Dynamic Momentum Index: A Volatility Adjusted RSI for Indian Traders

What the Dynamic Momentum Index actually does

The Dynamic Momentum Index (DMI), built by Tushar Chande and Stanley Kroll, fixes one weakness of the ordinary Relative Strength Index. A normal RSI always looks back the same number of days, usually 14. That fixed window is too slow when a stock turns volatile and too twitchy when it goes quiet. The DMI keeps the exact RSI maths but makes the lookback period breathe with volatility.

The logic is simple. When recent price swings are wider than usual, the DMI uses a shorter period so it reacts faster. When swings are narrower, it uses a longer period so it ignores small wiggles. The result still runs from 0 to 100 and still uses 70 and 30 as its lines, but is far less likely to lag a fast move or whipsaw you in a dull market.

This matters in Indian markets because volatility here is uneven. The same Nifty can crawl in a 60 point range for a week, then move 400 points on a budget day. A fixed 14 period RSI treats both regimes identically. The DMI does not, and that is the whole point.

The volatility adjusted RSI formula, step by step

Most pages describe the DMI in words and stop there. Here is the actual chain of calculations so you can reproduce it. There are three stages: measure volatility, convert it into an adaptive period, then run RSI on that period.

• Stage 1, raw volatility. Take the standard deviation of the closing price over the last 5 days. Call it StdDev5. This is your short term spread of prices.
• Stage 2, normalise it. Take the 10 day simple average of that StdDev5 value, call it ASD (average standard deviation). The Volatility Index is VI = StdDev5 divided by ASD. VI above 1 means today is more volatile than the recent norm, below 1 means calmer than the norm.
• Stage 3, adaptive period. Period = round(14 divided by VI). High VI makes the number smaller (faster RSI), low VI makes it larger (slower RSI). Clamp the result between 5 and 30 days so it never gets absurd.
• Stage 4, the RSI itself. Run a standard RSI over that adaptive period: average gain divided by average loss gives RS, then DMI = 100 minus (100 divided by (1 plus RS)).

A fully worked example on Reliance

Numbers below are illustrative and rounded for teaching. They are not a forecast and not a promise of any return. Verify live prices and contract specs on the NSE before trading. Suppose Reliance Industries posts these 14 most recent daily closes in rupees: 1402, 1410, 1396, 1421, 1438, 1455, 1447, 1462, 1450, 1471, 1489, 1478, 1495 and 1512. The stock has been climbing.

Step 1, short term volatility. The last 5 closes are 1471, 1489, 1478, 1495 and 1512. Their mean is 1489. The population standard deviation works out to about 14.21 rupees. So StdDev5 is roughly 14.2.

Step 2, the Volatility Index. Assume the 10 day average of that StdDev5 series (the ASD) is 9.0, meaning the stock has recently been calmer than today. Then VI = 14.21 divided by 9.0 = 1.58. Volatility is running well above its own recent average.

Step 3, the adaptive period. Period = round(14 divided by 1.58) = round(8.87) = 9 days. Because the stock got more volatile, the DMI automatically shrank the RSI window from 14 to 9. It will now react faster than a standard RSI.

Step 4, the RSI on 9 days. Over the last 9 price changes the up moves total 105 rupees and the down moves total 31 rupees. Average gain = 105 divided by 9 = 11.67. Average loss = 31 divided by 9 = 3.44. RS = 11.67 divided by 3.44 = 3.39. DMI = 100 minus (100 divided by 4.39) = 77.2. That is comfortably above 70, so the DMI is flagging Reliance as overbought.

The calm market case, for contrast

Now flip the regime. Suppose Reliance goes quiet and the 5 day standard deviation falls to 5.4 rupees while the 10 day ASD is still 9.0. Then VI = 5.4 divided by 9.0 = 0.60. The adaptive period becomes round(14 divided by 0.60) = round(23.3) = 23 days. The DMI has lengthened its window to filter out small noise.

This is the behaviour you want. In a dull market a 9 or 14 period RSI keeps tagging 70 and 30 on tiny moves. By stretching to 23 days the DMI demands a bigger, sustained move before it calls anything overbought or oversold. Same indicator, opposite reaction, driven purely by volatility.

Turning a DMI signal into a real Indian F&O trade

A reading is not a trade. Say the DMI at 77 convinces you Reliance is stretched, so you buy one lot of an at the money weekly put. Reliance F&O lot size is 500 shares (confirm the current lot on NSE, as exchanges revise lot sizes). You pay 22 rupees per share, so the outlay before costs is 22 times 500 = Rs 11,000.

Reliance pulls back over two sessions and the put premium rises to 34 rupees. You exit. Gross profit is (34 minus 22) times 500 = Rs 6,000. Your real take home is after costs, and on options those costs are small but not zero.

