Our medium-long term model predicts significant corrections. We call the rallies between significant corrections “big rallies”. Our model cannot predict small corrections. We call the rallies between small corrections “small rallies”.
Update: as of July 5, it has been 258 trading days since the last “small correction”. Only 3 out of 79 historical “small rallies” have lasted longer. If the S&P doesn’t make a 6%+ decline by the end of October 2017, this will be the longest “small rally” in history.
It has been 225 trading days (not calendar days) since the last small correction in June 2016 (on Brexit). Out of the 79 small rallies since 1962 (excluding the current one), only 7 small rallies have lasted longer. Here they are, ranked by number of trading days:
*There are approximately 21 trading days in a month.
The current small rally has lasted longer than 91% of all historical small rallies. So purely from a TIME perspective, bullish investors in the U.S. stock market face strong short term headwinds. This bearish risk is why we are sitting on 100% cash right now. Yes, the stock market can still go up despite the TIME extreme. But it’s a game of probability. Make smart investment decisions in which the odds are on your side. When the odds are stacked against you, take your chips off the table. Reduce your risk.
Let’s compare the above cases in which the S&P rallied for more than 225 trading days. How were they similar/different from the current small rally?
The current rally’s magnitude is 20.7%
October 5, 1992 – January 31, 1994 (335 trading days)
This is the longest case. Here is the daily bar chart.
The S&P rose 23% from the beginning of the rally to the end. Hence, this historical rally’s magnitude is similar to the current rally’s magnitude. The S&P went up in a very choppy fashion with many 3%, 4%, and 5% pullbacks.
After this small rally was over, the S&P 14% correction in 1994. Our model predicted this significant correction.
December 9, 1994 – February 13, 1996 (297 trading days)
This was the rally that started the internet boom. The S&P soared 49.9%! The S&P made a small correction after this HUGE rally. Here is a chart.
That historical case is not similar to today’s situation. That rally was driven by pure euphoria and belief in a brand new era – the internet era. There is nothing comparable today that can drive investor sentiment so high. Earlier this year, we thought that Trump’s pro-growth policies will make 2017 comparable to 1995. But Trump’s pro-growth policies are mostly dead right now. Whatever he does pass through Congress will not be big enough to generate a rally of that magnitude (almost 50%).
November 16, 1988 – January 3, 1990 (284 trading days)
Overall, the S&P went up 37.1%. The S&P was climbing out of the shadow of 1987’s significant correction. But immediately after this small rally ended, a small correction and then a big correction followed. Here’s a chart.
October 24, 1962 – October 29, 1963 (255 trading days)
This historical rally lasted barely longer than the current small rally. The S&P rose 43% because it was climbing out of the shadow of 1962’s massive correction. Obviously, this is not comparable to today because the S&P isn’t climbing out of the shadow of a massive correction right now. Here’s a chart.
March 12, 2003 – March 5, 2004 (248 trading days)
The S&P soared 47% because this was the first real rally from the 2000-2003 bear market. The first rally in every bull market is ALWAYS very fierce because there’s a strong mean-reversion theme going on. Obviously, this historical case cannot be applied to today. We are nowhere near the bottom of a bear market. Here’s the chart.
After this rally, the S&P made a small correction that was very long (84 trading days).
July 25, 1984 – July 17, 1985 (247 trading days)
Coming out of the 1983-1984 significant correction (which our model predicted), the S&P rallied 33%. Then it made a small correction. Here’s the daily bar chart.
September 29, 1986 – August 25, 1987 (229 trading days)
The S&P soared 48% during this rally. Then the S&P made a significant correction that more than gave up all of its prior gains (the crash of 1987, which our model predicted). Our model does not predict a significant correction right now, so this historical case should not be compared to today.
In addition, the U.S. economy deteriorated a little in 1987, which is not the case today. The U.S. economy is growing right now at a fair pace.
The S&P will begin a small correction soon. None of the very extreme historical cases (rallies that lasted more than 300 trading days) are similar to today.