Model Risk in Risk Analysis for No-Negative-Equity-Guarantees

Abstract

Understanding the risk for No-Negative-Equity-Guarantees (NNEGs) requires the proper modeling of the housing return, interest rate, and mortality rate dynamics. This article investigates the model risk for the risk measures of NNEGs by calculating the Value-at-Risk (VaR) and Conditional-Tail-Expectation (CTE) from the provider perspective, with an emphasis on the housing price return model. Therefore, we propose a jump ARMA-GARCH model, according to nationwide house price return data in the UK. Interest rate and mortality rate dynamics are assumed to follow the CIR model (Cox et al. 1985) and the CBD model (Cairns et al. 2006) respectively. Our numerical analyses reveal that the housing price risk, interest-rate risk, and longevity risk can affect the VaR and CTE of NNEGs, with the impact being as significant as that for housing risk. The VaR and CTE of NNEGs will be greater for female borrowers than for male borrowers, essentially because females have a longer life expectancy. The proposed framework can help financial institutions manage the major three risk factors for NNEGs and assist in meeting the regulator’s concerns. TOPICS:Quantitative methods, real estate, VAR and use of alternative risk measures of trading risk Key Findings ▪ We investigate model risk in risk analysis for No-Negative-Equity-Guarantees. ▪ Our numerical analyses reveal that the housing price risk, interest-rate risk, and longevity risk can affect the VaR and CTE of NNEGs, with the impact being as significant as that for housing risk.