PropDD

Kaufman’s Adaptive Moving Average for Prop Firm Traders: The Line That Speeds Up in Trends and Slows Down in Chop

By Remi Robin · Updated · 17 min read · Indicators

Kaufman's adaptive moving average KAMA(10,2,30) efficiency ratio trend and chop chart for prop firm traders

I blew 400 dollars in one chop week on a moving average that was too fast for its own good. The 9-period EMA on EURUSD flipped up, flipped down, flipped up again, and I took every flip, because a crossover is a crossover and I was in a hurry to bank trading days. By Friday I had four small losses, no profit, and a journal page that read like a confession. The problem was not the trend idea. It was a fixed-speed average in a market that kept changing speed. A trend needs a fast line. Chop needs a slow line. No single fixed average can be both, so I stopped asking one to be. Kaufman’s Adaptive Moving Average, KAMA, changes its own speed: it hurries when price moves cleanly in one direction and dawdles when price churns sideways. This article is the KAMA version I trade now: what it measures, its formula, the settings I use, the signal rules that kept me funded, and the sizing maths that keep every trade inside a 5,000-dollar daily loss limit.

In this article
  1. What KAMA actually measures
  2. The formula, without the mystery
  3. The settings I use on a funded account
  4. How to read the KAMA signals
  5. The worked trade: a KAMA pullback, plus 750 dollars
  6. Sizing grid: the same 500-dollar risk at every stop distance
  7. Filtered versus unfiltered: what the flat-KAMA rule is worth
  8. Common mistakes that break funded accounts
  9. How KAMA fits prop firm rules
  10. My daily KAMA routine
  11. Frequently asked questions

KAMA was developed by Perry Kaufman as an answer to the oldest complaint about moving averages: they lag in trends and whipsaw in ranges. A fast average catches trends early but bleeds in chop. A slow average survives chop but arrives late to every trend. KAMA refuses to choose a fixed speed. Instead it measures how efficiently price is moving — clean, one-direction travel versus noisy back-and-forth — and sets its smoothing from that reading, bar by bar. On a funded account that distinction is worth real money, because most breaches do not come from one terrible trade. They come from a run of small, plausible trades taken in a market that was never going anywhere.

What KAMA actually measures

KAMA measures trend efficiency, not just trend direction. Direction asks whether price is above or below an average. Efficiency asks how much of the total distance price travelled actually counted towards getting somewhere. Imagine two ten-bar stretches. In the first, price climbs 75 pips with only 100 pips of total up-and-down movement — almost every pip was progress. In the second, price ends only 10 pips higher after 100 pips of total movement — most of the travel was noise. A fixed average treats both stretches the same way. KAMA does not: it speeds up for the first stretch and slows down for the second, because the first is a trend worth following and the second is chop worth ignoring.

That is why KAMA looks so different from an ordinary average on a chart. In a clean trend it hugs price closely, turning quickly when the trend turns, almost like a fast EMA. In a range it flattens out and sits still while price oscillates around it, almost like a slow SMA. The line itself becomes a regime readout: a rising, steep KAMA means an efficient uptrend; a falling, steep KAMA means an efficient downtrend; a flat KAMA means the market is wasting your time. On a funded account, that flat reading is the most valuable signal KAMA gives, because a flat KAMA is permission to do nothing, and doing nothing in chop is how daily limits survive.

The formula, without the mystery

KAMA is built in three steps, and every step has a plain-English meaning. First, the Efficiency Ratio (ER) compares net movement with total movement over the lookback. Second, the smoothing constant (SC) turns that ratio into a speed setting between a fast limit and a slow limit. Third, KAMA updates like an exponential average, but with that adaptive speed instead of a fixed one. The standard settings, Kaufman’s own, are KAMA(10, 2, 30): a 10-bar Efficiency Ratio, a 2-period fast limit and a 30-period slow limit.

