I learned what the Jurik Moving Average really is on a funded account from a week where my EMA kept telling me the truth too late. It was a choppy EURUSD week, price kept spiking above the 21 EMA and falling back through it, and I took five of those crosses. Three stopped out, two barely broke even, and the week cost me 1,400 dollars — not because the trend read was wrong, but because by the time a normal moving average confirmed the move, the move was already half over and my stop was sitting exactly where the noise lived. A friend who trades a funded account sent me his chart that Friday: one smooth line, no jagged steps, hugging price through the same chop without flipping every hour. That line was the Jurik Moving Average, JMA for short, and it changed how I filter trades — not because it predicts anything, but because it answers the one question a funded trader actually needs answered: is this a real turn, or just noise that will eat my daily limit?
In this article
- What the Jurik Moving Average actually is
- Why the formula is proprietary — and what public versions do instead
- What JMA measures that EMA and SMA miss
- Funded settings: length, phase and the pairs they suit
- How to read its signals without chasing
- A worked trade, step by step
- The sizing grid: what each stop distance allows
- I tested it: twelve filtered trades against fifteen unfiltered
- Mistakes that cost funded traders their JMA accounts
- Frequently asked questions
This guide is written for a funded trader, not a retail hobbyist. It covers what JMA measures, why its exact formula is proprietary and what public versions do instead, the settings that work under a daily loss limit, how to read its signals without chasing, a worked trade with every number shown, the sizing grid for every stop distance you are likely to use, and a twelve-trade test of one simple filter. The account in every example is a 100,000-dollar funded account with a 5,000-dollar daily loss limit and a 500-dollar risk per trade, which is 0.5 percent of the account. If your account is a different size, the percentages still hold — divide your own daily limit by ten and you have your per-trade risk.
What the Jurik Moving Average actually is
The Jurik Moving Average is an adaptive moving average developed by Mark Jurik of Jurik Research in the 1990s. Jurik came from signal processing rather than charting, and JMA was built the way a navigation system tracks a moving target: follow genuine changes quickly, ignore noise, and never jerk around. In practice that means JMA is smoother than an EMA of the same responsiveness and faster than an SMA of the same smoothness. It is the rare indicator that tries to be both at once, and on a funded account that combination matters, because lag costs you entries and whipsaw costs you stops — and both costs come out of the same 5,000-dollar daily limit.
What JMA is not is a signal generator on its own. It does not tell you to buy or sell. It tells you what the underlying trend looks like once the market’s random jitter has been filtered out, and it tells you sooner than a conventional average does. The trading decisions — where to enter, where to put the stop, how much to risk — still come from price structure and your rules. JMA just stops you from making those decisions on a line that is still catching up with yesterday.
Why the formula is proprietary — and what public versions do instead
Here is the honest part, and it matters for a funded trader: the exact JMA algorithm has never been published. Jurik Research sells it commercially, and the versions you find on public charting platforms are reverse-engineered approximations built to match its documented behaviour. Two JMA scripts on the same chart can print slightly different values on the same candles, and neither is “wrong” — they are different reconstructions of a closed formula.
That sounds like a reason not to use it, but it is actually a reason to use it correctly. You should never trade JMA values to the fourth decimal place, never build a rule that depends on JMA being exactly 1.0943 rather than 1.0941, and never compare your JMA reading with another trader’s and argue about a two-pip difference. What the public versions agree on — and what you can rely on — is the behaviour: JMA tracks price with very low lag, stays smooth through chop, and turns decisively when the underlying move genuinely changes. The LuxAlgo JMA library page explains that documented behaviour well, including why public builds are approximations of a proprietary formula. The table below shows how the public approximations are built, in plain language.
| Stage | What it does | Why it matters on a funded account |
|---|---|---|
| Volatility measure | Measures how noisy recent price has been, bar by bar | Sets how much smoothing the line applies right now |
| Adaptive factor | Speeds the line up when price moves decisively, slows it in chop | Less lag in real moves, fewer flips in noise |
| Multi-stage smoothing | Passes price through several smoothing stages instead of one | Removes jagged steps without adding the lag of a long average |
| Final JMA line | The smoothed, low-lag line plotted on the chart | The line you read for slope, pullbacks and cross-backs |
The formula box below is the honest version of the maths. It is not Jurik’s proprietary code — nobody outside Jurik Research has that — it is the structure every public approximation follows, and it is enough to understand what your platform is doing.
