Barrie & Hibbert

Knowledge Base » An Algorithm for Imputing Missing Values in Multivariate Data

An Algorithm for Imputing Missing Values in Multivariate Data

Document ID: 2000-314 (previously 2000/013)
Published on: 1st August 2000
Author: Jonathan Rougier, Phil Mowbray

This note describes a method for imputing missing values in a multivariate data set that is consistent with its covariance structure and also returns its noisy characteristics.

• Model
  • 2 Factor Black Karasinski Model
  • 2 Factor Vasicek Model
  • Constant Volatility Model
  • Credit Model
  • Equity Mean-Reversion Model
  • Exchange Rate Model
  • General Equity Model
  • Inflation Model
  • Libor Market Model
  • Local Volatility Model
  • Mortality Model
  • Mortgage-Backed Securities Model
  • Other Models
  • Stochastic Volatility & Jump Diffusion Model
  • Time-Varying Deterministic Volatility Model

• Asset Class
  • Alternative Assets
  • Credit
  • Joint Distribution
  • Exchange Rate / Currency
  • Derivatives
  • Equity
  • Inflation
  • Interest Rates
  • Property

• Product Areas
  • Capital Measurement
  • Product Communication & Advice
  • Product Design & Governance
  • Risk & Capital Management

• Sectors
  • Retail Distribution
  • Asset and Wealth Management
  • Consultancy
  • Insurance
  • Pension Fund

Other categories

• Calibration Methodology & Techniques
• Calibration Notes
  • Market Consistent
  • Real World
• Calibration Targets
  • Target Methodology
  • Target Updates
  • Target Validation
• Events
• Insights
• iESG
  • Version 6.2
• Monte Carlo Methodology
  • Correlation Matrices
  • Random Number Generation
  • Variance Reduction
• Numerical Methods
• Outputs & Validation
• Archive

Frequently Asked Questions

View all Frequently Asked Questions »