CCI — the Commodity Channel Index
The answer
CCI measures how far the current price sits from its own average over the last N bars, expressed in units of that price's own normal wandering. A reading of +100 means price is unusually far above its recent average. It does not mean price is too high.
Why this costs you money
CCI has 2 lines drawn on it by default, at +100 and −100, and those lines are the trap. They look like boundaries. They are not boundaries. CCI has no upper or lower limit. It can reach +300 and keep rising for 6 weeks.
Here is the loss, and it is one of the most reliable ways to destroy a trading account. A stock breaks out. CCI crosses +100. You short it, because "overbought" sounds like a reason. The stock rises another 12% over the next month while CCI sits between +150 and +250 the entire time. You add to the short at +200, because now it is "even more overbought". You are removed from the market in a single week.
Selling strength because a number is high is not a strategy. It is a bet against the one condition in which markets make their largest moves.
The mirror version costs less but happens more often. A stock falls. CCI reads −180. You buy, because "oversold". The stock falls another 20%. Nothing was wrong with the reading. The reading meant "this is moving unusually fast in one direction", and you interpreted it as "this is about to stop".
How it works
CCI is built in 4 steps, and each one is worth naming because the last one is where the confusion comes from.
Step 1. Typical price. For each bar, add the high, the low and the close, and divide by 3. This is a single summary number for the bar.
Step 2. The average. Take a simple moving average of the typical price over N bars. The default N is 20 on most platforms.
Step 3. Mean deviation. Measure, on average, how far the typical price has been from that average over the same N bars. This is the instrument's own normal wandering distance.
Step 4. Scale it. Divide the current distance from the average by the mean deviation, then divide by a constant of 0.015.
That constant is the part almost nobody explains. Donald Lambert chose 0.015 so that roughly 70% to 80% of readings would fall between −100 and +100. It is a presentation choice, not a law. The lines at +100 and −100 are the edges of "the usual", drawn where they are because a man in 1980 wanted a readable scale.
So a CCI of +100 means: price is currently further above its 20 bar average than it usually is. That is all. It is a statement about the distribution of past distances, not about the future.
Lambert developed CCI for commodity futures, and he set the period from the length of the cycle he was trading, using roughly one third of the cycle length . Despite the name, the arithmetic contains nothing specific to commodities and it is used on shares and indices everywhere.
What it tells you, and what it does not
It tells you the current distance from the recent average, scaled to the instrument's own behaviour. That scaling is the useful part, and it is the same idea that makes ATR useful. A move of ₹40 means something different on a stock that usually wanders ₹8 and one that usually wanders ₹60. CCI normalises that.
It tells you when a stretch is unusual for that instrument. This is real information about the present.
It does not tell you the move is over. There is no mechanism inside CCI that produces exhaustion. High readings are what strong trends look like from the inside. If you take a chart of any large sustained move and read CCI across it, you will see the indicator pinned at high values for the entire best part of the move.
It does not have limits. RSI is bounded between 0 and 100, so a reading of 95 at least tells you there is not much room left on the scale. CCI has no such ceiling. This single difference is why CCI extremes are more dangerous to fade than RSI extremes.
And it does not contain anything price did not already contain. CCI is computed from the high, the low and the close. It is a rearrangement of price. What it adds is legibility, and what it costs is a delay of roughly half the averaging period.
The decision rule
CCI has 2 opposite uses. Pick one before you look at the chart, and pick it based on the market condition, not the reading.
In a range — highs and lows overlapping for weeks, ADX below 20 — a move beyond +100 or −100 is a stretch that usually returns. Fading it is defensible. Your stop goes beyond the recent extreme, not at the CCI level.
In a trend — clear higher highs and higher lows, ADX above 25 — a cross above +100 is a momentum entry in the direction of the trend, and a cross back below +100 is the exit. Fading it here is the mistake described at the top of this article.
If you cannot tell which condition you are in, you have no CCI trade. That is not a failure of nerve. That is the correct conclusion.
Notice that the same reading produces opposite actions depending on a judgement CCI does not make. This is true of every oscillator, and it is the reason oscillators are dangerous in isolation.
Try this now
Five minutes. Use a stock you hold, so the answer is about your own money.
- Open a daily chart, 1 year. Add CCI with the default period of 20.
- Mark every day CCI crossed above +100. Count them.
- For each one, look at what price did over the next 10 trading days. Mark it "continued" if price rose, and "reversed" if price fell.
- Write the 2 counts down. Now do the same for every cross below −100.
- Last step, and the important one. Go back to the 5 largest sustained price moves on the chart. Note the CCI reading at the start of each move.
What you should see. Two things, and they usually surprise people.
First, on most liquid stocks the "continued" count after a +100 cross is close to the "reversed" count, or higher. The signal that beginners treat as a sell is not a sell.
Second, and this is the payload: on nearly every chart, the largest moves of the year began with CCI pushing through +100 or −100. The reading you were taught to fade is the reading that marks the start of the moves that pay for a year of trading.
If your chart shows the opposite — extremes reliably reversing — check the shape of the year. You are probably looking at a stock that ranged, and you have just learned that the rule depends entirely on the environment.
Three real cases
1. Donald Lambert, 1980 — the constant nobody reads Lambert developed CCI and published it in Commodities magazine. He introduced the 0.015 divisor specifically to make most readings land inside a ±100 band. That is documented design intent about presentation. Two generations of traders have since read those lines as statistical thresholds with predictive meaning. The lines are a ruler, not a verdict, and the source says so.
