“Quant Regime Oscillator” reads overextension, momentum, and market regime in one pane

“Quant Regime Oscillator” is an open-source oscillator published by JackOfAllTrades. *On TradingView, the author goes by officialjackofalltrades.

It rolls two things into a single line bounded between −100 and +100: how far price has stretched from its mean, and how strong the momentum is. On top of that, the background color tells you whether the market is actually trending in one direction or just chopping back and forth.

There are three things to read. The line’s position tells you how overextended price is, crosses with the white signal line show shifts in momentum, and the background color shows the character of the market. You also get BUY/SELL labels that only print when every condition lines up, divergence dots, and a dashboard in the top-right corner.

Bitcoin 4H chart with Quant Regime Oscillator on default settings in TradingView. Overextension, momentum, and regime are all visible in a single pane.

When I first loaded it on Bitcoin’s daily, 4H, and 15m charts, I was honestly confused because not a single label showed up. The dashboard header stayed at “◆ FLAT” on every timeframe. Thinking I had messed up a setting, I dug into how it works and found it wasn’t a bug at all. The signal conditions are just extremely strict. In this article I lay out why, and which settings actually get signals to print, using the numbers I counted myself.

The “Quant” and “Regime” in the name

“Quant” refers to putting a number on the market with statistical yardsticks, such as how many standard deviations price is from its mean. “Regime” means the market phase, and this indicator sorts it into three buckets: trend, mean reversion (Mean-Rev), and random. Measure overextension with statistics, then change how you read it depending on the regime. That design is literally the name.

Start here: what’s in the lower pane

“Quant Regime Oscillator” draws nothing on the price chart. It runs in its own pane below. Here’s the whole thing in one view.

“Quant Regime Oscillator” layout at a glance — Shown on default settings. The line and fill change color at zero, and the background turns blue in uptrends, purple in downtrends, and gray in choppy markets. Bars that fit none of these get no background.

The line and fill switch between blue and magenta at the zero line. You can see the line flip to blue on the bar that crosses above zero in Figure ①. The fill is a gradient: the closer to +100, the bluer it gets, the closer to −100, the deeper the magenta, with a purple-ish blend around zero.

The white line is the signal line. It’s a 9-period EMA of the oscillator, and as Figure ② shows, it trails slightly behind the main line. Crosses between the two are your cue for momentum shifts.

The background shows the market’s character. Bars classified as an uptrend are shaded blue as in Figure ③, downtrends get purple, and choppy back-and-forth markets get a faint gray. Bars that fall in between get no color at all.

Element	Appearance	Meaning
Oscillator	Line that’s blue at 0 and above, magenta below 0	Combined value of distance from the mean and momentum (−100 to +100)
Gradient fill	Between the line and 0. Bluer when high, more magenta when low	Helps you read the line’s position at a glance
Signal line	White line	EMA of the oscillator (default 9 periods)
±55 dashed lines	Gray dashed lines	Overbought/oversold thresholds. Used in the signal conditions
±80 dotted lines	Very faint dotted lines	Rough guide for extreme levels. Not used in any logic
Background color	Blue, purple, gray, or none	Regime classification
Divergence dots	Small blue (bullish) and magenta (bearish) circles	Price and oscillator moving in opposite directions
BUY/SELL	Blue and magenta labels	Only print on bars where all three conditions line up

The ±80 dotted lines are hard to see

The author describes the area beyond ±80 as an exhaustion zone. But the dotted lines are extremely faint, and on a dark theme you really have to squint to see them. On top of that, ±80 isn’t used in any of the logic. The overbought/oversold calls are made off the ±55 dashed lines.

Under the hood, the line is an average of “distance from the mean” and “momentum”

The oscillator value is built from two components. Let’s go through them one at a time. Neither one is shown on the chart, so I overlaid them as dotted lines in the figure.

The line is an average of “Z-Score” and “momentum” — Both components cap ±3σ and map it to ±100. Late in the selloff, the momentum component stays pinned at −100 while only the Z-Score component recovers to the −40s. That combination is why the plotted line flatlines around −70.

