Though marking a deceleration and disappointing expectations, the data does not suggest a hard landing is imminent. Rather, it remains commensurate with a soft landing; a gradual deceleration in the growth rate, to be expected after years of enviable growth. The most likely trigger for a hard landing would be a financial crisis, and plenty of criticism over excessive debt levels has been levelled at China. But as Japan demonstrates, it is possible to have high debt and no crisis. What matters is how that debt is funded. On this basis, China’s financial system still looks sound. Lending is overwhelmingly funded by very stable domestic deposits. To avoid a future hard landing, credit growth must be curtailed, or else banks will come to rely on flighty wholesale and international funding. This requires an acceptance of lower growth, which would challenge the ruling Communist Party’s legitimacy. Yet a hard landing would be politically disastrous. The incentive to adapt is there.
Jamie Thompson, head of macro scenarios at Oxford Economics, says No.
If the Chinese government is to meet its ambitious growth targets, it will have to rely on further stimulus and rapid credit growth – a point illustrated by the substantial decline in demand growth in July. But the current pace of credit growth is clearly unsustainable. One option is to relax growth targets and gradually rein in credit expansion, accompanied by reforms. In the absence of such a strategy, however, a China hard landing becomes a much more material concern. This is reflected by the results of the latest Oxford Economics Global Risk Survey, which canvases the views of clients, including some of the world’s largest companies. The Chinese economy is the most commonly cited risk to global growth, even against a backdrop of Brexit and a potential Trump presidential victory. And with good reason. Our modelling suggests that a China hard landing would have stark implications – not least for global growth and inflation, the stance of monetary policy and key yield differentials.