Website of Different Investment News and Analysis Under One Roof
Why Do Bear Market Rallies Occur?
In last week’s discussion of bear market rallies, Forbes looked at some past bear markets and the rallies that interrupted their prevailing downtrends. Examining bear markets going back to the 1930's, the rallies generally fail to exceed key Fibonacci retracement levels. Specifically, these rallies often exceed the 50% retracement resistance from the bull market high, but do not close above the 61.8% resistance level.
On October 11, 2007, the S&P 500 had a high of 1576, but then closed at 1554. This marked the bull market high. By March 27, 2008, the S&P 500 had a low of 1257.
Ten days after the March 27 low, the 38.2% resistance level at 1379 was overcome, and on May 2the 50% resistance at 1417 was also overcome. On May 19, the S&P 500 had a high of 1440, just below the 61.8% retracement level, and then closed at 1426. As it turned out, the bear market rally was over, having lasted 45 days.
One question I often get from my articles about bear markets (list at end of article) is why bear market rallies occur at all? That is, what makes them an essential part of the bull-bear stock market cycle. In my observations, the key function of a bear market rally is to change investor sentiment, convincing them they should no longer fear a further decline. This briefly injects capital into the market, raising stock prices sufficiently enough for a wave of selling to take place, leading to further decline.
To illustrate this function, Forbes looks at price and sentiment data at the same time. I have taken the daily chart of the S&P 500 from 2007-08, and labeled various highs and lows with the percentage of investors that reported being bullish that week (Bullish%). My preferred data set is the weekly survey of the American Association of Individual Investors (AAII).
At the October 11 high (point 1), 54.6% of investors in their survey were bullish. By the January 23 low (point 2) only 25.4% were bullish. Over the course of the decline, bullish sentiment tends to decrease. After a 24-day, 9% rebound, the Bullish% had improved to 34.2% (point 3). On the next decline, the S&P 500 made a new low during the week of March 10, and the Bullish% declined to 20.4% (point 4).
On March 17, the S&P 500 made a new correction low of 1257. The strong close the next day was the start of a more meaningful rally. By April 7, the S&P 500 had reached a high of 1386, and the Bullish% had more than doubled from the March low, now at 45.8%. This reading was followed by a sharp market decline and the Bullish% the next week was 30.4%.
The market rebound resumed, and by May 1, the Bullish% had risen to 53.3% (point 5). The next week the Bullish% was also high at 52.8%. The week of the May 19 high, the Bullish% had dropped to 46.3%. By the end of the week, the S&P 500 had closed below the support from the March low. So how low was the S&P 500 expected to go?
One can use the rally from 1257 to 1440, a move of 183 points, to calculate downside targets. One Fibonacci level that I use to calculate price targets is 1.272, or 127.2%. If the decline is 127.2% of the rally, then it should drop the S&P 500 down 232 points from the May high, to give a downside target of “1440 – 232”, or 1207.
On July 15, the S&P 500 had a low of 1200 (point 1), before it began a further rebound. The rally peaked at 1313 on August 11 (point 2), which was just below the 50% retracement resistance of the latest decline, which was 1320. The 1.618 (161.8%) downside target at 1143 was reached on September 18 (not shown).
I have also applied the same projection analysis to the Nasdaq NDAQ Composite in 2000. The Composite made a high in March at 5132, before falling to the May low at 3042. The rebound from the May low peaked at 4289, which was below the 61.8% retracement resistance at 4334.
The 1.272 downside target of 2703 was reached at the start of December 2000. The 1.618 target of 2272 was reached in the first week of January 2001, as the Nasdaq Composite had a low of 2251. In October 2002, the bear market low was 1108, as the market approached but did not reach the 2.618 target at 1026.
In terms of sentiment, the AAII Bullish% had a high of 75% on January 6, 2000 (point 1), which was an extremely high reading. The week of March 10, the Nasdaq reached its high (point 2), and the Bullish% was 58.3%. On May 11, the Bullish% had dropped to 30.2%, making a new low, after the Nasdaq had dropped 37% in just six weeks. The Nasdaq made its low two weeks later. By July 20, the Bullish% had rebounded to 66.7% in conjunction with the bear market rally. As the 1.618 target was reached at the end of the year, the Bullish% had dropped back to 31.1% (not shown), a reading that set the stage for another rally on the horizon.
With the strong rally last week, the S&P 500 exceeded the 50% retracement resistance at 2792 on Thursday. The 61.8% resistance at 2934 is the next key resistance level to watch, as a daily close above this level would support the case for a resumption of the bull market. As I pointed out last week there were no strong signs that the bear market rally was over.
Though the fear in the market was at panic levels in the first few weeks of March, the Bullish% had a low of 29.7% the week of March 12 (point 1). Having watched these numbers for decades, I expected to see the Bullish% to drop to the 20% level or lower to match the market’s decline. The highest Bullish% reading so far this year has been 45.6%, which occurred the week of January 23.
For context, at the March 2009 bottom, the Bullish% had a low of 18.9% before the bear market ended The stock market also corrected sharply in early 2016, with the Bullish% making a low of 17.9%. On December 13, 2019, the Bullish% had a low of 20.9%, which preceded the S&P 500 price low by two weeks.
So far, we have not seen the very low Bullish% readings that typically accompany a market bottom. The S&P 500 made its low just thirteen days ago. As the past bear market examples have shown, bear market rallies often last much longer. If the bear market continues, I recommend using Fibonacci retracement levels in conjunction with sentiment data in order to anticipate the timing and extent of bear market swings.