Direct Answer
The Chande Momentum Oscillator (CMO) is a momentum oscillator developed by Tushar Chande. It's structurally similar to the Relative Strength Index (RSI): both are built from the same raw inputs, the sum of an asset's positive price changes and the sum of its negative price changes over a lookback period, but CMO normalizes that relationship differently.
Key Takeaways
- CMO = 100 × (SumUp − SumDown) ÷ (SumUp + SumDown), where SumUp and SumDown are the sums of positive and negative price changes over a lookback period (commonly 14).
- Unlike RSI, which is bounded 0-100, CMO is bounded -100 to +100 and crosses through zero as net momentum flips between positive and negative.
- CMO is structurally similar to RSI, both start from the same up-move/down-move inputs, but the different normalization means the two lines can diverge even on identical price data.
- Like any momentum oscillator built from past price changes, CMO confirms a shift after it starts rather than predicting it, and can whipsaw around zero in a sideways market.
What Is the Chande Momentum Oscillator (CMO)?
The Chande Momentum Oscillator (CMO) is a momentum oscillator developed by Tushar Chande. It's structurally similar to the Relative Strength Index (RSI): both are built from the same raw inputs, the sum of an asset's positive price changes and the sum of its negative price changes over a lookback period, but CMO normalizes that relationship differently. Instead of RSI's average-based 0-to-100 scale, CMO expresses the difference between total up-moves and total down-moves as a percentage of their combined total, which produces a symmetric -100 to +100 scale that crosses through zero rather than through a midpoint like RSI's 50 level.
That zero crossing is the main structural difference worth remembering: a positive CMO reading means the sum of recent gains outweighs the sum of recent losses over the lookback window (net upward momentum), while a negative reading means the reverse. Because CMO uses the raw sum of price changes rather than an averaged calculation, it can react somewhat differently than RSI to the same sequence of price moves, even though both are commonly calculated over the same 14-period lookback.
The Formula
Over a lookback period (commonly 14 periods):
- SumUp = the sum of all positive price changes over the lookback period.
- SumDown = the sum of all negative price changes over the lookback period, taken as a positive number.
CMO = 100 × (SumUp − SumDown) ÷ (SumUp + SumDown)
The denominator, SumUp + SumDown, is the total absolute price movement over the lookback period regardless of direction. The numerator, SumUp − SumDown, is the net directional movement. Dividing net movement by total movement and multiplying by 100 produces a reading that's bounded between -100 (every price change in the lookback period was negative) and +100 (every price change was positive), with zero marking a period where positive and negative changes exactly offset.
Worked Example
Hypothetical example, for education only.
Suppose a stock is tracked over a 14-period lookback. Summing every positive daily price change over that window gives a SumUp of $18.40. Summing every negative daily price change over the same window, expressed as a positive number, gives a SumDown of $9.20.
CMO = 100 × (18.40 − 9.20) ÷ (18.40 + 9.20)
CMO = 100 × 9.20 ÷ 27.60 = 33.33
A reading of roughly +33 says that, summed across the 14-period window, gains outweighed losses by about a third of the total price movement, consistent with net positive momentum, though well short of the more extreme readings some traders treat as stretched. If SumDown had instead exceeded SumUp, the same formula would produce a negative value, with CMO crossing below zero as net momentum turned down.
How CMO Is Commonly Used
Zero-line crossovers
Because CMO is centered on zero rather than 50, a cross from negative to positive is commonly read as a shift toward net positive momentum, and a cross from positive to negative as a shift toward net negative momentum. As with any momentum indicator, a zero crossing confirms a change already reflected in recent price action rather than predicting one in advance.
Overbought and oversold zones
Some traders watch informal threshold levels, figures such as +50 and -50 are commonly cited, as zones where momentum may be historically stretched, similar in spirit to RSI's 70/30 convention. CMO has no single standardized threshold the way RSI does, so any fixed level should be checked against the specific instrument's own historical CMO range rather than assumed to transfer from another oscillator or another asset.
Divergence
As with RSI and other momentum oscillators, some traders watch for divergence, price making a new high or low that CMO doesn't confirm. A divergence flags a disagreement between price and the indicator's underlying momentum reading; it is not, by itself, a confirmed reversal.
Lookback Period Settings
| Lookback period | Responsiveness | Common use |
|---|---|---|
| 9 | Faster, more signals | Shorter-term/intraday trading |
| 14 | Balanced (most commonly cited default) | General-purpose swing/trend analysis |
| 20+ | Slower, smoother | Longer-term trend filtering |
14 periods is the figure most commonly cited for CMO, mirroring the conventional RSI lookback, but there is no single universally correct setting. A shorter lookback reacts faster to new price changes and produces more zero crossings and threshold touches; a longer lookback smooths the line and reduces whipsaws at the cost of reacting later. Verify the default your specific charting platform uses before comparing readings across sources.
Why CMO Produces False Signals
- Sideways or choppy markets, with no sustained directional move, the sums of up-moves and down-moves over a rolling window can stay close to each other, causing CMO to cross zero and any threshold level repeatedly with little follow-through.