The Z-Score component: how many standard deviations from the 30-bar mean

The first is the Z-Score. It measures how many standard deviations the close is from the 30-bar simple moving average. +2 means 2σ above the mean, −1 means 1σ below. The value is capped at ±3σ and then rescaled to ±100, so anything beyond +3σ is simply treated as +100. The yellow dotted line in Figure ① is this component. It spikes right after the surge and drifts lower as the rally cools off.

The momentum component: rate of change divided by its own dispersion

The second is momentum. It takes the rate of change versus 14 bars ago (ROC), divides it by the standard deviation of that ROC over the last 14 bars, then caps it at ±3 and rescales it to ±100 as well. The idea is to let you compare assets with very different price levels and volatility on the same yardstick.

In practice, though, this component has a quirk. When price moves cleanly in one direction, the ROC barely varies, which means the denominator shrinks. The value then shoots straight to its cap and gets pinned at +100 or −100. In the back half of the selloff, you can see the momentum component flatlined at −100 around Figure ②. When I counted on Bitcoin from the 15m up to the daily, this pinning happened on roughly 10–13% of all bars. The Z-Score component only hits its cap about 2% of the time, so when the line is pressed against an extreme, it’s usually momentum doing the pushing.

Averaging the two, lightly smoothing, and clamping to ±100

By default the two components are averaged with equal weight. That average is lightly smoothed with a 2-period EMA and clamped to ±100, and that’s the line you see. In the far-right stretch, the pinned momentum component (−100) and the recovering Z-Score component (the −40s) are averaged, so the line hovers around −70, as Figure ③ shows. That combination is why the line is slow to turn up even after price stops falling.

How stretched is ±55, really?

If both components are swinging equally, ±55 works out to roughly ±1.65σ on the Z-Score. Under a normal distribution, that’s a level you’d only hit about 5% of the time on each side. Yet when I actually counted on Bitcoin, the line spent 15–21% of the time at +55 or higher and 14–15% at −55 or lower. Because the momentum pinning gets added in, it reaches those levels far more often than the numbers suggest. Keep that in mind and you won’t read too much into a simple push past ±55.

The background color is a test of whether price is moving in a straight line

The background color is driven by Kaufman’s Efficiency Ratio (ER). It’s the same measure used in the adaptive moving average KAMA: the net price change over the last 20 bars divided by the total path, meaning the sum of every bar-to-bar move.

The background is set by how straight price moved over 20 bars — The 20 bars that climbed in a straight line score 0.57 and are classified as an uptrend (blue), while the 20 bars that chopped near the highs score 0.15 and are classified as Mean-Rev (gray). Values between 0.25 and 0.45 count as Random and get no background. Up or down is decided by whether the close is above or below the 30-bar SMA.

If price climbs in a straight line for 20 bars, the net change and the path are nearly equal and ER approaches 1. If price whipsaws and ends up back where it started, the path is long but the net change is close to zero, so ER approaches 0. Put numbers on it and the difference is obvious. The rally in Figure ① reads 0.57 and is classified as Trending, while the chop near the highs in Figure ② reads 0.15 and is classified as Mean-Rev.

The default cutoffs are 0.45 and above for Trending, 0.25 and below for Mean-Rev, and Random for everything in between. Random bars get no background, as shown in Figure ③.

ER can’t tell you direction. “Quant Regime Oscillator” calls it an uptrend when the close is above the 30-bar simple moving average and a downtrend when it’s below, and shades the background blue or purple accordingly. That moving average is the same line used as the Z-Score baseline.

Once it’s on the chart, the background is almost always gray

The first thing I noticed was how rare blue and purple backgrounds are. I counted the regime breakdown on each Bitcoin timeframe.

Timeframe	Bars checked	Trending (blue/purple)	Mean-Rev (gray)	Random (no color)
15m	~30,000	9.1%	64.4%	26.5%
1H	~20,000	11.8%	60.2%	28.0%
4H	~12,000	14.6%	57.7%	27.7%
Daily	~3,300	16.4%	55.4%	28.2%

The sample periods start in November 2025 for the 15m, June 2024 for the 1H, April 2021 for the 4H, and August 2017 for the daily, all running through September 26, 2026. Mean-Rev shows up more on lower timeframes, most likely because small noise stretches the path.