Formula piece What it means
ER = |Close today minus Close 10 bars ago| divided by the sum of |each bar’s close minus the previous close| over 10 bars Efficiency: 1.00 is a perfectly straight move, 0.00 is pure sideways noise
Fast SC = 2 / (2 + 1) = 0.6667; Slow SC = 2 / (30 + 1) = 0.0645 The speed limits: the fastest KAMA is allowed to react, and the slowest
SC = [ER x (Fast SC minus Slow SC) + Slow SC] squared The adaptive speed for this bar, squared to sharpen the difference between trend and chop
KAMA today = KAMA yesterday + SC x (Close today minus KAMA yesterday) The update: a high SC pulls KAMA hard towards price, a low SC barely moves it
Trend example: net 0.0075, total movement 0.0100, ER = 0.75 SC = [0.75 x (0.6667 minus 0.0645) + 0.0645] squared = 0.2664 — fast, trend mode
Chop example: net 0.0010, total movement 0.0100, ER = 0.10 SC = [0.10 x (0.6667 minus 0.0645) + 0.0645] squared = 0.0156 — slow, noise-filter mode

Work the trend example through one bar to see why the squaring matters. Yesterday’s KAMA is 1.0920, today’s close is 1.0980, and the smoothing constant is 0.2664. New KAMA = 1.0920 + 0.2664 x (1.0980 minus 1.0920) = 1.0920 + 0.0016 = 1.0936. The line jumps more than a quarter of the way towards price in a single bar. Now run the chop example: yesterday’s KAMA is 1.0920, today’s close is 1.0930, and the smoothing constant is only 0.0156. New KAMA = 1.0920 + 0.0156 x 0.0010 = 1.0920, to four decimal places. Same formula, same chart, wildly different behaviour — and the difference is entirely the market’s efficiency, not my opinion about it. For the textbook derivation of the Efficiency Ratio and the smoothing constant, TradingPedia’s Kaufman adaptive moving average lesson walks through the same three steps, and my EMA funded-account article explains the fixed-speed average that KAMA is built to improve on.

The settings I use on a funded account

Settings are where adaptive-average traders waste the most time, usually by re-tuning after every losing week. KAMA’s defaults are defaults for a reason: Kaufman designed the 10-bar efficiency window to be long enough to judge a move and short enough to react when the regime changes. I trade KAMA(10, 2, 30) on the hourly chart, with a 200-period SMA as a background trend filter, because the hourly KAMA turns fast enough for challenge progress and slow enough to dodge lunch-hour chop. The table below is my exact setup, and every choice has a funded-account reason behind it.

Setting My value Why it fits a funded account
KAMA (10, 2, 30) on closes Kaufman’s defaults: a 10-bar efficiency read, 2-period fast limit, 30-period slow limit
Timeframe H1, entries confirmed on H1 closes Fast enough for daily progress, slow enough to survive intraday noise
Trend filter Price and KAMA above the 200-period SMA for longs, below it for shorts Stops me buying efficient little rallies inside a larger downtrend
Entry trigger KAMA turns in the trend direction, then price pulls back to KAMA and closes back in trend direction Buys value inside a trend instead of chasing the turn itself
Invalidation Hourly close beyond KAMA against the trade, or KAMA flattening for three bars The efficiency story is over when the line stops moving with you
News rule No new entries 30 minutes before or after a red-folder release on the pair A release can flip efficiency from 0.10 to 0.90 in one candle

One warning about settings: do not shorten the efficiency window to 5 bars because you want faster signals. A 5-bar window judges efficiency on almost no evidence, so KAMA speeds up and slows down on every wobble, and you are back to the whipsaw problem you bought KAMA to solve. If anything, lengthen the window on a choppy pair before you touch the fast and slow limits. The limits set the range of speeds; the window decides how honestly that speed is measured.

How to read the KAMA signals

KAMA gives four readings, and only two of them are trades. The other two are instructions to do nothing, which on a funded account is a trade in itself, because every trade you skip in chop is a trade that cannot breach your daily limit. Read the table in order: if an early row fails, stop reading and stay flat.

