Indicator Management
One single place for everything to do with indicators: looking up what each one does and deciding the values it starts from. Open it from Indicators → Manage Indicators and has two views.
Catalog
The card for each indicator: its formula, what it measures, a drawing of the pattern where there is one, and a guide to when to use it. This is the view for learning, not for changing anything — the parameters at the bottom are read-only.
Settings
The grid with the default values for all 78 indicators: source, period, offset, operators, threshold, availability and lock. These values are inherited by the whole program: the Strategy Lab, AQ Genesis, the Edge Scanner and the Walk-Forward all start from them.
One indicator, not its outputs
Bollinger Bands are not three indicators: they are one indicator with three lines, and all three share source and period because they are the same calculation. That is why they appear as a single entry, with a badge saying how many lines it governs. Whatever you change applies to all of them: they cannot get out of sync.
Bollinger Bands 3 → BB_Upper · BB_Mid · BB_Lower
The same goes for the Keltner Channel, the Donchian Channel, the MACD and the Stochastic. The line names still exist: you can write BB_Upper(C,20) in the AQ Editor, and they show up in the exports just as always.
The ⇆ chip — crossing between lines
By default, the lines of one indicator are only compared with each other when they share a configuration: BB_Upper(20) with BB_Lower(20), not with BB_Lower(35).
The chip lifts that restriction. It is not that the other pairings are meaningless —a short channel lifting off above a long one does say something— but for every crossing that reads well there are dozens of trivial ones, and the cost is not small:
The problem is not computing time: it is that with 461 million tests you are going to find something spectacular by pure chance. That is exactly what the PBO/DSR module measures. Check the combination counter before you launch.
Period and Offset are not the same thing
The size of the window. Lowest(C,20) looks at twenty bars, and it always ends on today's bar.
How many bars back it reads. Lowest(C,20)[1] looks at those twenty ending on the previous bar.
The distinction matters more than it looks, because there is a whole family of strategies that can only be expressed with the offset: the channel breakout.
Close > Highest(C,20)[1]
// With [0] it is NEVER true: Highest includes today's bar,
// so Close can never beat it.
CLOSE, OPEN, HIGH and LOW have no period: they are a single value per bar, there is no window to widen. Their only parameter is the offset, which is why their Type and Period columns show «—». A range of 0 to 5 on CLOSE generates CLOSE[0], CLOSE[1] … CLOSE[5]: the same price at six different moments.
Available
Turn it off and the indicator stops being offered in the Strategy Lab, AQ Genesis, the Scanner and AQ Mentor. Strategies you already saved that use it keep working: it is simply no longer offered.
Lock 🔒
It stops you from overshooting the period and threshold ranges in the Strategy Lab. Source and operators stay free: there are no absurd values there, only decisions. A 30,000-period SMA on 6,686 bars returns nothing, and no error warns you about it.
Modified ●
The chip next to the counter tells you how many entries differ from the factory settings in the category you are looking at, and one click leaves only those. The arrow on each row returns it to its original value.
The defaults are a starting point
The catalog comes preconfigured indicator by indicator, but those are suggestions, not obligations: in the Strategy Lab you can change any of them for that particular sweep without touching the general settings. And the other way round, whatever you set here is inherited by every module, including the ones that do not show those settings —AQ Genesis or the Edge Scanner—, so this is the place where one decision applies to the whole program.
Each of the four series reads its own price
HIGH[3] returns the high of three bars ago, not the close. It sounds obvious, but this is exactly where a backtest engine can get it wrong without ever revealing itself: if all four series returned the same thing, the numbers would still be consistent between the backtest and the exports, and nothing would look wrong.
The symptom would not help either: conditions between two series would simply never be true —HIGH > LOW would be CLOSE > CLOSE, always false—, so those strategies would never appear in the results and there would be no error to give it away. Only their absence.
The threshold accepts zero
A threshold of 0 to 0 is a legitimate request, and it is the canonical use of several indicators: the zero crossing of the MACD, of Momentum, of the ROC or of the TRIX. On Momentum and TRIX it is in fact the only threshold that makes sense, because they are in price units: a «30» means 30 points on an index and 30 dollars on a stock — it is not the same question.
This also sorts out the candlestick patterns, which are worth 1 or 0: otherwise they would reach the Strategy Lab asking for Doji = 30, a condition that is never true. They would burn combinations and return zero trades without warning. The same goes for the calendar filters (DayOfWeek 0..6, HourOfDay 0..23).
AniQuant can be tried free for 30 days, with every module and no card.