Volatility and market risk: Value-at-Risk (VaR) applied to the GOBIXDR index

What it is about

Research on the market risk of government bonds issued and traded in the local market, using the GOBIXDR index as a benchmark. Volatility is used to assess the performance of portfolios and investment strategies and to price the capital consumption of exposures.

The GARCH methodology is used to find the optimal model, then test and validate it; value at risk (VaR) is quantified with filtered historical simulation and the Monte Carlo method; capital consumption is estimated for trading, available-for-sale and held-to-maturity portfolios; and, trained on 2022 data, the Stress-VaR scenario is presented. The results apply to trading risk management and to the management of the equity position.

Session summary

The speaker presents research on the volatility of the Dominican stock exchange’s government bond index (GOBX-DR), with about ten years of data that already include the 2022 monetary tightening cycle. He explains what volatility is and what it is used for (assessing risk-adjusted performance, costing capital consumption, valuing derivatives and making investment decisions), distinguishes realized from implied volatility and reviews its empirical regularities: mean reversion, clustering and persistence, downside asymmetry and jumps. He defines value at risk and conditional value at risk and the reasons for using them as market risk measures.

He compares parametric, historical simulation and Monte Carlo methodologies, and applies GARCH-family models with filtered historical simulation, evaluated with coverage backtests. He concludes that, even before the 2022 cycle, the index made it possible to anticipate larger drops than those already seen; that in turbulent periods value at risk is multiplied several times; and that decisions on which portfolio to classify investments in have a significant effect on capital consumption. He warns about model risk and the rigidity of generic tools. The questions covered VaR versus conditional VaR, the representativeness of the index given the weight of the off-exchange market, and its sensitivity to inflation and the exchange rate.

Topics covered

• What volatility is and what it is used for in risk management
• Realized and implied volatility and their empirical regularities
• Value at risk and conditional value at risk as market risk measures
• Methodologies: parametric, historical simulation and Monte Carlo
• GARCH models applied to the GOBX-DR bond index
• Results of the 2022 cycle and capital consumption by portfolio
• Questions: VaR versus conditional VaR, off-exchange market and inflation