On the Chinese rebalancing timetable: a reply to professor Pettis

 

Posted by and on

Professor Michael Pettis has recently published another brilliant post on the Chinese rebalancing issue. We regularly follow Pettis’ views on the Chinese economy, which we consider valuable if one wants to understand the macro management dilemmas China will have to face. In his post, Pettis does not address how the rebalancing process should be ideally done (something which he has explained on several occasions), but rather how much time China has in order to accomplish it successfully. Pettis says that:

Credit growth in China is too high as are current debt levels, and the sooner Beijing gets credit growth under control, the better. This latter statement in itself is not controversial of course, but my simple debt model shows just how urgent it is for Beijing to get credit growth under control. It clearly does not have ten years or even seven years. It might have five years, but only if the markets – Chinese investors, businesses, and savers, both wealthy and middle class – are convinced that it is moving in the right direction.

In other words, the current high levels of debt can derail the rebalancing process if it is not done quick enough. In order to understand the link between debt levels and GDP, Pettis proposes a simple model that captures their dynamics over the long run (up to 2023). In his baseline scenario, Pettis assumes (following the trend of the last few years) that nominal debt initially grows twice as fast as nominal GDP (notice the use of nominal values), but gradually converging in a linear way to the growth rate of GDP by the end of 2023 – at that moment the economy reaches a steady-state position, and GDP and debt grow at the same pace. Depending on the GDP growth rate assumed, Pettis’ model projects debt-to-GDP ratios from 251% (with a 3% growth in GDP) to 274% (with a 6% growth) by the end of the period, too high in comparison to other economies. He then proposes alternative scenarios, but the result is the same: unless Beijing advances more radical measures to curb debt growth and improve the efficiency of the financial system, the growth in debt will derail the rebalancing process.

And because China does not have too much time, Pettis concludes that:

In every one of its economic policymaking choices, Beijing must ultimately choose between higher debt, higher unemployment, or higher transfers of wealth from the state sector to the household sector. Every single policy results in some combination of the three. The time frame within which this must be resolved is set by debt capacity limits, and as my model shows, Beijing probably has no more than five years, perhaps much less, within which to resolve the rebalancing if it wants to avoid a disruptive rebalancing.

Although Pettis’ model captures vividly the long-run dynamics between debt and growth, we think that his model suffers from one major drawback: his model simply tracks the ratio of total debt to GDP. However, nothing is said about how this debt is spread across sectors. In particular, it is obvious that how much debt each sector holds should play a central role in any analysis of the rebalancing process, because if China has to rebalance it can only be, by definition, through huge economic transfers from the corporate sector to the household sector. But given the already high amount of debt in the balance-sheets of corporations, these transfers will create further strain in firms’ financial statements. Therefore, although the debt-to-GDP ratio may be important from a long-run perspective, we believe that it is the corporate debt-to-cash flow ratio the one that is crucial for the rebalancing process. In our story, corporations are the weakest link in the Chinese rebalancing.

On the other hand, although Beijing could engineer (political considerations aside) a process to transfer resources to the household sector as a way to speed up the rebalancing as Pettis argues, we believe that, in the end, if China has still any option for a smooth rebalancing (which we highly doubt), it will have to come about through high levels of inflation (probably higher than 4-5%), eroding thus the current pile of debt. In our model corporations have an easier life during the rebalancing with high nominal GDP growth rates, and because such rates will unlikely be unattainable with high real GDP growth rates, high inflation is the only way to square the circle.

In the rest of the post we will go through our simple model with corporate debt, discussing carefully at the same time the assumptions of the model.

The rebalancing timetable for corporate debt: a baseline scenario

Instead of assuming that debt will grow at a certain pace in relation to GDP as Pettis does, we rather start with assumptions about the behaviour of the main GDP components if any meaningful rebalancing has to be achieved at the end of the forecasting period (which in our case is, for convenience, 2025). In other words, we simply project the main GDP components using average growth rates that ensure that China reaches a value of around 70% for the share of consumption (private plus public) to GDP and a value of around 30% of investment (private plus public) to GDP. As we explained in a previous post, we regard these values as the minimum acceptable threshold for the completion of the rebalancing process. Although the evidence for developed and developing economies suggests lower investment-GDP shares than our target share for China, we prefer to be conservative in our exercise and to ask for a less demanding rebalancing target.

