Manual · 3 of 5

Indicator reference

Every feed carries the same 110 indicators, computed from its own candles with fixed parameters. This page tells you what each one measures, in which unit it is stored, how to read it, and the range it has actually taken for BTC and ETH since 2020.

How to read this page

Name. Column = <feed>_<indicator>, e.g. bn_spot_rsi. Pick the feed in the explorer to see the exact names.
Timing. Every indicator value on a row is the value as of the previous completed bar. It was known before this bar opened. Raw open / high / low / close / volume are the row's own bar. See Data & integrity.
Unit. The badge names one of the unit classes below. Price-level indicators are stored as a % distance from the close (dev %), so “price above the moving average” is bn_spot_sma<0.
Observed range. min … max over the full history (1 Jan 2020 → 3 Sep 2026, a survey of every bar), with the average and the standard deviation. They describe reality, not the formula's limits; use them to pick thresholds that actually occur.
Null %. Share of bars without a value: the warm-up window at the start of the history, after an exchange gap, or “no event yet” for the event-distance columns. Null never matches a comparison.
Example. The small link under each name runs a ready-made condition on the selected feed.

Four decimals

All served values are rounded to four decimals. Thresholds finer than 0.0001 are meaningless, and exact equality on a continuous indicator (bn_spot_rsi=50) rarely matches — use ranges.

Unit classes

Badge	Range	How to read
dev %	open, usually within ±3	(indicator ÷ close − 1) × 100. 0 = the line is exactly at the close. NEGATIVE = the line is below the price, i.e. the price is ABOVE it. Positive = the line is above the price.
% of close	open, signed	indicator ÷ close × 100. The sign is the indicator's own (a MACD above zero stays positive); the size is a percentage of the close, so it is comparable across coins, exchanges and years.
% of close	0 and up	indicator ÷ close × 100; never negative.
%	open, signed	A ratio multiplied by 100. +1 means +1 %.
%	0 and up	A ratio multiplied by 100; never negative.
%	depends on the indicator	The indicator's own percentage definition (TA-Lib already computes it in percent).
0–100	0 to 100	Classic oscillator scale. 50 is neutral for RSI-type indicators.
−100–100	−100 to 100	Zero is neutral; the sign gives the direction.
−100–0	−100 to 0	Williams %R scale: near 0 = close at the top of the range, near −100 = at the bottom.
CCI	±466.67 at most (14-bar CCI)	Unbounded in theory; ±100 are the usual hot / cold lines.
−1–1	−1 to 1	A balance or a correlation. 0 = balanced.
rank 0–1	1/window to 1	Percentile rank of the latest value inside its window. 0.95 = in the top 5 % of the window; 0.10 = bottom 10 %.
ratio	open	Dimensionless statistic; compare it with its own history rather than with a fixed line.
%³ / %⁴	open	Central moments of percentage returns (skew- and tail-type information in raw form).
exchange units	as served by the exchange	Not normalized. Prices are in the feed's quote currency (USDT, USD or KRW); volumes in base or quote units. Compare within one feed only.
ln(1+bars)	0 to 6.5117 (672 bars)	Natural log of (1 + bars since the event). 0 = the event was on the previous bar, 0.6931 = one bar before that, 1.3863 = 3 bars, 4.5747 = 96 bars (1 day), 6.5117 = the 672-bar cap.
bars	0 to 671	A plain bar count (15-minute bars).
0–1	0 to 1	A share. 0.55 = 55 %.

All 110 indicators (interactive)

FeedFamilySearch

Showing 110 of 110 indicators on Binance Spot (bn_spot). Ranges are observed values over the full history (1 Jan 2020 → 3 Sep 2026); null % is the share of bars where the value is missing (warm-up, exchange gaps or no event), and “since” is the first bar that carried a value.

