Metrics framework

WEETHAave V3 · Ethereum

Every row states its formula, its inputs, its source and when it was retrieved. Where there is no figure there is a reason, not a dash.

24 of 109 metrics carry a value. The rest state what they are missing. That is not a gap in the presentation: this application's data is a yield and capital series — no price, no order book, no holder distribution and no contract data.

Every metric in this section refers to the tvlUsd series over 905 observations. Metrics that need a high and a low per period cannot be computed over it, and no close-only substitute is produced.

Trend

SMA 20 — Show the explanation3,976,000,000 USD
Definition
The arithmetic mean of the last `period` observations.
Formula
mean(series[t-period+1 .. t])
Inputs
series, period
Reading
Smooths the series; it lags it by roughly half the window.
Limit
Needs `period` observations. With fewer there is no value, not a shorter average.
Method
Standard definition. · sma · TV-1.0 · EXPERIMENTAL
SMA 50 — Show the explanation3,599,000,000 USD
Definition
The arithmetic mean of the last `period` observations.
Formula
mean(series[t-period+1 .. t])
Inputs
series, period
Reading
Smooths the series; it lags it by roughly half the window.
Limit
Needs `period` observations. With fewer there is no value, not a shorter average.
Method
Standard definition. · sma · TV-1.0 · EXPERIMENTAL
SMA 200 — Show the explanation2,941,000,000 USD
Definition
The arithmetic mean of the last `period` observations.
Formula
mean(series[t-period+1 .. t])
Inputs
series, period
Reading
Smooths the series; it lags it by roughly half the window.
Limit
Needs `period` observations. With fewer there is no value, not a shorter average.
Method
Standard definition. · sma · TV-1.0 · EXPERIMENTAL
EMA 20 — Show the explanation3,962,000,000 USD
Definition
A weighted average with geometrically decaying weights, seeded with the SMA of the first full window.
Formula
alpha = 2 / (period + 1); ema_t = alpha * x_t + (1 - alpha) * ema_{t-1}; seed = SMA of the first `period` values
Inputs
series, period
Reading
Reacts faster than the SMA of the same length.
Limit
The seed affects the early values, so the seeding rule is part of the formula rather than left to the caller.
Method
Standard definition with an SMA seed. · ema · TV-1.0 · EXPERIMENTAL
EMA 50 — Show the explanation3,585,000,000 USD
Definition
A weighted average with geometrically decaying weights, seeded with the SMA of the first full window.
Formula
alpha = 2 / (period + 1); ema_t = alpha * x_t + (1 - alpha) * ema_{t-1}; seed = SMA of the first `period` values
Inputs
series, period
Reading
Reacts faster than the SMA of the same length.
Limit
The seed affects the early values, so the seeding rule is part of the formula rather than left to the caller.
Method
Standard definition with an SMA seed. · ema · TV-1.0 · EXPERIMENTAL
EMA 200 — Show the explanation3,386,000,000 USD
Definition
A weighted average with geometrically decaying weights, seeded with the SMA of the first full window.
Formula
alpha = 2 / (period + 1); ema_t = alpha * x_t + (1 - alpha) * ema_{t-1}; seed = SMA of the first `period` values
Inputs
series, period
Reading
Reacts faster than the SMA of the same length.
Limit
The seed affects the early values, so the seeding rule is part of the formula rather than left to the caller.
Method
Standard definition with an SMA seed. · ema · TV-1.0 · EXPERIMENTAL
Order of the averagesema20>ema50>ema200
ADX — Show the explanationMODEL_NOT_APPLICABLE · Model not applicable

Reason: series has no high/low per period

Definition
A measure of how directional movement is, independent of its direction.
Formula
+DM/-DM over `period` (Wilder), DX = 100 * |(+DI) - (-DI)| / ((+DI) + (-DI)), ADX = Wilder average of DX
Inputs
bars, period
Reading
Higher means movement was more directional. It does not say in which direction.
Limit
Requires a high, low and close per period, and at least 2 × `period` periods. The threshold 25 is convention, not a rule here.
Method
Wilder (1978). · adx · TV-1.0 · EXPERIMENTAL
  • The series carries one value per day — without a high and a low this metric is undefined.

