I lost 500 dollars on a green histogram bar, and the bar was telling the truth the whole time. It was a Tuesday afternoon on EURUSD, the Accelerator Oscillator had just printed its second green column above the zero line, and I bought with confidence — then price chopped sideways for six hours, printed three red columns, and stopped me out. The indicator had not lied. It measures whether momentum is speeding up or slowing down, and on that afternoon momentum was speeding up inside a range, which is not the same thing as a trend. I had traded acceleration as if it were direction, and a funded account charges full price for that confusion.
In this article
- What the Accelerator Oscillator actually measures
- The formula, worked with real numbers
- How to read its signals: the zero line, the colours and the bar counts
- Best settings for prop firm trading
- One worked trade, from entry to both targets
- I tested it: ten trades with the rules applied
- Common mistakes that cost funded traders
- How it fits prop firm rules
- My Accelerator routine: the checklist I run before every trade
- Frequently asked questions
That loss taught me the one sentence that now governs every Accelerator Oscillator trade I take: AC tells me when the push is getting stronger, and only the trend filter tells me which way I am allowed to push. This guide is the version I wish I had read first: what AC measures, its exact formula worked with real numbers, the two-bar and three-bar signal rules, the settings that survive a funded account, one worked trade, ten tested trades, and the routine that keeps a colourful histogram from eating a daily limit.
What the Accelerator Oscillator actually measures
The Accelerator Oscillator, usually written AC, was developed by Bill Williams, the same trader behind the Awesome Oscillator and the Alligator. His idea was a sequence: acceleration changes first, momentum changes second, and price changes last. Most indicators measure momentum, which is how fast price is moving. AC measures the change in that momentum, which is whether the move is speeding up or slowing down. That is why AC can turn while price is still drifting sideways, and why it can keep printing green bars while price barely moves: the push is strengthening, but the push has not yet become a trend.
AC is drawn as a histogram around a zero line. A bar is green when today’s AC value is higher than yesterday’s, and red when it is lower. The colour does not mean buy or sell on its own. Green above the zero line means upward acceleration is increasing. Green below the zero line means downward acceleration is weakening, which is a different statement entirely. Red below the zero line means downward acceleration is increasing. Red above the zero line means upward acceleration is weakening. Four combinations, four different meanings, and my Tuesday loss came from treating the first one as a buy signal without asking whether a trend existed at all. If you take one thing from this article, take the colour table in the signals section below, not a single colour.
AC is also a standard indicator in MetaTrader 5, under Insert, Indicators, Bill Williams, Accelerator Oscillator, so there is nothing to download and nothing that looks like automation. That matters on a funded account: some firms restrict expert advisors and trade copiers, but a built-in indicator that only draws bars on your own chart is ordinary chart analysis. I still check my firm’s current rules before relying on any tool, and you should too, but AC has never been a restricted tool on any account I have traded.
The formula, worked with real numbers
AC is built from the Awesome Oscillator, so the formula has two stages. First, take the median price of each candle, which is the high plus the low divided by two. The Awesome Oscillator is the 5-period simple moving average of the median price minus the 34-period simple moving average of the median price. Then AC is the Awesome Oscillator minus its own 5-period simple moving average. In symbols: median price = (high + low) / 2, AO = SMA(median price, 5) minus SMA(median price, 34), and AC = AO minus SMA(AO, 5). The standard settings are therefore 5, 34 and 5, and I use them unchanged, for reasons the settings section explains.
The table below works the second stage with real numbers, using Awesome Oscillator values from a EURUSD hourly chart. You do not need to calculate this by hand on a live chart — MetaTrader does it — but seeing the subtraction once stops the most common misunderstanding, which is reading the height of an AC bar as a price level. AC values are in price units, but they are differences of averages, so a bar of 0.00078 is not a price target. It is a measure of how much faster momentum is running than its own recent average.
| Bar | Awesome Oscillator (AO) | 5-period average of AO | AC = AO minus average | Bar colour and meaning |
|---|---|---|---|---|
| Day 4 | 0.00180 | 0.00124 | +0.00056 | Green, above zero: upward acceleration present |
| Day 5 | 0.00210 | 0.00150 | +0.00060 | Green, rising: upward acceleration increasing |
| Day 6 | 0.00260 | 0.00182 | +0.00078 | Green, rising again: acceleration still increasing |
Check the arithmetic on day 6: the five AO values ending on day 6 are 0.00140, 0.00120, 0.00180, 0.00210 and 0.00260, which sum to 0.00910 and average to 0.00182. Subtract that from the current AO of 0.00260 and you get AC of +0.00078. Because +0.00078 is higher than the previous bar’s +0.00060, the bar prints green, and because it is above zero, upward acceleration is in charge. Three rising green bars above zero is exactly the kind of sequence that tempts a trader to buy immediately. Whether that buy is allowed depends on the rules in the next section, and on the trend filter after that.
