Potter Stewart an Associate Justice in the US Supreme Court while ruling on an obscenity case once wrote, ‘pornography is hard to define, but "I know it when I see it”. The same logic applies to the concept of complexity. It is a very difficult concept to define and yet we know intuitively that complexity is greatest in systems where a large number of mutually interacting components are arranged in some intricate, difficult to understand pattern. Scientists also tell us that a large number of mutually interacting parts leads to another characteristic of complex systems namely that of critical points. That is large changes occur as a result of the accumulation of small stimuli—just as the accumulated weight of many small sand grains precipitate large-scale avalanches.

The human brain is one important example of a complex system. Nerves throughout the body function like telephone lines, enabling the brain to communicate with every part of the body via electrical signals. Normally, the brain continuously generates tiny electrical impulses in an orderly pattern. These impulses travel along a network of billions of nerve cells, called neurons. In epilepsy, a disorder of the brain, these nerve cells misfire and generate a sudden, uncontrolled surge of electrical activity in the brain resulting in an epileptic seizure. Few experiences match the drama of an epileptic seizure. A person having a severe seizure may cry out, fall to the floor unconscious, twitch or move uncontrollably, drool, or even lose bladder control. Within minutes, the attack is over, and the person regains consciousness but is exhausted and dazed. While research has led to many advances in understanding and treating epilepsy, there are many unanswered questions about how and why seizures develop and most importantly what warning signs precede them.

The global financial markets represent another kind of a complex system. This system consists of thousands of investors and traders whose collective interactions lead to the formation of market prices. Like the electrical impulses in the brain, prices move in an orderly fashion most of the time. But every three to four years, there is a sudden, uncontrolled movement in prices leading to systemic market crashes. While research has led to many advances in understanding financial markets, there are many unanswered questions about how and why these crashes develop and what warning signs precede them. For most investors these crashes seem to come out of the blue without prior warning.

This problem of predicting critical transitions from one state to another is not limited to systems like the human brain or the financial markets. Most other complex systems ranging from earth’s climate to wildlife populations and ecosystems also exhibit abrupt changes at tipping points. Up until now scientists did not know what warning signs precede such transitions. Most recently however, a study conducted by an international team of scientists has found that regardless of the details, the dynamics of each system near its tipping point shares similar symptoms.  

One such warning behavior or symptom they identify is the “slowing down” of one part of the system, which can be an indication that the system is seeking to establish a new equilibrium.  

In financial markets, this slowing down of the system manifests most prominently in the bifurcation of market breadth.  Market breadth is a way of measuring the moves in the stock market in general by comparing the number of stocks that rise compared to those that fall or comparing stocks making new highs to those making new lows.

Market advances accompanied by increases in the number of stocks reaching new highs in price are advances that are well confirmed by market breadth. Such advances are more likely to continue. However those advances that are not accompanied by increases in the number of stocks reaching new highs in price are not as well grounded in internal strength and are likely to reverse.

Similarly market declines accompanied by an increasing numbers of stocks falling to new lows are likely to continue. However failure of new lows to expand with price declines indicates the possibility of stock market reversal to the upside.

The slowing market breadth in March 2009 for example was an early signal that the stock market was close to forming a bottom. In the chart below we have plotted the difference between the number of 52 week highs and 52 week lows on the NYSE. The lower pane has the SP500 price chart. The lowest reading was registered in October 2008 with ~2500 net new lows. In November 2008 as the SP500 kept falling, net new lows declined to about 1300. The SP500 made a new low in March 2009 but by then the net new lows had dried up to about 800. This represented a positive divergence indicating a “slowing down” in selling pressure, a condition that precedes critical transition from one phase to another.

Similarly, there were clear signs of the market topping out in 2007-2008. The chart below covering the 2005-2007 time frame shows that new highs slowed down in December 2006 much before the subprime crisis. The market however peaked only in July 2007 by which time the number of new lows were in the 800-1000 range, a clear “get out of the market” signal. 

The 2002-2003 chart below also illustrates the same concept of system slowdown. The new lows spiked into August 2002, declined in October 2002 even as the SP500 made a new low and dried up by March 2003 although the market was still testing the October 2002 lows. That month, March 2003, happened to mark the bottom of the 2000-2002 bear market. 

Other major market reversals like the 1998 crash, the 1987 crash and many more all exhibited the same symptoms: namely the “slowing down” of breadth even as the indices continued. In fact, I am yet to come across a major market reversal that occurs without the slowing down of breadth. This is because trend reversals occur one stock at a time, just as large changes in complex systems occur as a result of the accumulation of small stimuli — the accumulated weight of many small sand grains precipitates large scale avalanches.

So the next time you check the market’s health make sure that the troops (the individual stocks) are leading the generals (the market indices) and not the other way round. Markets do not normally get into serious trouble until the generals have been leading the troops for a while. 

Nihar Dalal, CFA

References

·          Early-warning Signals for Critical transitions by Marten Scheffer et al

·          Revisiting Market Efficiency: The Stock Market as a Complex Adaptive System by Michael J. Mauboussin