Profiting From The Squeeze

You’d be hard-pressed to find a trader who has never heard of John Bollinger and his namesake bands. Most charting programs include Bollinger Bands — of all technical indicators — but although these bands are some of the most useful if applied properly, they are also among the least understood. One good way to get a handle on how the bands function is to read the book Bollinger on Bollinger Bands, in which Bollinger himself explains the whys and wherefores of using the bands.

According to Bollinger, there’s one pattern that raises more questions than any other aspect of Bollinger bands. He calls it the Squeeze. As he puts it, his bands are driven by volatility, and the Squeeze is a pure reflection of that volatility. Here we look at the squeeze and how it can help you identify breakouts.

A Brief Description

Bollinger Bands employ upper and lower standard deviation bands together with a center simple moving average band around price to identify high and low volatility points. While it can be a real challenge to forecast future prices and price cycles, volatility changes and cycles are relatively easy to identify. This is because equities alternate between periods of low volatility, followed by periods of high volatility, and so on — much like the calm before the storm and the inevitable inactivity afterward.

Here is the squeeze equation:

Bollinger Bandwidth = (Top Bollinger Band (20 periods))-Bottom Bollinger Band (20 periods))/ Simple Moving Average Close (20 periods) A squeeze candidate is identified when the bandwidth is at a six-month low-low value.

When Bollinger Bands are far apart, volatility is high, and when they are close together, it is low. A squeeze is triggered when volatility reaches a six-month low and is identified when Bollinger Bands reach a six-month minimum distance apart. Traders can use scans they have written themselves or find scans in the Bollinger Band Tool Kit to search hard drives for stocks that are squeeze candidates.

Determining Breakout Direction

The next step in deciding which way stocks will go once they breakout is somewhat more challenging. Bollinger suggests that, to determine breakout direction, it is necessary to look to other indicators for confirmation. He suggests using the relative strength index along with one or two volume-based indicators such as the intraday intensity index (developed by David Bostian) or the accumulation/distribution index (developed by Larry William).

If there is a positive divergence — that is, if indicators are heading upward while price is heading down or neutral — it is a bullish sign. For further confirmation, look for volume to build on up days. On the other hand, if price is moving higher but the indicators are showing negative divergence, look for a downside breakout, especially if there have been increasing volume spikes on down days.

Another indication of breakout direction is the way the bands move on expansion. When a powerful trend is born, the resulting explosive volatility increase is often so great that the lower band will turn downward in an upside break, or the upper band will turn higher in a downside breakout.

See Figure 1 below, which shows the May KBH breakout. Bandwidth reaches a minimum distance apart in May (indicated by the blue arrow in window 2), followed by an explosive breakout to the upside. Note the increasing relative strength index (shown in window 1), along with increasing intraday intensity (the red histogram in window 2) and the accumulation/distribution index (the green line in window 2), both of which (demonstrated by line A) are showing positive divergence with price (demonstrated by line B). Note the volume build that occurred beginning in mid-April through July.

Figure 1 – Weekly chart of KB Home showing the squeeze pattern setting up in the year leading up to May 2003. Chart provided by Metastock.com.

The Head Fake

A third condition to look out for is something Bollinger calls a head fake. It is not unusual for a security to turn in one direction immediately after the squeeze, as if to trick traders into thinking the breakout will occur in that direction, only to reverse course and make the true and more significant move in the opposite direction. Traders who act quickly on the breakout get caught offside, which can prove extremely costly if they do not use stop-losses. Those expecting the head fake can quickly cover their original position and enter a trade in the direction of the reversal.

In Figure 2, Amazon.com (AMZN) looked to be giving a squeeze set-up in early February 2004. Bollinger Bands were at a minimum distance apart, which had not been seen for at least a year, and there is a six-month low bandwidth (see line A in window II). There is negative divergence between the RSI (line 1 of window I), the intraday intensity (line 2 of window II), accumulation-distribution index (line 3 or window II) and price (line 4 of window III) — all of which point to a downward breakout. Breaking above the 50-day moving average (the orange line in the lower volume window) on drops in stock price, suggesting a build up in selling pressure, volume shows above normal values on downside price moves. Finally, the long-term trend line is breached to the downside in the first week of February. A downside breakout would be confirmed by a penetration in the long-term support line (line 5 of window III) and a continued increase in volume on downside moves.

The challenge lies in the fact that the stock had demonstrated a strong up trend, and one pillar of technical analysis is that the dominant trend will continue until an equal or greater force operates in the opposite direction. This means that the stock could very well make a head fake down through the trend line and then immediately reverse and breakout to the upside. It could also fake out to the upside and break down. While it looks set to breakout to the downside along with a trend reversal, one must await confirmation that a trend reversal has taken place and, in case there is a fake out, be ready to change trade direction at a moment’s notice.