Latest Metrics

Everything ByteTree has worked out about one asset as of the latest day: its four ByteTrend scores, its trend regime, and its risk and return figures. The full record has around a hundred fields, but most people need only a handful — the fields most people need are listed first, and the complete reference follows.

Currency pairs work too (EURUSD): when no asset metrics exist for a pair, the endpoint answers from exchange-rate data. For currency-only work, Get FX Pair is the dedicated endpoint; for every asset in one call, use Latest Metrics — All Assets.

Request

curl https://api.bytetree.io/api/metrics/AAPL/latest

# Crypto
curl https://api.bytetree.io/api/metrics/BTCUSD/latest

The fields most people need

Key Concepts explains the scores and labels in plain terms. Every field is described in full under Response.

Path Parameters

Response

The data object contains date, close, and all computed metric fields at the top level (see below).

All numeric fields default to adjusted close (split- and dividend-adjusted). Every time-series metric (MAs, slopes, rolling highs/lows, ByteTrend scores, deviations, drawdowns, returns, risk) and the quoted-state fields under Valuation (closeUsd, capr, marketCapUsd) are all computed on adjusted close. Three explicit raw counterparts (closeRawUsd, caprRaw, marketCapRawUsd) are available for forensic / market-printed-price views. See the Data Model page for the exact adjustment contract and the Changelog for cutover history.

ByteTrend Scores (0–5)

Four scores — two ways of measuring, each on two timescales:

• Local vs CAPR — local close vs CAPR (close ÷ MSCI World close)
• Weekly (200w) vs Daily (30d) — long-horizon weekly windows vs short-horizon daily windows

Each score has companion fields that share the same suffix:

• *Rules — the five binary rules that compose the score (booleans: aboveMaLong, aboveMaShort, maLongRising, maShortRising, lastTouchMax)
• change* — day-over-day change (range −5 to +5)
• trendRegime* — trend regime classification (see Trend Regime)
• consec*5 / consec*0 — consecutive days at score 5 or 0

Full set per score, e.g. for Local (Weekly): byteTrendLocal200w, byteTrendLocal200wRules, changeLocal200w, trendRegimeLocal200w, consecLocal200w5, consecLocal200w0. Three equivalent sets exist for Local30d, Capr200w, Capr30d.

Moving Averages

Long/short SMAs for each dimension and window. Every suffix denotes the actual window of that field. Slopes measure day-over-day MA change (trend acceleration).

Rolling Highs & Lows

Two daily windows (used by Daily score regime classification and lastTouchMax rule respectively) and two weekly windows (the 20-week rolling high/low and 200-week trend context).

Deviations

Trend Regime

Trend regimes classify assets by combining their ByteTrend score with whether the price — or, for the CAPR variants, the asset’s performance against MSCI World — is at its recent high or low: the 30-week high or low for the Weekly (200w) variants, the 30-day high or low for the Daily (30d) variants. This is more informative than the score alone: it captures whether a trend is making new highs or sitting at its lows.

Possible trend regime values — the same labels for all four scores, where “high” and “low” mean the 30-week extreme for Weekly variants and the 30-day extreme for Daily ones:

Trend Quality & Conviction

Historical divergence values — before the 2026-04-22 cutover divergenceLocal / divergenceCapr compared the Daily score against a VWAP-weighted weekly score. They now compare Daily against the true 200w-window weekly score. Keys and units are unchanged, but values pre- and post-cutover are not directly comparable in historical series.

Risk Metrics

Drawdown

Performance Returns (USD)

Relative Strength (vs MSCI World)

Valuation

The default trio represents the latest quoted state on adjusted close — matching the basis used by every time-series metric on this page. The *Raw trio exposes the same values on raw close for forensic or audit consumers; see Data Model.

Notes

• Returns the most recent record that has computed metrics; dates without a completed computation run are skipped
• ByteTrend scores range from 0 (fully bearish) to 5 (fully bullish), computed from 5 binary rules
• Sharpe ratio risk-free rate is configurable in engine config (default 3.5%)
• Returns { "symbol": "BTCUSD", "data": null } if no computed data exists for the symbol