Limitations & Trust Boundaries

This page is not a disclaimer. A disclaimer is a sentence at the bottom of a marketing page that exists to protect the author. A trust boundary is a working part of the manual that exists to protect the reader. The two look similar on the page. They do opposite work.

The list below names what Axiom BB CTX is reliably good at, what it cannot tell you, and the specific places where a careful reader needs to verify rather than assume. It is written so that a reader who studies this page can put the tool on a chart and know, in advance, which of their own instincts about it are going to be correct and which are going to be off by the width of an honest answer. None of it is hidden in the corners; all of it is in the open so you can decide what to lean on and what to leave alone.

What this tool is reliably good at

Within the postures it ships with, the indicator does the following well.

• Drawing several Bollinger Band reads on one chart, in their native timeframes, without you having to maintain separate panes. The slot model is honest: each slot is configured in one place, computes against its own configuration, and plots in its own color.

• Holding a higher-timeframe value steady through a chart-bar cycle when On Bar Close? is ON. Once the slot's higher-timeframe bar has closed, the plotted value is fixed for the duration of the next higher-timeframe bar. That stability is what lets a 1-minute chart carry a 60-minute Bollinger context without lying about it.

• Producing a weighted composite that uses a stated rule. The blend is sum(weight * value) / sum(weight) for basis, upper, and lower, restricted to enabled slots with a non-zero weight. Early warm-up bars are not reliable blend reads because missing slot values are normalized before the composite is drawn. The rule is not a black box.

• Naming the slot that violates the chart-timeframe guard. When the chart goes blank, the runtime error tells you which slot to fix.

• Bringing an alternate symbol's volatility envelope into chart price space at a stated, simple scaling. The cross-ticker reading lets you study a related market without leaving the chart.

• Surfacing every alert condition with a fixed name and a stated firing rule. No alert in the indicator is hidden or implicit; every condition you can wire is in the alert dialog.

What this tool cannot tell you

Within the same scope, the indicator does not do the following, and any reading that pretends it does is a misread.

• Direction of price. The basis-trend state is a fixed-lookback comparison of the basis against itself. It is not a regime classifier and it is not a forecast.

• Setup quality. A band touch, a basis flip, an alignment alert — they are inputs to your read, not graded recommendations. The indicator does not score "good touches" against "bad touches."

• Cross-symbol causality. The cross-ticker scaling expresses one symbol's envelope in another symbol's price space. It does not claim the alternate symbol leads, lags, or pulls the chart symbol in any direction.

• Whether your weights are right. The blend executes the weights you (or the defaults) gave it. If the weights describe a process you do not actually have, the blend will faithfully report a picture that does not match how you trade.

• Whether Length: or Basis Trend Length: is well-tuned for your symbol. Both are choices. Both have honest defaults. Neither has an objectively correct value for your workflow.

• Whether an alignment alert means "now is the moment." Alignment is one fact among many. The indicator confirms the fact. It does not interpret it.

• The MA-type semantics of ALMA, KAMA, FRAMA, Jurik, Laguerre, or VAMA. Those live in the Axiom MA Pro library. If a slot's basis curve does not look right after a Type: change, the answer is in that library's manual, not in this one.

Trust boundary: the blended band is a composition, not an answer

The blended band is the most over-trusted element on the chart. It is the most visually dominant — red lines, a translucent fill, a lime basis — and it sits on top of everything else. Readers naturally treat it as the conclusion the indicator is offering.

It is not. It is a weighted average of the slots you enabled with non-zero weight. It loses information by design (averages always do) and it is composed in proportions you set.

Three behaviors keep the blend honest in your read.

1. Always scan the contributing slots before you act on the blend. A narrow blended band could be agreement; it could be cancellation. The slots tell you which.

2. Weight by what you trust, not by what feels balanced. Equal weighting is a deliberate statement that every contributing slot is equally informative. If that is not what you mean, do not weight that way.

3. Treat the blend as a summary line, not a level. The blended basis is not "the" basis. It is a weighted average of bases, on a chart that already shows you those bases.

The fast diagnostic: when a decision feels like it is hanging on the blended band alone, hide the blended plot for a minute (Hide Blended BB Plot = true). If the per-slot picture leads you to the same conclusion, the blend was reflecting the slots and the decision is grounded. If hiding the blend leaves you uncertain, you were trusting the summary more than the underlying read.

Trust boundary: hidden does not mean excluded

Stated three times in this manual already, and stated once more here because it shows up in real losses.

Hide BB NN Plot is a visibility choice. It removes the slot's three lines from the chart. It does not remove the slot from the blend, from the alignment alerts, or from the hidden alert-count plots.

If you find yourself surprised that:

• the blend moved when nothing visible moved, or

• an alignment alert is silent when every visible slot agrees, or

• a basis-trend alert fired for a slot whose lines you cannot see —

look for an enabled-but-hidden slot. The fix is the right kind of exclusion for what you actually want.

• Out of the blend: Blended Weight = 0.

• Out of the alignment alerts: Enable BB NN = false.

• Out of the per-slot basis-trend alerts: also Enable BB NN = false.

Three controls, three different exclusion semantics. Pick the one that matches the result you want.

Trust boundary: cross-ticker scaling is a study, not a forecast

Setting Optional Ticker: on a slot tells the slot to read its Bollinger Band from that symbol, at the slot's timeframe, and then rescale the result into the chart symbol's price space so the lines land where you can see them.

The scaling preserves the relative shape of the alternate symbol's envelope. The bands' relative width and the basis's relative position inside that envelope come through to the chart. What does not come through is any claim about how the two symbols relate causally. A QQQ band rescaled into SPY's price space is a study of QQQ's volatility, expressed in SPY's units. It is not a leading indicator for SPY. It is not a confirmation for SPY. It is QQQ, scaled.

