Black Friday: a whole free December, until 27 November. See the offer →
SonarSignal
Analysis7 min read

How to Read Bollinger Bands: Why One Timeframe Is Not Enough

Bitcoin, the Dow and the S&P 500 compared: we measured Bollinger Bands on the weekly, monthly and quarterly charts, with a look back at 2014 and 2018. The result is less convenient than the headline.

Share:XTelegramWhatsAppRedditFacebookLinkedIn

"Bollinger Bands are the tightest ever." You read that sentence a lot right now, usually followed by the claim that a big move is coming. We measured it. As a record, it only holds on the quarterly chart; on the monthly chart the squeeze sits in the lowest tenth of its history. On the weekly candle, the one most traders open first, the bands are not tight at all.

What the bands measure

Bollinger Bands draw two lines around the moving average of the last 20 candles, two standard deviations above and two below. The distance between the lines, relative to price, is called bandwidth. A small bandwidth means the price has moved little recently. That is all it says. It contains no direction.

The same Bitcoin, three different pictures

We calculated the bands with the usual settings (20 candles, two standard deviations, as in TradingView) on three timeframes and compared the bandwidth with Bitcoin's own history. The percentile tells you what share of all earlier candles had a bandwidth that was smaller or equal. A low percentile therefore means: tight compared with its own past.

This is how it looks on November 8, 2026, with a closing price of about 75,700 US dollars:

– Week: bandwidth 47 percent, percentile 29 to 31. That is mid-range. The tightest weekly bands in the whole history of Bitcoin came in November 2025 at 16 percent, almost three times as tight as today.

– Month: bandwidth 80 percent, percentile 5 to 7. That is very tight, in the lowest tenth. The record, however, belongs to May 2026 at 68 percent.

– Quarter: percentile 2 to 5, but here the measurement loses its meaning. Twenty quarterly candles are five years, and the lower line sits practically at zero, in the third quarter even below it on paper. A band that falls below zero tells you nothing about the next step.

So anyone who says "tightest ever" means the quarter or the month. On the week, it is not true.

Three cards for week, month and quarter showing the percentile and bandwidth of Bitcoin's Bollinger Bands

Why you need several timeframes

That is exactly why the bands should be read in triplicate: week, month and quarter side by side. This is no gimmick; it is the only honest way to answer the question "tight or not?". A squeeze on all three timeframes at the same time is rare. For Bitcoin, all three sat in the lowest quarter only in the months June to December 2025 and May to September 2026, and even that yardstick wobbles: Bitcoin has become calmer over its history. Today's monthly and quarterly bands therefore look tight almost by default, because they are compared with the wild years before.

Today the three disagree: week mid-range, month very tight, quarter meaningless. That disagreement is the real information. It says there is no clean squeeze worth forming an opinion on.

Bitcoin candles with Bollinger Bands on the weekly, monthly and quarterly chart, one below the other

Looking back: October 2018 and November 2014

Both dates are often named as templates. We recalculated the bands for each. For context: in October 2014 and October 2018, Bitcoin stood roughly 67 to 70 percent below its previous high. Today the gap is 39 percent.

– October 2018: On the week the bands were tight (percentile 8). On the month (percentile 67) and the quarter (percentile 91) they were wide. In the following month the price fell 37 percent, over the next three months the low was 50 percent below the October close, and six months later the price was still 16 percent lower.

– November 2014: The bands were not tight at all: week percentile 70, month percentile 75. The price still fell again: three months later it was 33 percent lower, and the low was 60 percent below the November close.

What can be stated: 2014 was not a case of tight bands. 2018 was, but only on the weekly candle. Two cases are not statistics, and both happened when Bitcoin was much further below its all-time high than it is today. No forecast can be drawn from them.