Moving averages & trendlines

Where the price sits relative to a smoothed line. All of these are stored as a distance from the close in percent (dev %): negative means the price is above the line.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_sma` Simple moving average bn_spot_sma<0 and bn_spot_adx>=25 →	The plain 30-bar average of closes, as a % distance from the close. (SMA30 ÷ close − 1) × 100	30 bars · 7.5 h	dev % open, usually within ±3	-10.89 … 37.88 avg -0.0072 · σ 0.9878	-11.62 … 37.21 avg -0.0049 · σ 1.29	0.27 / 0.27 since 20200101.0730	< 0: price above its average (bullish bias). > 0: below. −1 = the average sits 1 % under the price.
`bn_spot_ema` Exponential moving average bn_spot_ema<0 and p:bn_spot_ema>=0 →	A 30-bar average that weights recent bars more (multiplier 2/31). (EMA30 ÷ close − 1) × 100	30 bars · 7.5 h	dev % open, usually within ±3	-10.51 … 33.27 avg -0.0073 · σ 0.8502	-11.18 … 33.28 avg -0.005 · σ 1.11	0.27 / 0.27 since 20200101.0730	Same reading as sma; reacts faster to new bars.
`bn_spot_wma` Weighted moving average bn_spot_wma<-1 →	Linearly weighted 30-bar average (newest bar weight 30, oldest 1). (WMA30 ÷ close − 1) × 100	30 bars · 7.5 h	dev % open, usually within ±3	-10.74 … 29.33 avg -0.0047 · σ 0.7634	-14.76 … 28.98 avg -0.0031 · σ 0.9959	0.27 / 0.27 since 20200101.0730	Same reading as sma.
`bn_spot_dema` Double exponential MA bn_spot_dema<0 and bn_spot_sma>0 →	A low-lag 30-bar average: 2·EMA − EMA(EMA). (DEMA30 ÷ close − 1) × 100	30 bars (58-bar warm-up)	dev % open, usually within ±3	-14.03 … 16.94 avg 1.99e-4 · σ 0.5836	-19.24 … 20.67 avg 2.58e-4 · σ 0.7594	0.47 / 0.47 since 20200101.1445	Hugs the price more closely than ema; its sign flips earlier.
`bn_spot_tema` Triple exponential MA abs(bn_spot_tema)<0.2 →	An even lower-lag 30-bar average: 3·EMA − 3·EMA(EMA) + EMA(EMA(EMA)). (TEMA30 ÷ close − 1) × 100	30 bars (87-bar warm-up)	dev % open, usually within ±3	-16.09 … 14.2 avg 2.88e-4 · σ 0.492	-18.77 … 16.63 avg 3.47e-4 · σ 0.6396	0.67 / 0.67 since 20200101.2200	Fastest of the MA family; small values are normal (std ≈ 0.5).
`bn_spot_trima` Triangular moving average bn_spot_trima<-1.5 →	A doubly-smoothed 30-bar average with the heaviest weight in the middle of the window. (TRIMA30 ÷ close − 1) × 100	30 bars · 7.5 h	dev % open, usually within ±3	-11.95 … 40.45 avg -0.0072 · σ 1.06	-12.74 … 38.88 avg -0.0047 · σ 1.38	0.27 / 0.27 since 20200101.0730	Smoothest of the simple family; good for 'is the price well above the middle of the last 7 hours'.
`bn_spot_t3` Tillson T3 bn_spot_t3<0 and p:bn_spot_t3>=0 →	A very smooth, low-lag average built from six chained EMAs. (T3 ÷ close − 1) × 100	5 bars, vfactor 0.7 (24-bar warm-up)	dev % open, usually within ±3	-17.31 … 15.65 avg -7.81e-4 · σ 0.4193	-16.63 … 18.63 avg -4.12e-4 · σ 0.544	0.24 / 0.24 since 20200101.0615	Small values (std ≈ 0.4); use it for sign changes rather than size.
`bn_spot_kama` Kaufman adaptive MA bn_spot_kama<-1 →	An average that speeds up in trends and slows down in chop (efficiency ratio). (KAMA30 ÷ close − 1) × 100	30 bars · 7.5 h	dev % open, usually within ±3	-15.43 … 19.1 avg -0.0029 · σ 0.8454	-20.76 … 26.34 avg -0.0081 · σ 1.11	0.28 / 0.28 since 20200101.0745	Flat in sideways markets, so \|kama\| stays small until a real move.
`bn_spot_mama` MESA adaptive MA bn_spot_mama>bn_spot_fama and bn_spot_ma… →	Ehlers' adaptive average that follows the dominant price cycle. (MAMA ÷ close − 1) × 100	fast 0.5 / slow 0.05 (32-bar warm-up)	dev % open, usually within ±3	-10.22 … 17.24 avg -0.0024 · σ 0.4575	-16.03 … 23.96 avg -0.0019 · σ 0.5964	0.29 / 0.29 since 20200101.0815	Bullish when the MAMA line is above the FAMA line: {P}_mama>{P}_fama.
`bn_spot_fama` Following adaptive MA bn_spot_mama>bn_spot_fama and p:bn_spot_… →	The slower companion line of MAMA. (FAMA ÷ close − 1) × 100	fast 0.5 / slow 0.05 (32-bar warm-up)	dev % open, usually within ±3	-10.97 … 30.45 avg -0.0107 · σ 0.8465	-17.08 … 31.82 avg -0.0096 · σ 1.1	0.29 / 0.29 since 20200101.0815	Cross of mama over fama = the MESA buy signal.
`bn_spot_sar` Parabolic SAR bn_spot_sar<0 and p:bn_spot_sar>=0 →	The parabolic stop-and-reverse level. (SAR ÷ close − 1) × 100	step 0.02, max 0.2	dev % open, usually within ±3	-27.38 … 22.81 avg -0.0489 · σ 1.17	-32.79 … 33.67 avg -0.0646 · σ 1.53	0.08 / 0.08 since 20200101.0030	< 0: SAR below the price = uptrend. > 0: downtrend. A sign flip is a reversal signal.
`bn_spot_midpoint` Midpoint of closes bn_spot_midpoint<-1 →	Halfway between the highest and lowest close of the last 14 bars. ((max close + min close) ÷ 2 ÷ close − 1) × 100	14 bars · 3.5 h	dev % open, usually within ±3	-13.85 … 24.73 avg -0.0138 · σ 0.612	-16.52 … 22.02 avg -0.0188 · σ 0.7974	0.16 / 0.16 since 20200101.0330	< 0: price in the upper half of its 3.5-hour close range.
`bn_spot_midprice` Midprice bn_spot_midprice>1 →	Halfway between the highest high and lowest low of the last 14 bars. ((max high + min low) ÷ 2 ÷ close − 1) × 100	14 bars · 3.5 h	dev % open, usually within ±3	-13.35 … 24.71 avg -0.0332 · σ 0.6128	-16.05 … 22.74 avg -0.0483 · σ 0.7967	0.16 / 0.16 since 20200101.0330	Same reading as midpoint, using the full bar range.
`bn_spot_ht_trendline` Hilbert trendline bn_spot_ht_trendline<0 and bn_spot_sma<0 →	TA-Lib's instantaneous trendline from the Hilbert transform. (HT_TRENDLINE ÷ close − 1) × 100	adaptive (63-bar warm-up)	dev % open, usually within ±3	-11.36 … 31.38 avg -0.0056 · σ 0.9341	-15.47 … 32.45 avg -0.0053 · σ 1.22	0.50 / 0.50 since 20200101.1600	A smooth trend proxy without a fixed period.
`bn_spot_linearreg` Linear regression value bn_spot_linearreg>0.5 →	The 14-bar least-squares line evaluated at the latest bar. (LINEARREG14 ÷ close − 1) × 100	14 bars · 3.5 h	dev % open, usually within ±3	-15.98 … 12.03 avg 8.87e-5 · σ 0.3815	-15.02 … 14.13 avg 1.47e-4 · σ 0.4951	0.16 / 0.16 since 20200101.0330	How far the close is from its own short trend line.
`bn_spot_linearreg_intercept` Regression intercept bn_spot_linearreg_intercept<-2 →	Where the 14-bar regression line started (its value 13 bars back). (intercept ÷ close − 1) × 100	14 bars · 3.5 h	dev % open, usually within ±3	-16.21 … 40.26 avg -0.0065 · σ 1.22	-22.11 … 42.42 avg -0.0045 · σ 1.59	0.16 / 0.16 since 20200101.0330	Big positive = the price fell over the window; big negative = it rose.
`bn_spot_tsf` Time-series forecast bn_spot_tsf<bn_spot_linearreg →	The 14-bar regression line extended one bar ahead. (TSF14 ÷ close − 1) × 100	14 bars · 3.5 h	dev % open, usually within ±3	-16.78 … 10.88 avg 5.93e-4 · σ 0.4093	-15.65 … 13.13 avg 5.04e-4 · σ 0.5316	0.16 / 0.16 since 20200101.0330	Compare with linearreg: tsf < linearreg means the line slopes down.
`bn_spot_adaptive_dma` Volatility-adaptive MA bn_spot_adaptive_dma<0 →	An average whose speed rises with the 7-day volatility rank (rv_rank). y = α·close + (1−α)·y₋₁, α = clip(0.03 + 0.27·rv_rank, 0.03, 0.30)	α 0.03–0.30 (768-bar warm-up ≈ 8 days)	dev % open, usually within ±3	-12.21 … 11.34 avg -0.0032 · σ 0.4787	-12.1 … 13.95 avg -0.0044 · σ 0.6328	5.00 / 5.00 since 20200109.0000	Tracks fast in volatile weeks and slowly in calm ones. Null for the first 8 days of a feed.

Bands

Distance from each band in percent of the close. Because they are distances, the upper band is normally positive and the lower band negative.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_bbands_upper` Bollinger upper band bn_spot_bbands_upper<0 →	SMA20 plus two standard deviations. (upper ÷ close − 1) × 100	20 bars · 5 h, 2σ	dev % open, usually within ±3	-4.46 … 61.46 avg 0.8409 · σ 1.21	-5.2 … 57.29 avg 1.12 · σ 1.57	0.20 / 0.20 since 20200101.0500	< 0: the close is ABOVE the upper band (stretched). Typical value ≈ +0.9.
`bn_spot_bbands_middle` Bollinger middle band bn_spot_bbands_middle<0 and p4:bn_spot_b… →	The 20-bar simple average (the band's centre). (SMA20 ÷ close − 1) × 100	20 bars · 5 h	dev % open, usually within ±3	-11.32 … 30.45 avg -0.0047 · σ 0.8021	-15.95 … 30.32 avg -0.0032 · σ 1.05	0.20 / 0.20 since 20200101.0500	< 0: price above the band centre.
`bn_spot_bbands_lower` Bollinger lower band bn_spot_bbands_lower>0 →	SMA20 minus two standard deviations. (lower ÷ close − 1) × 100	20 bars · 5 h, 2σ	dev % open, usually within ±3	-27.84 … 8.45 avg -0.8503 · σ 1.1	-32.1 … 10.31 avg -1.13 · σ 1.4	0.20 / 0.20 since 20200101.0500	> 0: the close is BELOW the lower band. Typical value ≈ −0.9.
`bn_spot_accbands_upper` Acceleration band upper bn_spot_accbands_upper<0 →	SMA20 of high × (1 + 4·(H−L)/(H+L)). (upper ÷ close − 1) × 100	20 bars · 5 h	dev % open, usually within ±3	-6.8 … 50.82 avg 1.02 · σ 1.17	-7.91 … 52.08 avg 1.34 · σ 1.51	0.20 / 0.20 since 20200101.0500	< 0: close above the acceleration band = breakout territory.
`bn_spot_accbands_lower` Acceleration band lower bn_spot_accbands_lower>0 →	SMA20 of low × (1 − 4·(H−L)/(H+L)). (lower ÷ close − 1) × 100	20 bars · 5 h	dev % open, usually within ±3	-26.29 … 14.85 avg -1.03 · σ 1.06	-33.46 … 17.55 avg -1.34 · σ 1.34	0.20 / 0.20 since 20200101.0500	> 0: close below the lower acceleration band.