There are three failure modes a careful reader watches for.

• Session mismatch. If the alternate symbol trades on hours the chart symbol does not (or vice versa), the scaled band can go stale during one symbol's session gap. The chart will keep drawing; the scaled lines may quietly stop reflecting current behavior.

• Volatility regime divergence. When two symbols are in very different volatility regimes, the scaled band can drift far from the chart's price. That distance is not a "pull" — it is just the math reflecting the divergence honestly.

• Repaint posture inheritance. A cross-ticker slot uses the slot's On Bar Close? posture. A cross-ticker slot under OFF can drift on the live higher-timeframe bar of either symbol. The scaling does not hide that; it inherits it.

If you use a cross-ticker slot, keep its Blended Weight at zero unless you have a specific, stated reason to let an alternate symbol's envelope steer your blend. The "weight-zero, visible" pattern is the right home for cross-ticker context most of the time.

Trust boundary: alerts confirm conditions, not actions

Every alert in Axiom BB CTX confirms exactly one thing: the named condition was true at the close of a chart bar, evaluated under the slot configuration the indicator carried at that moment.

That confirmation is real. It is also narrower than the things readers commonly want from an alert.

An alert here does not confirm:

• that price will continue in the trend the alert describes,

• that the higher-timeframe bar has closed (for OFF slots),

• that the read is good for execution,

• that the slot configuration that fired the alert is the configuration you have on the chart now,

• that adjacent timeframes agree (per-slot alerts only describe their own slot; alignment alerts only describe the all-enabled-slots vote).

The honest mental model: alerts here are notifications about a computed condition holding at a chart-bar close. They are useful for getting your attention. They are not a substitute for looking at the chart.

A specific anti-pattern worth naming: wiring a blended-trend alert and treating it as if it were the indicator's "verdict." The blended trend is a weighted vote across your contributing slots. It can flip on a small slot crossing its trend-length threshold, especially with skewed weights. A blended-trend alert is a description of a vote outcome, not a graded read.

Trust boundary: the defaults are a starting position, not a recommendation

The defaults exist so the indicator loads cleanly on a 1-minute chart of a liquid symbol. They are the canonical Bollinger settings (length 20, multiplier 2.0, SMA basis), three intraday timeframes (5m, 15m, 60m), equal weighting (33.3 across the three enabled slots), and ON for repaint posture across the board.

That configuration is a working baseline. It is not a curated stack. The pack does not claim that:

• 5m / 15m / 60m is the right ladder for your workflow,

• 20-bar / 2.0-multiplier is the right Bollinger geometry for your symbol,

• equal weighting reflects how much you trust each slot,

• SMA is the right basis for your read.

If your read suggests a different value, change it. If you do not know whether your read suggests a different value, hold the default until you do. The honest move is the deliberate one, not the default-preserving one.

Trust boundary: power-user MA parameters belong to the MA Pro library

Each slot's Type: selection determines which visible PU parameters actually matter. ALMA uses offset, sigma, and floor; KAMA and FRAMA use fast and slow lengths; Jurik uses phase and power; Laguerre uses alpha; VAMA uses a volatility lookback.

The semantics of those MAs — what each one is trying to accomplish, when each one is the right choice, what reasonable parameter ranges look like — are governed by the Axiom MA Pro Library manual. This pack does not re-teach them, and any answer this pack gave to "what should ALMA Sigma be?" would be a worse answer than the library's.

If a curve looks wrong after you change Type:, the most efficient debug path is to read the library's page on that MA before you treat it as an Axiom BB CTX problem. Most "the basis looks weird" reports trace back to a PU parameter mismatch, not a slot configuration error.

Trust boundary: this is a discretionary tool, not a system

Axiom BB CTX gives you a workbench: ten slots, a weighted composite, a per-slot repaint switch, alerts on basis-trend conditions, cross-symbol context. It does not give you a system.

A system says: "when these conditions are true, take this action." Axiom BB CTX never says that. The closest it comes is All BB Slot Bases Uptrend, which says "every enabled slot's basis is rising at this chart-bar close." That is a fact, not an instruction.

Building the system around the workbench is the reader's job. The pack does not do it for you, and it does not pretend to. That ownership is the whole point: you compose the read, you wire the alerts, you set the weights, you decide what the reading means.

If you want a tool that decides for you, this is not it. If you want a tool you can inspect, configure, verify, and own, this is it.

What you can actually rely on

Reduced to a single list, the durable trust statements are:

• A slot under On Bar Close? = ON does not change its plotted value within a higher-timeframe bar.

• A slot's TimeFrame: lower than the chart raises a named runtime error; nothing renders silently wrong.

• The blended band is sum(weight * value) / sum(weight) over enabled slots with a non-zero weight — for basis, upper, and lower. Do not rely on the first warm-up bars, where missing values have not resolved into real slot readings yet.

• The blended basis trend is the side whose contributing weight is larger, with ties resolving toward uptrend.

• Alerts fire on the close of the chart bar; the underlying basis-trend conditions are computed against Basis Trend Length bars of the slot's basis.

• Hidden plots stay in the math; disabled slots are out.

• Cross-ticker bands are scaled into chart price space at a stated, simple ratio; they inherit the slot's repaint posture.

Those are the statements you can defend to yourself in a bad moment. Every other reading on the chart depends on how you configured the tool and what you brought to it. The indicator will not make that part easier by pretending the harder parts are its job.

Where to go next

• For the live story behind ON vs. OFF, MTF & Repainting.

• For exactly how the blend composes its bands and resolves its trend vote, For the Geeks.

• For documented setup recipes that respect the boundaries above, Workflows.

• For the alert posture in detail, Alerts.