The Credit Impulse Explained
The objectives of this research paper
In late July we wrote a research paper projecting more difficult times ahead for the U.S. as well as the euro zone economy (see The Credit Impulse). Our prediction was based on a leading indicator called the Credit Impulse. At the time it raised a few eyebrows, because there were no obvious signs of an economic slowdown.
Admittedly, it is a relatively new and complex leading indicator that many are not yet entirely comfortable with. We continue to receive many inquiries as to what the credit impulse really is, and how the data should be interpreted. This paper will address those issues.
Adding to that, in recent days, both the IMF and the ECB have warned that we will likely see a slowdown over the autumn and winter. Those warnings have let us to re-visit our prediction from late July.
What exactly is the credit impulse?
Let’s begin by defining the credit impulse again. The credit impulse is a term coined by Michael Biggs, then an economist at Deutsche Bank, when in 2008 he defined it as the change in new credit issued as a % of GDP. He showed that, in most countries, private sector demand (C+I) correlates very closely with the credit impulse and he argued that the important credit variable in terms of forecasting GDP growth is the change in the flow of credit, not the change in the stock of credit. When you look at chart 1, it is hard to disagree with him. It is probably the most precise leading indicator at investors’ disposal today as far as predicting GDP growth is concerned.
What are the pitfalls?
The change in the flow of credit, rather than the change in the stock, is not the only distinction readers of the credit impulse should be aware of. There are other pitfalls as well.
The most important one is probably periodicity. Many credit impulse charts show the change in credit year-on-year, and we have also come across charts based on month-on-month data; however, history provides very compelling evidence on this point. The 6-month credit impulse has by far the best track record in terms of being a leading indicator of GDP growth. The logic is relatively simple. It obviously takes longer than one month for a newly issued bank loan to show up in economic activity and, in most cases, it takes less than a full year.
The other major issue is what types of credit to include in the credit impulse and, in that respect, the good news is that bank credits provide a very good proxy. One could argue that shadow banking should be included but, at least as far as developed markets are concerned, shadow banking doesn’t account for a meaningful percentage of total lending, so it is a theoretical more than a practical issue.
Of the major economies, only China has a meaningful degree of shadow banking, which is why one would need to be more careful interpreting Chinese credit impulse numbers. Having said that, there is a (part) fix. Most credit impulse stats are based on official lending figures published at regular intervals by the banking industry. If one instead uses money supply stats, one will catch shadow banking as well. That creates one or two other problems but, as a proxy, it provides a fairly accurate picture.
Some would also want to include the corporate bond market but, as bond issuance doesn’t affect money supply numbers, and it is the impact on money supply that affects economic activity, we don’t see any problems ignoring bond issuance.
The current reading
All in all, the picture is relatively bleak but not dramatically so. Many investors were very encouraged by the healthy rebound in total euro zone lending when the ECB published the numbers for the month of July (see the extreme right of chart 2).
However, when looking at recent trends in the 6-month credit impulse for the euro zone, it is obvious that the rebound was not as convincing as what first met the eye. The arrow is still pointing towards a slowing of economic activity (chart 3).
The U.S. picture is not dramatically different. Over there, and based on recent readings of the 6-month credit impulse, we would continue to expect some slowdown over the next several months (chart 4).
In China, the credit impulse based on official bank lending numbers paints a very bearish but, if one includes shadow banking by using money supply stats, the picture is altogether different, and much more positive (the chart has not been included in this paper). We therefore maintain our view that China is not likely to face a hard landing anytime soon.
It is still premature to say anything more exact about the timing or magnitude of the slowdown, but a (relatively) mild cyclical recession cannot be ruled out at this stage.
How will financial markets likely be affected? It obviously depends on the severity of the slowdown, but it also depends on expectations. In that respect the U.S. is much worse off the Europe. 2016 earnings expectation for S&P 500 companies are still uncomfortably high, and a slowdown will almost certainly lead to earnings disappointments.
Finally, it will depend on how the Fed reacts to all of this. We are increasingly convinced (but obviously not certain) that they will sit on their hands when they meet in about ten days. Should they do so, we don’t expect long bonds to be affected in a major way, as the consensus has moved in the direction of no Fed action in September; however, should they also choose to do nothing in December, long bonds will benefit.
Niels C. Jensen
4 September 2015