Here is an excerpt from a commentary posted at TSI on 30th August:
During bull-market years and bear-market years, it is not uncommon for the US stock market to experience a quick decline of 10% or more at some point. For example, there was at least one quick decline of 10% or more in 1994, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2007, 2008, 2009, 2010, 2011 and 2012. In other words, 15 out of the 19 years from 1994 to 2012, inclusive, had quick declines of 10% or more. Only two of these years (2001 and 2008) had declines that could reasonably be called crashes.
The periods from mid-2003 through to early-2007 and late-2012 through to mid-2015 were unusual because they did NOT contain any quick 10%+ declines. In other words, the 12.5% decline in the S&P500 Index (SPX) from its July peak to last Monday’s low was not extraordinary in an historical context, it only seemed extraordinary because the market had gone an unusually long time without experiencing such a decline. That is, it only seemed extraordinary due to “recency bias” (the tendency to think that trends and patterns we observe in the recent past will continue in the future). Furthermore and as noted in the email sent to subscribers late last week, this year’s July-August decline was significantly smaller than the July-August decline that formed part of a bull-market correction in 2011.
In summary, what happened over the past few weeks was not a crash by any reasonable definition of the word and was only extraordinary in the context of the unusually long period of low volatility that preceded it.
That being said, the recent market action could well have longer-term significance. Just as the sudden increase in volatility in 2007 following a multi-year period of exceptionally-low volatility marked the end of a cyclical bull market, the sudden increase in volatility over the past few weeks could be marking the end of a cyclical bull market. In fact, there is a better-than-even-money chance that this is the case.
Also, while the recent quick decline doesn’t meet a reasonable definition of a stock-market crash, it could be part of a developing crash pattern. Recall from previous TSI commentaries that a US stock-market crash pattern involves an initial sharp decline in the 7%-15% range (step 1) followed by a rebound that retraces at least 50% of the initial decline (step 2) and then a drop back to support defined by the low of the initial decline (step 3). A breach of support can then result in a crash. Step 1 of a potential crash pattern is complete and step 2 is now very close to being complete. Note, however, that even if steps 2 and 3 are completed over the next couple of weeks the probability of a crash will still be low, albeit much higher than it was a few weeks ago.