MACD & price differences

Differences of prices, expressed as a percentage of the close so that they mean the same thing at $20,000 and at $120,000.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_macd` MACD line bn_spot_macd>0 and p:bn_spot_macd<=0 →	EMA12 minus EMA26, in % of the close. (EMA12 − EMA26) ÷ close × 100	12 / 26 / 9 (33-bar warm-up)	% of close open, signed	-11.46 … 2.78 avg 0.0036 · σ 0.3615	-10.85 … 3.57 avg 0.0025 · σ 0.4744	0.30 / 0.30 since 20200101.0830	> 0: bullish regime; < 0: bearish. Crossing zero upward = golden cross, downward = death cross.
`bn_spot_macd_signal` MACD signal bn_spot_macd>bn_spot_macd_signal and p:b… →	EMA9 of the MACD line, in % of the close. EMA9(MACD) ÷ close × 100	12 / 26 / 9	% of close open, signed	-9.15 … 2.52 avg 0.0036 · σ 0.3401	-8.97 … 3.24 avg 0.0026 · σ 0.4469	0.30 / 0.30 since 20200101.0830	MACD above its signal (macd > macd_signal) = momentum improving.
`bn_spot_macd_hist` MACD histogram Streak(bn_spot_macd_hist@entity)>=3 →	MACD minus its signal line, in % of the close. (MACD − signal) ÷ close × 100	12 / 26 / 9	% of close open, signed	-2.63 … 2.4 avg -4.67e-5 · σ 0.1087	-2.83 … 2.96 avg -9.34e-5 · σ 0.1415	0.30 / 0.30 since 20200101.0830	Sign = which side of the signal line; growing bars = momentum building. Works well with Streak().
`bn_spot_plus_dm` +DM bn_spot_plus_dm>bn_spot_minus_dm*2 →	Wilder-smoothed upward directional movement. +DM14 ÷ close × 100	14 bars · 3.5 h	% of close 0 to …	0.0471 … 26.28 avg 1.17 · σ 0.9308	0.0705 … 32.65 avg 1.54 · σ 1.16	0.16 / 0.16 since 20200101.0330	Compare with minus_dm: plus_dm > minus_dm = buyers pushing the highs.
`bn_spot_minus_dm` −DM bn_spot_minus_dm>bn_spot_plus_dm*2 →	Wilder-smoothed downward directional movement. −DM14 ÷ close × 100	14 bars · 3.5 h	% of close 0 to …	0.0501 … 43.92 avg 1.25 · σ 1.15	0.0827 … 51.95 avg 1.66 · σ 1.51	0.16 / 0.16 since 20200101.0330	minus_dm > plus_dm = sellers pushing the lows.
`bn_spot_trange` True range bn_spot_trange>=1 →	The previous bar's true range in % of the close. max(H−L, \|H−prev C\|, \|L−prev C\|) ÷ close × 100	1 bar	% of close 0 to …	0 … 20.45 avg 0.4088 · σ 0.4149	0 … 26.26 avg 0.5359 · σ 0.5202	0.08 / 0.08 since 20200101.0030	A single-bar volatility. ≥ 1 is a big bar for BTC.
`bn_spot_linearreg_slope` Regression slope bn_spot_linearreg_slope>0.1 →	Slope of the 14-bar regression line, in % of the close per bar. LINEARREG_SLOPE14 ÷ close × 100	14 bars · 3.5 h	% of close open, signed	-2.69 … 1.36 avg 5.04e-4 · σ 0.0946	-2.6 … 1.71 avg 3.58e-4 · σ 0.1237	0.16 / 0.16 since 20200101.0330	> 0 rising trend, < 0 falling; 0.1 = +0.1 % per 15 minutes.
`bn_spot_stddev` Standard deviation bn_spot_stddev<0.2 →	One standard deviation of closes over 20 bars, in % of the close. STDDEV20 ÷ close × 100	20 bars · 5 h	% of close 0 to …	0.0167 … 15.51 avg 0.4228 · σ 0.4141	0.0278 … 13.82 avg 0.5639 · σ 0.5254	0.20 / 0.20 since 20200101.0500	Half the Bollinger band width. Low values = squeeze.
`bn_spot_avgdev` Average deviation bn_spot_avgdev>1 →	Mean absolute deviation of closes over 14 bars, in % of the close. AVGDEV14 ÷ close × 100	14 bars · 3.5 h	% of close 0 to …	0.0093 … 10.63 avg 0.2937 · σ 0.3054	0.0134 … 11.03 avg 0.3917 · σ 0.3878	0.16 / 0.16 since 20200101.0330	A robust cousin of stddev.