2. S&P 500, November 2023 to March 2024 — the trend that never gave the oversold reading The US index rose steadily from late October 2023 into March 2024 without a meaningful pullback. Through long stretches of that advance, a 20 period CCI stayed above +100. Every trader who read those values as "too far, too fast" and sold was correct about the arithmetic and wrong about the outcome, continuously, for months. This is the cleanest available demonstration that an extreme reading is a description of strength.
3. NIFTY 50, February and March 2020 — the downside version The Indian index fell from around 12,400 in January 2020 to roughly 7,500 on 24 March 2020. CCI reached extreme negative readings early in that fall and stayed extreme. Anyone buying the first −100 cross was buying near the top of the decline. There was no floor in the indicator because there is no floor in the formula. The same pattern appeared on the S&P 500 over the same weeks.
The question that resolves it
A novice sees CCI at +190 and asks: how much higher can it go?
An expert sees CCI at +190 and asks: is price making higher highs and higher lows?
The first question has no answer, because the scale is unbounded. The second question has an answer you can read off the chart in 5 seconds, and it decides whether +190 means "sell the stretch" or "the trend is accelerating, do not stand in front of it".
What would make this wrong
If CCI extremes predicted reversals, then crosses above +100 would be followed by falls more often than rises, across many instruments and many years. The counting exercise above is the test. Run it on 10 stocks. If the extremes reliably reverse on most of them, this article is wrong.
The honest limits are 3.
First, this article is about the default 20 period on daily charts. A 5 period CCI is a different instrument with far more crosses and far more noise. A 50 period CCI barely reaches ±100. Any claim about CCI without a period attached is incomplete.
Second, the range-versus-trend judgement recommended in the decision rule is itself imperfect. ADX lags. "Higher highs and higher lows" is clear in hindsight and ambiguous at the right edge of the chart. Do not pretend the classification is free.
Third, and most important for this cluster: CCI is largely redundant. It measures distance from a moving average, scaled by recent variability. Bollinger Bands measure distance from a moving average scaled by standard deviation. The two agree most of the time, because they are 2 arrangements of the same 2 facts. If you already use Bollinger Bands or RSI, adding CCI does not add an opinion. It adds a second voice repeating the first one. Article 18 in this cluster shows how to test that for yourself.
In India
CCI is available on every Indian broker chart and on the free charting sites Indian traders use. It appears frequently in intraday NIFTY and BANK NIFTY setups, usually paired with a moving average.
Three Indian features matter.
Price bands compress the inputs. CCI uses the typical price, which uses the high and the low. When a stock is limited to a 5% or 10% daily band, the recorded high and low are administrative, not natural. Both the distance from the average and the mean deviation are then understated. A stock locked at its upper circuit for 3 days can show a CCI that looks moderate during the most violent week of its year.
Weekly index option expiries create characteristic intraday patterns in NIFTY and BANK NIFTY that have nothing to do with any company. A 5 minute CCI on an expiry day is measuring positioning flows. Thin mid and small caps. In a stock where a single large order moves the price 4%, the mean deviation in the denominator is dominated by a handful of bars. CCI becomes unstable. Do not apply an oscillator designed for liquid futures to a stock that trades ₹2 crore a day and expect the scale to mean anything.
In the United States
CCI is standard on US platforms and is used on equities, index futures and commodities, which is where it started.
Three US features matter.
Full daily ranges. US equities have no daily band equivalent to India's, so the high and the low that feed the typical price are a fair record of the day . CCI readings are therefore more comparable across time.
Earnings gaps. Most US results are released outside trading hours. A gap lands in the typical price as a single jump. CCI reacts violently for a few days and then normalises as the gap enters the averaging window. Readings in the 5 sessions after a US earnings gap should be discounted.
Commodity futures, where it came from. CCI was built for commodities with seasonal cycles, and Lambert tied the period to the cycle length. That original use still exists in US markets. If you trade an instrument with a genuine seasonal pattern, the period choice has a principled basis. On a share, it does not, and the default 20 is a convention rather than a finding.
Where they differ, and what that tells you
The mathematical difference is in the denominator.
CCI divides by the mean deviation — the instrument's own recent wandering. That denominator is the entire reason the indicator is comparable across instruments. It is also the number Indian price bands damage most.
When an Indian stock spends several days at a circuit limit, its recorded deviations shrink. The denominator shrinks. From then on, ordinary moves produce large CCI readings, because the indicator now believes this stock barely moves. You get extreme readings from unremarkable days, immediately after the period when the stock was genuinely extreme.
In the US the same stock would have printed its full range, the denominator would have expanded, and the readings would have been damped exactly when they should be.
What that tells you is a rule you can apply today. On Indian charts, look at the CCI denominator indirectly: if the last 20 candles include any that closed at a circuit limit, treat the current CCI reading as inflated and do not trade off it. On US charts, apply the same suspicion after an earnings gap, for the opposite reason — there the denominator has been temporarily inflated by one abnormal bar, so genuine stretches will read as smaller than they are.
Both markets distort the same number. India distorts it downward by censoring ranges. The US distorts it upward with overnight gaps.
Carry this
- CCI measures distance from the average, in units of that instrument's own normal wandering.
- It is unbounded. There is no "too high". Fading it in a trend is the classic account-ending mistake.
- It largely repeats what Bollinger Bands and RSI already say. Check before you add it.