This study aims to capture volatility patterns using GARCH (1,1) models. It evaluates these models to obtain one-step-ahead forecastabilities by employing four major forecasting evaluation criteria, and compares two different currencies— the Pakistan rupee and the US dollar—as domestic and foreign currency-valued exchange rates, respectively. The results show that using an international vehicle currency is favorable in Pakistan’s context. However, the Kuwaiti dinar, Canadian dollar, US dollar, Singapore dollar, Hong Kong dollar, and Malaysian ringgit are found to be preferable when performing direct international transactions. Using the root mean square errors and mean absolute errors techniques, the study also assess the robustness of measuring one-step-ahead forecasts.