On the 15m chart, Quant Regime Oscillator's regime background is mostly gray (Mean-Rev). — On the 15m chart, Quant Regime Oscillator’s regime background is mostly gray (Mean-Rev).

If you wait for the background as a “trend signal,” you’ll be disappointed, because it barely ever shows up. That was me at first. These days I flip it around: when blue or purple appears, I read it as price is unusually moving in one direction right now.

The regime playbook the author recommends

The author suggests using the regime as a cue to switch your playbook.

When Trending, favor trend-following entries off pullbacks and rallies
When Mean-Rev, look to fade moves that stretch beyond ±55
When Random, stand aside or size down

“Mean-Rev” isn’t a mean reversion forecast

Despite the name, Mean-Rev only means the efficiency over the last 20 bars was low. It doesn’t say price is going to revert to its mean from here. The bar right before price breaks out of a range and takes off will also read Mean-Rev. If you’re going to use it as a reason to fade a move, confirm it against the highs and lows on the price chart too.

Three dashboard rows that are easy to misread

The dashboard is a small 2-column, 9-row table that sits in the top-right corner of the chart by default.

What the dashboard's 9 rows tell you — The values are an example from a daily chart. The right side of the header isn't the current direction but “the direction of the last signal,” and it stays at FLAT if no BUY or SELL has ever fired.

Most rows are self-explanatory. The one to watch is the right side of the header in Figure ①, which is easy to misread. It doesn’t show the current market direction. It shows the direction of the most recent BUY/SELL signal. Once a BUY fires, it stays at “◆ LONG” until a SELL prints, and it never goes back to FLAT. “◆ FLAT” means no signal has fired anywhere within the loaded history.

The other confusing one is Regime Str. The name makes you think it’s a confidence score for the regime call, but the value in Figure ② is just ER multiplied by 100. If you see 17% next to Mean-Rev, it means the trend efficiency is only 17%. It doesn’t measure how strong the mean reversion is.

The bottom two rows, Divergence and Signal in Figure ③, only fill in on the bar where the event happens. A divergence shows for just the one bar it’s confirmed on, and BUY/SELL only on the bar it prints, so most of the time you’ll see “—” and “waiting.”

Row	Example	What it shows
Composite	53.6	Current oscillator value. The cell’s blue/magenta shading also changes with the value
Z-Score	1.44σ	Distance from the 30-bar mean, before the ±3σ cap
Momentum	55.4	Momentum component (after rescaling to ±100)
Regime	Mean-Rev	One of Trending ▲, Trending ▼, Mean-Rev, or Random
Regime Str	17%	ER × 100
State	Neutral	Overbought, Oversold, or Neutral based on ±55
Divergence	—	Divergence confirmed on that bar
Signal	waiting	BUY/SELL on that bar

The Z-Score row drops the leading zero. 0.47σ displays as “.47σ” and −0.77σ as “-.77σ.” The first time I saw it I thought a digit was missing, but that’s just the format.

Why don’t BUY/SELL signals show up? I checked the conditions one by one

As I mentioned at the top, “Quant Regime Oscillator” on default settings will almost never print a label. A BUY only fires when all three of the following line up on the same bar. SELL is the mirror image.

1The oscillator crosses above the signal line

This is a bullish cross between the oscillator and the white line. On the Bitcoin 4H, crosses in either direction happen about 15 times per 100 bars, so on its own it’s very common.

2It dipped to −55 or lower within the last 6 bars

It’s satisfied if at least one bar within the last 6 bars, including the current one, was at −55 or lower. That lookback is controlled by OB/OS Recall Bars.

3It’s in an uptrend regime

ER is 0.45 or higher and the close is above the 30-bar mean. These are the bars with a blue background.

Each condition is common on its own. The problem is the combination, and the reason is simple. Price only drops below −55 when it’s stretched far below its 30-bar mean. For it to get back above the mean within 6 bars, and for the 20-bar efficiency to clear 0.45 on top of that, you need a rebound that’s far more violent than the selloff.