| Formula box: how a public JMA approximation works |
|---|
| Step 1: measure recent volatility from the range of recent bars. Step 2: convert volatility into an adaptive factor — high when price moves cleanly, low when it chops. Step 3: smooth price in stages, each stage applying that factor, so noise is filtered several times over. Step 4: plot the final stage as the JMA line. A rising JMA means the filtered trend is up; a flat JMA means the market is undecided. Rule for funded trading: read the slope and the distance of price from the line, never the fourth decimal place — two JMA builds can differ by a pip or two and both be right. |
To see the behaviour in numbers, take five EURUSD closes: 1.0900, 1.0920, 1.0910, 1.0940 and 1.0960. An illustrative public JMA build prints roughly 1.0900, 1.0912, 1.0914, 1.0928 and 1.0945 across those bars — it follows the rise, dips barely on the 1.0910 pullback, and ends close to price without ever jumping. A 20-period EMA on the same closes ends near 1.0915, a full 45 pips behind price, because it is still averaging in older, lower closes. That gap is the whole point: on a funded account, 45 pips of lag is the difference between a stop that sits under structure and a stop that sits in the middle of the noise. These JMA values are illustrative — your platform’s build will differ slightly, and that is expected.
What JMA measures that EMA and SMA miss
Every moving average answers the same question — where is the middle of recent price? — but they answer it with different trade-offs. The table below is the comparison that matters for a funded trader, because each row is really a statement about where your stop will sit and how often it will get hit by nothing.
| Average | Lag | Smoothness in chop | Funded-account weakness |
|---|---|---|---|
| SMA | High — every old bar counts equally | Good, but slow to turn | Entries come late; stops sit far from price |
| EMA | Medium — recent bars weighted more | Poor — flips on spikes | Whipsaw crosses eat the daily limit |
| JMA | Low — adapts to volatility | Very good — filters noise first | Values differ slightly between platforms |
| KAMA | Low in trends, slow in chop | Good | Can freeze during long sideways markets |
JMA is closest in spirit to the Kaufman Adaptive Moving Average, which also speeds up and slows down with market efficiency — but KAMA adapts on a single efficiency ratio, while JMA filters through several stages, which is why it stays smoother when KAMA starts to step. It is also worth distinguishing from the EMA guide on this site: EMA is the baseline every funded trader should understand first, and JMA is what you graduate to when EMA whipsaw starts costing you real money. And unlike the Zero Lag EMA, which removes lag by adding a correction term that can overshoot, JMA removes lag by filtering noise before it averages — so it does not spike past price on a fast bar.
Funded settings: length, phase and the pairs they suit
JMA has two inputs that matter: length and phase. Length is roughly how many bars the average looks back over; phase controls how responsive versus how smooth it is. Most public builds default to length 7 and phase 0, and those defaults are a reasonable starting point — but a funded trader should tune them to the job, not accept them blindly.
| Setting | Funded value | Why |
|---|---|---|
| Length | 7 to 14 on H1 and H4 | Short enough to track, long enough to filter hourly noise |
| Phase | -100 to 0 for entries, 0 to 100 for filters | Negative phase reacts faster; positive phase smooths harder |
| Timeframe | H1 for entries, H4 for bias | H1 gives tradeable pullbacks; H4 keeps you out of counter-trend chop |
| Pairs | EURUSD, GBPUSD, EURJPY | Liquid majors where JMA’s smoothing is not distorted by gaps |
My funded setup is JMA length 10, phase 0 on H1, with an H4 JMA for bias. I do not trade a JMA signal on H1 that fights the H4 JMA slope — that one rule alone removed most of my losing crosses. I also do not use JMA on thin, gappy instruments or on M1 scalping charts, where the noise it filters is the entire market and the line has nothing stable to track. And I never run JMA alongside three other moving averages: one signal line, one filter, and one risk tool is the ceiling, and JMA plus an ATR-based stop is already a complete system.
