Limitations & Trust Boundaries

Limitations & Trust Boundaries

Every tool has a region where its output is reliable, a region where it is approximate, and a region where trusting it will get you in trouble. This page maps those regions for Axiom BB Pro.

This is not a disclaimer section. These are the things you need to understand to use the indicator well — the places where the reading gets soft, the assumptions that are easy to forget, and the gap between what the chart shows and what you can actually conclude from it.

The trust fulcrum: On Bar Close

The single biggest determinant of whether you can trust what you see is the On Bar Close setting. Everything else in this page is secondary to this.

What you can trust when On Bar Close is on

• The band values on historical bars are the same values that were visible when those bars were live

• The historical indicator values match what the script could actually know live — no look-ahead, no future-leak, no silent revision of past values

• The bands update predictably: once per higher-timeframe candle close, not during the candle

• Alert conditions evaluated against these bands were evaluated against confirmed data

What changes when On Bar Close is off

• The bands update with every chart bar as the higher-timeframe candle builds

• The values you see during the candle are provisional — they reflect an incomplete candle and may change before the candle closes

• On historical bars, you see only the final settled value — not the provisional values that would have been visible live during the candle

• This means any backtest done with On Bar Close off is testing against information that was not actually available at the time

• The backtest will overstate accuracy because it shows settled values as if they were known in advance

This is not a subtle distinction. It is the difference between an honest chart and one that is flattering the past. If you turn On Bar Close off, know what you are getting and what you are giving up. See MTF & Repainting for the full mechanical explanation and a verification walkthrough.

The blended band: structural reference, not superior signal

The blended band creates a visual that looks authoritative. A single smooth envelope, wrapping price in a clean channel, drawn with heavier lines and a fill. It looks like it knows something the individual slots do not.

It does not. The blended band is a normalized weighted average. Upper bands are averaged with upper bands. Basis lines are averaged with basis lines. Lower bands are averaged with lower bands. Each slot contributes in proportion to its weight.

What you can trust about the blend

• It accurately represents the weighted middle ground between your configured slots

• If the weights are thoughtful and the slots are well-chosen, the blend provides a useful structural center-of-gravity reference

• Its arithmetic is correct: equal weights produce a midpoint, skewed weights shift proportionally, zero-weight slots are excluded

• You can verify this yourself by toggling slots on and off and checking that the blend moves as expected

Where the blend gets soft

Smoothing over disagreement. When a 5-minute BB says price is compressed and a 60-minute BB says price is extended, the blend splits the difference. The resulting "middle ground" may look calm when the underlying reality is tension. The most interesting structural information — the disagreement between timeframes — is exactly what the blend averages away.

Sensitivity to weight choices. The blend only reflects what you told it to reflect. If you set the 60-minute weight to 80 and the shorter-timeframe weights to 10 each, the blend is mostly the 60-minute BB with a little smoothing from below. That might be what you want, or it might be an accidental configuration that hides the shorter-timeframe information you thought you were seeing.

Hidden-slot contribution. A slot can be hidden on the chart but still contribute to the blend. This means the blended band can be shaped by bands you cannot see. If the blend looks unexpected, check whether hidden slots with non-zero weights are influencing it.

No independent analytical power. The blend does not compute anything new. It does not detect patterns the individual slots miss. It does not generate insights beyond what a human could see by visually estimating the average of the visible bands. Its value is convenience and precision in that averaging — not any deeper analytical capability.

The over-trust pattern

The most common way the blended band gets misused: a trader sees price touch or cross the blended upper band and treats it as a breakout, or sees price touch the blended lower band and treats it as a reversal level. The blend's visual weight — heavier lines, the fill, the way it sits on top of everything — reinforces the feeling that it is the definitive reading.

Here is what is actually happening when price crosses the blended upper band: price exceeded the weighted average of several upper bands at different timeframes. One of those upper bands might already have been crossed two hours ago. Another might still be well above price. The "cross" is an event in the averaged space, not in any specific structural space. It tells you less about the market than a cross of any individual slot's upper band at a timeframe you understand.

The self-check: when you notice yourself reacting to a blended-band event, stop and ask which individual slots price actually crossed. If price is above the blended upper but still inside the 60-minute upper, the "breakout" is an artifact of averaging, not a structural break at the timeframe that matters most to you. If price has crossed the upper band of every individual slot, the blended cross is confirming what the slots already told you, and the slots told you first.

Over-trust with the blend tends to build gradually. You start using the blend as a quick reference. You stop checking the individual slots as often. The blend becomes the primary reading. And then, on a day when the blend and the slots disagree, you follow the blend because you have been trained by weeks of using it as the main view. That is the pattern to notice before it sets in. The blend is the summary. The slots are the source. When they conflict, the source wins.

Cross-ticker scaling: an approximation, not a projection

When a slot uses a different symbol via the Optional Ticker setting, the BB values from that symbol are scaled into the chart symbol's price space using a close ratio. This makes the bands appear in a visually useful range on the chart.

What you can trust about cross-ticker scaling

• The scaling preserves the proportional structure of the reference symbol's BB — if the reference bands are wide relative to its price, they will appear wide relative to the chart price

• The ratio updates per bar, so it adapts as the relationship between the two symbols evolves

• The ratio respects the On Bar Close setting, so confirmed-close scaling is stable and consistent

Where it drifts

Close-ratio instability. The ratio is derived from the closing prices of both symbols. When that ratio changes — for example, if SPY rises 2% while the target stock drops 5% — the scaling shifts proportionally. The cross-ticker bands will appear to move even if the reference symbol's actual BB did not change, because the scaling multiplier changed.

