Are we on the verge of another 1929-like meltdown in stocks?
Lately, we have seen a spate of commentary by technicians and commentators on popular financial news channels such as CNBC, Bloomberg and the Financial Times pointing to the chart below, which tends to show a pattern. The pattern the market historians are pointing to is the "seeming" parallel between 1928-1929 period and 2012-2014 period in the Dow Jones Industrial Average.
I will say that at first glance, the overlaid charts are without a doubt eye catching, but the underlying question we ask is, is there any statistical significance between the two periods when you crunch the numbers? The human brain is wired to find patterns in noise due to the priming effect, in which our brain is likely to interpret stimuli according to a perceived model. An excellent article in the Scientific American discusses the paper, "The Evolution of Superstitious and Superstition-like Behaviour", by Harvard University biologist Kevin R. Foster and explains this phenomenon quite well.
But in this article, we will examine the statistical similarities and differences between the two periods in question to see if the perception is supported by mathematical fact or not.
Dow Jones Average Daily Closing Charts (5-16-1928 to 10-31-1929) And (7-2-2012 to 3-3-2014)
Source: Tiger Technologies LLC, Bloomberg
In the above chart, we have shown the daily closing price charts of the two periods in question. The top line is the current period from 2012-2014 and the bottom line is the chart from 1928-1929. In 1928-1929, the index was comprised of 11 stocks, while now it contains 30 stocks.
Returns, Volatility, Correlation
In the first study done on the two periods, we compare the returns, volatility and see the correlation between the two periods.
Daily serial correlation: -0.70%
We took the highest peak of 1929 to see what the run up in the market was in the 1928-1929 period and did the same for the 2012-2014 period. Clearly, the market in 1928-1929 had more than twice the returns of the 2012-2014 period with 1.8 times the volatility.
The size of returns is an important comparison as it often shows overbought or oversold conditions. Large outsized returns combined with increased volatility often lead to large moves in the opposite direction as we saw in 1929. The same cannot be said for 2014 as the returns over the same comparison period are smaller and the volatility has not picked up. The chart below shows the daily ranges of S&P 500 from 2002 to present.
Additionally, the serial daily correlation between the returns of the two periods is close to 0. So statistically, there is very little proof to support the hypothesis that we are in a similar period to 1929.
Comparison of Large Drop Periods
Next, we compared the three large drop periods in 1928-1929 to 2012-2014.
Once again, the size of the moves bears very little similarity between the two periods. We have already shown that the daily correlation is non-existent. Therefore, to compare the drop periods between the two time series we had to approximate the closest similar period between the two time series. While the chart makes it look like the two time series are coincident, it is a just the brain finding patterns in noise.
Next Drop in 2014?
Statistically, there is very little evidence that we are in a similar period to 1929 and that we will suffer a large 48% drop like we did in 1929. But there is also statistical evidence that the period from May to September is weaker than the rest of the year. In fact, the lowest number of positive months, lowest average returns and highest volatility has been recorded in the months of May through September. We will be publishing a study on this in a later article.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.