OpenBB/openbb_platform/extensions/technical/docs/user-guides/01-indicators.md

Single-Symbol Indicators

The bulk of the extension — 38 endpoints — computes a technical indicator on a single OHLC(V) series. They're grouped by family below. Every example assumes:

from openbb import obb

data = obb.equity.price.historical(
    symbol="SPY", start_date="2022-01-01", provider="yfinance"
).results

Overlays (15 endpoints)

Smooth or envelope the price series. Output overlays on the same axis as price.

sma = obb.technical.sma(data=data, length=50).to_df()
ema = obb.technical.ema(data=data, length=20).to_df()
bb = obb.technical.bbands(data=data, length=20, std=2.0).to_df()
ich = obb.technical.ichimoku(data=data).to_df()
st = obb.technical.supertrend(data=data, length=10, multiplier=3.0).to_df()

Moving-average variants: sma, ema, hma (Hull), wma (Weighted), zlma (Zero-Lag), tema (Triple-EMA), dema (Double-EMA), kama (Kaufman Adaptive), frama (Fractal Adaptive), vwma (Volume-Weighted).

Bands / channels: bbands (Bollinger), donchian, kc (Keltner).

Composites: ichimoku (5-line cloud), supertrend (ATR-anchored trailing band).

Oscillators (10 endpoints)

Bounded momentum indicators. Often consulted alongside an overlay for divergence signals.

rsi = obb.technical.rsi(data=data, length=14).to_df()
stoch = obb.technical.stoch(data=data).to_df()  # %K and %D
mfi = obb.technical.mfi(data=data, length=14).to_df()  # volume-weighted RSI

All 10: rsi, stoch (Stochastic), cci (Commodity Channel), fisher (Fisher Transform), cg (Centre of Gravity), williams_r, mfi (Money Flow Index), trix (Triple Smoothed ROC), ultimate_oscillator, awesome_oscillator.

Trend (5 endpoints)

Decompose the price series into direction + strength signals.

macd = obb.technical.macd(data=data, fast=12, slow=26, signal=9).to_df()
adx = obb.technical.adx(data=data, length=14).to_df()
aroon = obb.technical.aroon(data=data, length=14).to_df()

macd (Moving Average Convergence/Divergence), adx (Average Directional Index), di (Directional Indicator +DI/-DI), aroon, choppiness.

Volume (4 endpoints)

Use the volume column. Useful for confirming or invalidating price moves.

obv = obb.technical.obv(data=data).to_df()
ad = obb.technical.ad(data=data).to_df()
vwap = obb.technical.vwap(data=data, anchor="D").to_df()    # pandas offset alias

obv (On-Balance Volume), ad (Accumulation/Distribution), adosc (Chaikin A/D Oscillator), vwap (Volume-Weighted Average Price).

Volatility (4 endpoints)

atr = obb.technical.atr(data=data, length=14).to_df()

# All six realised-volatility estimators side by side
rv = obb.technical.realized_volatility_compare(data=data, window=30).to_df()

# Volatility cones — historical band of rolling vol vs. realised today
cones = obb.technical.cones(data=data, model="yang_zhang").to_df()

atr (Average True Range), realized_volatility (single estimator), realized_volatility_compare (all six estimators in one frame), cones (per-window realised vs. historical band).

Estimators available for realized_volatility and cones: std (close-to-close), parkinson (range-based), garman_klass (OHLC), hodges_tompkins (bias-corrected close-to-close), rogers_satchell (drift-invariant range), yang_zhang (minimum-variance composite).

Structure (3 endpoints)

Price-action levels (support/resistance/pivot points).

fib = obb.technical.fib(data=data, period=120).to_df()
dem = obb.technical.demark(data=data).to_df()
piv = obb.technical.pivot_points(data=data, method="classic", anchor="day").to_df()

fib (Fibonacci retracement), demark (sequential 9 setup), pivot_points (five methods: classic, fibonacci, woodie, camarilla, demark; three anchors: day, week, month).

Statistics (6 endpoints)

Distributional and time-series diagnostics on a price or return series.

clenow = obb.technical.clenow(data=data, period=90).to_df()
dd = obb.technical.drawdown(data=data).to_df()
stats = obb.technical.returns_stats(data=data, frequency="daily").to_df()
stat = obb.technical.stationarity(data=data).to_df()
hurst = obb.technical.hurst(data=data).to_df()
acf = obb.technical.autocorrelation(data=data, max_lag=20).to_df()

clenow (volatility-adjusted momentum), drawdown (peak-to-trough), returns_stats (skew, kurtosis, sharpe, sortino, etc.), stationarity (ADF + KPSS hypothesis tests), hurst (R/S and DFA exponents), autocorrelation (ACF + PACF with confidence bands).

Common parameters

Every data-consuming endpoint accepts:

• data — the price series (list[Data] / DataFrame / records).
• index — column to use as the time index (default "date").
• target — for endpoints that work on a single column, the column name (default "close").

Overlay/oscillator endpoints expose a length (or period) parameter for the rolling window. Most also accept offset to shift the output. See the per-endpoint docstring (help(obb.technical.<name>)) for the full parameter list.

Choosing an input shape

pandas.DataFrame is convenient if you're already in pandas:

import pandas as pd
df = pd.read_csv("spy.csv")
obb.technical.rsi(data=df, length=14)

list[dict] is the simplest when reading from CSV or JSON without pandas:

import csv
with open("spy.csv") as f:
    records = list(csv.DictReader(f))
obb.technical.rsi(data=records, length=14)

Both produce identical output.