Year of the Slaughtered Black Sheep

Short-term update: so far cited in both Abnormal Returns, and Seeking Alpha.

It is a natural question to ask, given the massive drop in China’s market recently.  If it could happen there, could it now happen anywhere?

Hedge fund titan Stanley Druckenmiller before the summer was similar to many other fund managers saying that he was "very intrigued" by China.  Very intrigued is surely one way to put it.  In a frightening, short-selling criminalized collapse, enduring more than a month from the June 12 peak, the Shanghai Stock Exchange (SSE) evaporated nearly 1/3 of its value and there was a rare and sudden forced currency devaluation.   Mr. Druckenmiller's amour with China lately?  Not so much.  Instead memories of synchronized, and more intense market crashes in 2008 -all over the world- now come charging to the fore.  In this article we point to the statistical similarities and differences between now, and that global financial crisis (GFC).  Leading TARP's analytics and adding insights to Paul Volcker’s proposals into Dodd Frank, provides an interesting perspective on how economics combined with probability can add value to the current China story.

We start with the idea that no one could predict when these crashes occur.  People are great at observing them and their losses with awe as they unfold, whenever they do (here, here).  Poor causal relationships always enter those retrospective conversations.  But a broader mathematical viewpoint shows that there is no significant relationship between the volatility in the Chinese stock market, and say the volatility in the largest domestic market (i.e., the American stock market).  The same has been true in the past for other large markets, such as for Japan.

One way to monitor this is to track the outer decile market volatility, occurring for the SSE and the largest 500 U.S. stocks (S&P).  We can then evaluate if they match up or not.  In the illustration below (in red), we see the number of high risk (beyond +2.9%) days per year for the SSE.  Geometrically computed to keep things symmetrical.  Things indeed have picked up this year.  While mathematically appropriate to see it as we show below, this annualized risk is at a 7-year high.

Now (in blue) we add on the same 10^th decile statistic days for the S&P (though now respectively using a daily threshold of beyond +1.7%).  Using the c² statistic we gather that there is less than a 1% probability that these two unmatched series are distinct over this time, by chance alone.  So with the exception of 2008 therefore, there is no confluence of risk across these countries.  Sure the U.S. may gyrate about a little for many rationales, but so far in 2015 it's far different from 2008 or what China is seeing in 2015.  We've noted in the past that the speed of the outburst has been more remarkable then the magnitude.

Looking below the line, at the 1^st decile statistics (within +0.1% on either index), we get the exact same result.  Here we mellow the colors versus those used above the line.  And if we add each exchanges’ 1^st and 10^th decile counts per year, then those two exchanges’ series also have the same less than 1% probability result.

Prominent, though smaller, aftershocks often occur when the economic circumstances seem similar to a recent risk event.  Let’s see the recent, and certainly worrying, trends in China that can help explain why this 2^nd largest economy is by itself suffering a market correction.  The first is that its economy has been shakier than expected in this recovery, versus prior to the GFC.  Contrast that with the current recoveries occurring both in the U.S., as well as around the world.

Another consideration is that the issues at the core of the GFC stemmed from leveraging up and then unraveling of the financial services sector (which has since improved thanks to government regulation post the crisis).  In China, the banks coming into this year have seen their critical metrics appear at risk, and in the face of the weak manufacturing confidence (PMI) measures.  Loan loss provisions are too low and not picking up quickly enough.  And return on equity has been good but sliding.  The fear is always great of being the fool again stuck with a bad overvalued investment.  Here are some examples for the past few years, with the large China banks:

Financial debt (a growing portion tied to unsold physical inventories) is also making up a larger component of the large debt to GDP ratio in China (pages 4, 75, 80 of this report).  The financial sector took on more than 1/3 weight of the MSCI All China Index (developed markets saw a 1/5 weight at the start of the GFC).  So again the situation in China versus elsewhere is different.  And we’ve been here before, with other large booming countries experiencing a financial tremor.  But probability analysis shows that there is not a risk correlation conflux across time.  What’s happening in China isn’t predetermined to export by spreading across to other larger shores anytime soon.  This type of analysis also proves the value of considering statistics in different ways.  And what we see in China’s economy (and their domestic stock market) is perilous but, as is typically in these situations, somewhat of a black sheep.