Momentum & oscillators

Bounded indicators with the standard reading levels traders use. The translator applies these defaults when you say 'overbought' or 'oversold'.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_rsi` RSI bn_spot_rsi<=30 →	Relative strength index — the balance of recent up-closes vs down-closes. RSI(close, 14), Wilder smoothing	14 bars · 3.5 h	0–100 0 to 100	4.23 … 96.7 avg 50.49 · σ 11.19	4.87 … 95.57 avg 50.54 · σ 11.4	0.17 / 0.17 since 20200101.0345	≥ 70 overbought, ≤ 30 oversold, 50 neutral.
`bn_spot_adx` ADX bn_spot_adx>=25 and bn_spot_plus_di>bn_s… →	Trend strength regardless of direction. ADX(14) = Wilder average of DX	14 bars (27-bar warm-up)	0–100 0 to 100	4.93 … 80.47 avg 26.12 · σ 11.48	4.82 … 81.54 avg 26.89 · σ 11.56	0.26 / 0.26 since 20200101.0700	≥ 25 trending, ≥ 40 strong trend, < 20 no trend. The direction comes from plus_di / minus_di.
`bn_spot_adxr` ADX rating bn_spot_adxr>=30 →	Average of the current ADX and the ADX 14 bars ago. (ADX + ADX₋₁₄) ÷ 2	14 bars (40-bar warm-up)	0–100 0 to 100	5.78 … 76.73 avg 26.12 · σ 10.38	5.93 … 78.18 avg 26.89 · σ 10.43	0.35 / 0.35 since 20200101.1015	A smoother ADX; same levels.
`bn_spot_dx` DX bn_spot_dx>=50 →	The unsmoothed directional index behind ADX. \|+DI − −DI\| ÷ (+DI + −DI) × 100	14 bars · 3.5 h	0–100 0 to 100	2.00e-4 … 97.43 avg 26.12 · σ 19.08	1.00e-4 … 97.99 avg 26.9 · σ 19.27	0.17 / 0.17 since 20200101.0345	Spiky; prefer adx for regimes.
`bn_spot_plus_di` +DI bn_spot_plus_di>bn_spot_minus_di and p:b… →	Positive directional indicator. +DM14 ÷ TR14 × 100	14 bars · 3.5 h	0–100 0 to 100	1.1 … 77.09 avg 20.9 · σ 7.81	1.02 … 83.12 avg 20.98 · σ 7.89	0.17 / 0.17 since 20200101.0345	plus_di > minus_di = bullish direction; a cross is a classic entry.
`bn_spot_minus_di` −DI bn_spot_minus_di>bn_spot_plus_di and bn_… →	Negative directional indicator. −DM14 ÷ TR14 × 100	14 bars · 3.5 h	0–100 0 to 100	0.7351 … 80.84 avg 21.48 · σ 7.63	0.5549 … 79.87 avg 21.73 · σ 7.91	0.17 / 0.17 since 20200101.0345	minus_di > plus_di = bearish direction.
`bn_spot_roc` Rate of change bn_spot_roc>=2 →	Percent change of the close over 10 bars. (close ÷ close₋₁₀ − 1) × 100	10 bars · 2.5 h	% depends on the indicator	-21.79 … 24.23 avg 0.0153 · σ 1.01	-25.58 … 28.37 avg 0.0212 · σ 1.32	0.14 / 0.14 since 20200101.0245	Already a percent: 2 = +2 % in 2.5 hours.
`bn_spot_ppo` Percentage price oscillator bn_spot_ppo>0.5 →	MACD-style spread as a percent of the slow average. (SMA12 − SMA26) ÷ SMA26 × 100	12 / 26, SMA	% depends on the indicator	-12.22 … 6.2 avg 0.0072 · σ 0.5036	-11.07 … 6.48 avg 0.0087 · σ 0.6579	0.24 / 0.24 since 20200101.0630	Same sign logic as macd, slightly different scale.
`bn_spot_aroon_up` Aroon up bn_spot_aroon_up>=70 and bn_spot_aroon_d… →	How recently the 14-bar high was set (100 = on the latest bar). (14 − bars since the high) ÷ 14 × 100	14 bars · 3.5 h	0–100 0 to 100	0 … 100 avg 46.99 · σ 35.92	0 … 100 avg 46.85 · σ 35.95	0.17 / 0.17 since 20200101.0345	≥ 70 with aroon_down ≤ 30 = fresh uptrend.
`bn_spot_aroon_down` Aroon down bn_spot_aroon_down>=70 and bn_spot_aroon… →	How recently the 14-bar low was set. (14 − bars since the low) ÷ 14 × 100	14 bars · 3.5 h	0–100 0 to 100	0 … 100 avg 45.46 · σ 35.69	0 … 100 avg 44.61 · σ 35.54	0.17 / 0.17 since 20200101.0345	≥ 70 with aroon_up ≤ 30 = fresh downtrend.
`bn_spot_aroonosc` Aroon oscillator bn_spot_aroonosc>=50 →	Aroon up minus Aroon down. aroon_up − aroon_down	14 bars · 3.5 h	−100–100 -100 to 100	-100 … 100 avg 1.53 · σ 60.9	-100 … 100 avg 2.23 · σ 60.85	0.17 / 0.17 since 20200101.0345	> 0 bullish, < 0 bearish; ±50 strong.
`bn_spot_cci` Commodity channel index bn_spot_cci<=-100 →	How far the typical price is from its average, scaled by the mean deviation. (TP − SMA(TP)) ÷ (0.015 · mean deviation), TP = (H+L+C)/3	14 bars · 3.5 h	CCI ±466.67 at most (14-bar CCI)	-466.7 … 466.7 avg 2.48 · σ 109.1	-466.7 … 462.7 avg 2.83 · σ 109.4	0.16 / 0.16 since 20200101.0330	≥ 100 hot, ≤ −100 cold. The 14-bar formula cannot exceed ±466.67.
`bn_spot_cmo` Chande momentum bn_spot_cmo<=-50 →	Net up-moves vs down-moves over 14 bars. (Σup − Σdown) ÷ (Σup + Σdown) × 100	14 bars · 3.5 h	−100–100 -100 to 100	-91.55 … 93.39 avg 0.9731 · σ 22.38	-90.26 … 91.14 avg 1.07 · σ 22.8	0.17 / 0.17 since 20200101.0345	≥ 50 overbought, ≤ −50 oversold.
`bn_spot_imi` Intraday momentum index bn_spot_imi<=30 →	Share of candle body that was green over 14 bars. Σ green bodies ÷ Σ all bodies × 100	14 bars · 3.5 h	0–100 0 to 100	0 … 100 avg 50.53 · σ 15.4	0 … 100 avg 50.61 · σ 15.67	0.16 / 0.16 since 20200101.0330	≥ 70 overbought, ≤ 30 oversold (like an RSI of candle bodies).
`bn_spot_mfi` Money flow index bn_spot_mfi<=20 →	A volume-weighted RSI using typical price × volume. MFI(14)	14 bars · 3.5 h	0–100 0 to 100	0 … 100 avg 50.04 · σ 17.1	0 … 100 avg 50.16 · σ 17.83	0.17 / 0.17 since 20200101.0345	≥ 80 overbought, ≤ 20 oversold.
`bn_spot_stoch_k` Slow stochastic %K bn_spot_stoch_k>bn_spot_stoch_d and p:bn… →	Where the close sits in the recent high-low range, smoothed. SMA3 of fast %K(5)	5, 3, 3 (8-bar warm-up)	0–100 0 to 100	0 … 100 avg 51.54 · σ 23.05	0.1705 … 100 avg 51.76 · σ 22.78	0.13 / 0.13 since 20200101.0215	≥ 80 overbought, ≤ 20 oversold; %K crossing above %D from below 20 is a classic buy setup.
`bn_spot_stoch_d` Slow stochastic %D bn_spot_stoch_d>=80 →	The signal line of the slow stochastic. SMA3 of slow %K	5, 3, 3	0–100 0 to 100	1.26 … 99.96 avg 51.54 · σ 21.12	1.3 … 99.27 avg 51.76 · σ 20.88	0.13 / 0.13 since 20200101.0215	Use with stoch_k for crosses.
`bn_spot_stochf_k` Fast stochastic %K bn_spot_stochf_k<=5 →	The raw stochastic: close position inside the 5-bar range. (C − min L5) ÷ (max H5 − min L5) × 100	5 bars	0–100 0 to 100	0 … 100 avg 51.54 · σ 28.31	0 … 100 avg 51.76 · σ 27.95	0.11 / 0.11 since 20200101.0145	Very jumpy; 0 = closed at the 5-bar low.
`bn_spot_stochrsi_k` Stochastic RSI %K bn_spot_stochrsi_k<=10 →	A stochastic applied to RSI — RSI's position inside its own recent range. StochRSI(14, 5, 3) %K	RSI 14, stoch 5 / 3 (20-bar warm-up)	0–100 0 to 100	0 … 100 avg 50.3 · σ 41.13	0 … 100 avg 50.5 · σ 41.07	0.21 / 0.21 since 20200101.0515	≥ 80 / ≤ 20; more extreme and faster than RSI.
`bn_spot_stochrsi_d` Stochastic RSI %D bn_spot_stochrsi_k>bn_spot_stochrsi_d an… →	The smoothed StochRSI line. SMA3 of stochrsi_k	RSI 14, stoch 5 / 3	0–100 0 to 100	0 … 100 avg 50.3 · σ 31.87	0 … 100 avg 50.5 · σ 31.87	0.21 / 0.21 since 20200101.0515	Cross of stochrsi_k over stochrsi_d = momentum turn.
`bn_spot_trix` TRIX bn_spot_trix>0 and p:bn_spot_trix<=0 →	One-bar percent change of a triple-smoothed EMA. ROC1 of EMA(EMA(EMA(close,30)))	30 bars (88-bar warm-up)	% depends on the indicator	-0.5279 … 0.2402 avg 9.81e-4 · σ 0.0352	-0.5549 … 0.3059 avg 0.0012 · σ 0.0463	0.67 / 0.67 since 20200101.2215	Sign = smoothed trend direction; tiny numbers (std ≈ 0.035).
`bn_spot_ultosc` Ultimate oscillator bn_spot_ultosc<=30 →	Buying pressure averaged over three windows. ULTOSC(7, 14, 28)	7 / 14 / 28	0–100 0 to 100	11.77 … 89.21 avg 51.11 · σ 8.59	13 … 86.11 avg 51.18 · σ 8.45	0.26 / 0.26 since 20200101.0715	≥ 70 overbought, ≤ 30 oversold.
`bn_spot_willr` Williams %R bn_spot_willr<=-80 →	Close position inside the 14-bar range, on a −100…0 scale. (max H − C) ÷ (max H − min L) × −100	14 bars · 3.5 h	−100–0 -100 to 0	-100 … 0 avg -47.89 · σ 27.01	-100 … 0 avg -47.62 · σ 26.7	0.16 / 0.16 since 20200101.0330	≥ −20 overbought, ≤ −80 oversold.
`bn_spot_bop` Balance of power bn_spot_bop>=0.8 →	How much of the bar's range the close won from the open. (C − O) ÷ (H − L)	1 bar	−1–1 -1 to 1	-1 … 1 avg 0.0058 · σ 0.5368	-1 … 1 avg 0.0077 · σ 0.5306	0.07 / 0.07 since 20200101.0015	+1 = opened at the low and closed at the high; −1 the reverse.
`bn_spot_natr` Normalized ATR bn_spot_natr>=1 →	Average true range as a percent of the price. ATR14 ÷ close × 100	14 bars · 3.5 h	% 0 to …	0.017 … 7.18 avg 0.4086 · σ 0.2888	0.0203 … 8.28 avg 0.5355 · σ 0.3602	0.17 / 0.17 since 20200101.0345	The everyday volatility gauge: ≈ 0.4 % is typical for BTC, ≥ 1 is a wild stretch.

Candle shape

The size and shape of the previous bar.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_range` Bar range bn_spot_range>=1 →	Height of the bar (high − low) as a % of the close before it. (H − L) ÷ prev close × 100	1 bar	% 0 to …	0 … 24.34 avg 0.4088 · σ 0.4139	0 … 25.6 avg 0.5357 · σ 0.5188	0.08 / 0.08 since 20200101.0030	≥ 1 is a big bar for BTC; the average is ≈ 0.4.
`bn_spot_body` Candle body bn_spot_body>=1 →	Close vs open as a signed percent. (C − O) ÷ O × 100	1 bar	% open, signed	-11.92 … 22.61 avg 0.0016 · σ 0.3367	-11.56 … 23.77 avg 0.0022 · σ 0.4325	0.07 / 0.07 since 20200101.0015	> 0 green, < 0 red; ≥ 1 = a big green candle.
`bn_spot_clv` Close location value bn_spot_clv>=0.8 →	Where the close sits inside the bar. (2C − H − L) ÷ (H − L); null when H = L	1 bar	−1–1 -1 to 1	-1 … 1 avg 0.0175 · σ 0.6203	-1 … 1 avg 0.0195 · σ 0.61	0.08 / 0.08 since 20200101.0015	+1 closed at the high, −1 at the low; \|clv\| ≥ 0.8 is an 'extreme' close.

