This past week, technical analysts were waiting with bated breath to see if the original Hindenburg Omen signal that occurred on August 12, 2010 would be confirmed (refer to my post dated August 15, 2010 for an explanation of the theory). This, in fact, occurred on August 20, 2010 as all of the theory’s criteria were again met. It is thought that the more signals that occur in quick succession, the more likely it is that we will witness steep stock market decline.
In his article dated August 21, 2010, Robert McHugh provides a chart of all Hindenburg Omen signal occurrences in the past 25 years, including the number of signals that occur within 36 days of the first signal and the subsequent drop in the Dow Jones Industrial Average. For some reason, his data only includes one instance of a lone signal occurring, even though he notes earlier in his article that there have been eight sole signal occurrences in the past.
During this time period, if there has been a confirmation of a Hindenburg Omen signal, the average subsequent drop in the stock market has averaged 12.11%. Of course, the extent of the drop is always measured in hindsight. When an investor is in the midst of a falling market, there is no means of knowing whether a drop will continue or whether a bottom has been reached. After all, these declines have occurred over periods as long as nine months and there have been significant market recoveries within these market drops. Thus, if an investor was shorting the market upon the occurrence of a Hindenburg Omen signal, the average returns he would realize would be lower than this figure.
This small data set does suggest that this Hindenburg Omen phenomenon is worth examining given its predictive success. Note, however, that the data also suggest that whether more signals occur or not is immaterial to the potential for an imminent stock market plunge. My regression analysis of the historical Hindenburg Omen data indicates that there is no significant relationship between the total number of signals that occur and the resulting market performance. For statistics fans, the R-squared value of the regression was less than 2% and the standard error was 6.84. Therefore, if you believe that the Hindenburg Omen is a good indicator of a stock market decline, there is no reason to wait for the occurrence of additional signals. It’s time to place your bets!
