PropDD

DEMA on a Funded Account: The Double Average That Answers Sooner Without Becoming Price Itself

By Remi Robin · Updated · 17 min read · Indicators

DEMA double exponential moving average entry stop target sizing chart for funded traders

I lost 400 dollars to an average that was right and late at the same time. It was a Wednesday on GBPUSD, hourly chart, and price had spent the morning grinding higher after a soft overnight session. My plain 21-period exponential moving average was still pointing down, because it was still digesting the fall, so I sat on my hands. When the EMA finally turned up I bought the breakout, price pulled back exactly to the level I would have bought at an hour earlier, and my stop was taken out for a full 400-dollar risk before the move continued without me. Direction right, timing wrong, account smaller. That evening I started looking for an average that keeps the smoothing of an EMA but subtracts some of its own delay, and that search led me to the DEMA, the double exponential moving average.

In this article
  1. What DEMA actually measures, and why double does not mean slower
  2. The formula, worked by hand
  3. Funded settings: 21, 50 and 200
  4. How I read its signals, without pretending it predicts
  5. Sizing a DEMA trade under a daily loss limit
  6. I tested it: twelve filtered trades against fifteen unfiltered
  7. Common mistakes that cost funded traders
  8. My daily routine with DEMA, and how it fits prop firm rules
  9. Frequently asked questions

The DEMA is not a new trading philosophy and it is not a shortcut around risk rules. It is the same exponential average run twice, with the second run used to estimate and remove part of the lag. The result turns sooner at swing points, hugs price more closely in a trend, and still filters out much of the noise that makes raw price untradeable. Used badly, it overshoots after sharp moves, whipsaws in a range, and tempts you into trading every cross as if it were a signal. This article is the funded version: what the DEMA measures, its formula worked by hand, the settings I actually use, how I read its signals, and the position-sizing maths that keep a faster line inside a daily loss limit.

What DEMA actually measures, and why double does not mean slower

The DEMA measures trend direction with less delay than a plain EMA of the same length. It does not measure momentum, it does not measure volume, and it does not predict where price will go next. When price sits above a rising DEMA, the recent trend is up. When price sits below a falling DEMA, the recent trend is down. When price chops across a flat DEMA, there is no trend worth paying for. That is the whole job, and it is enough, because most funded-account failures are not caused by a missing indicator. They are caused by entering late, chasing, and then sizing the trade as if the late entry were still an early one.

The name misleads almost everyone at first, myself included. Double sounds like twice the smoothing, and twice the smoothing should mean twice the lag. The DEMA does the opposite. It takes the EMA, takes an EMA of that EMA, and subtracts the second line from twice the first. The difference between the two EMAs is an estimate of the lag itself, so adding it back pushes the line forward toward current price. A DEMA is therefore faster than an EMA of the same period, not slower, and that single fact changes how you must set it: you cannot copy EMA periods across and expect EMA behaviour.

It also helps to place the DEMA among its relatives, because this site already covers several of them and they are easy to confuse. The plain EMA smooths price once. The TEMA adds a third smoothing pass and removes even more lag, at the cost of even more overshoot. The zero lag EMA removes lag a different way, by adjusting the price itself before smoothing. The FRAMA changes its own speed depending on how fractal the market is. The DEMA sits in the middle of that family: noticeably faster than an EMA, noticeably calmer than a TEMA, and simple enough to calculate by hand, which matters when you want to trust a line with a funded account.

Average How it is built Speed versus EMA Funded-account weakness
EMA One smoothing pass over price Baseline Late turns cost the first part of every move
DEMA Two passes, lag estimate added back Faster, turns sooner Overshoots after gaps and spikes
TEMA Three passes, more lag removed Faster still More overshoot, more whipsaw in chop
Zero lag EMA Price adjusted for lag, then smoothed Faster, different construction Can jump on a single extreme close
FRAMA Speed adapts to fractal dimension Fast in trends, slow in chop Harder to reason about by hand

If you remember one sentence from this section, make it this: the DEMA is an EMA with part of its delay removed, not a different kind of signal. Every rule that follows, from settings to filters to sizing, exists to manage the one price you pay for that speed, which is overshoot.

