Definition and stable mechanics of CMO
CMO usually refers to the Chande Momentum Oscillator (CMO), a momentum-based oscillator that compares upward and downward price changes over a specified lookback window. The key idea is stable: over a period, the oscillator measures the balance between total gains and total losses derived from price changes, then scales that balance to an oscillator range.
To verify information about CMO, start with the definition that is not tied to any broker, platform, or live market feed: what it measures, which price change series it uses (e.g., close-to-close differences), the lookback length, and how gains and losses are aggregated.
How CMO works: inputs, assumptions, and a reproducible check
A verification-friendly way to understand CMO is to treat it as a deterministic calculation once you choose inputs.
Step 1: Fix the calculation assumptions Write down, before calculating:
- the price source (for example, closing prices)
- the step used to compute changes (for example, difference between consecutive closes)
- the lookback length (often denoted as n)
- how you treat zero changes (they typically contribute neither gain nor loss)
Step 2: Compute gains and losses inside the window For each bar in the lookback window, compute the change from the previous price.
- If the change is positive, add it to total gains.
- If the change is negative, add the absolute value of that negative change to total losses.
- If the change is zero, add nothing to either side.
Step 3: Compute the oscillator from the totals The CMO value is then determined from the difference between total gains and total losses, divided by their sum (with a scaling factor). Even if different resources describe the oscillator in slightly different words, a correct source hierarchy should match the same mechanics: same totals, same scaling, same handling of edge cases.
Step 4: Recompute with the same data and compare Take a small, manually chosen historical slice (for example, 30 consecutive bars). Calculate CMO using the written assumptions and confirm whether a second calculator or platform implementation matches. A mismatch often indicates a difference in one of the assumptions (price field, lookback length definition, or rounding).
Evidence and example: what you can independently verify
Because CMO is formula-driven, verification does not require real-time data.
Reproducible verification tasks
- Definition check: Confirm that each source describes CMO as a momentum oscillator based on the imbalance of gains versus losses over a period.
- Formula check: Confirm the calculation uses (gains − losses) over (gains + losses), with consistent scaling.
- Implementation check: Confirm details such as the price field (close vs. typical price), window boundary (how many bars), and treatment of zero changes.
- Edge-case check: Test a window where total gains + total losses is very small. Different implementations may produce different values or handle division-by-zero differently.
If you are comparing two sources that disagree, treat the difference as evidence that at least one source has a different assumption. The most reliable resolution is to match both sources to your written assumptions and recompute.
Limitations, risks, and failure modes to verify
Verification includes checking where the concept can break down in practice.
Material limitations
- Dependence on the input series: CMO uses price changes; changing the chosen price field or bar calculation method can change results.
- Sensitivity to costs and execution is not built in: Even though CMO is computed from prices, any real trading use would be affected by costs, execution, and jurisdiction rules. Those factors are external to the indicator calculation.
- Noise and flat ranges: When gains and losses are both small or cancel out, the oscillator can be unstable, and small input differences can create larger relative changes.
- Historical relationships do not guarantee future behavior: A correlation between CMO readings and outcomes in past data does not establish predictable future results.
Verification or next question: what to do when information is conflicting
When you see a claim about CMO (for example, its range, formula, or calculation steps), verify it in this order:
- Mechanics first: Does the source match the same gain/loss accumulation and the same normalization idea? 2. Assumptions second: Does it specify which price series and window length are used? 3.