On default settings, the oversold memory expires before the trend classification arrives — A BUY needs three things on the same bar: a cross above the signal line, a reading of −55 or lower within the last 6 bars, and an uptrend regime. Here the oversold stretch ended 20 bars before the cross, so the 6-bar memory had long expired.

In this example, after dropping below −55 in Figure ①, price bounces and even earns an uptrend classification. In Figure ③, the oscillator crosses above the signal line inside the blue background, which looks like a textbook buy-the-dip setup. Still no BUY. As Figure ② shows, the oversold memory expires after 6 bars, and by the time of the cross, that oversold dip was 20 bars in the past.

Zero signals across three assets and four timeframes

Is it really that rare? I counted on real data. These are the results on Bitcoin, Ethereum, and PAXG (a token that tracks the price of gold), from the 15m up to the daily, all on default settings. As with the earlier table, the data runs through September 26, 2026.

Asset	15m (~30k bars)	1H (~20k bars)	4H (~12k bars)	Daily
Bitcoin	0/0	0/0	0/0	0/0 (3,328 bars)
Ethereum	0/0	0/0	0/0	0/0 (3,328 bars)
PAXG (gold-backed)	0/0	0/0	0/0	0/0 (2,221 bars)

The numbers are BUY/SELL counts. Across nearly 200,000 bars in total, not a single signal fired. Looking at the intermediate conditions, the Bitcoin 4H had 293 bars where a bullish cross and a sub−55 reading within the last 6 bars overlapped, but the number of those bars that also had an uptrend classification was zero.

For the record, I verified that the calculations behind these counts are correct. On the daily, 4H, and 15m charts in TradingView, the dashboard’s Composite, Z-Score, Momentum, and Regime Str values, plus the signal line value, matched down to the decimal.

Want to see signals? Increase OB/OS Recall Bars

I tried a few ways to loosen the conditions, and the one that made the clearest difference was OB/OS Recall Bars. Keep the overbought/oversold memory around longer, and the setup in the figure becomes a signal as-is.

Set OB/OS Recall Bars to 30 and the same cross prints a BUY — Because the oversold memory now lasts 30 bars, the cross 20 bars later still qualifies. The BUY locks in once the bar closes, and the dashboard header switches to “◆ LONG.”

The price action is identical to the previous figure. The only thing changed is the memory length in Figure ①. If it remembers back 30 bars, the cross 20 bars later still qualifies, and a BUY prints as Figure ② shows. Here’s how the counts compare on Bitcoin.

Settings (BUY/SELL counts)	15m	1H	4H	Daily
Default	0/0	0/0	0/0	0/0
OB/OS Recall Bars 20	5/3	1/5	6/3	0/0
OB/OS Recall Bars 30	15/18	13/17	21/11	2/0
Recall 30 + Trending Threshold 0.35	39/32	33/34	31/20	6/2
Overbought/Oversold changed to ±35	0/0	0/0	0/0	0/0

Even with OB/OS Recall Bars maxed out at 30, the 4H gives about 30 signals over five and a half years, roughly one every two months. Only when you also drop Trending Threshold to 0.35 does the count visibly pick up. On the other hand, tightening Overbought Level and Oversold Level in to ±35 left the count at zero. The bottleneck is the memory length and the timing of the trend classification.

More signals doesn’t mean better results

All I counted here is how often a label printed. I didn’t evaluate how price behaved afterward. Stretching the memory to 30 bars means it starts catching second crosses that happen after price has already recovered from a dip well in the past. Treat it as changing the character of the signal away from the “reversal out of oversold” the author intended. The sample is also too small to say anything statistically meaningful about win rates.

Divergence dots print away from the trough, at the height of “5 bars later”

Divergence works by confirming an oscillator trough (or peak) with 5 bars on each side, comparing it to the previous one, and plotting a dot when price moved the opposite way. Bullish is blue, bearish is magenta. This is another quirk I only discovered by actually using it.

