Dependence
A few weeks ago we wrote about probability and how standard statistical tools could not account for large movements in a stock market (Probability Suggests 20th July 2017 ); they are seen as anomalies.
To deal with these variations the rules of statistics are often adjusted and theories such as Efficient Market Hypothesis assume factors such as all investors being rational, all information has the same impact and returns in one period are totally independent of the previous period. This allows standard statistics to apply and fulfils our human needs to standardise and create rules.
However, empirical evidence suggests this is not the case and in our previous round up we showed an example of how returns deviated by 6x from their mean over 30 times since the early 60’s and not once every 1.4 million years!
We feel its important to think about investment markets differently and understand how they work; the key is that investors do not behave rationally and do not react the same way to different information. This is because of the different types of investors within the market and their different investment time horizons – in Edgar E Peters book Fractal Market Analysis he eloquently proposes that markets only remain stable when trading is balanced.
For example short term traders may have large losses on their day’s trades and want to exit their positions (sell); whilst their losses are large for them the longer term investor sees it as an opportunity and steps in becoming the buyer – the market is balanced. It is when long term perceptions of fundamentals change and long term investors start heading to the exit door and liquidity dries up (there are no buyers of what everyone is selling) that large volatility occurs.
Many people have written and proposed alternatives, but unfortunately by and large remain in the minority.