How to read its signals: the zero line, the colours and the bar counts
Bill Williams’ original rules use the zero line and a bar count, and the count is the part most traders skip. When you trade with the driving force — buying above zero or selling below zero — you need two consecutive bars of the right colour. When you trade against it — buying below zero or selling above zero — you need three, because you are asking the market to turn, and a turn deserves extra evidence. The table below is the whole signal system, and I keep it printed beside my screen because every AC mistake I have made was a violation of one row of it.
| Signal | Where AC sits | Bars needed | What it means |
|---|---|---|---|
| Buy with the force | Above the zero line | 2 consecutive green bars | Upward acceleration is increasing; the push is with you |
| Buy against the force | Below the zero line | 3 consecutive green bars | Downward acceleration is fading; a turn needs one more bar of proof |
| Sell with the force | Below the zero line | 2 consecutive red bars | Downward acceleration is increasing; the push is with you |
| Sell against the force | Above the zero line | 3 consecutive red bars | Upward acceleration is fading; a turn needs one more bar of proof |
Two warnings belong here, because both cost me money before I learned them. First, AC has no overbought or oversold levels. It is not RSI, and a high AC reading does not mean price must fall. AC can stay above zero for days in a strong trend, printing green bar after green bar, while traders waiting for an “overbought” reversal get run over. Second, a zero-line cross on its own is not a signal. Williams’ rules require the confirming bars after the cross, and the cross itself often happens mid-chop. I trade the bars, not the cross, and I trade them only when the trend filter in the settings section agrees.
Divergence is the one advanced use worth mentioning, and I use it as a warning, not an entry. When price makes a new high but AC makes a lower high, upward acceleration is weakening even though price looks strong. I do not short on that alone — fading a trend on divergence is how funded accounts die — but I stop adding to longs, I tighten stops, and I refuse to take a fresh buy signal until AC prints green again. Divergence tells me the push is tired. It does not tell me the trend is over.
Best settings for prop firm trading
The default settings are 5, 34 and 5, and on a funded account I keep them. I tested a faster 3, 21 and 3 setting on hourly charts, and it produced roughly twice as many colour flips, most of them inside ranges, which doubled my losing trades without improving my winners. A slower 8, 55 and 8 setting filtered the chop better but entered trends so late that the stop distances grew past my risk cap. Defaults are defaults because they balance those two failures, and on a funded account the cost of extra signals is not just spread — it is daily-limit room, which is the one resource you cannot buy back.
| Setting | What I use | Why it fits a funded account |
|---|---|---|
| AO fast / slow / AC smoothing | 5 / 34 / 5 (defaults) | Balanced signal count; faster settings double chop trades, slower settings enter too late |
| Timeframe | H1 for entries, H4 for trend direction | Hourly bars filter most range noise; H4 keeps me from fighting the bigger trend |
| Trend filter | Price above or below the Alligator lines, or above or below the 50-period EMA | AC alone cannot tell a trend from a range; the filter does that job |
| Session filter | London and New York hours only | AC flips most in thin Asian hours; I refuse those signals outright |
| News filter | No new signal within 30 minutes of red-folder news | News spikes print perfect-looking AC bars that reverse in minutes |
The trend filter deserves emphasis because it is the difference between my losing Tuesday and my winning weeks. I take AC buy signals only when price is above the Alligator lines — the jaw, teeth and lips all pointing the same way — or, if I am not running the Alligator, when price is above the 50-period exponential moving average. I take sells only on the mirror image. Read my Alligator guide for the full filter, and my Awesome Oscillator guide for the indicator AC is built from: AC without AO context is a histogram without a story.
One worked trade, from entry to both targets
My funded account is 100,000 dollars with a 5,000-dollar daily loss limit, and I risk 500 dollars per trade, which is 0.5 percent of the account and exactly one tenth of the daily limit. That ratio is deliberate: a single loss never threatens the day, and ten consecutive full losses would breach it, a run I have never had with this system. The trade below happened on EURUSD hourly, with price above the Alligator lines, AC printing its second consecutive green bar above zero, and no red-folder news within thirty minutes. Every filter passed, so the signal was allowed.
| Step | Price | Money on 1.00 lot |
|---|---|---|
| Entry, second green AC bar above zero, trend filter green | 1.0950 | — |
| Stop, below the recent swing low | 1.0900 (50 pips) | Risk 500 dollars (50 pips x 10 dollars per pip) |
| First target, half the position closed | 1.1000 (+50 pips) | Profit 250 dollars on the half lot |
| Stop moved to breakeven, rest of the position held | 1.0950 | Remaining risk zero |
| Second target, rest of the position closed | 1.1050 (+100 pips) | Profit 500 dollars on the remaining half lot |
| Total result | — | Profit 750 dollars, minus spread and fees |
The sizing is the part that makes the trade funded-account safe, and it is the same formula I use for every trade: risk divided by stop distance in pips multiplied by pip value. At 500 dollars of risk, a 50-pip stop on EURUSD at 10 dollars per pip per lot gives exactly 1.00 lot. The grid below shows how the same 500-dollar risk changes with stop distance, which is why I never trade a fixed lot size with AC: a tight trend-continuation setup and a wide reversal setup are different trades, and fixed lots would quietly turn a wide setup into a daily-limit threat. Run your own numbers through the max daily drawdown calculator before you trade, so you see the trade’s share of the day before the market shows it to you.
| Stop distance | Lots at 500 dollars risk | What it usually means |
|---|---|---|
| 25 pips | 2.00 lots | Tight trend-continuation entry on H1 |
| 40 pips | 1.25 lots | Standard pullback entry with room for noise |
| 50 pips | 1.00 lot | The worked trade above, swing-low stop |
| 80 pips | 0.60 lots (rounded down from 0.62) | Wide reversal setup, only with the trend filter passed |
I tested it: ten trades with the rules applied
I ran the full rule set — default settings, trend filter, London and New York sessions only, no news-window signals, 500-dollar risk cap, size rounded down — across ten EURUSD hourly trades in my journal. The table below is the honest log, losers included, because a system that only shows its winners is not a system. Three of the ten trades were against-the-force signals that needed three bars, and two of those three lost, which is exactly why the extra bar exists: turns are harder than continuations, and the rules charge more evidence for them.
| Trade | Signal | Filter | Result |
|---|---|---|---|
| 1 | Buy above zero, 2 green bars | Price above Alligator | +2R (+1,000 dollars) |
| 2 | Sell below zero, 2 red bars | Price below 50 EMA | +1R (+500 dollars) |
| 3 | Buy below zero, 3 green bars | Price below Alligator | -1R (-500 dollars) |
| 4 | Buy above zero, 2 green bars | Price above Alligator | +2R (+1,000 dollars) |
| 5 | Sell above zero, 3 red bars | Price above 50 EMA | -1R (-500 dollars) |
| 6 | Buy above zero, 2 green bars | Price above 50 EMA | +1R (+500 dollars) |
| 7 | Sell below zero, 2 red bars | Price below Alligator | +2R (+1,000 dollars) |
| 8 | Buy above zero, 2 green bars | Price above Alligator | -1R (-500 dollars) |
| 9 | Sell below zero, 2 red bars | Price below 50 EMA | +1R (+500 dollars) |
| 10 | Buy above zero, 2 green bars | Price above Alligator | +1R (+500 dollars) |
| Total across ten trades | +7R (+3,500 dollars) | ||
Seven winners and three losers, with every single loss capped at 500 dollars, produced a net gain of 3,500 dollars — and, more importantly for a funded account, the best single day in the run was 1,000 dollars, which is 28.6 percent of the net gain and inside the 30 percent consistency ceiling most firms apply. That last number is the one I check first in any journal review, because a system can be profitable and still fail a consistency rule if one day does all the talking. Read my consistency rule guide if that ceiling is new to you: it is the rule that quietly ends more funded accounts than the daily loss limit.
For contrast, I also logged twenty raw colour flips taken without any filter — every green bar bought, every red bar sold — over the same period. That unfiltered run produced eight wins and twelve losses for a net of minus 4R, or minus 2,000 dollars at the same risk. The indicator did not change between the two runs. The filters did. AC is a good acceleration gauge and a terrible standalone entry system, and the difference between those two statements is worth 5,500 dollars across thirty trades.
Common mistakes that cost funded traders
The mistakes below are all mine, made across my first month with AC, and each one maps to a row of the signal table or a line of the settings table that I ignored. The pattern is consistent: AC punishes traders who want it to be simpler than it is. It is not a traffic light, it is not an overbought gauge, and its bars are not price levels. Treat it as any of those three things and it will take your daily limit one plausible-looking signal at a time.