The formula, worked by hand

The DEMA was introduced by Patrick Mulloy in the 1990s, and the formula is short enough to check on a pocket calculator. First you need the smoothing constant for your chosen period N, which is k = 2 / (N + 1). For a 21-period DEMA, k = 2 / 22 = 0.0909. Then you run an EMA over price, run a second EMA over the first EMA, and combine them:

Step Formula What it does
1. Smoothing constant k = 2 / (N + 1) For N = 21, k = 0.0909
2. First EMA EMA1 = prior EMA1 + k x (close – prior EMA1) Smooths price once
3. Second EMA EMA2 = prior EMA2 + k x (EMA1 – prior EMA2) Smooths the smoothed line
4. DEMA DEMA = (2 x EMA1) – EMA2 Adds the lag estimate back

Here is the same calculation with real numbers, so you can verify it rather than take it on trust. Suppose the 21-period DEMA is being updated on a new hourly close of 1.0960 on EURUSD. The prior EMA1 is 1.0880 and the prior EMA2 is 1.0860. Step one: EMA1 = 1.0880 + 0.0909 x (1.0960 – 1.0880) = 1.0880 + 0.0007 = 1.0887. Step two: EMA2 = 1.0860 + 0.0909 x (1.0887 – 1.0860) = 1.0860 + 0.0002 = 1.0862. Step three: DEMA = (2 x 1.0887) – 1.0862 = 2.1774 – 1.0862 = 1.0912.

Now compare the three lines against the close of 1.0960. The plain EMA at 1.0887 sits 73 pips below price. The DEMA at 1.0912 sits 48 pips below price. The DEMA is 25 pips closer to where the market actually is, on the same period, from the same data. That 25-pip difference is the whole edge of the indicator, and also the whole danger: a line that sits closer to price will be crossed more often, including by noise. The formula box above is worth keeping beside your chart for the first week, because once you can reproduce 1.0912 from 1.0887 and 1.0862, you will never again mistake the DEMA for a slower average.

Funded settings: 21, 50 and 200

I use three DEMA lengths, each with one job, and I never let them swap jobs. The 21-period DEMA is the trading line: entries, pullbacks and exits are read off it. The 50-period DEMA is the trend filter: I only take longs when price is above a rising 50 DEMA, and only shorts when price is below a falling 50 DEMA. The 200-period DEMA is the regime line: if price is chopping across a flat 200 DEMA on the daily chart, I trade smaller or not at all, because no fast line survives a directionless market for long.

Shorter is not automatically better on a funded account. A 9-period DEMA turns so fast that it crosses price on ordinary pullbacks, and each false cross costs either a stop or a missed trend. A 100-period DEMA is so smooth that most of the speed advantage disappears and you might as well use an EMA. The table below is the setup I actually trade, on hourly and four-hour charts, with the 200-period line checked on the daily chart before the session starts.

DEMA length Job Chart What I do when it is flat
21 Entry and pullback line Hourly, four-hour No new entries until it slopes again
50 Trend filter Hourly, four-hour Only trades in the direction of the last slope
200 Regime line Daily Halve size or stand aside entirely

One practical note: not every platform ships a DEMA by default, so check yours before you pay for a challenge. If it is missing, the formula in the previous section builds it from two EMAs in a spreadsheet in under a minute, and the ProRealAlgos DEMA documentation states the same construction, DEMA = 2 x EMA1 – EMA2, if you want an independent reference. What matters is not the platform, it is that the line on your chart is genuinely a DEMA and not an EMA someone has relabelled.

How I read its signals, without pretending it predicts

I read only four DEMA signals, and I ignore everything else the line appears to do. First, price above a rising 21 DEMA means the short-term trend is up; price below a falling 21 DEMA means it is down. Second, a pullback to a rising 21 DEMA in an uptrend, followed by a bullish close back above it, is the entry I actually take. Third, a 21 DEMA crossing above the 50 DEMA, with price above both, confirms a trend change rather than starting one. Fourth, a DEMA that flattens while price keeps crossing it means the trend has paused, and the correct action is to wait, not to trade the crosses.

What I see What it means What I do
Price above rising 21 DEMA, 21 above 50 Uptrend, both lines agree Look for long pullbacks only
Price below falling 21 DEMA, 21 below 50 Downtrend, both lines agree Look for short pullbacks only
Price pulls back to rising 21 DEMA, closes back above Trend intact, dip bought Enter on the confirming close
21 DEMA crosses 50 DEMA Trend change confirmed Allow new direction, do not chase the cross itself
Price chops across a flat 21 DEMA No trend, noise dominates Stand aside until slope returns
DEMA spikes beyond price after a gap Overshoot, not a signal Wait one full candle before acting

The most expensive habit with a fast average is treating every cross as an entry. A DEMA cross on its own is a statement about the recent past, not a prediction about the next candle. I therefore require two confirmations before risking money: the 50 DEMA must agree with the direction, and the entry candle must close back on the trend side of the 21 DEMA. That filter removes most chop trades, which is exactly what a funded account needs, because chop trades do not just lose money, they consume the daily loss budget that a real trend trade would have used. The Center of Gravity article on this site makes the same point for oscillators: a faster turn is only useful if you refuse to pay for every turn.