Returns

Log returns over several horizons, already in percent. ret_log is the 15-minute return of the previous bar; the short / mid / long names are 1 h, 4 h and 8 h.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_ret_log` 1-bar return abs(bn_spot_ret_log)>=1 →	Log return of the previous bar. ln(C ÷ C₋₁) × 100	1 bar · 15 min	% open, signed	-12.69 … 20.4 avg 0.001 · σ 0.3364	-12.39 … 21.32 avg 0.0012 · σ 0.4323	0.08 / 0.08 since 20200101.0030	±1 is a sharp 15-minute move for BTC (std ≈ 0.34).
`bn_spot_ret_log_short` 1-hour return bn_spot_ret_log_short>=2 →	Log return over the last 4 bars. ln(C ÷ C₋₄) × 100	4 bars · 1 h	% open, signed	-20.1 … 25.94 avg 0.0041 · σ 0.6581	-24.79 … 31.83 avg 0.005 · σ 0.8536	0.10 / 0.10 since 20200101.0115	±2 = a sharp hour.
`bn_spot_ret_log_mid` 4-hour return bn_spot_ret_log_mid<=-2 →	Log return over the last 16 bars. ln(C ÷ C₋₁₆) × 100	16 bars · 4 h	% open, signed	-41.57 … 26.19 avg 0.0163 · σ 1.28	-40.41 … 30.23 avg 0.0197 · σ 1.67	0.18 / 0.18 since 20200101.0415	std ≈ 1.3.
`bn_spot_ret_log_long` 8-hour return bn_spot_ret_log_long>=3 →	Log return over the last 32 bars. ln(C ÷ C₋₃₂) × 100	32 bars · 8 h	% open, signed	-45.6 … 35.05 avg 0.0325 · σ 1.78	-47.87 … 37.47 avg 0.0389 · σ 2.33	0.29 / 0.29 since 20200101.0815	std ≈ 1.8.
`bn_spot_weighted_ret_by_volume` Volume-weighted mean return bn_spot_weighted_ret_by_volume>0.05 →	Average 1-bar return over the day, weighting bars by their volume. Σ(ret·vol) ÷ Σvol × 100	96 bars · 1 day	% open, signed	-2.05 … 1.13 avg -0.0067 · σ 0.1169	-2.47 … 1.93 avg -0.0128 · σ 0.1633	0.73 / 0.73 since 20200102.0015	> 0: the heavy-volume bars were up bars.
`bn_spot_signed_return_run_sum` Run return bn_spot_signed_return_run_sum>=2 →	Cumulative return of the current same-sign run of bars. Σ ret_log while the sign persists; resets on a sign change, a zero or a gap	current run	% open, signed	-21.18 … 29 avg 0.0052 · σ 0.6342	-24.79 … 29.6 avg 0.0113 · σ 0.8254	0.08 / 0.08 since 20200101.0030	How far the current push has already gone: ≥ 2 = a 2 % run without a red bar.
`bn_spot_ret_median` Median return bn_spot_ret_median>0.02 →	Median 1-bar return of the last day. median(ret_log over 96)	96 bars · 1 day	% open, signed	-0.3619 … 0.4075 avg 0.0011 · σ 0.0313	-0.5058 … 0.3603 avg 0.0028 · σ 0.0409	0.73 / 0.73 since 20200102.0015	Tiny numbers (±0.03); the sign tells the day's typical bar.
`bn_spot_ret_q10` Return 10th percentile bn_spot_ret_q10<=-0.6 →	The 10 % worst 1-bar return of the last two days. quantile(ret_log, 0.10) over 192	192 bars · 2 days	% open, signed	-2.78 … -0.0385 avg -0.2918 · σ 0.1803	-3.24 … -0.0411 avg -0.3886 · σ 0.2403	1.38 / 1.38 since 20200103.0015	Typically −0.3; ≤ −0.6 marks a turbulent two days.
`bn_spot_ret_q90` Return 90th percentile bn_spot_ret_q90>=0.6 →	The 10 % best 1-bar return of the last two days. quantile(ret_log, 0.90) over 192	192 bars · 2 days	% open, signed	0.044 … 1.95 avg 0.2962 · σ 0.1745	0.0481 … 2.63 avg 0.3897 · σ 0.2258	1.38 / 1.38 since 20200103.0015	Typically +0.3.
`bn_spot_finite_decay_ret` Decay-weighted return bn_spot_finite_decay_ret>=0.2 →	An average of recent returns that forgets old bars quickly (weights 0.9ⁱ). Σ 0.9ⁱ·ret₋ᵢ ÷ Σ 0.9ⁱ × 100	32 bars · 8 h, factor 0.9	% open, signed	-2.55 … 1.21 avg 0.001 · σ 0.076	-2.61 … 1.83 avg 0.0012 · σ 0.0992	0.29 / 0.29 since 20200101.0815	A fast momentum reading; ≥ 0.2 is strong.

Volatility & return distribution

How wide and how skewed the recent return distribution is.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_rv_short` Realized volatility 4 h bn_spot_rv_short>=1 →	Standard deviation of 1-bar returns over 4 hours, in %. std(ret_log over 16) × 100	16 bars · 4 h	% 0 to …	0.0124 … 7.12 avg 0.2526 · σ 0.2072	0.0147 … 8.09 avg 0.3324 · σ 0.2559	0.18 / 0.18 since 20200101.0415	Average ≈ 0.26; ≥ 1 is very volatile.
`bn_spot_rv_long` Realized volatility 1 d bn_spot_rv_long<=0.15 →	Standard deviation of 1-bar returns over a day, in %. std(ret_log over 96) × 100	96 bars · 1 day	% 0 to …	0.0278 … 3.81 avg 0.2784 · σ 0.1857	0.0314 … 4.22 avg 0.3638 · σ 0.2291	0.73 / 0.73 since 20200102.0015	Average ≈ 0.28. rv_rank ranks this value over 7 days.
`bn_spot_ret_skew` Return skew bn_spot_ret_skew<=-1.5 →	Skewness of the day's 1-bar returns. unbiased sample skew over 96	96 bars · 1 day	ratio open	-9.01 … 8.07 avg -0.0383 · σ 1.2	-8.94 … 8.43 avg -0.0469 · σ 1.23	0.73 / 0.73 since 20200102.0015	> 0: a few big up bars; < 0: a few big down bars.
`bn_spot_ret_kurt` Return kurtosis bn_spot_ret_kurt>=10 →	Excess kurtosis (fat tails) of two days of returns. unbiased excess kurtosis over 192 (normal = 0)	192 bars · 2 days	ratio open	-0.5075 … 166.8 avg 7.13 · σ 10.26	-0.608 … 163.4 avg 6.99 · σ 10.85	1.38 / 1.38 since 20200103.0015	High values mean a couple of outlier bars dominate the window.
`bn_spot_ret_ir_short` Information ratio 8 h bn_spot_ret_ir_short>=0.3 →	Mean return divided by its volatility over 8 hours. mean(ret) ÷ std(ret) over 32	32 bars · 8 h	ratio open	-0.913 … 0.7148 avg 0.0075 · σ 0.16	-0.7103 … 0.6614 avg 0.0083 · σ 0.1635	0.29 / 0.29 since 20200101.0815	A smooth-trend score: ≥ 0.3 = a steady climb.
`bn_spot_ret_ir_long` Information ratio 1 d bn_spot_ret_ir_long<=-0.2 →	Mean return divided by its volatility over a day. mean(ret) ÷ std(ret) over 96	96 bars · 1 day	ratio open	-0.3715 … 0.3682 avg 0.0066 · σ 0.0935	-0.3555 … 0.4596 avg 0.007 · σ 0.0955	0.73 / 0.73 since 20200102.0015	≤ −0.2 = a steady slide.
`bn_spot_ret_moment3` 3rd central moment bn_spot_ret_moment3<-5 →	Third central moment of % returns (raw skew information). E[(r − r̄)³] × 100³	192 bars · 2 days	%³ / %⁴ open	-20.23 … 44.65 avg 0.0155 · σ 0.9689	-19.87 … 61.1 avg 0.0177 · σ 1.45	1.38 / 1.38 since 20200103.0015	Sign like skew; use ret_skew unless you need the raw moment.
`bn_spot_ret_moment4` 4th central moment bn_spot_ret_moment4>=20 →	Fourth central moment of % returns (raw tail information). E[(r − r̄)⁴] × 100⁴	192 bars · 2 days	%³ / %⁴ 0 to …	0 … 1,272 avg 1.54 · σ 35.05	0 … 1,622 avg 2.42 · σ 45.23	1.38 / 1.38 since 20200103.0015	Always ≥ 0; spikes with crash bars.
`bn_spot_ret_l2` Return energy 8 h bn_spot_ret_l2>=4 →	Root of the sum of squared returns over 8 hours. √Σ ret² × 100	32 bars · 8 h	% 0 to …	0.0997 … 31.3 avg 1.52 · σ 1.15	0.1249 … 34.47 avg 1.99 · σ 1.42	0.29 / 0.29 since 20200101.0815	Average ≈ 1.6; ≥ 4 is a stormy 8 hours.
`bn_spot_range_l2` Range energy 8 h bn_spot_range_l2>=6 →	Root of the sum of squared bar ranges over 8 hours. √Σ range² × 100	32 bars · 8 h	% 0 to …	0.1261 … 47.69 avg 2.65 · σ 1.95	0.1664 … 54.8 avg 3.46 · σ 2.42	0.29 / 0.29 since 20200101.0815	Average ≈ 2.7.