So a clean DMI overbought signal that played out gave roughly Rs 5,924 net on one Reliance put lot in this illustration. Charges ate only about 1.4 percent of the gross because option premiums are small. With Reliance futures the picture differs sharply, since STT and brokerage there are charged on the full contract value, not a tiny premium.

Tuning the DMI for Nifty, Bank Nifty and single stocks

The defaults (5 day standard deviation, 10 day average, base of 14, clamp 5 to 30) suit most large caps and the headline indices. But the instrument changes how aggressive you should be. Bank Nifty swings harder than Nifty, so its VI spikes higher and its period collapses to single digits more often, meaning more signals and more noise. Treat a Bank Nifty DMI of 75 with more caution than a Nifty DMI of 75.

• Nifty 50: default settings work well. The index is broad and mean reverts, so 70 and 30 lines are reliable. Lot size 65.
• Bank Nifty: consider widening the bands to 75 and 25 because high volatility keeps the period short and the oscillator pinned near extremes. Lot size 30.
• FinNifty and Sensex: sit between Nifty and Bank Nifty; default bands are fine. FinNifty lot 60, Sensex lot 10.
• Volatile single stocks (PSU banks, new listings): consider raising the lower clamp from 5 to 7 or 8 so the DMI does not turn hyperactive on every gap.
• Steady FMCG and IT large caps (HUL, TCS, Infosys): the DMI naturally sits on longer periods, so trust its overbought and oversold calls a little more.

Reading signals and avoiding the obvious traps

The standard signals are band crossings. A move up through 30 hints selling pressure is exhausting (a possible long), and a move down through 70 hints buying is exhausting (a possible short or exit). Because the DMI adapts, these crossings arrive earlier in volatile phases than a fixed RSI gives them, and cluster less in calm phases. Both are benefits.

The classic trap is treating an overbought reading as an automatic sell. In a strong trend the DMI can sit above 70 for many sessions while price keeps rising. Our Reliance DMI of 77 was actionable only because we had a reason to expect a pause. Shorting just because an oscillator is high is how traders get run over by momentum.

Combining the DMI with other tools

The DMI tells you about momentum exhaustion, not trend direction. That gap is best filled by a trend tool. Pairing the DMI with a moving average or with MACD lets you take signals only when they agree with the larger picture. If the DMI hits 70 while price is below its 50 day average, a short goes with the trend, which is far safer than fighting it.

• DMI plus moving average: only take oversold longs above the 50 EMA, only take overbought shorts below it.
• DMI plus volume: a DMI reversal backed by a volume spike beats one on thin volume.
• DMI plus support and resistance: an overbought DMI right at a known resistance is a stronger short than one in open space.
• DMI plus India VIX: when India VIX is elevated the period stays short and signals get noisier, so demand extra confirmation.

Taxes and rules that decide your real result

The DMI is a chart tool, so it is never taxed. The trade it triggers is. How that trade is taxed in India depends entirely on what you traded and how long you held it, and getting this wrong quietly eats a chunk of the profit the indicator helped you earn.

• F&O (futures and options): profit is treated as business income and taxed at your income tax slab. Our Rs 5,924 Reliance put gain adds to business income, not to capital gains. You can offset it against other business expenses and losses.
• Intraday equity: speculative business income, also taxed at slab rates.
• Delivery, held under 12 months: short term capital gains at 20 percent (the rate effective from 23 July 2024) plus cess.
• Delivery, held over 12 months: long term capital gains at 12.5 percent on gains above Rs 1.25 lakh in the year, plus cess.
• STT always applies and is charged automatically by the broker. On options it is 0.15 percent of the sell side premium, which is why our charges stayed tiny.

Where the DMI fails

No adaptive indicator removes risk. In a strong trend the DMI repeatedly calls overbought or oversold while price ignores it, because momentum can stay extreme far longer than an oscillator suggests. In a thinly traded counter the 5 day standard deviation can jump on one odd print, distorting the VI. And around results, budget or RBI policy days, the volatility spike can collapse the period so far that the DMI fires constantly with little meaning.

The defence is ordinary discipline. Use the DMI as one input, confirm with trend and structure, size positions so a single wrong call is survivable, and always trade with a defined stop. It is a sharper RSI, not a crystal ball.

Sources and further reading

For authoritative data and further reading, refer to Zerodha Varsity, Investopedia and NSE India. Always confirm current rules, rates and contract specifications on the official source before you trade.

Sources and Further Reading

For authoritative data and further reading on this topic, refer to Zerodha Varsity, Investopedia and NSE India. Always confirm current rules, rates and contract specifications on the official source before you trade.