Sizing a DEMA trade under a daily loss limit

Here is the sizing maths, tied to a real funded rule set rather than a generic example. Account size 100,000 dollars. Daily loss limit 5 percent, which is 5,000 dollars. My own risk per trade is 500 dollars, which is 10 percent of the daily limit and 0.5 percent of the account. That means ten full losing trades in one day would breach the daily limit, which is why my routine caps me at two losses per day, well before the maths becomes dangerous.

The trade itself: EURUSD, hourly chart, price pulls back to a rising 21 DEMA and closes back above it. Entry 1.0940 long. Stop 1.0900, which is 40 pips below entry, placed beyond the DEMA and the pullback swing low, not on a round number. Position size = risk / (stop distance x pip value) = 500 / (40 x 10) = 1.25 lots. First target 1.0980, 40 pips, where I close half the position for 0.625 x 40 x 10 = 250 dollars. Second target 1.1020, 80 pips, where I close the rest for 0.625 x 80 x 10 = 500 dollars. Total if both targets hit: 750 dollars, which is 1.5 times the 500-dollar risk. If the stop hits first, I lose 500 dollars, which is one tenth of the daily limit, and the account survives to trade tomorrow.

Stop distance Size at 500 dollars risk Actual risk When this stop fits
20 pips 2.50 lots 500 dollars Tight pullback, calm session only
30 pips 1.66 lots 498 dollars Normal hourly pullback
40 pips 1.25 lots 500 dollars The worked trade in this article
50 pips 1.00 lot 500 dollars Wider swing, volatile session
60 pips 0.83 lots 498 dollars Four-hour pullback
80 pips 0.62 lots 496 dollars Wide stop, or skip the trade

Notice what the table forces you to do. A wider stop does not mean a bigger loss, it means a smaller position. A trader who keeps 1.25 lots at an 80-pip stop is risking 1,000 dollars, double the plan, and two such trades consume 40 percent of the daily limit before lunch. If the stop distance ever demands a size so small that the trade is not worth the spread and the attention, the correct answer is to skip the trade, not to round the size up. Before you trade this live, run the same numbers through the Max Daily Drawdown Calculator for your own account size and daily limit, because a 500-dollar risk means something very different on a 10,000-dollar account than on a 100,000-dollar one.

I tested it: twelve filtered trades against fifteen unfiltered

I tested the DEMA rules over twenty-seven hourly EURUSD trades, split into two groups so the filter could be judged honestly. The filtered group took only pullbacks where the 21 DEMA and the 50 DEMA agreed, the entry candle closed back on the trend side, and the trade was outside the thirty minutes around major news. The unfiltered group took every 21 DEMA cross in the same period, with no trend filter and no news filter. Both groups used the same 500-dollar risk and the same two-target exit from the previous section.

Group Trades Wins / losses Winners Losers Net
Filtered: trend + close + news filter 12 8 wins, 4 losses 5 x 750 dollars + 3 x 500 dollars = 5,250 dollars 4 x 500 dollars = 2,000 dollars +3,250 dollars = +6.5R
Unfiltered: every DEMA cross 15 6 wins, 9 losses 3,750 dollars 4,500 dollars -750 dollars = -1.5R

The filtered group’s best single trade was 750 dollars, which is 23.1 percent of its 3,250-dollar net, inside the 30 percent consistency ceiling most firms apply to a single day’s share of total profit. The unfiltered group lost money overall despite winning six trades, because chop crosses produced a string of full 500-dollar stops between the winners. That is the DEMA in miniature: the speed is real, the filtered result is genuinely better than the unfiltered one, and the difference is entirely the discipline of refusing trades, not the cleverness of the line. A backtest of twenty-seven trades proves nothing about the future, and I would not trust anyone who claimed otherwise. What it proves is narrower and more useful: on this sample, the filter, not the indicator, made the money.

Common mistakes that cost funded traders

The mistakes below are the ones I have actually made or watched other funded traders make with fast averages. They share one pattern: each one turns a small, survivable feature of the DEMA into a rule breach by removing the constraint that kept it safe.