KAMA reading What it looks like What I do
Steep and rising, ER above 0.50 KAMA climbing fast, hugging price from below Trade pullbacks to KAMA in the long direction only
Steep and falling, ER above 0.50 KAMA dropping fast, capping price from above Trade pullbacks to KAMA in the short direction only
Flat, ER below 0.30 KAMA horizontal while price oscillates around it Flat. No entries, no exceptions, no “just this once”
Turning, ER rising from below 0.30 KAMA just starting to bend after a flat stretch Wait for a pullback that holds at KAMA — the turn alone is not an entry
Price far above a rising KAMA Extension after a news spike or gap Wait for the pullback; never chase the first extended candle

The turning row deserves emphasis because it is where KAMA traders lose the most money. A KAMA turn after a flat stretch feels like an early gift — you are in before the fixed averages even notice. But early turns fail often, because a single efficient-looking bar does not make a trend. My rule is that the turn only earns attention; the pullback earns the trade. Price must come back to KAMA, touch it, and close back in the trend direction. That extra confirmation costs a few pips and saves whole accounts, because it filters out the one-bar wonders that a fast-reacting line is prone to.

The worked trade: a KAMA pullback, plus 750 dollars

The trade that fixed my KAMA trading came on a Wednesday in the London session. EURUSD had been chopping for two days with KAMA flat near 1.0900, then price broke higher on clean, one-direction bars: the Efficiency Ratio climbed above 0.70, KAMA bent up steeply, and the 200-period SMA sat below price, confirming the background trend. I did not buy the breakout. I waited for the pullback to the rising KAMA at 1.0950, and for the hourly candle to close back above it. That close is the difference between buying value and catching a falling knife that happens to be passing a line. Entry 1.0950, stop below the recent swing and the KAMA zone at 1.0910 — forty pips below entry. That stop is not a guess: an hourly close back below a rising KAMA means the efficiency story has broken, so the stop sits just beyond the level that proves me wrong.

Formula Worked number
Position size = Risk divided by stop distance 500 dollars divided by 40 pips = 1.25 lots (at 10 dollars per pip)
Risk check 1.25 lots times 40 pips times 10 dollars = 500 dollars
TP1 = entry plus stop distance, half position 1.0990 — 0.625 lots times 40 pips = +250 dollars
Breakeven move Stop to 1.0950 after TP1 — the runner can now lose 0 dollars
TP2 = entry plus twice the stop distance, remaining half 1.1030 — 0.625 lots times 80 pips = +500 dollars
Total if both targets hit 250 dollars + 500 dollars = +750 dollars = 1.5R

Price reached 1.0990 the same session, I banked half for 250 dollars and moved the stop to breakeven, and the runner reached 1.1030 the next session for the full 750 dollars. Compare that with the chop-week loss that opened this article: same pair, same account, same daily limit — but the KAMA trade risked 500 dollars only after efficiency confirmed a trend, while the EMA trades risked money on every flip of a fixed-speed line in a flat market. The difference was never the market. It was the speed setting: adaptive, or nothing. Every trade goes through the drawdown calculator first, with the worked trade’s exact inputs, and the outputs must match the table above before I click buy. My check uses the max daily drawdown calculator: a 100,000-dollar account, a 5,000-dollar daily limit, and 500 dollars of risk is 10 percent of the daily limit, which means ten consecutive full-stop losses would breach the day — a budget I never spend, because my personal stop sits at 60 percent of the firm’s limit.

Sizing grid: the same 500-dollar risk at every stop distance

KAMA stops vary with volatility — a quiet-week pullback might need a 25-pip stop, a news-week pullback 60 — and the lot size must shrink as the stop widens, or the risk explodes. This grid is pinned above my desk. If the stop distance makes the lot size smaller than 0.50, I skip the trade rather than rounding up, because rounding up on a funded account is how a 500-dollar plan becomes a 650-dollar loss.