Not a correlation measure. The scaling does not adjust for correlation between the symbols. It maps one price space into another using a ratio. If the two instruments diverge structurally (one trends up while the other trends down), the projection becomes less meaningful — you are mapping a structure that no longer has a stable relationship to the chart instrument.

Large price-level divergence. When the chart symbol and the reference symbol have very different price levels (a $500 instrument vs. a $10 instrument), the ratio is large, and small percentage changes in either symbol produce proportionally large shifts in the projected bands. The scaling works best when the two symbols have a relatively stable proportional relationship.

What not to treat as support or resistance

Cross-ticker band levels are not the same as native BB levels on the chart symbol. They are projections. The reference symbol's BB levels exist in the reference symbol's price space — the scaling maps them into the chart's space for visual context, but the chart symbol's price has no intrinsic reason to respect those levels. Treat them as directional and structural context, not as price levels with their own gravitational pull.

Alignment: a filter, not a verdict

The alignment state (all slots above basis or all below basis) is one of the more useful readings the indicator provides. It tells you that all your configured timeframes agree on which side of the structural center price sits.

What alignment tells you

• The near-term and higher-timeframe structures currently agree on positioning

• The possibility space is narrowed — the set of structural configurations you might be facing is smaller when all timeframes agree than when they disagree

• Full alignment is relatively uncommon, which makes it worth noticing when it occurs

What alignment does not tell you

How stable the agreement is. Alignment can break on the very next bar. A 5-minute slot can cross its basis while the 60-minute slot remains positioned — and alignment is gone. The fact that alignment existed a moment ago does not predict how long it will last.

How deep the positioning is. All slots could be above basis, but the lowest-timeframe slot might have just barely crossed while the highest-timeframe slot has been above basis for days. "All above" does not distinguish between early agreement and deep agreement.

Whether the agreement means anything directionally. Some of the most dangerous moments in trading occur at the peak of alignment. The sequence is familiar: the higher-timeframe slots have been positioned for a while. The mid-range follows. And then the lowest-timeframe slot finally crosses its basis to complete the set. At that moment, every configured timeframe agrees, and the feeling is unmistakable — this looks like confirmation. Everything points the same way.

But what actually happened is that the last lagging timeframe caught up at the point of maximum extension. The reversal that follows can be sharp, because the structural agreement that felt like strength was actually the final stretch. This does not happen every time alignment forms. But it happens often enough that alignment should reduce your uncertainty about the current state without reducing your caution about what comes next. Agreement tells you where things stand. It does not tell you how long they will stand there.

Backtesting reliability

Axiom BB Pro can be combined with other tools and strategies for backtesting purposes. The reliability of the backtest depends almost entirely on the On Bar Close setting.

Safe for backtesting

• All slots set to On Bar Close = true

• The bands on historical bars reflect the last confirmed HTF candle — no future information is embedded

• What you see on the historical chart is what a live user would have seen at the time

Unreliable for backtesting

• Any slot with On Bar Close = false

• Historical bars show the settled final value of the HTF candle, not the provisional values that were visible during the candle

• This means the backtest looks better than reality — it appears to show a clean reading that was never actually available during the live candle

• The degree of overstatement depends on how volatile the HTF candle was during its formation. Calm candles barely change from provisional to settled. Volatile candles can change significantly.

There is no way to backtest the provisional values. TradingView does not store the mid-candle state of request.security() calls. The historical bars always show the settled result. This means you cannot retroactively evaluate what On Bar Close = false would have looked like in real time. The only way to evaluate it is to watch it live — which is time-consuming and difficult to do systematically. This is one of the reasons the default is On Bar Close = true: it makes the backtestable version the one you are most likely to use, which means the version you test and the version you trade are the same. When those diverge, trust erodes in ways that are hard to trace back to the cause.

The gap between indicator output and tradeable conclusions

This is the broadest trust boundary, and it applies to every reading the indicator produces.

Axiom BB Pro shows you where price sits relative to volatility envelopes at multiple scales. It shows you when timeframes agree or disagree. It shows you whether the structural centers have been crossed. It does all of this honestly, given its settings.

What it does not do:

• It does not tell you whether a move will continue or reverse

• It does not tell you whether a band touch is a bounce or a breakout

• It does not tell you whether alignment will hold or break

• It does not tell you whether cross-ticker context is relevant to the chart instrument's next move

• It does not tell you when to enter, when to exit, or how much to risk

The indicator provides structure. You provide the judgment. The structure is useful — it saves you the work of running multiple BB indicators and doing the timeframe comparison in your head. But the judgment cannot be outsourced to the tool.

The boundary is subtle, and it shifts depending on the pressure you are under. When you are calm and analytical, you read the indicator as structural context — data that informs your process. When you are stressed, underwater on a position, or desperate for confirmation, the same visual can start to feel like an answer. The blended band looks like it is confirming your view. The alignment alert feels like a green light. The cross-ticker overlay seems to be agreeing with you. The indicator did not change. Your relationship to it did.

Noticing that shift in yourself is harder than understanding any of the mechanics on this page. But it is more important. The indicator is an instrument. It shows you where things are. It does not tell you what to do, and the moment it starts feeling like it does, you are no longer reading — you are looking for permission. That distinction is the real trust boundary.