Are Bollinger Bands Useful?

Bollinger Bands are an often-used technical indicator and if they are used properly, they can be a powerful indicator.

The problem with Bollinger Bands is that sometimes, when the stock or index makes a move outside of the Bollinger Bands, the move is the start of something bigger and other times, the move is followed-up by mean-reversion (e.g. a trading range).

I have noticed that when Bollinger Bands "open their mouth," any move outside of the Bollinger Bands tends to have legs.

When I say that the bands "open their mouth," I mean that the upper-band is higher than it was the previous day and the lower-band is lower than it was previous day (this is a specific type of "band widening" because it calls for both bands to move in a widening direction).

The average 1-day return on the S&P 500 (from 01/02/62 until 02/07/07) is .03%. When the Bollinger Bands open their mouth and the S&P 500's current daily high is greater than the upper-band (i.e. the S&P 500 makes a thrust higher while the Bollinger Bands open their mouth), the average 1-day return is .11%. Those results are statistically significant at the 99% level.

The average 5-day return for the S&P 500 is .16%. When the Bollinger Bands open their mouth and make a thrust higher, the average 5-day return is .29%. Those results are statistically significant at the 95% level.

The average 10-day return for the S&P 500 is .31%. When the Bollinger Bands open their mouth and make a thrust higher, the average 10-day return is .44%. Those results are statistically significant at the 90% level.

In short, when there is a market-thrust accompanied by higher market volatility, expect the S&P 500 to move in the direction of the thrust (especially if the thrust is higher).

Market Sentiment

Taking a look at the 10-day moving average of the ISE call % (the call % is a derivative of the ISE Sentiment Index that I use because it gives a more symmetrical representation of how many calls and puts are being traded in relation to one another).

After normalizing the data over one-year (I did not use the standard % Range calculation; rather I used a method that takes standard deviation into account), I divided the stats into 5 quintiles:

0% - 20% (the fewest number of calls in relation to puts)
20% - 40%
40% - 60%
60% - 80%
80% - 100% (the highest number of calls in relation to puts)

Put another way, the higher the reading, the more bullish the ISE option traders are.

From 10/13/2003 until the present, the average 10-day return of the S&P 500 has been .41% and the average 21-day return has been .84%.

When there is a reading from 0% - 20%, the average 10-day and 21-day returns have been .72% and 1.41%.

When there is a reading from 20% - 40%, the average 10-day and 21-day returns have been .38% and .42%.

When there is a reading from 40% - 60%, the average 10-day and 21-day returns have been .45% and .65%.

When there is a reading from 60% - 80%, the average 10-day and 21-day returns have been -.35% and .25%.

When there is a reading from 80% - 100%, the average 10-day and 21-day returns have been .41% and .92%.

The current reading of the normalized ISE call % is in the 0% - 20% range (in fact, it's 1.93%), which should be a positive for stocks over the next 10 and 21 trading days. That being said, my gut feeling is that there is more selling to come.

What To Expect Going Forward

After today's weak bounce and given the fact that the statistics favor a sell-off on days 2 through 10 after "the big one" (like we experienced on Tuesday), I would not be a buyer of this market here.

The sell-off still has legs.