Divergence dots are plotted at “the value 5 bars later,” not at the trough — A trough is confirmed by checking 5 bars on each side, so the dot appears 5 bars after the trough. Its horizontal position goes back to the trough bar, but its vertical position is the oscillator value at confirmation. The faster the rally, the higher the dot floats above the trough.

Price makes a lower low in Figure ①, while the oscillator makes a higher low in Figure ②. A classic bullish divergence. But the blue dot floats well above the trough, as in Figure ③. That one caught me off guard.

A trough is only confirmed once 5 bars have printed to its right. The dot’s horizontal position is drawn back at the trough bar, but its vertical position uses the oscillator value at the moment of confirmation. If price rips higher after the low, the dot ends up far away from the trough.

Quant Regime Oscillator's divergence dots are confirmed 5 bars after the trough and plotted at the value from that moment. — Quant Regime Oscillator’s divergence dots are confirmed 5 bars after the trough and plotted at the value from that moment.

I confirmed this on the actual Bitcoin daily chart too. The June 30, 2026 trough had the oscillator at −59.6, yet the blue dot sits at +6.2. If you just look at the dot and read it as “divergence near zero,” you’re reading it wrong. Take the dot as a marker that says “there was a trough around this bar”, and check the depth of the trough on the line itself.

What it actually compares: the low of the bar where the oscillator bottomed

One more detail. According to the author’s description, oscillator troughs and peaks are compared with price troughs and peaks. What’s actually compared is the low of the bar where the oscillator bottomed (or the high, for peaks). It doesn’t look for a separate swing low in price. The offset is usually small. But when the actual price low and the oscillator trough are a few bars apart, the result can disagree with what you’d see by connecting the lows on the chart.

Also, it only compares against the single previous trough (or peak), and it doesn’t care whether that’s above or below zero. You can get a bullish divergence above zero. Dots never move on confirmed bars, but on the bar that’s still forming, a dot can pop in and out depending on how that bar trades.

Parameters and settings by use case

The settings are split into five groups, with 23 inputs in total. The input names are shown in English, so I’ll refer to them by those names as-is.

01 · Engine (the oscillator itself)

Setting	Default	Suggested	Effect
Z-Score Length	30	30	Length of the SMA and standard deviation used as the Z-Score baseline. Longer values measure distance from a slower mean and calm the line’s swings. The regime’s up/down call also uses this mean
Momentum (ROC) Length	14	14	How many bars back to measure momentum against. The same length is used for the dispersion calculation
Weight · Z-Score	1.0	1.0	Weight of the Z-Score component. Set to 0 and the line becomes pure momentum
Weight · Momentum	1.0	1.0 (0.5 if the pinning bothers you)	Weight of the momentum component. Lowering it reduces how much the ±100 pinning affects the line
Oscillator Smoothing	2	2〜4	EMA length used to smooth the line. At 4, crosses with the signal line drop by about 25%
Signal Line Length	9	9	EMA length of the signal line. Longer means fewer crosses but more lag

In this group, Oscillator Smoothing is about the only one I touch. On the 4H, raising it from 2 to 4 cut signal-line crosses from 15.0 to 11.2 per 100 bars. The line gets less jagged, which makes it easier to read if you like using crosses as a cue. Setting Signal Line Length to 14 also cuts crosses to 12.8, but the signal line lags a lot more.

02 · Regime (regime classification)

Setting	Default	Suggested	Effect
Efficiency Ratio Length	20	20	Lookback for measuring ER. Longer means fewer Trending bars, shorter means more
Trending Threshold	0.45	0.45 (0.35 if you prioritize signals)	ER at or above this is Trending. Lower it to get more blue/purple background and more signals
Mean-Revert Threshold	0.25	0.25	ER at or below this is Mean-Rev. Keep it lower than Trending Threshold

Efficiency Ratio Length has a big impact. Comparing the share of Trending bars on the Bitcoin 4H, the default 20 gave 14.6%. At 40 it drops to 4.9%, and at 10 it jumps to 29.8%. Go shorter if you want to see the background more often, and longer if you only want color on truly strong trends.