- Lagging construction, like RSI and other lookback-based oscillators, CMO is calculated entirely from price changes that have already happened, so it confirms a momentum shift after it begins rather than forecasting it.
- Short lookback sensitivity, a shorter period reacts faster but can amplify normal short-term noise into signals that don't hold up over the following bars.
Common Mistakes
- Treating CMO's overbought/oversold levels as fixed like RSI's 70/30, CMO has no single standardized threshold; a level worth watching on one instrument may not carry over to another.
- Assuming CMO and RSI will always agree, because the two indicators normalize the same SumUp/SumDown inputs differently, they can diverge on identical price data.
- Trading every zero crossing as a standalone signal, a crossing confirms a shift that's already underway; most traders pair it with trend or price-structure context first.
- Comparing CMO readings across different lookback periods without noting the setting, a 9-period CMO and a 20-period CMO on the same chart can show meaningfully different values at the same moment.
Limitations
CMO is a lagging indicator: every input is a sum of already-realized price changes, so it confirms a momentum shift after it starts rather than predicting it in advance. Its bounded -100 to +100 scale makes extreme readings easy to spot, but an extreme or crossing reading describes recent price behavior, not a guarantee about what happens next. As with any single oscillator, CMO works best evaluated alongside trend context or other independent evidence rather than in isolation.
What You Get From an Oscillator That Does Not Smooth
The Chande Momentum Oscillator computes its ratio from raw up and down sums without the smoothing that similar oscillators apply. The result reacts faster and is noisier, and whether that is an improvement depends entirely on whether your method suffers more from delay or from false signals.
The practical consequence is that thresholds borrowed from smoother oscillators do not transfer. This indicator reaches extreme readings more frequently, so a level that identifies genuinely unusual conditions elsewhere identifies ordinary ones here. Any threshold needs to be established from the instrument's own readings rather than adopted.
The mistake is running it alongside a smoothed oscillator and treating agreement as confirmation. They measure the same underlying quantity with different filtering, so they agree by construction most of the time, and the disagreements are about smoothing rather than about the market.
The oscillator also shares the limitation of the whole momentum family: a reading can remain at an extreme for as long as the trend that produced it persists, and there is no level at which continuation stops being possible.
CMO FAQs
What is the Chande Momentum Oscillator (CMO)?
The CMO is a momentum oscillator developed by Tushar Chande that measures the difference between the sum of recent up moves and the sum of recent down moves over a lookback period, expressed as a percentage of their total. It ranges from -100 to +100 and crosses through zero as momentum shifts between net positive and net negative.
How is CMO different from RSI?
Both use the sum of up moves and down moves over a lookback period, but RSI normalizes that relationship onto a 0-100 scale using an average-based formula, while CMO divides the difference between SumUp and SumDown by their total, producing a -100 to +100 scale that crosses through zero. Because the two normalizations differ, CMO and RSI can diverge even when calculated from the same price series and lookback.
What is a good CMO period setting?
14 periods is the most commonly cited starting point, matching the conventional RSI lookback, but there is no universally best setting. A shorter period reacts faster and generates more signals; a longer period smooths the line and reduces whipsaws. Test nearby values against the instrument and timeframe being traded rather than assuming one setting transfers everywhere.
Does CMO have overbought and oversold levels?
Traders commonly watch levels such as +50 and -50 as informal overbought and oversold zones, but unlike RSI's widely cited 70/30 convention, CMO has no single standardized threshold. Any fixed level should be verified against the specific instrument's own historical CMO range rather than assumed from another oscillator.
Why does CMO produce false signals?
CMO is built entirely from a rolling window of past price changes, so it reacts after a move has already started rather than predicting it. In sideways or choppy markets it can cross zero and its own threshold levels repeatedly with no sustained follow-through, and a short lookback period can amplify that noise.
Can CMO be used with other indicators?
It can be paired with trend or volume evidence rather than used alone, similar to how RSI or MACD are commonly combined with other tools. Pairing indicators that measure genuinely different things, momentum plus trend direction, for example, tends to add more information than stacking two oscillators built from the same closing prices.
Why does the oscillator range from negative to positive rather than from zero upward?
It compares the sum of gains against the sum of losses as a difference divided by their total, which produces a symmetric scale centred on zero. This differs from oscillators built as a ratio, which are bounded between zero and one hundred with a midpoint at fifty. The symmetric scale makes the zero line a natural reference for whether gains or losses have dominated the lookback.
How does the oscillator behave in a market with almost no movement?
When both gains and losses are very small, the denominator shrinks and the reading becomes erratic, because small absolute changes produce large relative swings. This is a general property of ratio-based momentum measures and is most visible in quiet, low-volume periods. Filtering out readings taken when the underlying movement is negligible avoids acting on noise amplified by the formula.
Can the oscillator be used to measure trend strength rather than momentum direction?
Its absolute value has been used that way, on the reasoning that a reading far from zero in either direction indicates that one side has dominated the period. Some adaptive moving average designs use exactly this idea as an efficiency input. Used this way it is a measure of directional consistency rather than of magnitude, so a slow steady trend can register as strongly as a fast one.