XRPL RESEARCH LABACTIVE VERIFIED-DATA EXPERIMENTS
● LOADING
ACTIVE VERIFIED-DATA RESEARCH PROGRAM

Rolling XRP Downside Classification

Hypothesis: current Kraken market-pressure features improve one-hour downside classification beyond momentum and volatility controls.

WINDOW ROLLING HORIZON 1 HOUR FREQUENCY 15-MIN CANDLES SCHEDULE EVERY 15 MIN
ACTIVE EXPERIMENT

CHECKING VERIFIED INPUTS

Loading the newest chronological holdout test…

--
OPERATIONAL LIVE PATH

XRPL Validated-Ledger Observatory

CONNECTING
LEDGER INDEX--
CLOSE TIME--
TRANSACTIONS--
NATIVE XRP PAYMENTS--
PAYMENT VOLUME--
≥100K XRP TRANSFERS--
LARGEST TRANSFER--
VALIDATION--
Validated XRPL telemetry is used as a live evidence-freshness gate. It is not inserted into historical candle rows and exchange ownership is never inferred.

Regime-Specific Holdout AUC

Verified market-pressure model versus momentum + volatility controls
CALCULATING
Calculating the newest chronological holdout.

Verified Predictor Effects

Standardized training-sample effect weights; not causal coefficients
CALCULATING
Calculating verified-data effect weights.

Evidence Coverage

Quality gates for a publishable result
Observations--
Market coverage--
Nonzero volume candles--
Holdout downside rate--

Verified Inputs & Controls

Every feature comes from named public data
  • Control: XRP one-hour momentum
  • Control: recent realized volatility
  • Candidate: Kraken volume pressure
  • Candidate: close versus Kraken VWAP
  • Candle range and trade-count pressure
  • Validated XRPL freshness gate

Decision Rule

The stated result is tested, never assumed
Overall ΔAUC--
High-vol ΔAUC ≥ 0.02--
Low-vol ΔAUC ≥ 0.02--
ConclusionNOT TESTED

Methodology & Limitations

Audit notes carried with every experiment
Candidate predictors

Kraken volume pressure, VWAP deviation, candle range, and trade-count pressure beyond the control model.

Outcome

Whether XRP/USD closes lower four completed 15-minute candles later.

Validation

Chronological 70/30 holdout with a four-candle purge and no future predictors.

Gated study

The separate three-year XRPL exchange-inflow study remains blocked until authoritative exchange-label provenance is supplied.