03 · Signals (signals and divergence)

Setting	Default	Suggested	Effect
Overbought Level	55	55	Overbought threshold. Used for the SELL condition and the dashboard’s State
Oversold Level	-55	-55	Oversold threshold. Used for the BUY condition and the dashboard’s State
OB/OS Recall Bars	6	6 (30 if you want to see signals)	How many bars the memory of crossing a threshold lasts. Has the biggest impact on signal count
Min Bars Between Signals	10	10	Minimum spacing between BUY/SELL signals. Shared by BUY and SELL, so no SELL can fire right after a BUY
Show Divergences	On	On	Shows the divergence dots
Divergence Pivot Len	5	5	Bars on each side used to confirm troughs and peaks. The dots are also delayed by this many bars

On the Bitcoin 4H, setting Divergence Pivot Len to 3 increased the number of dots by about 40%, while 8 cut them by about 20%. Smaller values fire earlier but pick up even minor troughs. Larger values filter dots down, but the confirmation delay grows by the same number of bars.

04 · Visuals and 05 · Dashboard (display)

Setting	Default	Suggested	Effect
Gradient Fill	On	On	Gradient between the line and 0
Oscillator Line	On	On	The oscillator line. The fill stays even when turned off
Signal Line	On	On	The white signal line
Regime Background	On	On	Regime background color
Signal Markers	On	On	BUY/SELL labels. Turning them off doesn’t change the alert conditions
Show Dashboard	On	On	Shows the dashboard
Position	Top Right	Your preference	Dashboard position (choose from 8 spots)
Size	Small	Small	Dashboard text size (Auto, Large, Normal, Small, Tiny)

Example settings by goal

Goal	What to change	Why
Read it as an oscillator	Keep the defaults	The share of time beyond ±55 barely changes across assets and timeframes, so you can compare directly
Also see BUY/SELL	OB/OS Recall Bars at 30, and Trending Threshold at 0.35 if needed	The combination that produced signals in testing. Even then, expect one every few weeks to few months
Calm the line down	Oscillator Smoothing at 3–4	Fewer crosses and a less jagged line
More trend background	Efficiency Ratio Length at 10–15	The shorter the lookback, the easier it is to classify as Trending
Filter divergences	Divergence Pivot Len at 8	Fewer dots, but they print 8 bars late

Same yardstick across assets

I checked 12 combinations across Bitcoin, Ethereum, and PAXG from the 15m to the daily. The line spent 15–24% of the time at +55 or higher and 12–16% at −55 or lower. Being able to compare different assets and timeframes straight out of the box is a clear strength of this indicator.

How I use it: set direction elsewhere, then use QRO to gauge pullback depth

Since signals are so rare, I felt it would be a waste to sit around waiting for labels from “Quant Regime Oscillator.” So what I’ve been testing is handing the directional call to a long-term moving average and using QRO only to gauge the depth and timing of pullbacks and rallies. For the partner, I went with the 200 EMA, which a lot of traders watch.

Set direction with the 200 EMA, gauge pullback depth with QRO — “Quant Regime Oscillator” alone doesn't print a BUY here, because the uptrend classification (Efficiency Ratio 0.45 or higher) rarely kicks in right after a pullback. Handing direction to the 200 EMA lets you use the oscillator's oversold readings as a dip gauge. The candidate marker is an annotation, not something the indicator plots.

1Set direction with the 200 EMA

If price is above the 200 EMA as in Figure ①, I only look for longs. Below it, only shorts. Since direction is decided here, I don’t rely on QRO’s regime classification.

2Wait for QRO to pull back to −55 or lower

In an uptrend, a drop below −55 like Figure ② is a deep pullback, with price stretched well below its 30-bar mean. I pass on shallow dips.

3Flag a candidate on the cross above the signal line

When the oscillator crosses above the signal line in Figure ③, that’s a buy-the-dip candidate. Place your stop on the price chart, for example just below the pullback low.

There’s one goal here. What this approach replaces is the trend classification in “Quant Regime Oscillator.” Because ER looks at the efficiency of the last 20 bars, it’s always going to be low right after a deep pullback. The better the buy-the-dip setup, the more likely the uptrend classification is to be missing, and that’s baked into how it works. Handing direction off to a longer-period moving average sidesteps that conflict.

