TY - JOUR T1 - Time Series Modeling of Daily Log-Price Ranges for CHF/USD and USD/GBP JF - The Journal of Derivatives SP - 39 LP - 59 DO - 10.3905/jod.2007.699045 VL - 15 IS - 2 AU - Celso Brunetti AU - Peter M. Lildholdt Y1 - 2007/11/30 UR - https://pm-research.com/content/15/2/39.abstract N2 - Volatility of the underlying asset is a critical input to modern option pricing models. How to get the best volatility estimate is, therefore, a matter of great importance to options traders. One technique that has been known for a long time, although it is less commonly used, involves not just daily closing prices, but the daily high and low. This actually introduces quite a bit more information into the calculation, since the high and low are functions of the entire intraday price path for the underlying, not just the price at a single point in time. In contrast, models that allow volatility to vary over time, notably GARCH, have become popular as a way to capture time-variation in the volatility process. This article introduces a GARCH-type Multiplicative Error Model (MEM) for the daily price range. Like other GARCH specifications, the MEM can be modified to accommodate different densities for the stochastic shocks, different orders of dependence on past shocks and volatilities, and even fractional integration to produce a long-memory process. The authors fit the model to the daily price ranges for the CHF/USD and USD/GBP exchange rates and show it to be quite effective in producing accurate (and profitable) volatility predictions.TOPICS: Options, simulations, technical analysis ER -