Momentum

RSI — Show the explanation70.34
Definition
The ratio of upward to downward movement, smoothed over `period` and mapped to 0–100.
Formula
Wilder smoothing of gains/losses over `period`; RSI = 100 - 100 / (1 + avgGain / avgLoss)
Inputs
series, period
Reading
Describes how one-sided the movements in the window were.
Limit
No threshold is stored here as a signal: "70/30" is convention, not a property of the measure. With no downward move in the window the value is 100 and says only that.
Method
Wilder (1978), New Concepts in Technical Trading Systems. · rsi · TV-1.0 · EXPERIMENTAL
MACD — Show the explanation162,300,000 USD
Definition
The distance between two EMAs of different length, its own EMA, and the difference of the two.
Formula
macd = EMA(fast) - EMA(slow); signal = EMA(macd, signalPeriod); histogram = macd - signal
Inputs
series, fastPeriod, slowPeriod, signalPeriod
Reading
Describes whether the shorter average sits above or below the longer one, and whether that is changing.
Limit
In the series' own units, so not comparable across assets. A sign change here is a sign change, not a signal.
Method
Appel; the standard parameters 12/26/9 serve only as a default, not as a recommendation. · macd · TV-1.0 · EXPERIMENTAL
Signal line — Show the explanation181,400,000 USD
Definition
The distance between two EMAs of different length, its own EMA, and the difference of the two.
Formula
macd = EMA(fast) - EMA(slow); signal = EMA(macd, signalPeriod); histogram = macd - signal
Inputs
series, fastPeriod, slowPeriod, signalPeriod
Reading
Describes whether the shorter average sits above or below the longer one, and whether that is changing.
Limit
In the series' own units, so not comparable across assets. A sign change here is a sign change, not a signal.
Method
Appel; the standard parameters 12/26/9 serve only as a default, not as a recommendation. · macd · TV-1.0 · EXPERIMENTAL
Distance to the signal line — Show the explanation-19,070,000 USD
Definition
The distance between two EMAs of different length, its own EMA, and the difference of the two.
Formula
macd = EMA(fast) - EMA(slow); signal = EMA(macd, signalPeriod); histogram = macd - signal
Inputs
series, fastPeriod, slowPeriod, signalPeriod
Reading
Describes whether the shorter average sits above or below the longer one, and whether that is changing.
Limit
In the series' own units, so not comparable across assets. A sign change here is a sign change, not a signal.
Method
Appel; the standard parameters 12/26/9 serve only as a default, not as a recommendation. · macd · TV-1.0 · EXPERIMENTAL
ROC — Show the explanation0.01557
Definition
Relative change over `period` periods.
Formula
(x_t - x_{t-period}) / x_{t-period}
Inputs
series, period
Reading
A difference between two points in time; what happened between them is not in it.
Limit
Undefined when the base value is 0.
Method
Standard definition. · roc · TV-1.0 · EXPERIMENTAL

Variation

ATR — Show the explanationMODEL_NOT_APPLICABLE · Model not applicable

Reason: ATR requires a high, low and close per period; a close-only series has no true range

Definition
The average true range per period, smoothed over `period`.
Formula
TR_t = max(high-low, |high-prevClose|, |low-prevClose|); ATR = Wilder average of TR over `period`
Inputs
bars, period
Reading
In the series' own units. A width of movement, not a direction.
Limit
Requires a high, low and close per period. Over a close-only series ATR is not computable, and no close-to-close substitute is produced here.
Method
Wilder (1978). · atr · TV-1.0 · EXPERIMENTAL
  • The series carries one value per day — without a high and a low this metric is undefined.
Historical variation — Show the explanation1.328
Definition
The annualized standard deviation of log period returns of the observed series.
Formula
stdDev(logReturns) * sqrt(periodsPerYear)
Inputs
series, periodsPerYear
Reading
Describes the measured series' variation over the measured window.
Limit
Annualizing assumes equal-length periods and independent returns. Over an APY or TVL series this is not price volatility — the result states which series it is.
Method
Standard definition; square-root-of-time annualization. · realizedVolatility · TV-1.0 · EXPERIMENTAL
Realized variation — Show the explanation1.13
Definition
The annualized standard deviation of log period returns of the observed series.
Formula
stdDev(logReturns) * sqrt(periodsPerYear)
Inputs
series, periodsPerYear
Reading
Describes the measured series' variation over the measured window.
Limit
Annualizing assumes equal-length periods and independent returns. Over an APY or TVL series this is not price volatility — the result states which series it is.
Method
Standard definition; square-root-of-time annualization. · realizedVolatility · TV-1.0 · EXPERIMENTAL
Upper band — Show the explanation4,348,000,000 USD
Definition
A moving average with a band at a distance of k standard deviations of the same window.
Formula
middle = SMA(period); upper = middle + k * stdDev(window); lower = middle - k * stdDev(window)
Inputs
series, period, k
Reading
The band width describes the dispersion within the window.
Limit
k = 2 covers about 95% under normally distributed returns — returns rarely are.
Method
Bollinger. · bollingerBands · TV-1.0 · EXPERIMENTAL
Middle band — Show the explanation3,976,000,000 USD
Definition
A moving average with a band at a distance of k standard deviations of the same window.
Formula
middle = SMA(period); upper = middle + k * stdDev(window); lower = middle - k * stdDev(window)
Inputs
series, period, k
Reading
The band width describes the dispersion within the window.
Limit
k = 2 covers about 95% under normally distributed returns — returns rarely are.
Method
Bollinger. · bollingerBands · TV-1.0 · EXPERIMENTAL
Lower band — Show the explanation3,605,000,000 USD
Definition
A moving average with a band at a distance of k standard deviations of the same window.
Formula
middle = SMA(period); upper = middle + k * stdDev(window); lower = middle - k * stdDev(window)
Inputs
series, period, k
Reading
The band width describes the dispersion within the window.
Limit
k = 2 covers about 95% under normally distributed returns — returns rarely are.
Method
Bollinger. · bollingerBands · TV-1.0 · EXPERIMENTAL

Structure and regime

A turning point counts as confirmed only once enough periods follow it — the most recent move is necessarily absent. A description of the past, not a forecast and not a signal.

Higher highyes
Higher lowyes
Lower highno
Lower lowno
Beyond the rangeno
Below the rangeno
RegimeINSUFFICIENT_DATA — inputs insufficient for any rule
Rule appliedinputs insufficient for any rule

To the market →To the analysis framework →To this framework's methodology →

Data as of: Oct 5, 2026, 3:45 PM UTC