Percentile ranks

Where the latest value sits inside its own recent window, from 0 to 1. Ranks are the easiest way to say 'unusually high' without picking a number.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_ret_rank_short` Return rank 1 d bn_spot_ret_rank_short>=0.95 →	Percentile of the latest 1-bar return within the day. rank(ret_log) ÷ 96, ties averaged	96 bars · 1 day	rank 0–1 0 to 1	0.0104 … 1 avg 0.5044 · σ 0.2885	0.0104 … 1 avg 0.5046 · σ 0.2885	0.73 / 0.73 since 20200102.0015	≥ 0.95 = one of the five best bars of the day.
`bn_spot_ret_rank_long` Return rank 2 d bn_spot_ret_rank_long<=0.05 →	The same over two days. rank(ret_log) ÷ 192	192 bars · 2 days	rank 0–1 0 to 1	0.0052 … 1 avg 0.5021 · σ 0.2889	0.0052 … 1 avg 0.5021 · σ 0.2887	1.38 / 1.38 since 20200103.0015	≤ 0.05 = a bottom-5 % bar over two days.
`bn_spot_volume_rank_short` Volume rank 1 d bn_spot_volume_rank_short>=0.9 →	Percentile of the latest volume within the day. rank(volume) ÷ 96	96 bars · 1 day	rank 0–1 0 to 1	0.0104 … 1 avg 0.4976 · σ 0.2987	0.0104 … 1 avg 0.4977 · σ 0.2972	0.72 / 0.72 since 20200102.0000	≥ 0.9 = a heavy bar for today.
`bn_spot_volume_rank_long` Volume rank 7 d bn_spot_volume_rank_long>=0.95 →	Percentile of the latest volume within the week. rank(volume) ÷ 672	672 bars · 7 days	rank 0–1 0 to 1	0.0015 … 1 avg 0.4984 · σ 0.2935	0.0015 … 1 avg 0.497 · σ 0.2944	4.43 / 4.43 since 20200108.0000	≥ 0.95 = a volume spike by weekly standards (the stricter one).
`bn_spot_range_rank` Range rank 1 d bn_spot_range_rank>=0.95 →	Percentile of the latest bar height within the day. rank(range) ÷ 96	96 bars · 1 day	rank 0–1 0 to 1	0.0104 … 1 avg 0.499 · σ 0.2963	0.0104 … 1 avg 0.4993 · σ 0.2951	0.73 / 0.73 since 20200102.0015	≥ 0.95 = one of the biggest bars today.
`bn_spot_clv_rank` CLV rank 1 d bn_spot_clv_rank>=0.9 →	Percentile of the latest close-location value within the day. rank(clv) ÷ 96	96 bars · 1 day	rank 0–1 0 to 1	0.0104 … 1 avg 0.505 · σ 0.2884	0.0104 … 1 avg 0.5051 · σ 0.2882	0.73 / 0.73 since 20200102.0000	High = closes near the high are unusual today.
`bn_spot_rv_rank` Volatility rank 7 d bn_spot_rv_rank<=0.15 →	Percentile of today's realized volatility (rv_long) within the week. rank(rv_long) ÷ 672	672 bars (768-bar warm-up)	rank 0–1 0 to 1	0.0015 … 1 avg 0.4851 · σ 0.3105	0.0015 … 1 avg 0.4859 · σ 0.3119	5.00 / 5.00 since 20200109.0000	≤ 0.15 squeeze (calm), ≥ 0.9 explosion.
`bn_spot_emv_rank` EMV rank 1 d bn_spot_emv_rank>=0.9 →	Percentile of ease-of-movement within the day. rank(emv) ÷ 96	96 bars (121-bar warm-up)	rank 0–1 0 to 1	0.0104 … 1 avg 0.5029 · σ 0.2999	0.0104 … 1 avg 0.5025 · σ 0.2997	0.90 / 0.90 since 20200102.0630	High = price moving easily on little volume.

Regression & correlation

Trend fits of the log price and the relationship between returns and volume changes.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_logprice_slope` Log-price trend slope bn_spot_logprice_slope>=0.02 →	Slope of ln(close) against time over the last day, in % per bar. LINEARREG_SLOPE(ln C, 96) × 100	96 bars · 1 day	% open, signed	-0.5391 … 0.2278 avg 9.67e-4 · σ 0.0353	-0.5861 … 0.3044 avg 0.0012 · σ 0.0465	0.72 / 0.72 since 20200102.0000	0.02 ≈ +2 % per day of steady drift; the sign = trend direction.
`bn_spot_logprice_resid` Log-price trend residual bn_spot_logprice_resid<=-2 →	How far the close is above (+) or below (−) its own 1-day trend line, in %. (ln C − LINEARREG(ln C, 96)) × 100	96 bars · 1 day	% open, signed	-22.79 … 14.8 avg -0.0014 · σ 1.08	-25.55 … 20.51 avg -0.0017 · σ 1.42	0.72 / 0.72 since 20200102.0000	≥ 2 overheated vs its trend; ≤ −2 washed out.
`bn_spot_ret_dlogvol_corr` Return–volume correlation bn_spot_ret_dlogvol_corr>=0.3 →	Correlation between 1-bar returns and changes in log volume over a day. CORREL(ret_log, Δln volume, 96)	96 bars · 1 day	−1–1 -1 to 1	-0.6266 … 0.6196 avg -0.0196 · σ 0.2076	-0.6831 … 0.6091 avg -0.0299 · σ 0.214	0.73 / 0.73 since 20200102.0015	> 0: volume rises on up bars (buyers active); < 0: volume rises on down bars.
`bn_spot_ret_dlogvol_partial_corr` Partial correlation (range-controlled) bn_spot_ret_dlogvol_partial_corr<=-0.2 →	The same correlation after removing the effect of bar range. partial r(ret, Δln vol \| range) over 192	192 bars · 2 days	−1–1 -1 to 1	-0.4011 … 0.3691 avg -0.0173 · σ 0.1116	-0.4541 … 0.3327 avg -0.0278 · σ 0.1082	1.39 / 1.39 since 20200103.0015	Same reading, less driven by volatility.
`bn_spot_ret_dlogvol_slope` Return-on-volume beta bn_spot_ret_dlogvol_slope>0.1 →	How much return a unit change in log volume 'buys', in %. β of ret_log ~ Δln volume over 96, × 100	96 bars · 1 day	% open, signed	-2.28 … 1.35 avg -0.0193 · σ 0.1528	-2.64 … 1.51 avg -0.032 · σ 0.1875	0.73 / 0.73 since 20200102.0015	Positive = volume expansion coincides with up moves.
`bn_spot_ret_dlogvol_resid` Return-on-volume residual abs(bn_spot_ret_dlogvol_resid)>=1 →	The part of the last return that volume does not explain, in %. ret − β·Δln vol − α	96 bars · 1 day	% open, signed	-11.9 … 21.68 avg -2.17e-4 · σ 0.3286	-11.8 … 22.26 avg -3.12e-4 · σ 0.4223	0.73 / 0.73 since 20200102.0015	Large \|value\| = a move that volume did not accompany.
`bn_spot_ret_multivar_resid` Multivariate return residual abs(bn_spot_ret_multivar_resid)>=1 →	Return unexplained by volume change, bar range and CLV together, in %. residual of ret ~ [1, Δln vol, range, clv] over 192	192 bars · 2 days	% open, signed	-6.27 … 18.75 avg 0.0014 · σ 0.2376	-5.84 … 18.77 avg 0.0027 · σ 0.3004	1.39 / 1.39 since 20200103.0015	Surprise moves relative to the usual microstructure.

