Fair value in investment portfolios: implications for market risk management

What it is about

Challenges and best practices in measuring fair value in portfolios with complex or illiquid instruments, and their impact on market risk models.

Session summary

The speaker organizes the topic around three axes: business model and investment portfolios, market prices and valuation, and market risk. He reviews how valuation has evolved in the Dominican market, explains why moving to fair-value accounting changes treasury management and argues that the business model should determine the accounting classification, not the other way around. He details which characteristics (horizon, turnover, risk appetite, concentration and the weight of the portfolio) point an institution toward an active, mixed or passive model.

On prices, he stresses that for a long time there will be more ‘mark-to-model’ than mark-to-market valuation, that data quality and cleaning are key, and that institutions should monitor valuation, have a pricing committee, a technical format for challenges and, if possible, an independent price verification. He presents an example of filtering transaction data with machine learning and of estimating yield curves. On market risk, he warns that noisy price series distort risk measures, backtesting and capital calculations. The questions covered recommendations for internal audit, the difference between market value and model value, small institutions with a passive portfolio, the governance of the price provider and hedging without a derivatives market.

Topics covered

• Evolution of valuation and of the Dominican debt market
• The business model determines accounting classification and fair value
• Market and model prices: data quality and cleaning
• Pricing committee, challenges and independent price verification
• Data filtering and yield curves with machine learning
• Market risk, capital and risk-based supervision
• Questions: audit, small institutions, price provider and lack of derivatives