Technical Analysis With Bollinger Bands

If you think this metaphor is going the usual way, think again. Technical analysis is like a magnifying glass. It is my firm belief that we should not use technical analysis to magnify market information, but rather to concentrate it on a single, tiny point—just as a scout would use it to concentrate light from the sun and start a fire.

As any seasoned analyst will tell you, thousands of indicators claim just that, adding to the quest for the holy grail of technical tools—a trap many analysts have fallen into early in their careers. Nevertheless, some indicators have survived long enough to become standards, such as moving averages, relative strength indexes (RSIs), and Bollinger Bands.

BOLLINGER BANDS: A BRIEF INTRODUCTION

John Bollinger’s bands looked to answer the question of whether a stock was expensive or cheap relative to its recent history. They worked by establishing a fixed number of standard deviations from the mean of past closing prices. To illustrate the concept, think of a river with banks. When the river is calm, the banks are clearly defined; when it is faster flowing, they become harder to define due to the erratic nature of the water. As you move away from them, it becomes increasingly improbable that you will be reached by water – unless something extraordinary happens.

A river and its banks are an apt metaphor for average price and its standard deviations. During times of low volatility, its banks will contract; in moments of high volatility, it is only natural that the bands expand. The bands signify a safe distance away from the main current—usually meaning prices won’t stray far from the average. Consequently, when it reaches the upper band, this signals that prices are above the average, and if it reaches the lower band, this implies they are lower than normal.

The price flow in the markets is like a river, except that it stops at the edge of our screens. We can gauge its movement by monitoring whether it moves within or beyond established deviations from the average. If it pushes against the outer limits, we can surmise that there is a chance it will either revert back or exhibit enough momentum to stay in that direction.

Bollinger Bands provide us with valuable context to make better decisions, as they can alert us to price extremes or recent volatility. Nonetheless, it is important not to rely on them as a predictor of future movements. We should instead use them as areas of support and resistance and wait for price to confirm our expectations before taking any steps. This way, we increase our chances of achieving the desired outcome.

INDICATOR BOLLINGER BANDS

We can use Bollinger Bands beyond pricing data. Technical analysis often employs various averages such as simple, exponential, linear-weighted, and volume-weighted. Additionally, several indicators display an average value to provide support or smoother the results in comparison to current values. This is why I believe in using Bollinger Bands for other indicators over price; it is especially effective at showing when activity reaches historically extreme levels and dynamically defining them.

Indicator data, as in the case of relative strength index (RSI) and moving average convergence/divergence (MACD), provides valuable information regarding price. Traditionally, Bollinger Bands have been used to detect shifts in volatility and momentum, which can be applied here; indicating when activity becomes excessive or inadequate. For example, a rising RSI accompanied by higher closes suggests an incoming bull market, with values above 70 considered overbought and below 30 indicating oversold. However, shorter lookback periods are recommended for better signal production.

It is unlikely that the RSI will reach the traditional overbought level if it has been moving slowly for a while and then makes a sharp move upward. Due to a lack of volatility, the bands will contract, which will cause the price to exceed the standard deviation from the long-term average. By showing this was indeed noteworthy, it would otherwise go unnoticed because built-in levels do not adapt to past behavior as well as bands do.

The same concept can be applied when considering indicators such as the money flow index (MFI), which uses a similar technique to calculate values yet with more consideration given to trading volume. In addition, this applies to any indicator that produces one single value, whether it is confined or unrestricted. Applying some interpretation is necessary here: When used in conjunction with rate of change (ROC) indicator, bands help display when there has been an atypically large gap or shift; when paired with on-balance volume (OBV), it reveals when volumes going either into or out of a market reach a point of extremity.

The combination of bands and average true range (ATR) allows for better stop-loss and profit target setting. When price is close to the lower band, the ATR can be used to anticipate a possible increase in volatility. Close proximity to the upper band signals decreasing volatility and gives an opportunity for tighter stops accordingly. In either case, this combination provides an efficient way to filter out markets that are uneventful or about to become highly volatile.

FORMING THE BANDS

For this article, I used MetaQuotes’ MetaTrader 5 with a datafeed from IC Markets. It’s as easy to set up a Bollinger Band within any mentioned indicator on the platform – just drag and drop it into the window. To access this, press Ctrl+N → Indicators → Examples → BB (Bollinger Bands). This will allow you to customize each line’s width, color and style. Don’t forget to go to the parameters tab and select “first indicator’s data” (see Figure 3).

The length of the moving average, paired with your desired number of standard deviations, will depend on the settings you are using for the indicator in question. A longer average relative to what you’ve specified may result in fewer signals, but be more reliable; two standard deviations usually suffice. As a general rule, aim for an average that is 4 to 4.5 times the indicator value. To make comparison easier, you can always plot these alongside built-in levels (if applicable).

In my own trading, I multiply a Fibonacci number by the golden ratio (phi: 1.618) raised to the third power (4.235)—roughly 4 to 4.5 times the indicator value. Thus, both my RSI and ATR use 13 periods and a simple moving average of 55; MFI uses 21 periods and an 89-period simple moving average. For each of these, I use two standard deviations. Indicators such as OBVs and ROCs can be used as references.

BANDS: HOW TO USE THEM

I use the MFI and RSI with the settings mentioned above, to identify applicable levels. Once they exceed or dip beyond the bands’ borders, record the highest/lowest price within that span as a resistance/support line. Then draw a 1.5 – 2 times multiplied ATR line which is derived from the market’s volatility for setting a tolerance zone for these boundaries. Stop-losses are set slightly below it and if exceeded, wait for a new support/resistance level to be created.

When looking for entries, I strive to get as close as possible to the support or resistance line, which can be at the formation of the level or during a retest of the area. The maximum stop-loss should not exceed three times the average ATR value, and this strategy works best on timeframes from one hour to daily charts.

A ratio of at least 1.5x the risk is recommended, but price targets will depend on how long you plan to hold the position. Simply look for the previous confirmed level of support & resistance in your current or higher timeframe to determine whether price can reach this minimum reward potential. If you work along with the main trend, the probability is likely to be in your favor most frequently.

The levels discussed here do not predict the movements of the markets, rather they tell us what is less likely to happen. In combination with evaluating the fundamentals like the current economic/sector/company condition, we will be able to make some informed decisions and use John Bollinger’s techniques and recognition of standard deviations in technical analysis more effectively.