time-specific standard devs [keypoems]

6月4日

Time-Specific Net Change Percentage Standard Deviations Extensions

A candlestick is the story of how price moved during a single period: it opens, explores a high and a low, and finally closes.

The indicator asks one question:

“By what percentage did price change from open to close?”

That single percentage value – called the net-change % – is the raw material for everything that follows.

Why net-change % matters

If you collect that percentage for thousands of finished candles (daily candles, 1-hour, 4-hour – whatever you close) you get a probability distribution: most candles move only a little, a few move a lot. When you plot those percentages you get the familiar bell-curve. From any bell-curve you can read two key statistics:

Mean – the “typical” net-change.
Standard Deviation (σ) – how far a candle normally wanders from that mean.

A well-known property of a bell-curve is that ≈68 % of all observations stay between –1 σ and +1 σ, and ≈95 % stay within ±2 σ.

For traders that translates into a probabilistic map of where an unfinished candle is likely to close.

What the indicator draws

Open-anchored levels
Every box is centred on the candle’s open price.
The indicator draws boxes that mark +σ, +½ σ, –½ σ, –σ … as selected in the Inputs.

Time-specific calculation

Instead of mixing all 1-hour candles together, the script isolates only the 09:00-10:00 New-York hour (or any other slot you are at that moment) and builds a bell-curve from that subset.

The result is a set of expectations truly tailored to that exact hour’s behaviour.

Generic and Specific Std Dev Box

Indicator computes both generic levels – the classic “all candles aggregated” standard deviations, and specific levels – the refined, hour-by-hour (or 4-hour-by-4-hour, etc.) deviations.
The indicator shows Boxes that represent both aggregated values and "per period" standard deviation projection levels.

Mirrored levels (optional)

Negative σ levels are drawn automatically so you have symmetrical upside & downside projections.

Math behind the scenes
For each chosen higher-timeframe (HTF):

Collect the last N net-change % values (user-defined look-back, default 5 000 candles).
Compute:

μ = mean(net-change %)
σ = stdev(net-change %)
Project prices:

LevelPrice = Open × (1 + k × σ)
where k is 0 , ±0.5 , ±1 , ±1.5 … as enabled in the inputs.

Because everything is done in percentage space the technique adapts automatically to instruments that trade at very different absolute prices (equities, futures, crypto, FX).

How to read it in real time

When a developing candle races straight into +1 σ with plenty of time left, statistics say “68 % of the time it will finish back inside the box.”
A tag of +2 σ is an even rarer event (~5 % probability) – a natural place to look for exhaustion or take-profit cues.
Conversely, if price refuses to reverse at +1 σ and instead sticks above that edge, it is signalling an unusually strong session and may warrant trend-continuation tactics.

Key Inputs

Higher Timeframe Choose 5m to H8 reference frame
Show levels for last X periods
Look-back length Sample size for σ
StdDev Fibs Fib1 … Fib9 toggles & values: Decide which σ multiples to plot
Mirror levels Draw matching negative projections
Extend boxes Keep boxes visible beyond their candle

Best-practice usage
Combine with structure – treat σ edges as context, not automatic entry.
Mind the sample size – if you choose extremely narrow windows (e.g., Sunday 18:00-19:00 futures open) make sure your look-back still captures hundreds of cases.

Disclaimer
Trading involves risk. The author and this indicator simply display historical probability ranges; they do not assure profits or prevent losses. Always perform your own due diligence and, if necessary, consult a licensed financial professional before dealing in any financial instrument.

6月4日

リリースノート

Changelog: Apologies for the bump the screenshot was broken.