can we say about China’s latest “deleveraging” campaign? Although our data sample is only up to 2016, and the campaign did not start until around 2015, we can still learn a lot from our core empirical findings to shed light on this campaign. In particular, they suggest this campaign could be a mixed blessing in terms of achieving sustainable deleveraging.

The campaign’s major objectives included curbing “shadow borrowing” and “shadow banking”. To do so, it offered local governments some legal access to the domestic bond market, while tightening regulations over certain heavily indebted industries. One campaign device has been to swap some LGFV corporate debts into local government bonds. This should facilitate headline corporate deleveraging, but keep government obligations elevated (see exhibit 9).

Finally, we take care of other technical issues such as trend property (stationarity), possible two-way interactions among variables, alternative statistical approaches, etc.