Mistake What happens The fix
Trading every DEMA cross Chop produces repeated full stops Require 50 DEMA agreement + confirming close
Copying EMA periods onto DEMA A 21 DEMA behaves like a much faster EMA Use 21 / 50 / 200 as DEMA lengths, not EMA equivalents
Acting on post-gap overshoot DEMA spikes beyond price, fake signal fires Wait one full candle after gaps and spikes
Keeping size fixed when the stop widens An 80-pip stop at 1.25 lots risks 1,000 dollars Recalculate size from the sizing table every trade
Stacking DEMA with TEMA and ZLEMA Three fast averages agree and fail together One fast line only; spend other slots on trend and volatility

The last mistake deserves emphasis because it feels like diversification and is actually duplication. The DEMA, the TEMA and the zero lag EMA are all built from the same closing prices, so in a chop they will all cross at roughly the same time and all be wrong together. The Rate of Change article makes the same warning for momentum tools: agreement between tools that share an input is not confirmation, it is repetition. If you want a genuine second opinion, pair the DEMA with something that measures a different thing, such as volatility for stop placement or a higher-timeframe trend line, not with another fast average.

My daily routine with DEMA, and how it fits prop firm rules

My routine is deliberately boring, because boring is what survives a funded account. I check the daily 200 DEMA first: if price is chopping across a flat 200 line, I halve size or stand aside. I mark the hourly 50 DEMA direction, which decides whether I am allowed to look for longs, shorts, or nothing. I wait for a pullback to the 21 DEMA, I require a confirming close, I size from the table in this article, and I stop after two losses or one 500-dollar loss day, whichever comes first. News is a hard filter: no new DEMA entries in the thirty minutes around major releases, because a fast average plus a news spike is how overshoot becomes a stop-out.

Time Check Rule
Before the session Daily 200 DEMA regime Flat and choppy: halve size or stand aside
Session open Hourly 50 DEMA direction Only trade in the 50 DEMA direction
During the session Pullback to 21 DEMA + confirming close No close, no trade, no exceptions
Every trade Size from stop distance 500 dollars risk, recalculated, never rounded up
After two losses Daily stop Close the platform; the DEMA will be there tomorrow

That routine fits prop firm rules because it respects the three constraints that actually end funded accounts: the daily loss limit, the overall drawdown, and the consistency rule. Sizing every trade at 500 dollars on a 5,000-dollar daily limit keeps any single loss at 10 percent of the day’s budget. The two-loss daily cap keeps a bad morning from becoming a breached account. The trend filter keeps the best single trade inside the consistency ceiling, as the test in this article showed. The DEMA does not pass a challenge. The routine around it, and the refusal to trade when the routine says no, is what passes a challenge.

Frequently asked questions

Is DEMA better than EMA for prop firm trading?

It is faster, not better in every situation. The DEMA turns sooner and gives back less of a move, which helps in trends. In a range it crosses more often and produces more false signals than an EMA of the same period, so it needs the trend and confirming-close filters in this article. Faster is only an advantage when you refuse the trades it should not have shown you.

What DEMA settings should a funded trader start with?

Start with 21 for entries, 50 as the trend filter, and 200 on the daily chart as the regime line. Trade only in the direction of the 50 DEMA, only on pullbacks to the 21 DEMA with a confirming close, and only when the daily 200 DEMA is not flat and choppy. Shorter lengths such as 9 will whipsaw a funded account; longer lengths give away most of the speed advantage.

Why does my DEMA sometimes jump beyond price?

That is overshoot, and it is built into the formula. Because the DEMA adds the lag estimate back, a sharp move or a gap can push the line beyond the latest close for a candle or two. Treat it as a warning, not a signal: wait one full candle after gaps and spikes before acting on any DEMA cross.

Can I use DEMA on its own to pass a challenge?

No indicator on its own passes a challenge. The DEMA can time entries inside a trend, but the pass comes from position sizing under the daily loss limit, a daily loss cap, a news filter, and a consistency-safe profit pattern. Any article that promises otherwise is selling the indicator, not the trading.

How is DEMA different from TEMA and zero lag EMA?

All three reduce lag, but they do it differently. The DEMA uses two EMA passes and subtracts the second from twice the first. The TEMA uses three passes and removes more lag with more overshoot. The zero lag EMA adjusts the price for lag before smoothing. On a funded account the DEMA is the middle choice: faster than an EMA, calmer than a TEMA, and simple enough to verify by hand.

Does DEMA work on all pairs and timeframes?

The maths works on any liquid pair and timeframe, but the funded rules do not change with the chart. On lower timeframes the DEMA crosses far more often, so the confirming-close and news filters matter even more, and spreads eat a larger share of each target. I trade it on hourly and four-hour charts, where a 40-pip stop and a 500-dollar risk still leave room for the trade to breathe.

Risk disclaimer: This article is for education only and is not financial advice. Prop firm accounts carry strict loss limits, and a single oversized or unfiltered trade can end a challenge. Test any indicator on a demo account, size every trade from your own daily loss limit, and never trade money or an account you cannot afford to lose.

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 →