3.2.1. Robustness tests
We perform several additional tests on the sample of positive exposures to examine
the robustness of our results. First, we examine whether our results depend on
the three-year time interval (1992–94) that we use to estimate firm exposures. We
therefore estimate exposures using a longer (five-year) time interval (1991–95).
Regression 1 in Table 3 presents the results of this test. The results are very similar
to those of the base-case regression (Regression 3, Table 2). Consistent with our
hypothesis, we find a positive and significant relation between foreign sales and
exposure, and a negative and significant relation between foreign currency derivative
use and currency exposure.