The “candidate bar” in the figure isn’t something the indicator plots

The “candidate bar” in the figure is an annotation for explanation only. In this situation, “Quant Regime Oscillator” doesn’t display a BUY. If you want notifications, a practical split is to set an alert for when the oscillator crosses above the signal line, then check where price is relative to the 200 EMA on the chart once the alert comes in.

As for timeframes, I mostly use it on the 1H and 4H. It works on the 15m as well. The line behaves the same, but Mean-Rev makes up over 60% of bars, so the background adds very little information. Even on an asset like gold, where rallies and pullbacks tend to be clearly separated, the numbers I counted on PAXG came out almost identical to Bitcoin’s.

What you can rely on, and what to watch out for

What you can rely on

See overextension and momentum in one line, using just one pane
Values behave almost the same across assets and timeframes, making comparisons easy
The ER background lets you check trend efficiency by color and by number
Confirmed bars never change after the fact
The dashboard lets you check the Z-Score and Momentum breakdown

What to watch out for

On default settings BUY/SELL almost never fires, and the dashboard stays at FLAT
The background is mostly gray, with few trend classifications
Divergence dots print at a height away from the trough, which is easy to misread
The momentum component pins at ±100 easily, pushing the line to the extremes in trends
The ±80 dotted lines are very faint and aren’t used in any logic

What to know before using it in real time

Confirmed bars don’t repaint. The forming bar moves

The oscillator, signal line, and background color are calculated using only values up to each bar’s close. There’s no higher-timeframe data and nothing that looks ahead. The display on confirmed bars never changes after the fact.

The bar that’s still forming is a different story. The values move every time the close moves. A BUY/SELL label can appear mid-bar and disappear before the bar closes. The author also states that signals lock in on the bar close.

Set alerts to “Once Per Bar Close”

Three alert conditions are available: QRO — Long, QRO — Short, and QRO — Any Signal. To avoid catching signals that appear and vanish on the forming bar, the safe choice is to set the trigger to Once Per Bar Close. According to TradingView’s help center, if you change the indicator’s settings after creating an alert, the alert keeps running on the settings it was created with. If you change OB/OS Recall Bars or anything else, recreate the alert.

One more important point. As we saw in the previous section, the conditions are almost never met on default settings. It’s easy to end up with an alert that sits there and never goes off, so keep that in mind.

“◆ FLAT” only applies to the loaded history

The dashboard header and past BUY/SELL signals are calculated over the bars loaded on the chart. How many bars you can load depends on your plan, with the free Basic plan reportedly capped at 5,000 bars (as of 2026/9/26). BUY/SELL labels on the chart are limited to the most recent 60, and older ones are removed automatically.

Use Bar Replay to see how it behaves

You can’t see divergence dots appearing 5 bars late, or the moment a signal fires, just by scrolling back through history. Stepping forward one bar at a time with Bar Replay lets you check it in an order close to real time. On the free plan, Replay is limited to the daily and higher, and intraday Replay and how far back you can go depend on which paid plan you have (as of 2026/9/26).

Less a buy/sell signal, more a market thermometer worth keeping on your chart

At the core of “Quant Regime Oscillator” are a single line that combines distance from the mean with momentum, and an ER-based read on the market’s character. Both are designed to be read on the same yardstick across assets and timeframes, and in 12 different combinations the values behaved almost identically. That consistency is a big draw if you want to compare across markets.

On the flip side, the BUY/SELL signals, one of its headline features, almost never fire on default settings. If you load it expecting signals, you’ll probably be puzzled by a dashboard stuck on FLAT. I know I was. If you want to see them, increase OB/OS Recall Bars, and use them knowing that both the frequency and the character of the signals will change. This is where opinions will differ.

Personally, I treat the labels as a bonus and keep it on the chart as an oscillator that measures “how deep did it pull back” and “has momentum come back” relative to a direction I’ve set with something like the 200 EMA. Just make sure you know from the start that divergence dots print at a height offset from the trough. Get that down, and it’s a genuinely handy oscillator that lets you check overextension, momentum, and market character in a single pane.