Skip to main content

Overview

  • Dataset ID: 0411
  • Table: 0411_liquidity_pools_yields
  • What it is: Rolling yield/ROI predictions per pool over fixed horizons (1D, 3D, 7D, 14D, 30D, 90D, 365D), based on historical swap fees (0303_liquidity_pools_swap_fees) and current reserves (latest 0301 snapshot).
  • Grain: 1 row per (pool, as-of event triple, horizon).
  • Primary location fields: pool_uid, block_number, block_time (UTC), tx_index, log_index
  • Join keys:
    • Pool: pool_uid0203_liquidity_pools.uid
    • Provenance: (block_number, tx_index, log_index)in0103_logs
This dataset emits aligned arrays (one element per pool token). Use tokens[] to line up current_reserves[], volumes, fees, and ROI predictions.

Core concepts

  • Target horizon vs observed history
    • target_period_days: horizon you requested (1, 3, 7, 14, 30, 90, 365)
    • observed_period_days: how many days of history were available/used
  • Extrapolation
    • When observed_period_days < target_period_days, the dataset marks the row as extrapolated and scales observed values by:
      • (extrapolation_factor = target_period_days / observed_period_days)
  • ROI
    • For each token index i:
      • (roi_predicted[i] = user_fees_predicted[i] / current_reserves[i])

Columns

ColumnTypeDescription
idBIGINTSurrogate identity for the yield record (auto-incremented).
pool_uidBYTEAPool identifier (internal).
block_numberBIGINTBlock height of the as-of snapshot anchor.
block_timeTIMESTAMPTZUTC timestamp of the as-of snapshot anchor.
tx_indexINTEGERTransaction index within the block.
log_indexINTEGERLog index within the transaction.
target_period_daysSMALLINTTarget horizon in days. Allowed: 1, 3, 7, 14, 30, 90, 365.
observed_period_daysSMALLINTHistory actually used (must be > 0 and <= target_period_days).
is_full_periodBOOLEANtrue when observed_period_days == target_period_days.
is_extrapolatedBOOLEANtrue when the row is scaled from shorter history.
extrapolation_factorNUMERIC(18,6)Scaling factor applied when extrapolating (must be > 1 when extrapolated; 1 when full period).
window_start_timeTIMESTAMPTZStart time of the observed window.
window_end_timeTIMESTAMPTZEnd time of the observed window.
tokensBYTEA[]Pool token addresses (aligned arrays index by this order).
current_reservesNUMERIC(78,18)[]Current reserves per token (decimals-adjusted).
volume_observedNUMERIC(78,18)[]Observed traded volume per token in the window.
volume_predictedNUMERIC(78,18)[]Predicted traded volume per token for the target horizon.
user_fees_observedNUMERIC(78,18)[]Observed user/LP fees per token in the window.
user_fees_predictedNUMERIC(78,18)[]Predicted user/LP fees per token for the target horizon.
roi_predictedNUMERIC(78,18)[]Predicted ROI fraction per token for the target horizon.
_tracing_idBYTEADeterministic BlockDB lineage identifier for the yield record.
_genesis_tracing_idsBYTEA[]Tracing IDs for the raw artifacts that seeded this record.
_parent_tracing_idsBYTEA[]Tracing IDs for upstream derived records referenced during computation.
_created_atTIMESTAMPTZRecord creation timestamp.
_updated_atTIMESTAMPTZRecord update timestamp.

Use Cases

  • Pool-level yield and ROI forecasting across standardized horizons
  • Ranking pools by predicted fee yield (token-wise and pool-wise)
  • Inputs for portfolio allocation and liquidity mining analytics
  • Validating fee models by comparing predicted vs realized forward windows