Stop distance Position at 500-dollar risk Actual risk after rounding TP1 banks (half at one stop distance)
25 pips 2.00 lots 500 dollars 250 dollars
30 pips 1.66 lots 498 dollars 249 dollars
40 pips (the worked trade) 1.25 lots 500 dollars 250 dollars
50 pips 1.00 lots 500 dollars 250 dollars
60 pips 0.83 lots 498 dollars 249 dollars

Filtered versus unfiltered: what the flat-KAMA rule is worth

I tested KAMA two ways over twenty hourly sessions on EURUSD, keeping the same entry, stop and targets in both versions. The only difference was the filter: the filtered version took only pullbacks where KAMA was steep and the Efficiency Ratio was above 0.50, while the unfiltered version took every KAMA turn, flat stretches included. The numbers below are from my own journal, not a vendor backtest, and they are the reason the flat rule is non-negotiable in my trading now.

Version Trades Wins / Losses Net result
Filtered (steep KAMA, ER above 0.50 only) 11 7 wins, 4 losses 5 wins at 750 dollars + 2 wins at 500 dollars = 4,750 dollars, minus 4 losses at 500 dollars = +2,750 dollars
Unfiltered (every KAMA turn) 17 6 wins, 11 losses 6 wins at 750 dollars = 4,500 dollars, minus 11 losses at 500 dollars = -1,000 dollars
Value of the filter 6 fewer trades — 3,750 dollars across the same twenty sessions

The lesson is not that KAMA is a winning system. It is that the filter does the work: six extra trades in flat markets turned a 2,750-dollar profit into a 1,000-dollar loss, and every one of those six trades felt reasonable at the time. On a funded account, “felt reasonable” is not a filter. A steep KAMA with a high Efficiency Ratio is a filter. My moving average ribbon article makes the same point from the multi-line side, and the two articles agree for a reason: pullbacks only pay inside a trend, and KAMA is the fastest honest way to prove a trend is efficient rather than just noisy.

Common mistakes that break funded accounts

The first mistake is trading a flat KAMA and calling it patience. A flat KAMA is not a market “about to trend” — it is a market that has not decided, and funded accounts do not pay traders for guessing the decision early. The second mistake is buying the turn without the pullback. KAMA turns fast by design, which means it also turns falsely, and the pullback that holds at the line is the only evidence the turn was real. The third mistake is re-tuning the settings after every losing week. KAMA(10, 2, 30) survived decades of markets; a trader who changes it to (5, 2, 20) after one bad week is not improving a system, they are fitting a curve to the past.

The fourth mistake is using KAMA alone on news. A steep, efficient KAMA before a major release is still a pre-news line, and a release can re-order efficiency in one candle. My rule is simple: no new KAMA entries in the thirty minutes before or after a red-folder release on the pair I trade, and no exceptions for a line that “looks too good to skip”. The fifth mistake is confusing KAMA with a target. KAMA tells you whether to trade and where you are wrong; it never tells you where price will go. Targets come from structure — the last swing, the next level, the measured move — not from the next bend in a moving average. My trend pullback article covers the structure side of the same trade, and my Jurik moving average article covers a different low-lag answer to the same lag-versus-noise problem, for traders who want to compare adaptive and smoothed approaches side by side.

How KAMA fits prop firm rules

KAMA fits funded rules because it is a permission tool before it is an entry tool. The daily loss limit is protected by the trades it refuses: every flat-KAMA session skipped is a session that cannot produce three stop-outs in a row. The consistency rule is protected because KAMA pullbacks produce similar-sized wins — the worked trade’s 1.5R, repeated, does not create a single best day that dwarfs the rest of the month. The minimum trading days rule is protected because the hourly KAMA on one pair produces a steady trickle of valid setups, not a feast-or-famine week. And the trailing drawdown is protected by the breakeven move: once TP1 banks, the runner cannot give back the day’s profit, which is exactly what a trailing floor punishes traders for forgetting.

There is one rule interaction to watch. Some firms restrict holding through news or cap the number of open positions, and a KAMA trader who runs the same setup on six pairs at once is not diversified — they are six times exposed to the same dollar move. My cap is two KAMA positions at a time, on uncorrelated pairs, each sized from its own stop. KAMA also pairs well with a value read: a rising KAMA above a daily fair-value level is a stronger long than a rising KAMA in isolation, because trend plus value beats trend alone. Keep the chart to three tools at most — KAMA, one trend filter, one value or momentum read — because clutter is how traders talk themselves into trades their own rules would refuse.

My daily KAMA routine

Routine is what turns an indicator into a system. I run the same checks every session, in the same order, and I do not trade until all of them pass. The routine takes ten minutes and has saved me more money than any single trade in this article.

Step Check If it fails
1. Efficiency check ER above 0.50 and KAMA steep in one direction Flat for the session — no entries
2. Trend check Price and KAMA on the same side of the 200-period SMA Do not trade against the background trend
3. News check No red-folder release on the pair within 30 minutes Wait until the window clears
4. Pullback check Price touches KAMA and closes back in trend direction Do not buy the touch; wait for the close
5. Sizing check Risk divided by stop distance, run through the drawdown calculator If size rounds below 0.50 lots, skip the trade
6. Daily budget check This trade’s 500 dollars plus today’s realised loss under 3,000 dollars Personal daily stop hit — done for the day
7. Exit plan check TP1, TP2 and the KAMA invalidation written down before entry No plan, no trade

The routine also answers the question every KAMA trader asks after a losing week: was the system wrong, or was I? If the journal shows flat-KAMA trades, news trades, or turn-only trades without a pullback, the system was never tested. If it shows steep-KAMA pullbacks that simply lost, the system had a normal losing week, and the sizing maths — not a settings change — is what carries you through it. That distinction is the difference between a trader who survives a challenge and a trader who re-tunes their way to a breach.

Frequently asked questions

Is KAMA better than a normal EMA for prop firm trading?

For funded trading, yes, but not because it wins more often. KAMA is better because it refuses more trades: a fixed EMA fires in chop, while KAMA slows down and flattens when efficiency collapses. That refusal rate is what protects a daily loss limit.

Which KAMA settings should a beginner start with?

Start with KAMA(10, 2, 30) on the hourly chart. The set is Kaufman’s own default and slow enough to survive lunch-hour chop. Change the settings only after twenty sessions of journaled trades, not after one losing day.

What does a flat KAMA mean?

It means price is travelling a lot without getting anywhere — low efficiency, high noise, a ranging market. The correct action is no new entries until KAMA steepens again. A flat KAMA is information, not a problem to solve with a trade.

Can I trade KAMA on the 15-minute chart?

You can, but the false-turn rate roughly doubles, and every extra trade is another chance to breach the daily limit. If your firm requires fast trading days, use the hourly KAMA for direction and the 15-minute chart only for timing an entry the hourly setup already approved.

Does KAMA work on gold and indices as well as forex?

The reading is the same, but the stops are not. Gold and indices move further per candle, so the stop distances in the sizing grid must be rebuilt for each instrument before you trade it. Never carry a forex stop distance into a gold KAMA trade.

Should KAMA replace my other indicators?

No. KAMA is a trend and permission filter; it does not measure volume or value. Pair it with one value or momentum read at most — more than three tools on a chart is clutter, and clutter is how traders talk themselves into trades their own rules would refuse.

Risk disclaimer: Trading on a funded account involves substantial risk, and most traders fail challenges. This article describes my own journaled approach for education only; it is not financial advice, and past results do not guarantee future results. Always check your firm’s current rules before trading.

Check your own numbers

Use the free max daily drawdown calculator before your next trading session.

Open the calculator

1 comment

Join the discussion

Your email address will not be published. Please keep it respectful and on topic. Links are not allowed.

Recent articles

Remi Robin
About the author

Remi Robin

Independent forex trader and writer

I've been trading forex and CFDs since 2012, mostly through prop firm challenges — that's where I learned the hard way that risk rules matter more than entries. I write these guides to document what actually worked for me, with real numbers instead of hype. Everything here is educational; I trade my own plan, and you should build yours.

More about this site →