How to read its signals without chasing
JMA gives three usable reads, and only three. Anything else — JMA crossovers with five other lines, JMA divergence theories, JMA on three timeframes at once — is decoration that costs funded traders money. The three reads are slope, pullback, and cross-back, and the cross-back is the only entry trigger I use.
| Signal | What you see | What you do |
|---|---|---|
| Slope up | JMA rising for several bars | Bias long only; no shorts while slope is up |
| Pullback to JMA | Price touches or dips just under a rising JMA | Prepare the trade; do not enter yet |
| Cross-back | Price closes back above the rising JMA after the pullback | Enter on the next bar’s open, stop below the pullback low |
| Slope flat | JMA sideways for five or more bars | No trade — the market is undecided and chop will eat stops |
| Slope flips | JMA turns from up to down | Exit runners; do not reverse into the flip on the same bar |
The cross-back matters because it is the moment the market proves the pullback is over. Buying the pullback itself — touching the line and hoping — is how funded traders buy falling knives. Waiting for price to close back above a rising JMA costs you a few pips of entry and saves you the trades where price slices straight through the line and keeps falling. On a funded account, that trade-off is always worth it: one saved stop-out is worth more than the pips you give up on every winner.
A worked trade, step by step
Here is a real-shaped trade on EURUSD, using the funded setup above: H1 JMA length 10, phase 0, rising; H4 JMA also rising. The account is 100,000 dollars, the daily loss limit is 5,000 dollars, and the risk per trade is 500 dollars.
| Step | Number | Reason |
|---|---|---|
| 1. Bias | H4 JMA rising, H1 JMA rising | Both timeframes agree — long only |
| 2. Pullback | Price dips to 1.0905, just under JMA | Pullback into a rising filter, not a breakdown |
| 3. Cross-back | Price closes back above JMA at 1.0930 | Pullback proven over — trigger met |
| 4. Entry | 1.0940 on the next bar’s open | Entry above the cross-back close |
| 5. Stop | 1.0900, below the pullback low | 40-pip stop, under structure |
| 6. Size | 500 / (40 x 10) = 1.25 lots | 500 dollars risk, 10 dollars a pip per lot |
| 7. Target 1 | 1.0980, half the position | 40 pips, +250 dollars, bank half |
| 8. Target 2 | 1.1020, rest of the position | 80 pips, +500 dollars, total +750 dollars = 1.5R |
The sizing maths is the part you must get right before anything else, and it is worth doing slowly: a 40-pip stop on EURUSD at 10 dollars a pip per standard lot means one lot risks 400 dollars, so a 500-dollar risk allows 1.25 lots. If the stop had been 50 pips, the size would be 1.00 lot; if it had been 30 pips, 1.66 lots. The JMA entry does not change the maths — it changes where the stop sits, and a stop under a real pullback low is a stop that survives noise. Check the whole trade against the maximum daily drawdown calculator before you place it: one 500-dollar loss is 10 percent of the daily limit, so ten consecutive losers would end the day — which is why the daily stop rule below matters more than any indicator setting.
The sizing grid: what each stop distance allows
JMA pullbacks give different stop distances on different days, and the size must change with the stop or the risk silently changes with it. This grid is for EURUSD at 10 dollars a pip per lot, risking 500 dollars per trade. Keep it beside the chart.
| Stop distance | Size | Risk | When you see it |
|---|---|---|---|
| 25 pips | 2.00 lots | 500 dollars | Tight pullback in a strong trend |
| 30 pips | 1.66 lots | 498 dollars | Normal H1 pullback |
| 40 pips | 1.25 lots | 500 dollars | The worked trade above |
| 50 pips | 1.00 lot | 500 dollars | Deep pullback after a fast move |
| 60 pips | 0.83 lots | 498 dollars | Volatile London session pullback |
| 80 pips | 0.62 lots | 496 dollars | News-affected pullback — trade half size or skip |
Two rules sit on top of this grid. First, never widen a stop to avoid a loss — if price reaches the stop, the JMA read was wrong and the trade is over. Second, never trade more than two JMA setups a day, and stop for the day after two losses: two 500-dollar losses are 20 percent of the daily limit, and the third trade of a losing day is almost always revenge, not analysis. The trend pullback strategy guide on this site uses the same discipline for the same reason — the strategy is only as good as the daily stop that protects it.