Because we want to focus on corporate debt and not on the debt of the economy as a whole, we need numbers for the debt in the corporate sector and not total social financing (TSF). For convenience, we start with a corporate debt level (as a share of GDP) of 180%, which includes both non-financial and financial corporations. This number has been taken from a McKinsey report. In any case, there is a huge uncertainty about the real value of such a ratio; for instance, data from the BIS shows that the ‘total credit to non-financial corporations’ reached 163.1% in the second quarter of 2015, which would imply higher levels for the corporate sector (with financials now) than those given by McKinsey. Again, we want to be conservative and that’s why we choose the lower end of the estimates.

In Exhibit 1 we present our baseline scenario for the rebalancing timetable (for interested readers, the model is available upon request in a friendly spreadsheet). For convenience, we have normalised the values setting GDP to be 100 in 2015. The shares of the various flow components for 2015 have been taken from the most recent Chinese statistics when possible, but most of them come from the 2012 version of the flow-of-funds accounts – alas, the most recent one. We have projected the Levy-Kalecki profit equation for the Chinese corporate sector to gauge the amount of cash Chinese corporations will have at their disposal during our hypothetical rebalancing process (see here for details on the LK equation). Because in our Levy-Kalecki profit equation we are using gross corporate investment (instead of net investment, as it should be if one is interested in net profits), we are effectively getting a cash measure, the relevant measure for debt repayment issues. In our model, we assume that corporations will only issue debt for covering the gap between outlays for investment projects and internal resources. We know, again, that this is a conservative assumption (a quick glance at the flow-of-funds tells us that corporations have employed their resources for other uses), but will be enough for our purposes – i.e. the debt numbers will be very high even with conservative assumptions. Beyond that, we have implicitly assumed that dividend payments will be matched by the issue of new shares (and that’s why the evolution of debt can be expressed as the difference between investment and internal resources). All in all, our baseline scenario is as follows:

baseline

We have assumed that nominal GDP grows at 5% for the rest of the decade (alternatively, we could have assumed some sort of linear trend, as Pettis does, but we have preferred to keep things simple). It should be stressed again that nominal GDP is used instead of real GDP, as debt -our subject of interest- is a nominal variable. We use the same rate for the growth of the main government expenditure components, because we want to project no change in the government financial position relative to GDP. On the other hand, the trade balance (net exports) simply fills the gap between the imposed GDP growth rate and the consumption and investment growth rates needed to rebalance the economy. Because we have assumed that private investment will not grow in nominal terms, it logically implies that private consumption will grow at 10% on average for the rest of the decade. Finally, we have assumed that wages grows at 7%, higher than GDP but lower than private consumption: this implies that households will be reducing their savings rate, which will be a boost for corporate profits.

We should remark that the use of a 5% nominal GDP growth rate (and the same applies for the rest of our assumptions) in our baseline scenario does not imply by any means that we believe this is the most likely rate at which China will grow over the next decade. It rather should be interpreted as a counterfactual exercise that tries to ask: what conditions have to be met if China has to grow at 5%? The question then allows us to assess whether these conditions are realistic or not.

The baseline scenario shows crystal clear that the rebalancing in the GDP expenditure components necessarily implies a shift in income distribution: Levy-Kalecki profits fall from 21.5% to 10.7% over the simulation. The reduction in the cash-flow generation (both in absolute and in relative terms), coupled with constant investment needs, increase firms’ debt burden. However, by the end of the simulation, if you look at the corporate debt-to-GDP ratio, you might conclude that firms’ financial situation is not so bad: they have been able to go through the rebalancing process with even a reduction in that ratio (from 180% to 176.8%). But that would be misleading: firms have to service their debt (and interest payments) with actual cash-flows, not with GDP. For that reason, we have created an additional ratio (debt to cash-flows) that captures that. The ratio is already very high for the ratios observed at the firm level in developed economies, 8.4x, but it soars to 16.5x by the end of the simulation. It is clear that Chinese corporations will not have to reach that point, because their dire financial situation will become unsustainable much earlier, but this is what the rebalancing process would logically imply in the end for the Chinese corporate sector.

