Smith, Pareto and the divergence debate

Life is all about making sense of information. This information could be personal, economic, or societal kind.

When we comprehend information, we make a decision, which we assume to be positive. So at a certain level we trying to understand performance and how to perform as individuals and make performant choices? Performance assumes a kind of order. We try to create order in minds from disorder in the external world. This disorder could also be called divergence.

Divergence is a phenomenon seen in nature and markets. It is something that needs explaining and can be ineffable at times. It is linked with change or rate of change, fast and faster change. How fast can a price asset grow or decay? Divergence is also assumed to have a non normal aspect. It happens less than normal. Normality assumes that markets and 90 per cent of its components move higher together or vice versa. Divergence is different from normality and could indicate change, a potential reversal. Divergence is also known as non confirmation in technical analysis. Divergence also creates news, as something that is non normal is strange and worth talking about. Divergence can be seen in information, data, and patterns. Divergence is studied and researched. It is a system that can be built as a strategy.

Adam smith (1723-1790) talked about the invisible hand, what we don't know, something unpredictable, a kind of divergence. Vilfredo Pareto (1848-1923) talked about non homogeneous wealth allocation. 80 per cent of the wealth is with 20 per cent of people. This unequal allocation was a divergence. Charles Dow (1851-1902) talked about a basic tenet of markets in his now famous Dow Theory. If Dow industrials and Dow transports were not moving together, it was a non confirmation, a case of divergence.

Dow Industrials made higher high in Jan 2000, while Dow transports did not. Markets changed trend after divergence. Ralph N Elliott (1871-1948) talked about truncation, double extension, throw over as rare (divergent, non normal) rare formations in classic Elliott. Edward Norton Lorenz (1917 – 2008) talked about Chaos and how small changes in early conditions lead to totally new behavior, leads to the butterfly effect.

We can see cases of divergences in current times. John Murphy talked about inverse movement of gold and dollar and when prices diverge from this anticipated relationship, it’s called as an intermarket failure. Sam Stovall talked about sector performance divergences. Sector performances vary as the economic cycle changes. Robert Shiller talked about divergence between market data and fundamental value.

Shiller calls it fluctuations (a divergence case). Mandelbrot talked about fractals and about extremities. He said “bell curve is nonsense” in his book 'the (mis)behaviour of markets: a fractal view of risk, ruin and reward' what was Mandelbrot doing. He was trying to justify large divergences more than explained by Gaussian curve. Nassim Taleb says black swan is random, rare, unpredictable, an extreme divergence, divergence again. Robert Arnott talks about how growth diverges from value and creates cyclical opportunities. Robert Prechter has done extensive work on social mood divergences; rising hem of skirts, films we watch, sugar we consume, social mood is patterned and diverges from one extreme to other.

The inefficient markets school of thought is based on large divergence. The efficient markets school of thought based on small divergence. Coming to look at it, the biggest research debates of all time is built around divergence. If it is small it is normal, if it's large, it's inefficient. Large divergence is a reality between high, low, negative correlated assets, 60% between gold and palladium from July 2009 till Jan 2010, 55 per cent between Brent and Exxon from July to Oct 2008. Even 100 per cent divergences between sector peers are not too rare. They happen regularly.

Can we define large? Is small divergence as normal as large divergence? Is small divergence any less important than the case of large divergence? Sam Stovall sector rotation outperformance and underperformance can be seen as divergence of 12 per cent between Dow Industrials and Dow Financials from Oct to Nov 2009. But did Stovall talk about small divergences between Dow Industrials and S&P 500 of 5 per cent or 6 per cent in 6months. These small consistent repeating small divergences are never discussed. But they are as important as large 100 per cent repeating divergences between two sectors. Behavioral finance talks about divergences as anomalies that are tough to capture. Is it because divergences are tough to isolate, measure, explain, quantify?

Robert Shiller explains why markets are inefficient using various time series of present value of dividends and plotting them against stock prices illustrating more than normal fluctuations. His proof that markets are inefficient is based on the fact that large divergences can't be explained. Should the debate not have been about divergence and not about efficiency or inefficiency? Divergences are everywhere, in capital markets, economics, sciences and we still consider them non forecastable, extremities, errors, non normal behavior, non order, and chaos. Is this really true?

Email the writer: