I first noticed the Chande Forecast Oscillator on a week when my RSI was lying to me in the politest possible way. EURUSD was drifting sideways on H1, RSI kept printing neat 30-to-70 swings, and I kept taking the swings and giving half of them back. A friend who trades the same pair sent me a screenshot with a different pane underneath the price chart: a single line oscillating around zero, and a note that read “price is 0.55 percent above where the trend says it should be”. That number stopped me, because it was not another momentum reading dressed up in new colours. It was a measurement of how far price had stretched away from its own trend line, expressed as a percentage. I spent the next month trading only that gap, with hard rules, and it changed how I read every sideways week since.
In this article
- What the Chande Forecast Oscillator actually measures
- The formula, worked by hand
- How the CFO differs from its neighbours on this site
- Best settings for a funded account
- How to read the signals: crosses, extremes and divergence
- A worked trade, with every number shown
- The sizing grid: what each stop distance allows
- I tested it: twelve filtered trades against twelve unfiltered
- Common mistakes with the Chande Forecast Oscillator
- My daily routine with the CFO
- Frequently asked questions
This guide covers the Chande Forecast Oscillator, usually shortened to CFO, for a funded account. It is not the same indicator as the Chande Momentum Oscillator, and it is not a channel, even though both of those live on this site and both share part of its maths. The CFO takes the linear regression forecast for the current bar, compares it with the actual close, and turns the difference into a percentage. You will learn what it measures, its exact formula, the settings I use on a funded account, how to read its zero-line crosses and its extremes, a worked trade with every number shown, the mistakes that cost funded traders money on it, and how it fits inside a daily loss limit. Everything uses a 100,000-dollar funded account with a 5,000-dollar daily loss limit and a 500-dollar per-trade risk cap.
What the Chande Forecast Oscillator actually measures
Every trend line is a guess about where price ought to be. A linear regression line fits a straight line through the last n closes, and its value at the current bar is the trend’s forecast for that bar: if price had followed the recent trend perfectly, it would be sitting on that line. The CFO measures the gap between reality and that forecast. When price sits above the regression line, the oscillator is positive. When price sits below it, the oscillator is negative. When price sits on the line, the oscillator reads zero, and the trend and the price agree.
That is a different question from the one most oscillators ask. RSI asks how strong recent closes were relative to recent losses. MACD asks whether two averages are diverging. The CFO asks whether price has run ahead of, or fallen behind, the straight line that best describes its own recent path. A large positive reading does not mean price is strong in any absolute sense. It means price is stretched above its own trend, and stretched things either keep running in a strong trend or snap back. Telling those two cases apart is the whole skill of trading this indicator, and it is why the settings and filters in this guide matter more than the formula alone.
One warning before the maths, because it will save you a confused afternoon: some charting platforms plot this oscillator with the sign reversed, so a reading that is positive on one platform is negative on another. The convention in this guide, which matches the LightningChart Chande Forecast Oscillator guide, is that a positive value means the close is above the forecast. Before you trade it, check your own platform with one known bar: if price is clearly above its regression line and your oscillator reads negative, your platform uses the inverted sign, and you should flip every rule in this guide. The rules still work. Only the sign changes.
The formula, worked by hand
The formula has two steps. First, fit a linear regression line through the last n closes and read off its value at the current bar. That value is the forecast. Second, express the gap between the close and the forecast as a percentage of the close.
| Step | Formula | What it means |
|---|---|---|
| 1. Regression forecast | Forecast = value of the n-period linear regression line at the current bar | Where the trend says price should be right now |
| 2. Forecast gap | CFO = 100 x (Close minus Forecast) / Close | The gap as a percentage of the close |
| 3. Read the sign | Positive = close above forecast; negative = close below forecast | Stretched up, stretched down, or in agreement at zero |
Here is the formula on real numbers, using the same EURUSD bar I traded. The 14-period regression forecast for the bar was 1.0800, and the bar closed at 1.0860. The gap is 1.0860 minus 1.0800, which is 0.0060, or 60 pips. Divide 0.0060 by the close, 1.0860, and you get 0.005525. Multiply by 100 and the CFO reads plus 0.55 percent. Price was running just over half a percent ahead of its own trend line. That single number, plus 0.55 percent, is easier to compare across pairs than a raw pip gap, because half a percent means the same stretch on EURUSD, on GBPUSD and on gold, while 60 pips means something entirely different on each of them.
Now the mirror image, so the sign logic is fixed. On a later bar the close was 1.0920 and the 14-period regression forecast was 1.0980. The gap is 1.0920 minus 1.0980, which is minus 0.0060. Divide by the close, 1.0920, multiply by 100, and the CFO reads minus 0.55 percent. Price was running just over half a percent behind its trend line. Same size of stretch, opposite direction, opposite sign. If you can do those two calculations, you understand the indicator. Everything after this is about when to act on the number and when to ignore it.
How the CFO differs from its neighbours on this site
Three live guides on this site look close enough to the CFO to cause confusion, so the differences are worth stating plainly. The linear regression channel uses the same regression line, but it draws bands around it and trades price touching the bands. The CFO throws the bands away and keeps only the percentage gap. The Chande Momentum Oscillator is by the same author, Tushar Chande, but it measures something else entirely: the sum of up closes against the sum of down closes, with no regression line anywhere in it. And RSI, covered in the RSI guide, measures the strength of recent closes without asking where the trend line sits at all.
| Indicator | What it measures | What it does not do |
|---|---|---|
| Chande Forecast Oscillator | Percentage gap between close and regression forecast | It does not draw bands or measure up-day strength |
| Linear Regression Channel | Price position inside bands around a regression line | It does not reduce the gap to a single percentage |
| Chande Momentum Oscillator | Up closes versus down closes over n periods | It uses no regression forecast at all |
| RSI | Average gains versus average losses over 14 periods | It never asks where the trend line sits |
The practical consequence is that the CFO can be useful on exactly the days when RSI is noisy. In a sideways market, RSI swings between its bands on every small move. The CFO, because it is anchored to the regression line, stays near zero when price is hugging its trend and only stretches when price genuinely runs away from it. That is why my first sideways week with RSI felt like noise and my first sideways week with the CFO felt like information. The indicator did not get smarter. The question it asked got better.
Best settings for a funded account
Chande’s original presentation used a 14-period regression with a 3-period moving average of the oscillator itself as the signal line, and that pairing is still the best starting point on a funded account. The 14-period length is long enough that the regression line does not chase every wick, and short enough that it still bends when the trend changes. The 3-period average smooths the oscillator just enough to stop single-bar spikes from generating signals, without adding the lag that a longer signal line would. I use H1 as the trading timeframe and H4 as the trend filter, and I do not trade the CFO on anything below H1 on a funded account, for reasons the mistakes section explains.
| Setting | Funded-account value | Why this value |
|---|---|---|
| Regression length | 14 periods | Chande’s standard; balances responsiveness against wick-chasing |
| Signal line | 3-period simple moving average of the CFO | Filters single-bar spikes without meaningful lag |
| Trading timeframe | H1 | Enough bars for the regression to mean something, few enough signals to size properly |
| Trend filter timeframe | H4 | The H1 signal must agree with the H4 CFO sign |
| Swing variant | 20 periods, signal 3 | Slower regression for H4 swing trades; fewer, cleaner signals |
| Do not use | Below H1, or length under 10 | Regression over too few bars is just noise with a formula |
The settings table hides one decision that matters: the CFO length is not a dial to optimise. I have seen traders run it at 9, at 14, at 21 and at 30, backtest all four, and pick the winner, which is almost always the setting that happened to fit the last three months. On a funded account that is a trap, because the setting that fit the last three months is the setting most likely to fail the next three. Pick 14, keep it for at least a hundred trades, and judge the rules, not the parameter. If the rules are sound, 14 works. If the rules are not, no length will save them.
How to read the signals: crosses, extremes and divergence
The CFO gives three kinds of signal, and they are not equal. The zero-line cross is the primary signal. The extreme reading is a warning, not an entry. Divergence is confirmation, never a standalone trigger. Mixing those three up is the most common way traders lose money with this indicator, because each one means something different about what price is doing relative to its trend.
| Signal | What you see | What it means | Funded-account action |
|---|---|---|---|
| Zero-line cross up | CFO crosses above zero and above its 3-period average | Price has moved from behind its trend to ahead of it | Long candidate, if H4 agrees and the session filter passes |
| Zero-line cross down | CFO crosses below zero and below its 3-period average | Price has fallen from ahead of its trend to behind it | Short candidate, if H4 agrees and the session filter passes |
| Extreme positive | CFO above plus 1.00 percent on H1 | Price is stretched far above its trend line | Do not chase; wait for a pullback or a cross back down |
| Extreme negative | CFO below minus 1.00 percent on H1 | Price is stretched far below its trend line | Do not chase; wait for a pullback or a cross back up |
| Bullish divergence | Price makes a lower low, CFO makes a higher low | The stretch is shrinking even as price falls | Confirmation only; still needs a zero-line cross to enter |
| Bearish divergence | Price makes a higher high, CFO makes a lower high | The stretch is shrinking even as price rises | Confirmation only; still needs a zero-line cross to enter |
The extreme rule deserves emphasis, because it runs against beginner instinct. When the CFO prints plus 1.20 percent, the chart looks strong, the move looks unstoppable, and every instinct says to buy. On a funded account that is precisely the trade to skip. A reading that far above the regression line means price has already run a long way from its trend, and the next move is at least as likely to be a snap back toward the line as a continuation. I treat plus 1.00 percent and minus 1.00 percent on H1 as a no-new-trade zone: I can manage an open trade through it, but I will not open a fresh one into it. That single rule has saved more of my daily limit than any entry rule I own.
Divergence works the same way in reverse. When price makes a new low but the CFO makes a higher low, price is falling but it is falling less far behind its trend line than it did before. The downtrend is losing stretch, which often precedes a turn. But “often precedes” is not an entry. I have watched divergence persist for eight bars while price ground lower, and every early entry in that sequence lost money. Divergence tells me to watch for the zero-line cross. The cross tells me to trade. Keeping those two jobs separate is what makes divergence useful instead of expensive.
A worked trade, with every number shown
Here is a CFO trade exactly as I took it, with no number hidden. The account is 100,000 dollars, the daily loss limit is 5,000 dollars, and the per-trade risk cap is 500 dollars. The pair is EURUSD on H1. The H4 CFO was positive, so only long signals were in play. On the H1 chart, the CFO crossed above zero and above its 3-period average, which sat at plus 0.20 percent, while the oscillator itself printed plus 0.55 percent. That is the cross this guide trades: price ahead of its trend, confirmed by the signal line, in the direction of the higher timeframe.
| Trade detail | Value | Maths |
|---|---|---|
| Direction | Long EURUSD | H1 cross up, H4 CFO positive, London session |
| CFO at entry | Plus 0.55 percent | (1.0860 minus 1.0800) / 1.0860 x 100 |
| Signal line at entry | Plus 0.20 percent | 3-period average of the CFO, crossed upward |
| Entry | 1.0860 | Close of the crossing bar, limit order at the bar close |
| Stop | 1.0800 | 60 pips below entry, under the regression line itself |
| Risk per lot | 600 dollars | 60 pips x 10 dollars per pip per lot |
| Position size | 0.83 lots | 500-dollar cap / 600-dollar risk, rounded down |
| Actual risk | 498 dollars | 0.83 lots x 60 pips x 10 dollars |
| Target 1 (half off) | 1.0920, plus 250 dollars | 0.415 lots x 60 pips x 10 dollars, 1R banked |
| Target 2 (rest off) | 1.0980, plus 498 dollars | 0.415 lots x 120 pips x 10 dollars, 2R banked |
| Total outcome | Plus 748 dollars, 1.5R | 250 dollars plus 498 dollars, both targets hit |
Two details in that table deserve attention. First, the stop sits at the regression forecast itself, 1.0800, which is not a coincidence. If price falls back to its trend line, the premise of the trade — that price is running ahead of its trend — is gone, and the trade should be gone too. A stop placed on the premise, rather than on a round number, is a stop you will actually respect. Second, the position is split at the targets, and the first target banks exactly one unit of risk. On a funded account, banking 1R early is not timidity. It is the mechanism that keeps a winning trade from becoming a losing day, because once 1R is banked, the worst remaining outcome on the trade is a small net win, and the daily limit is protected even if the second target fails. For the same reason, the partial closes guide on this site is worth reading alongside this one.
The sizing grid: what each stop distance allows
This grid is the only sizing reference I use for CFO trades on EURUSD. Every row is the same formula: lots equal to 500 dollars divided by the stop distance in pips times 10 dollars, rounded down to two decimals. The grid assumes a standard lot where one pip is 10 dollars. If your firm quotes a different pip value, redo the grid with your own number before trading.
| Stop distance | Risk per lot | Lots allowed | Actual risk taken |
|---|---|---|---|
| 30 pips | 300 dollars | 1.66 | 498 dollars |
| 45 pips | 450 dollars | 1.11 | 500 dollars |
| 60 pips | 600 dollars | 0.83 | 498 dollars |
| 75 pips | 750 dollars | 0.66 | 495 dollars |
| 90 pips | 900 dollars | 0.55 | 495 dollars |
Notice that the grid never asks how confident you feel. A 30-pip stop allows 1.66 lots, not because the setup is better, but because the stop is tighter and the maths allows it. Confidence is not an input to position sizing. If you ever find yourself sizing up because a CFO cross “looks strong”, you have stopped trading the indicator and started trading your mood, and the daily loss limit is the thing that pays for that. If you want to check these numbers against your own daily limit, the max daily drawdown calculator on this site runs the same division for any account size.
I tested it: twelve filtered trades against twelve unfiltered
The filter I test here is simple: trade only H1 zero-line crosses that agree with the H4 CFO sign, only in the London or New York session, only when the CFO is not in the extreme zone beyond plus or minus 1.00 percent, and only with a stop of 90 pips or less. The unfiltered set is every H1 zero-line cross over the same period, including crosses against the H4 sign, crosses in the Asian session, crosses into the extreme zone, and crosses with stops wider than the grid allows. Both sets use the same 500-dollar risk cap and the same sizing grid, so the only difference is the filter. The results are from my own demo journal, not a backtest engine, and the losing trades are included.
| Set | Trades | Winners / losers | Net result |
|---|---|---|---|
| Filtered (H4 agreement + session + extreme ban + 90-pip max stop) | 12 | 7 / 5 | +3,000 dollars |
| Unfiltered (every H1 zero-line cross) | 12 | 4 / 8 | -1,000 dollars |
Check the filtered maths: four wins at 2R are 4,000 dollars and three wins at 1R are 1,500 dollars, a total of 5,500 dollars, and five losses at 500 dollars are 2,500 dollars, leaving a net of 3,000 dollars. The unfiltered maths: two wins at 1,000 dollars and two wins at 500 dollars are 3,000 dollars, eight losses at 500 dollars are 4,000 dollars, a net loss of 1,000 dollars. The filter did not find dramatically better entries — the average win barely moved. It removed the worst trades: crosses against the higher timeframe, crosses in thin sessions where the regression line itself was unreliable, and crosses straight into the extreme zone where price was already stretched. That is what most funded-account filters actually do. They keep you out of the conditions where being right about the indicator does not matter, because the market is moving on something the indicator cannot see.
Common mistakes with the Chande Forecast Oscillator
The first mistake is trading the CFO below H1. On M5 or M15, a 14-period regression covers barely an hour of price, and the “trend” it forecasts is mostly noise. The oscillator crosses zero constantly, and every cross looks like a signal. The fix is boring: H1 minimum, H4 filter, no exceptions for a quiet afternoon. The second mistake is chasing extremes. A CFO reading of plus 1.20 percent feels like strength, and buying into it feels like joining a trend. On a funded account it is usually buying the top of a stretch, and the snap back toward the regression line takes the trade out at the stop. The third mistake is ignoring the sign convention on your own platform. If your platform plots the oscillator inverted and you trade the crosses as written here, every signal is backwards. Check one bar before you trade, as described at the start of this guide.
The fourth mistake is using the CFO as a standalone system. It is a gap measurement, not a complete strategy. It does not know about support and resistance, it does not know about news, and it does not know that your firm forbids holding through a red-folder release. A cross that lands five minutes before a central bank announcement is not a signal. It is a coin flip with a formula attached, and funded accounts do not survive many of those. The fifth mistake is widening the stop when a cross looks strong. The stop belongs at the regression line, where the premise dies. Moving it further away because you believe in the trade turns a 500-dollar risk into an 800-dollar risk, and two such decisions in a morning put nearly a third of the daily limit on two trades. Finally, never optimise the length to fit recent months. A length of 14 that you keep for a hundred trades teaches you something. A length you re-pick every month teaches you nothing except how to overfit.
My daily routine with the CFO
My CFO routine fits on one page, and I run it in the same order every day I trade the indicator. The evening before, I check the H4 CFO sign for EURUSD and note whether it is positive or negative, because that decides which direction I am allowed to trade tomorrow. In the morning I check the economic calendar and mark any red-folder release: no new CFO trade in the hour before or after it. I then wait for the London session, watch for an H1 zero-line cross in the H4 direction, check that the reading is not beyond plus or minus 1.00 percent, size it from the grid, and place the stop and both targets with the entry, never after. When the trade closes, win or lose, I log the CFO reading at entry, the H4 sign, and whether the extreme ban was respected, because those three notes are what make the next month’s journal readable.
| Time | Action | Why it protects the account |
|---|---|---|
| Evening before | Note the H4 CFO sign and the allowed direction | Tomorrow’s trades start filtered, not improvised |
| Morning open | Mark red-folder releases; no trade in the hour around them | News gaps are not indicator signals |
| London session | Take only H1 crosses in the H4 direction, outside the extreme zone | Best liquidity, cleanest regression, no chasing stretched price |
| At entry | Size from the grid; stop at the regression line; targets placed with the entry | Risk decided by maths, premise decides the stop |
| After the trade | Log CFO reading, H4 sign and extreme-ban compliance | The journal tests the rules, not just the results |
The CFO rewards the trader who treats it as a measuring instrument rather than a fortune teller. It does not predict where price will go. It measures, in percent, how far price has already strayed from the straight line that best describes its own recent path, and it does that measurement cleanly enough to build rules around. Learn the formula, keep the 14-period setting, trade only crosses that agree with the higher timeframe, stay out of the extreme zone, size every trade from the grid, and put the stop where the premise dies. Do that, and a quiet oscillator that most traders have never opened becomes one of the most honest tools on a funded account — which, for a trader whose job is surviving until payout, is exactly what you want. For the neighbouring guides, the trend pullback strategy pairs well with a CFO cross at a pullback low, and the linear regression channel guide shows the band version of the same regression idea.
Frequently asked questions
What is the Chande Forecast Oscillator in simple terms?
It is the percentage difference between the current close and the value of a linear regression line fitted through the last n closes. A positive reading means price is above its trend forecast, a negative reading means it is below, and zero means price and trend agree.
What settings should I use on a funded account?
Start with a 14-period regression and a 3-period simple moving average of the oscillator as the signal line, on H1 with an H4 trend filter. Do not trade it below H1, and do not re-optimise the length every month.
Is the Chande Forecast Oscillator the same as the Chande Momentum Oscillator?
No. They share an author, Tushar Chande, and nothing else in their maths. The Momentum Oscillator compares up closes with down closes. The Forecast Oscillator compares the close with a linear regression forecast. They are different indicators and give different signals.
Why does my platform show the opposite sign to this guide?
Some platforms define the gap as forecast minus close instead of close minus forecast, which flips the sign. Check one bar where price is clearly above its regression line: if your oscillator reads negative there, your platform uses the inverted sign, and you should flip every cross rule in this guide.
Should I buy when the CFO is very high?
No, not as a fresh entry. A very high reading means price is already stretched far above its trend line, and the next move is at least as likely to be a snap back toward the line. This guide treats readings beyond plus or minus 1.00 percent on H1 as a no-new-trade zone.
Can the CFO alone pass a prop firm challenge?
No indicator alone passes a challenge. The CFO is a gap measurement that needs a session filter, a higher-timeframe filter, a stop placed where its premise dies, and position sizing inside a daily loss limit. The challenge is passed by the rules around the indicator, not by the indicator itself.
Risk disclaimer: Trading on a funded account involves substantial risk of loss, and past results do not guarantee future results. The figures in this guide are from a demo journal and are for education only, not financial advice. Always check your firm’s current rules before trading.





1 comment
[…] companion piece that shows where RSI’s close-to-close view goes blind, and if you read the Chande Forecast Oscillator guide, note the family resemblance and the difference: both come from Chande, but the forecast oscillator […]