Does a 20% drop always mean a bear market?

All week I have heard mainstream media commentators say that because major indices have dropped 20%, this means we are in a bear market. A few add the phrase ‘a sustained drop’ but most don’t.

Let’s look at the last 20% drop in 2011.

S&P500 Index

Between May and October, the index dropped 21.6% but then rallied to a new high.

Many other national stock indices did likewise, including the FTSE100 Index. Other examples include the Indian and Brazilian national stock indices in 2004 and 2006, which had 20% drops in the middle of strong bull markets.

How can we tell which 20% drops really are bear markets?

There are three ways to look at this but they are manifestations of the same thing: falling participation in the bull market

1. Divergences on cumulative breadth indicators

Bull market peaks form with breadth divergences. The number of stocks participating in the rally drops off when the index peaks.

In June 2015, I showed how the cumulative advance-decline line for the NYSE Composite Index had formed a divergence at the May 2015 high.

The chart is updated below:

Contrast this with 2011 when the late April high was confirmed by the cumulative breadth indicator.

2. Divergences on range bound breadth indicators

In November, I showed that the percentage of stocks above their 200 day moving average on the SPDR Consumer Discretionary (XLY) had formed a very pronounced divergence.

The chart is updated below:

SPDR Consumer Discretionary (XLY) and % above 200 day moving average

The ETF has since fallen from $81.47 to $73.31. Note that plenty of minor divergences did not end the trend: the more pronounced the divergence, the more likely it will have an effect.

3. Sector indices rolling over

As a bull market matures, whole sectors drop out of contributing to the uptrend. I took a sample of 50 US sector indices (not completely scientific but a sample of the 67 Industries from the S&P GICS framework).

Here is how many had rising 200 day moving averages at various dates:

29 April 2011 (closing high on the S&P500 Index then)
46 out of 50 rising

21 May 2015 (closing high on the S&P500 Index)
39 out of 50 rising

20 July 2015 (retest the May high, only 0.1% lower)
29 out of 50 rising

03 November 2015 (retest the May high, only 1.0% lower)
34 out of 50 rising

22 January 2016
7 out of 50 rising

• At the peak, there were 7 fewer rising moving averages than in April 2011
• By July, although the index was almost unchanged, another 10 sectors had dropped out
• By the November peak, the number had recovered (to my surprise) but still one third of sector indices had a falling moving average
• At the April 2011 peak, there had not been a deterioration at the sector level, so there was a pull back within an uptrend rather than the start of a bear market

Conclusions

• There is enough breadth deterioration to mean that indices have topped
• Don’t let the media tell you that every time indices drop 20%, this automatically means that it is a bear market