| Mistake | What happens | The rule that prevents it |
|---|---|---|
| Trading every colour change | Range chop prints alternating green and red bars; each flip costs spread plus a stop | Trend filter first — no filter pass, no trade, whatever the colour |
| Buying the first green bar | One green bar is noise; Williams requires two with the force, three against it | Count the bars before you click, every single time |
| Treating high AC as overbought | AC stays above zero for days in a strong trend; fading it bleeds the account | AC has no overbought level — use RSI if you want one |
| Trading into red-folder news | News spikes print perfect AC bars that reverse within minutes | Thirty-minute news blackout around every major release |
| Sizing from the bar height | A tall bar feels strong, so traders size up — but bar height is not a risk measure | Size comes only from the stop distance and the 500-dollar cap |
| Using AC on thin hours | Asian-session chop prints the most signals and the worst fills | London and New York sessions only, no exceptions |
How it fits prop firm rules
AC fits funded trading well, with one condition: it must stay a filter inside a rule-based plan, not become the plan. The daily loss limit is protected by the 500-dollar risk cap and the sizing grid, not by the indicator. The consistency rule is protected by taking many similar-sized trades, so no single day dominates the month — my ten-trade run’s best day sat at 28.6 percent of net, inside the ceiling, precisely because the losers were capped and the winners were not allowed to balloon. The minimum-days style rules are unaffected, because AC trades are ordinary manual trades with no holding-time games. And because AC is a built-in chart indicator, not an automated system, it does not trip the automation rules that end some funded accounts.
The one rule AC tests is discipline, and it tests it through boredom. Most days, AC prints signals that fail the trend filter or the session filter, and the correct action is nothing. A trader who cannot sit through five rejected signals will eventually take the sixth, which is usually the choppiest one of the day. I treat rejected signals as the system working, not as missed money: every signal I refuse in a range is daily-limit room I still own at the New York open. If you pair AC with a trend strategy, my trend pullback guide shows how a pullback entry and an AC confirmation can share one stop and one risk cap without doubling the risk.
My Accelerator routine: the checklist I run before every trade
| Check | What I do | Why it matters with AC |
|---|---|---|
| Trend check | Price above Alligator lines for buys, below for sells (or 50 EMA) | AC cannot separate trend from range; this check does |
| Zero-line check | Note whether AC is above or below zero before counting bars | The side of zero decides whether I need two bars or three |
| Bar-count check | Count consecutive green or red bars out loud before entry | One bar is noise; the count is the signal |
| Session check | London or New York hours only | Thin hours print the most and worst AC signals |
| News check | No red-folder release within thirty minutes either way | News bars look perfect and reverse in minutes |
| Sizing check | Risk 500 dollars, stop in pips, size rounded down | Keeps every loss at one tenth of the daily limit |
I run it in the same order every time: trend first, zero line second, bar count third, and only then the entry. For the theory behind the indicator, the Admiral Markets Accelerator Oscillator guide covers the formula and the two-bar and three-bar rules in the same form I use here. The routine travels to any pair and any funded account; only the stop distance and the lot size change.
Frequently asked questions
What does the Accelerator Oscillator measure?
It measures the acceleration or deceleration of market momentum, not price direction itself. It is calculated as the Awesome Oscillator minus a 5-period moving average of the Awesome Oscillator, so it shows whether the driving force behind a move is speeding up or slowing down. That is why it can turn while price is still moving sideways.
What are the best settings for the Accelerator Oscillator?
The defaults: 5, 34 and 5. The Awesome Oscillator uses 5-period and 34-period averages of the median price, and AC subtracts a 5-period average of the Awesome Oscillator. Faster settings double the false signals in ranges, and slower settings enter trends too late for a funded risk cap. I use defaults on H1 entries with an H4 trend read.
How many green bars do I need to buy?
Two consecutive green bars when AC is above the zero line, and three when it is below the zero line. The extra bar below zero exists because you are trading against the driving force and asking the market to turn, which deserves more evidence. The mirror rule applies to sells: two red bars below zero, three red bars above it.
Can the Accelerator Oscillator stay above zero for a long time?
Yes, for days in a strong trend, and that is normal. AC has no overbought or oversold levels, so a long run above zero is not a sell signal. Traders who fade a persistent AC reading because it “looks too high” are fighting the trend with an indicator that was never designed to call tops.
Does the Accelerator Oscillator work on its own?
No, not on a funded account. My own journal comparison showed raw colour flips losing 4R across twenty trades, while the same indicator with a trend filter, session filter and news blackout made 7R across ten trades. AC is an acceleration gauge that belongs inside a rule-based plan, with a trend filter deciding direction and a risk cap deciding size.
Is the Accelerator Oscillator allowed on prop firm accounts?
It is a standard built-in indicator in MetaTrader 5, not an automated system, so it is ordinary chart analysis on every firm I have traded with. Firms restrict expert advisors, trade copiers and certain strategies, not built-in indicators. Still, always read your own firm’s current rules before relying on any tool, because rules change.
Risk disclaimer: Trading involves substantial risk of loss and is not suitable for every investor. Funded account rules vary by firm and change over time — always check your firm’s current rules before trading with any indicator, and never risk money you cannot afford to lose.





1 comment
[…] behind it, and the Gator tells me whether the trend behind that bar still has energy. My Accelerator Oscillator guide covers the third Bill Williams histogram, for traders who want acceleration read alongside the […]