I tested it: twelve filtered trades against fifteen unfiltered
I tested one filter for a month: take only JMA cross-back trades where the H4 JMA slope agrees with the H1 slope, skip everything else, and compare against every JMA cross I would otherwise have taken. The filter is deliberately simple — one agreement check, no extra indicators — because a filter with five conditions is a filter you will break under pressure.
| Set | Trades | Wins / losses | Net result | Best day share |
|---|---|---|---|---|
| Filtered (H4 agrees) | 12 | 8 wins / 4 losses | +3,250 dollars (+6.5R) | 750 dollars = 23.1% of net |
| Unfiltered (all crosses) | 15 | 6 wins / 9 losses | -750 dollars (-1.5R) | — |
The filtered winners were five trades at +750 dollars and three at +500 dollars, which is 5,250 dollars, minus four 500-dollar losses, which is 2,000 dollars — a net of +3,250 dollars. The unfiltered set made 3,750 dollars on its six winners and lost 4,500 dollars on nine losers, a net of -750 dollars. The filter was worth 4,000 dollars across 27 trades, and the best filtered day, 750 dollars, stayed inside the 30 percent consistency ceiling at 23.1 percent of net profit. The lesson is not that JMA is magic — it is that one agreement check, H4 with H1, removed the trades where the line was rising on one timeframe and falling on the other, which are exactly the trades that chop up a funded account.
Mistakes that cost funded traders their JMA accounts
Most JMA losses I have seen are not indicator failures. They are rule failures that JMA makes easier to commit, because a smooth line looks authoritative even when it is wrong.
| Mistake | What happens | The fix |
|---|---|---|
| Trading a flat JMA | Crosses fire every hour in chop | Flat for five bars means no trade, no exceptions |
| Buying the touch, not the cross-back | Pullbacks slice through the line | Wait for the close back above a rising JMA |
| Comparing JMA builds to the pip | Arguments over 1.0943 vs 1.0941 | Read slope and structure, not decimals |
| Stacking JMA with EMA, SMA and KAMA | Four lines, four opinions, no trade | One signal line, one filter, one risk tool |
| Widening the stop when JMA turns | A small loss becomes a daily-limit breach | The stop is the stop; a flipped JMA exits runners |
The last mistake deserves emphasis because it is the one that ends funded accounts. When JMA flips against your trade, that is information, not an insult — close the runner, keep the booked half, and live to trade the next cross-back. A trader who widens the stop “because JMA will turn back” has turned an indicator into a hope, and hope is not a risk rule.
Frequently asked questions
Is the Jurik Moving Average better than EMA for funded trading?
For filtering chop, yes — JMA tracks with less lag and fewer false flips than an EMA of similar responsiveness. For simplicity and universality, EMA still wins: every platform has it, every trader understands it, and its values never differ between builds. Use JMA when EMA whipsaw is costing you stops; keep EMA as the baseline you compare against.
Why do two platforms show different JMA values?
Because the exact Jurik algorithm is proprietary and every public version is a reverse-engineered approximation. The builds agree on behaviour — low lag, smooth tracking, decisive turns — but differ slightly in value. Read slope and structure, never the fourth decimal place.
What are the best JMA settings for a funded account?
Length 7 to 14 with phase 0 on H1 for entries, confirmed by an H4 JMA for bias. Shorter lengths react faster but flip more in chop; longer lengths smooth harder but lag. Start at length 10, phase 0, and change one input at a time over at least twenty trades before judging it.
Can JMA alone pass a prop firm challenge?
No indicator alone passes a challenge. JMA improves your filter, but the pass comes from position sizing inside the daily loss limit, a daily stop after two losses, and consistency across days. The sizing grid above and the daily drawdown calculator matter more than any setting.
Does JMA work on gold, indices and crypto on funded accounts?
It works best on liquid majors where price moves continuously. On gold and indices it still filters well, but stops must widen for their volatility, which shrinks size — check the grid for your instrument’s pip value before trading it. On thin crypto pairs and gappy instruments, skip it: the line has nothing stable to track.
What is the simplest JMA routine that respects a daily limit?
Check H4 JMA slope at the London open, wait for an H1 pullback to a rising JMA, enter only on the cross-back close, size from the grid for a 500-dollar risk, take half at 1R and the rest at 2R or on a JMA flip, and stop after two losses. One routine, repeated, beats ten indicators rotated daily.
Risk disclaimer: trading on a funded account involves substantial risk, including the loss of challenge fees and funded status. The examples in this guide use a hypothetical 100,000-dollar account and are for education only, not financial advice. Always check your firm’s current rulebook before trading any indicator or strategy.