Weekly Bitcoin charts with Bollinger Bands: October 2018 with tight bands, November 2014 with wide bands

The cross-check with the Dow Jones and the S&P 500

The indices have a much longer history, the Dow Jones from 1992 and the S&P 500 from 1950. There the picture today is the reverse of Bitcoin's:

– Weekly bands: Dow 6 percent bandwidth (percentile 6), S&P 500 7 percent (percentile 11). Tight.

– Monthly bands: Dow percentile 59, S&P 500 percentile 69. Rather wide.

Weekly and monthly charts of the Dow Jones and the S&P 500 with Bollinger Bands

Then we counted what followed tight bands (percentile 15 or lower) in the past. We measured the change in the closing price over the next 13 weeks or six months:

– S&P 500, month: 15 episodes, 8 times higher, 7 times lower. The typical move afterwards was 5.1 percent, smaller than usual (7.7 percent).

– Dow, month: 6 episodes, 4 times higher, 2 times lower, typically 5.0 percent against 6.9 percent otherwise.

– S&P 500, week: 56 episodes, 34 times higher, 22 times lower, typically 4.2 percent against 4.9 percent otherwise.

– Bitcoin, week: 13 episodes, 10 times higher, 3 times lower, typically 27.9 against 26.8 percent. Bitcoin, month: 5 episodes, 3 times higher, 2 times lower, 46.7 against 41.2 percent. That is too few cases for a verdict.

The individual cases show how widely this scatters: tight monthly bands on the S&P 500 in May 2008 turned into a loss of 36 percent within six months, while in the fourth quarter of 1994 (quarterly bands) they led to a gain of 34 percent within a year. Overall, our measurement shows that in this data tight bands neither announced a direction nor reliably a bigger move. The slogan "squeeze, it is about to blow" only half holds. That more cases went up than down comes from the general upward trend of the markets, not from the bands.

The better variant: the GannSonar Trendlage

The bands only answer the question "how much is the price moving?". Anyone who wants to trade asks first: where is the trend? That is exactly what the GannSonar Trendlage in the Trading Desk shows. It colors every candle by the position of the trend, with ten positions from strong uptrend through "the market is still deciding" to downtrend. The "Squeeze" switch also marks the phases in which the zones contract, and the breakout from them. Both appear in one chart, on a timeframe of your choice from 5 minutes to 1 week.

The picture for Bitcoin on November 8, 2026:

– Weekly chart: uptrend weakening, since mid-August (last closed weekly candle). No squeeze marking.

– Daily chart: strong uptrend. No squeeze marking.

That matches the band comparison above: no clean squeeze on the week. And it adds something the bands never say: the trend on the week is intact but losing strength, while the day is still strong. Like the bands, the Trendlage is a status display. It describes where the market stands, not where it is going.

How to read Bollinger Bands in practice

– Tight means quiet, not direction. Bandwidth alone gives neither a long nor a short.

– Always name the timeframe. "The bands are tight" without a timeframe is not a statement.

– Measure against its own history. Whether something "looks tight" is a feeling. The percentile of the bandwidth is a number.

– Read week, month and quarter side by side and check whether they contradict each other.

– Get the direction elsewhere: from the Trendlage, from the structure of the chart, from your plan with stop and target.

Our analysis method: We use Bitcoin against the US dollar as daily candles from Bitstamp since August 2011, the Dow Jones from January 1992 and the S&P 500 from January 1950 from Yahoo Finance. From these we build weekly, monthly and quarterly candles. We calculate the bands with the moving average over 20 candles and two standard deviations. The percentiles refer to the history of each timeframe, and the data runs to November 8, 2026. For Bitcoin the number of cases is small.

Risk warning: This article explains how an indicator can be read and shows historical measurements. It is not investment advice and not a forecast. Crypto markets are highly volatile; you can lose all or part of the capital you invest.

Not investment advice. This post is for information and education. It is not investment advice and not a recommendation to buy or sell. Trading cryptocurrencies and leveraged products can lead to the total loss of the capital employed.
Share:XTelegramWhatsAppRedditFacebookLinkedIn

More posts