Eigen-COS VaR: Real-Time Exact Risk for High-Dimensional Portfolios
Unreviewed draft. This paper has not been human-reviewed.
Mathematical claims may be unverified. Use with appropriate caution.
Summary
Value-at-Risk (VaR) and Expected Shortfall (ES) for non-normal portfolios typically require a trade-off: slow Monte Carlo simulations or inaccurate parametric approximations (Cornish-Fisher). We present the Eigen-COS VaR method, which computes exact risk measures for sums of correlated lognormals in milliseconds.
Length
307 words
Status
Outline
Target
Journal of Risk / Risk Magazine