Volume & money flow

Who is doing the trading. ad, adosc and obv are cumulative in exchange units, so compare their CHANGE, not their level.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_ad` Accumulation / distribution line bn_spot_ad-p16:bn_spot_ad>0 →	Chaikin's running sum of CLV × volume. Σ clv × volume	cumulative	exchange units as served by the exchange	-145,290 … 267,994 avg 11,489 · σ 32,555	-1,519,000 … 2,395,000 avg 110,451 · σ 295,155	0.07 / 0.07 since 20200101.0015	Never compare the level; use differences: {P}_ad-p16:{P}_ad>0.
`bn_spot_adosc` Chaikin oscillator bn_spot_adosc>0 and p:bn_spot_adosc<=0 →	Fast minus slow EMA of the A/D line. EMA3(AD) − EMA10(AD)	EMA 3 / 10 (9-bar warm-up)	exchange units as served by the exchange	-24,177 … 22,853 avg 52.62 · σ 694.5	-116,387 … 92,586 avg 512 · σ 5,627	0.13 / 0.13 since 20200101.0230	> 0 accumulation, < 0 distribution (exchange units, feed-specific size).
`bn_spot_obv` On-balance volume bn_spot_obv-p4:bn_spot_obv>0 →	Volume added on up bars and subtracted on down bars. Σ ±volume	cumulative	exchange units as served by the exchange	-443,805 … 446,912 avg 1,506 · σ 46,463	-2,376,000 … 1,684,000 avg -6,944 · σ 350,538	0.07 / 0.07 since 20200101.0015	Use changes: {P}_obv-p4:{P}_obv>0 = net buying volume in the last hour.
`bn_spot_volume_return_split_imbalance` Volume split imbalance bn_spot_volume_return_split_imbalance>=0.3 →	Volume on the 16 best-return bars vs the 16 worst, over a day. (vol_top16 − vol_bottom16) ÷ (vol_top16 + vol_bottom16)	96 bars, top / bottom 16	−1–1 -1 to 1	-0.7156 … 0.7952 avg -0.0091 · σ 0.1596	-0.7357 … 0.8172 avg -0.0174 · σ 0.1806	0.73 / 0.73 since 20200102.0015	> 0: the day's volume concentrated on up moves.
`bn_spot_emv` Ease of movement bn_spot_emv>0.5 →	How far the price moves per unit of volume. tdx EMV(14)	14 bars (26-bar warm-up)	ratio open	-2.5 … 2.42 avg 1.67e-4 · σ 0.0831	-3.02 … 2.94 avg -1.68e-4 · σ 0.1057	0.25 / 0.25 since 20200101.0645	Large positive = rising easily on light volume.
`bn_spot_vr_log` Log volume ratio bn_spot_vr_log>=0.5 →	Log ratio of up-bar volume to down-bar volume over 12 hours. ln((2·up + flat + ε) ÷ (2·down + flat + ε))	48 bars · 12 h	ratio open	-2.07 … 1.78 avg -0.0015 · σ 0.3627	-1.96 … 2.09 avg -0.0028 · σ 0.3872	0.40 / 0.40 since 20200101.1215	> 0 more volume on up bars.
`bn_spot_vr_imbalance` Volume imbalance bn_spot_vr_imbalance<=-0.3 →	Up-bar volume minus down-bar volume as a share of all volume, over 12 hours. (up − down) ÷ (up + down + flat)	48 bars · 12 h	−1–1 -1 to 1	-0.7766 … 0.7107 avg -7.03e-4 · σ 0.175	-0.7531 … 0.7796 avg -0.0014 · σ 0.1864	0.40 / 0.40 since 20200101.1215	≤ −0.3 heavy selling volume; ≥ 0.3 heavy buying volume.

Event distance & streaks

How long ago something happened, stored as ln(1 + bars) so a query can say 'just now' (= 0) or 'more than a day ago' (≥ 4.5747). Null means no such event inside the 672-bar window.

Indicator	What it measures	Window	Unit	BTC observed	ETH observed	Null % B / E	How to read
`bn_spot_bars_since_volume_spike` Bars since volume spike bn_spot_bars_since_volume_spike=0 →	Time since the last volume spike (volume_rank_long ≥ 0.95). ln(1 + bars), cap 672	672-bar window and cap	ln(1+bars) 0 to 6.51	0 … 5.25 avg 3.01 · σ 1.29	0 … 5.25 avg 2.98 · σ 1.29	15.47 / 13.49 since 20200108.0015	0 = the spike was the previous bar; null = none in the last 7 days.
`bn_spot_bars_since_range_spike` Bars since range spike bn_spot_bars_since_range_spike<=1.6094 →	Time since the last bar-height spike (range rank over 672 ≥ 0.95). ln(1 + bars), cap 672	672-bar window and cap	ln(1+bars) 0 to 6.51	0 … 5.26 avg 2.91 · σ 1.29	0 … 5.25 avg 2.89 · σ 1.29	13.48 / 11.99 since 20200108.0015	≤ 1.6094 = within the last hour.
`bn_spot_bars_since_high_breakout` Bars since high breakout bn_spot_bars_since_high_breakout=0 →	Time since the close last broke above the prior 96-bar high. ln(1 + bars), cap 672	672-bar cap	ln(1+bars) 0 to 6.51	0 … 6.51 avg 4.21 · σ 1.35	0 … 6.51 avg 4.21 · σ 1.37	4.64 / 5.09 since 20200108.0015	0 = breakout on the previous bar.
`bn_spot_bars_since_low_breakout` Bars since low breakout bn_spot_bars_since_low_breakout=0 →	Time since the close last broke below the prior 96-bar low. ln(1 + bars), cap 672	672-bar cap	ln(1+bars) 0 to 6.51	0 … 6.51 avg 4.38 · σ 1.35	0 … 6.51 avg 4.41 · σ 1.33	6.43 / 6.09 since 20200108.0015	0 = breakdown on the previous bar.
`bn_spot_trend_streak` Trend streak bn_spot_trend_streak>=3.4965 →	Consecutive bars with the close above EMA20. ln(1 + run length), cap 672	EMA 20, cap 672	ln(1+bars) 0 to 6.51	0 … 4.74 avg 1.06 · σ 1.21	0 … 4.96 avg 1.07 · σ 1.23	0.20 / 0.20 since 20200101.0500	≥ 3.4965 = 32 bars (8 h) above the EMA. For arbitrary runs use Streak().
`bn_spot_positive_return_streak` Green streak bn_spot_positive_return_streak>=1.7918 →	Consecutive bars with a positive 1-bar return. ln(1 + run length), cap 96	cap 96	ln(1+bars) 0 to 4.57	0 … 2.71 avg 0.4928 · σ 0.5498	0 … 2.77 avg 0.4958 · σ 0.5521	0.08 / 0.08 since 20200101.0030	1.7918 = 5 in a row.
`bn_spot_bars_since_clv_extreme` Bars since extreme close bn_spot_bars_since_clv_extreme=0 →	Time since a bar closed at an extreme (\|clv\| ≥ 0.8). ln(1 + bars), cap 672	672-bar window and cap	ln(1+bars) 0 to 6.51	0 … 3.99 avg 1.08 · σ 0.8374	0 … 3.89 avg 1.12 · σ 0.8308	4.43 / 4.43 since 20200108.0015	0 = the previous bar closed at its high or low.
`bn_spot_barssincen_volume_spike` Bars since volume spike (raw) bn_spot_barssincen_volume_spike<=4 →	The same distance as bars_since_volume_spike, but as a plain count. bars since volume_rank_long ≥ 0.95, 0–671	672 bars	bars 0 to 671	0 … 191 avg 101.2 · σ 42.39	0 … 191 avg 104.1 · σ 42.31	15.47 / 13.49 since 20200108.0015	≤ 4 = within the last hour.

