forex-scalping.com

The Chande momentum oscillator on M1

Tushar Chande's momentum oscillator, CMO, is close to RSI but with a key difference in the calculation. RSI uses smoothed averages of gains and losses. CMO uses the raw sums over the period: the sum of up moves minus the sum of down moves, divided by the total of both. The result runs from minus 100 to plus 100.

Because there's no smoothing, CMO reacts faster than RSI and moves to extremes more readily. That makes it noisy on most timeframes, but on M1 with a sensible period it gives a clear read of very short term momentum.

With a period of 9 on M1, CMO shows whether the last nine minutes have been mostly up, mostly down or balanced. A reading above plus 50 means the up minutes have greatly outweighed the down minutes. Below minus 50, the opposite.

I don't use the overbought and oversold levels for reversal trades on M1. They're hit too often. Instead I watch for zero line behaviour during pullbacks. In an uptrend on M5, a pullback on M1 that brings CMO down near zero and then sees it turn up again often marks the end of the pullback. If CMO drops well below zero instead, the pullback has more behind it and I wait.

It's a timing tool for entries within a trend I've already identified on a higher timeframe.

One practical setup is CMO on M1 with a period of 9, a 50 EMA on M5 for direction, and nothing else. When price is above the M5 average, I only look at CMO turns up from near zero. Below it, only turns down from near zero. It keeps the chart clean, and every entry has a reason that can be written in the journal in one line.

CMO reacting faster than RSI is useful on M1, as long as the thresholds are set wider. Plus or minus 50 is a common choice, but on very short periods the line reaches those levels often. Looking at how long CMO stays beyond the threshold, rather than the moment it crosses, filters out short spikes. If the reading stays above 50 for several bars, the momentum is real. A single bar above 50 followed by a drop back is usually just a tick burst.

Because CMO is unsmoothed, a single large candle can swing it a long way. On M1, a news spike can push it to an extreme in one bar and it then stays there until that bar drops out of the period. That matters around releases. Either leave it alone during news or accept that its readings will be distorted for the next several minutes, until the spike bar has rolled out of the lookback. Practical example: skip the CMO for a few minutes after a high impact release, or until the news candle is outside the lookback, whichever comes later. Then use it normally once it reflects ordinary trading again.

Some traders add a moving average of CMO as a signal line, like MACD's signal line. A cross of CMO over its average gives a slightly earlier signal than a zero cross. On M1 this produces a lot of signals. It might suit people who scalp very short moves, but for pullback entries I find the zero line simpler and less likely to trigger on noise. If you want a signal line, a short exponential average of three or four periods works better on M1 than a longer one. Longer averages lag too much for one minute charts.

Chande also developed a variable moving average, VIDYA, that uses CMO to adjust its speed. When CMO is strongly positive or negative, the average speeds up. When it's near zero, it slows down. That's an interesting idea for scalpers who want a moving average that follows price closely in trends and flattens in ranges. It's available on some platforms as a custom indicator. On MT4 and MT5, VIDYA is available in some indicator collections, and the code is short enough to write yourself. It's not a magic tool, but it handles the shift between trending and ranging conditions more gracefully than a standard EMA.

The original period suggested by Chande was 20, and many platforms default to that. On M1, a period of 20 covers twenty minutes, which makes it slower and closer to RSI in behaviour. Shorter periods like 9 or 10 keep it quick. I'd test both on your pair and see which one matches the length of the pullbacks you usually trade. If your typical pullback lasts about five minutes, a period near 9 captures it. If they last longer, a period closer to 20 might suit better. Matching the period to the swings you trade is more useful than any default setting.