Volatility Calculation - Methodology
How we measure risk and volatility
Table of Contents
Overview
The CoinRisqLab 80 Volatility system calculates risk metrics for both individual cryptocurrencies and the CoinRisqLab 80 Index portfolio. Our methodology uses logarithmic returns and implements a rolling window approach to measure historical volatility.
For portfolio-level calculations, we use proper market-cap weighting and full covariance matrix calculations to account for correlations between constituents, providing a comprehensive view of portfolio risk.
This approach is based on Modern Portfolio Theory and uses the same statistical methods employed by professional financial institutions to measure and manage risk.
Glossary
Volatility
A statistical measure of the dispersion of returns for a given security or market index. Higher volatility means higher risk and larger price swings.
Logarithmic Returns
The natural logarithm of the ratio of consecutive prices. Logarithmic returns have better statistical properties than simple percentage returns.
Log_Return = ln(Price_today / Price_yesterday)Standard Deviation
A measure of the amount of variation in a set of values. In finance, the standard deviation of returns is used as a measure of volatility.
Annualization
The process of converting a daily volatility measure to an annual equivalent by multiplying by the square root of the number of trading periods in a year.
Annual_Volatility = Daily_Volatility × √365Rolling Window
A fixed-size time period (e.g., 90 days) that slides forward in time. Each calculation uses the most recent N observations, providing a moving view of volatility.
Covariance Matrix
A square matrix showing the covariance between pairs of assets. Used to capture how different assets move together, essential for portfolio risk calculations.
Portfolio Volatility
The volatility of a portfolio that accounts for both individual asset volatilities and their correlations. Usually lower than the weighted average of individual volatilities due to diversification.
Trading Days
For annualization purposes, we use 365 trading days per year. Unlike traditional financial markets that close on weekends and holidays, cryptocurrency markets operate 24/7, 365 days a year.
Bessel's Correction
When calculating variance from a sample of data (like 90 days of returns), we divide by n-1 instead of n to get an unbiased estimate. This is applied to all our variance and covariance calculations.
Risk Level Classification
We classify volatility into four risk levels using rigorous analytical standards calibrated for cryptocurrency markets. This classification covers both annualized and daily volatility, providing an intuitive way to understand and compare risk across different time scales.
| Risk Level | Annualized | Daily | Color | Description |
|---|---|---|---|---|
Low Risk | < 10% | < 0.52% | Green | Stable mature cryptos, low risk |
Medium Risk | 10% - 30% | 0.52% - 1.57% | Yellow | Moderate volatility, established assets with fluctuations |
High Risk | 30% - 60% | 1.57% - 3.14% | Orange | High volatility, speculative assets or unstable phases |
Extreme Risk | ≥ 60% | ≥ 3.14% | Red | Extreme volatility, high-risk altcoins or strong market turbulence |
Application Usage:
- Portfolio risk level indicators on the dashboard
- Individual cryptocurrency volatility badges
- Risk contribution analysis in portfolio breakdown
- Volatility gauge on the main dashboard
Note: These thresholds are calibrated for cryptocurrency markets, which typically exhibit higher volatility than traditional financial assets. A "low risk" crypto asset (5% annualized volatility) would still be considered moderate to high risk in traditional equity markets.
Base Parameters
| Parameter | Value | Description |
|---|---|---|
| Window Period | 90 days | Rolling window for volatility calculations |
| Return Type | Logarithmic | Natural logarithm of price ratios |
| Annualization Factor | √365 ≈ 19.10 | Assumes 365 trading days per year (crypto markets 24/7) |
| Minimum Data Points | 90 observations | Required for volatility calculation |
Calculation Pipeline
The volatility calculation follows a three-stage pipeline, where each stage builds upon the previous one:
Log Returns Calculation
Calculate daily logarithmic returns for all cryptocurrencies
Individual Crypto Volatility
Calculate rolling volatility for each cryptocurrency
Portfolio Volatility
Calculate index-level volatility using covariance matrix
Logarithmic Returns
The first stage calculates daily logarithmic returns for all cryptocurrencies, which serve as the foundation for all subsequent volatility calculations.