Raw candle columns on this feed

The row's own bar, in exchange units (USDT). Only taker_buy_ratio is lagged like an indicator.

Column	Meaning	Bar	BTC observed	ETH observed	Null % B / E
`bn_spot_open` Open	First trade price of the bar.	this bar	3,889 … 126,011 avg 49,705 · σ 30,873	88.44 … 4,943 avg 2,162 · σ 1,124	0.07 / 0.07 since 20200101.0000
`bn_spot_high` High	Highest trade price of the bar.	this bar	4,241 … 126,200 avg 49,797 · σ 30,914	97.66 … 4,957 avg 2,168 · σ 1,127	0.07 / 0.07 since 20200101.0000
`bn_spot_low` Low	Lowest trade price of the bar.	this bar	3,782 … 125,648 avg 49,610 · σ 30,832	86 … 4,934 avg 2,157 · σ 1,121	0.07 / 0.07 since 20200101.0000
`bn_spot_close` Close	Last trade price of the bar.	this bar	3,882 … 126,011 avg 49,705 · σ 30,873	88.35 … 4,943 avg 2,162 · σ 1,124	0.07 / 0.07 since 20200101.0000
`bn_spot_volume` Volume	Traded amount in the base coin (BTC or ETH).	this bar	0 … 40,371 avg 712.1 · σ 1,163	0 … 256,455 avg 6,313 · σ 8,169	0.07 / 0.07 since 20200101.0000
`bn_spot_quote_volume` Quote volume	Traded amount in the quote currency (USDT / KRW).	this bar	0 … 1.09e+9 avg 23,540,000 · σ 30,650,000	0 … 578,900,000 avg 12,250,000 · σ 17,200,000	0.07 / 0.07 since 20200101.0000
`bn_spot_trades` Trades	Number of trades in the bar.	this bar	0 … 1,755,000 avg 27,493 · σ 32,347	0 … 1,269,000 avg 18,040 · σ 26,713	0.07 / 0.07 since 20200101.0000
`bn_spot_taker_buy_base` Taker buy volume	Volume bought at market (aggressive buys), base units.	this bar	0 … 19,926 avg 351.6 · σ 581.1	0 … 128,878 avg 3,136 · σ 4,079	0.07 / 0.07 since 20200101.0000
`bn_spot_taker_buy_quote` Taker buy quote volume	The same in quote currency.	this bar	0 … 561,900,000 avg 11,610,000 · σ 15,400,000	0 … 324,800,000 avg 6,084,000 · σ 8,590,000	0.07 / 0.07 since 20200101.0000
`bn_spot_taker_buy_ratio` Taker buy ratio	taker_buy_base ÷ volume of the previous bar: the share of volume that was market buying (0–1). ≥ 0.55 = buyers dominant.	previous bar	0.0772 … 0.9429 avg 0.4888 · σ 0.0887	0.0862 … 0.931 avg 0.4968 · σ 0.0877	0.08 / 0.08 since 20200101.0015

Cross-market columns

These have no feed prefix. They are computed from two feeds at the same bar and lagged like indicators.

Column	Meaning	Range	Example
`x_binance_basis` Binance basis	Binance perpetual vs spot close, in %. Positive = futures trade above spot (contango / long demand).	observed −9.6 … +1.5, typically ±0.07	x_binance_basis>0.1
`x_okx_basis` OKX basis	OKX perpetual vs spot close, in %.	observed −7.0 … +7.5	x_okx_basis<0
`x_krw_premium` Kimchi premium	Upbit price (converted to USDT with the Upbit KRW/USDT rate) vs Binance, in %. Negative = 'reverse premium'.	observed −10.3 … +6.8; available from 2024-06-07	x_krw_premium<0

Identity and calendar columns

Derived from the bar's UTC start time — never from a clock — so they are identical in history and live.

Column	Meaning	Values	Example
`date` Bar start (UTC)	YYYYMMDD.HHMM as a number. The bar that opened at 09:30 UTC on 1 Sep 2026 is 20260901.0930 (results show 20260901.093 because trailing zeros are dropped).	20200101.0000 and up	date>=today-7
`entity` Coin	BTC or ETH — one row per coin per bar.	BTC, ETH	entity=ETH
`day` Weekday	Monday … Sunday in UTC. Quote it.	Monday–Sunday	day in ('Saturday','Sunday')
`day_num` Day of month	1–31.	1–31	day_num=1
`month` Month	1–12.	1–12	month=12
`year` Year	Four-digit year.	2020 and up	year>=2024
`hour_progress` Minute of the hour, %	0, 25, 50 or 75: which quarter of the hour the bar starts in. 0 = the bar on the hour.	0 / 25 / 50 / 75	hour_progress=0
`day_progress` Time of day, %	Minutes since 00:00 UTC ÷ 1440 × 100. 0 = the midnight bar; 50 = the 12:00 bar; each bar adds 1.0417.	0 … 98.96	day_progress=0
`month_progress` Progress through the month, %	Minutes since the 1st ÷ minutes in the month × 100.	0 … <100	month_progress<5
`year_progress` Progress through the year, %	Minutes since 1 January ÷ minutes in the year × 100.	0 … <100	year_progress>=95
`interval` Interval	Always 15 (minutes).	15	interval=15

What each feed carries

	bn_spot	bn_perp	okx_spot	okx_perp	cb_spot	up_spot
110 indicators	✓	✓	✓	✓	✓	✓
open · high · low · close · volume	✓	✓	✓	✓	✓	✓
quote_volume	✓	✓	✓	✓	—	✓
trades · taker_buy_base · taker_buy_quote · taker_buy_ratio	✓	✓	—	—	—	—
price currency	USDT	USDT	USDT	USDT	USD	KRW

The internal up_krwusdt feed (Upbit KRW/USDT rate, from 7 June 2024) exists only to build x_krw_premium; it is not selectable in the app.

Standard reading levels

When a plain-language question says “overbought”, “spike” or “squeeze” without a number, the translator uses these conventional levels. They are starting points, not rules — the observed ranges above show what each indicator really does.

Indicator	Levels
`rsi`	≥ 70 overbought · ≤ 30 oversold
`mfi`	≥ 80 · ≤ 20
`stoch_k / stoch_d / stochrsi_*`	≥ 80 · ≤ 20
`willr`	≥ −20 · ≤ −80
`cci`	≥ 100 · ≤ −100
`cmo`	≥ 50 · ≤ −50
`ultosc / imi`	≥ 70 · ≤ 30
`adx`	≥ 25 trending · ≥ 40 strong · < 20 no trend
`volume_rank_long / volume_rank_short`	≥ 0.95 spike · ≥ 0.9 heavy
`rv_rank`	≤ 0.15 squeeze · ≥ 0.9 explosion
`taker_buy_ratio`	≥ 0.55 buyers dominant · ≤ 0.45 sellers dominant
`body / range`	≥ 1 big candle (1 % of price)
`ret_log`	≥ 1 jump · ≤ −1 drop (15 minutes); ret_log_short ±2 for an hour
`macd`	> 0 golden regime · < 0 death regime; a sign change = cross
`sar`	< 0 uptrend · > 0 downtrend; a sign change = reversal
`bbands_upper / bbands_lower`	upper < 0 = above the band · lower > 0 = below the band

ln(1 + bars) lookup

The event-distance columns store the natural log of one plus the number of bars since the event, capped at 672 (96 for positive_return_streak).

Bars since	Stored value	Meaning
0	0	event on the previous bar
1	0.6931	1 bar earlier
2	1.0986	2 bars (30 min)
3	1.3863	3 bars (45 min)
4	1.6094	4 bars (1 h)
8	2.1972	8 bars (2 h)
16	2.8332	16 bars (4 h)
32	3.4965	32 bars (8 h)
96	4.5747	96 bars (1 day)
192	5.2626	192 bars (2 days)
672	6.5117	672 bars (7 days, the cap)

Reading it in a condition

bn_spot_bars_since_high_breakout=0 — the breakout was on the previous bar.
bn_spot_bars_since_volume_spike<=1.6094 — a volume spike within the last hour.
bn_spot_trend_streak>=3.4965 — the close has been above EMA20 for 32 bars or more.
For arbitrary “in a row” questions, Streak() in the query language is simpler than converting logs.