Value Commentary

The Impact of Volatility on MBS Valuation

Implied volatility in the swaption market has been on the decline recently, illustrated in Chart 1 on the next page. In this environment, it is useful to consider the impact that volatility can have on the value of a portfolio. The sensitivity of an instrument to volatility is known as its vega. For this study, we used a Hull-White (normal) term structure model to define vega as the percentage change in price for a 100 basis point change in absolute volatility.

The data, gathered from the April 17 market analysis, is published on the Andrew Davidson & Co., Inc. website. On this date, calibrating the interest rate process implied volatilities in the swaption market showed that the best constant volatility for the Hull-White model was 132 basis points, with mean reversion of 3.5%. This was the starting point for the analysis. We then shocked the volatility up and down by 50 basis points to test what effect this would have on the value of the TBAs used for the market analysis.

Tables 1 and 2 show the results of the analysis for FNMA and GNMA MBS, respectively. In these tables, base price is the quoted price of the TBA on Bloomberg. Down price is the computed price, given constant OAS, of the TBA with a volatility of 82 basis points, and up price is the computed price at 182 basis points volatility. Vega is defined as 100 * (down_price - up_price) / base_price.

As higher volatility always increases the option cost, the value of most MBS will decline when volatility rises. The results of this analysis show that the effect is strongest for the lowest coupons. This is not surprising, as the prepayments of high premiums are fairly rate-insensitive. The lower coupons, however, show a lot of sensitivity to volatility. The Ginnie 5, for example, would theoretically fall in value by 3.8% if volatility rises by 100 basis points.

In conclusion, the applied volatility assumption can have a large impact on your analysis. The Andrew Davidson & Co., Inc. Interest Rate Process can calibrate itself to the swaption market, ensuring a realistic volatility in your analysis.

>>>