Other scenarios

We can, of course, run the model as much as we want with different assumptions. We will precisely do that in order to consider the effects of different GDP growth rates on the delicate balance between the rebalancing process, its timing, and the evolution of corporate debt.

The first alternative scenario assumes that the nominal GDP growth rate will be around 3% on average over the next ten years. If one thinks that Chinese real GDP growth rates will not be higher than 2% and that in addition Beijing will have to live in a near zero-inflation environment, then the 3% assumption is not as crazy as it seems. Under these conditions, an (again) if the rebalancing has to take place, we have assumed private consumption growth rates of 7.5% and private investment growth rates of roughly -2%. In a low growth environment we find difficult to swallow the assumption of consumption growth rates higher than 7.5%, so private investment in this environment will have to grow at negative rates if it has to be less than 30% of GDP by 2025.

Scenario 1

The main takeaway from this scenario is that the corporate sector will find much more difficult to handle high levels of debt. Even with our conservative assumptions on how the debt is generated, corporations would add another 25% debt points, whereas the Levy-Kalecki multiple would reach levels close to 20x (by the way, the fact that Levy-Kalecki cash-flows reach 10.5% as in the baseline scenario is not a coincidence, but rather it shows the nature of the rebalancing process).

Finally, we have crunched some more numbers in order to produce a scenario with GDP nominal growth rates of 7%. Make no mistake, we think that the only feasible way China will attain that growth rate is with high levels of inflation, and not through high real GDP growth rates – as many international analysts still seem to imply. But even in this scenario, corporations would still suffer the rebalancing process, and some additional measures by Beijing in form of transfers to the private sector would be needed – as Pettis suggests. These transfers would show up in our scenario as a positive source of corporate profits. In this case, a back-of-the-envelope calculation shows that in order to keep constant firms’ leverage at 8.4x, the government balance should move from zero (the assumption we chose) to roughly a 2.5% deficit (and increasing as the time goes by). But because the current debt levels are already too high, additional government intervention will be needed if the corporate sector has to return to more normal leverage ratios by the end of 2025.

Scenario 2

In order to compare the different results, we have drawn in a couple of graphs the paths of both leverage measures under the different scenarios. As we have remarked, all of them show the difficulties the corporate sector will have to face during the rebalancing process:

Chinese corporate debt-to-GDP under different scenarios

Chinese corporate debt-to-LK under different scenarios

Summing up

If you have been so kind to put up with us so far, then the conclusions are pretty clear. Otherwise, we repeat them here for convenience (with some comments):

  • The Chinese economy will rebalance (with private consumption rising from 36% of GDP to roughly 60%) but the question is how. We believe that if not handled properly, current corporate debt levels are high enough to preclude a smooth rebalancing process. This process implies by definition a transfer of economic resources from corporations (and government) to households, but given a high corporate leverage these transfers will put additional pressure on firms’ balance sheets.
  • The rebalancing of the Chinese economy – in our opinion – will not proceed smoothly. Our scenarios show serious difficulties with the service of corporate debt – and the related problems for the banking sector. As investment is curtailed the cash flow generation to service debt diminishes – and, moreover, the incidence of the lack of investment will mostly fall on the commodity and capital goods industry.
  • A serious financial crisis may emerge as loans are not repaid, credit is curtailed and investment further reduced. This would rebalance the economy faster, but at the cost of rising unemployment – something that Beijing cannot tolerate for long.
  • Inflation is badly needed if the rebalancing process has to be done in a ‘painless’ way. Even in this scenario, substantial government transfers to the private sector will be required if corporate leverage has to be kept under control. Most likely, Chinese government will have to take on new debt in excess of 60% of GDP.
  • Beijing may continue to postpone the rebalancing with additional and preplanned doses of investment; it is doing that now intermitently, but eventually it will have to give up.

Leave a Reply

Your email address will not be published. Required fields are marked *