What are Log Returns?
Logarithmic returns measure the continuously compounded rate of return between two periods. They have several advantages over simple percentage returns:
- Time-additive: Returns over multiple periods sum algebraically
- Symmetric: A +10% gain and -10% loss produce equal absolute log returns
- Better statistics: More suitable for normal distribution assumptions
- Approximation: For small changes, log returns ≈ percentage changes
Calculation Formula
For each consecutive pair of days:
Where ln() is the natural logarithm function.
Data Selection
We select the latest price for each day:
- End-of-day snapshot (latest timestamp per day)
- Only positive prices (price_usd > 0)
- Ordered chronologically
Individual Crypto Volatility
The second stage calculates rolling volatility for individual cryptocurrencies using the log returns from Stage 1.
Rolling Window Setup
For each cryptocurrency with sufficient data (≥90 log returns):
The window slides forward one day at a time, always containing exactly 90 consecutive daily log returns.
Statistical Calculations
For each 90-day window, we calculate:
a) Mean Return
b) Daily Volatility (Standard Deviation)
c) Annualized Volatility
Why Multiply by √365?
The square root of time rule applies under the assumption of independent and identically distributed returns. Since cryptocurrency markets operate 24/7 throughout the year, we use 365 days to convert daily volatility to an annual measure that's comparable across different assets and time periods.
Portfolio Volatility
The third stage calculates the volatility of the CoinRisqLab 80 Index portfolio using market-cap weights and the full covariance matrix to account for correlations between constituents.
Why Use a Covariance Matrix?
Simply taking a weighted average of individual volatilities would overestimate portfolio risk. The covariance matrix captures how assets move together:
- Assets that move in opposite directions reduce portfolio risk
- Imperfect correlation provides diversification benefits
- Portfolio volatility is typically lower than weighted average of individual volatilities
Step 1: Weight Calculation
Weights are based on market capitalization:
Important: Weights must sum to 1.0 (100%)
Step 2: Covariance Matrix Construction
Build the covariance matrix for all constituent pairs:
The covariance matrix is an n×n symmetric matrix where:
- Diagonal elements: variances of individual assets
- Off-diagonal elements: covariances between asset pairs
- Captures the correlation structure of the portfolio
Step 3: Portfolio Variance Calculation
Using modern portfolio theory:
σ²_portfolio = w' × Σ × w Where: w = column vector of weights [w₁, w₂, ..., wₙ]' Σ = covariance matrix (n×n) w' = transpose of weight vector
Expanded form:
Step 4: Annualization
Calculation Examples
Example 1: Individual Crypto Volatility
Given: Bitcoin (BTC) with 90-day log returns
Example 2: Portfolio Volatility (Simplified)
Given: 2-asset portfolio for simplicity
Covariance matrix:
Σ = [σ₁² ρ×σ₁×σ₂ ]
[ρ×σ₁×σ₂ σ₂² ]
= [0.0009 0.00084 ]
[0.00084 0.0016 ]Portfolio variance:
σ²_p = [0.6 0.4] × [0.0009 0.00084] × [0.6]
[0.00084 0.0016 ] [0.4]
= [0.6 0.4] × [0.000876]
[0.001144]
= 0.000983Diversification Benefit
The portfolio volatility calculation demonstrates a key principle of Modern Portfolio Theory: diversification reduces risk.
The Diversification Effect
From Example 2 above:
The portfolio volatility (59.8%) is 5.1% lower than the weighted average (64.9%), demonstrating the benefit of diversification!
Mathematical Relationship
The general principle:
This inequality holds as long as assets are not perfectly correlated. The benefit is greater when:
- Correlations between assets are lower
- The portfolio is more diversified (more constituents)
- Asset weights are more balanced