Morgan Stanley Is Wrong: Goldman Warns China's Credit Impulse Collapse Will "Drag On Growth" This Year

Just over a month ago - in what seemed to be an effort to keep the dream of a global synchronous recovery narrative alive - Morgan Stanley attempted to show that the link between China's (declining) credit impulse and the global economy (which we are constantly told is ebullient) has now been severed and all is well in the world.

Their Chief Asia Economist Chetan Ahya began by confirming that "if you had been able to reliably pick the key global macro variable over 2012-16, China’s credit impulse would have been your choice" and explains why (this should be obvious to regular readers): 

The incredibly tight link between the credit impulse and China’s growth cycle, emerging markets (EMs) exports, global growth and commodity prices meant that it would have accurately predicted the direction of almost all other global macro variables that mattered, with about six months’ lead time.

He further explains the "simple" - in retrospect of course - reason behind this observation:

China’s credit impulse – or its leverage cycle – was the only game in town back then. With global aggregate demand weak as developed markets (DMs) were deleveraging and EMs were adjusting, the change in China’s credit impulse was the most significant driver of the global economy

However, in a striking claim which breaks with precedent and which, if correct suggests a historic change in the relationship between China's credit creation and its impact on global markets and economies, the Morgan Stanley economist then writes that the link between China's credit impulse and the global economy "has now been broken" and justifies his answer as follows: 

China’s tightening has not had a material impact on the growth cycle either in China or globally, even though its credit impulse began to weaken about 24 months ago. As deleveraging and adjustment headwinds recede, the recoveries in domestic demand in both DMs and EMs have emerged as additional global growth engines.

Ahya used the following chart to prove his thesis...

However, Goldman Sachs disagrees, warning that things are aboutto get a lot more problematic for that global synchronous recovery narrative as China's collapsing credit impulse wriggles its way thru the economy...

Since 2008, there have been two and half distinct cycles in China credit. We see these cycles more clearly in the chart below, which plots net credit flows from lenders to borrowers against nominal IP growth in China.

Lower net credit flows from lenders to borrowers weigh on future activity growth

We can visually see how peaks/troughs in net credit flows from lenders to borrowers tend to systematically predict growth outperformance/ underperformance over a short period of time, but then growth underperformance/ outperformance over a longer horizon. For example, the peak in net credit flows from lenders to borrowers in late 2009, early 2013 and mid-2016 was systematically followed by peaks in nominal activity 2-3 quarters after, and then – in the case of peaks in 2009 and 2013 – a trough in nominal activity growth 2.5-3 years after. In our empirical estimations, we attempt to measure this predictable lag between peaks in net credit flow from lenders to borrowers and nominal activity growth.

Credit boost to China activity peaked in mid-2017 and has declined since. With the results of our modeling exercise in hand, we can simulate the net impact of changes in the net credit flow on past and future activity growth. Armed with the stylized impact of shocks to credit flow on activity growth, we can sum up the effects from the historical path of changes in the net credit flow to generate an “aggregate growth impact of net credit flow” for every period. The results of this exercise for nominal growth are shown in the chart below.

Net credit impulse to activity growth likely to decelerate over the coming year...

This suggests that net credit flows were a substantial drag on growth in late 2014 and 2015, as new borrowings decelerated and debt service costs rose. Thereafter, as net credit flows rose in mid-2015 as new borrowings accelerated and debt service costs fell (maturity extensions and falling interest rates), this provided a meaningful boost to activity growth. The positive credit impulse on nominal activity peaked in mid-2017 , about a year after the peak in net credit flows and has declined since. Results for real activity growth are very similar.

The credit impulse should continue to push activity growth modestly weaker in our base case of a stabilization in new borrowings at current levels. What can we expect for activity growth going forward? While the pre-specified path of debt service already bakes some impacts in the cake, we evaluate some scenarios on the future path of new borrowings in the economy, to estimate the net activity impact going forward (chart below).

In our base case, where new borrowings stabilize at current levels, net credit flows would gradually decline as debt service costs rise, dragging nominal growth lower by 70bp (from +30bp above trend to -40bp below trend) and real growth lower by 20bp (from +5bp above trend to -15bp below trend) over the next 12 months. In practice, this scenario is in line with our expectation on the policy stance - i.e. that the government will continue to tighten policies in targeted areas (such as on shadow banking products, local government irregular borrowings etc) while keeping the broad monetary policy accommodative and lowering interest rates to offset some of the targeted tightening effect. The growth impacts in our base case are broadly in line with our current forecasts of modestly slower nominal and real GDP growth over the next year.

... with more negative impacts on activity should new borrowings continue to decelerate at the same pace since mid-2016. In a bearish case, new borrowings continue to decelerate at the same pace they have done since mid-2016, and net credit flows decline more sharply than in our base case. In practice, this case could emerge if the government overtightens on shadow banking products or local governments’ irregular borrowings and does not sufficiently substitute these borrowings with other standard credit. In this base case, we would expect the drag on nominal growth to increase to 140bp and real growth to 40bp, over the next 12 months. Even in a more optimistic case in which new borrowings gradually increase to offset rising debt service costs over the next year, stabilizing net credit flows at current levels vs. GDP , real and nominal activity growth should decelerate although more modestly than our base case.

The key takeaway from this exercise is that, barring meaningful increases in new borrowings over the next year, activity growth in China is likely to decelerate, as the credit impulse to activity growth fades and debt service costs rise.