Adaptive Quantile Aggregation
Rather than a simple weighted average, OCRR uses adaptive quantile aggregation.
We take the central 80% of the distribution (trimming extreme outliers) and compute
a weighted mean.
HOW QUANTILE TRIMMING WORKS
RATE DISTRIBUTION
12.5%
8.2%
5.8%
5.3%
4.9%
4.7%
4.5%
4.2%
3.9%
3.6%
1.8%
0.5%
◄ 10th %
80%
90th % ►
STEP 1: SORT & IDENTIFY
Rates sorted lowest to highest. Calculate 10th and 90th percentiles for central 80% boundaries.
STEP 2: TRIM OUTLIERS
Top 10% and bottom 10% excluded. Removes edge cases and temporary anomalies.
STEP 3: WEIGHTED AVERAGE
Apply √TVL weights to central 80% and compute weighted average for final OCRR rate.
OCRR = Σ(wi × ri) / Σwi
Why Not Simple Average?
A simple average can be skewed by temporary anomalies. Quantile trimming ensures
the rate reflects the core market, and is not overly impacted by edge cases.
This methodology mirrors how SOFR (Secured Overnight Financing Rate) handles outliers
in traditional finance, a proven, robust, and transparent methodology.