We propose a new variant of RESET that is appropriate for distributed lag models. Monte Carlo evidence on size and power strongly supports the use of the new variant instead of the traditional RESET.

We propose a new variant of RESET that is appropriate for distributed lag models. Monte Carlo evidence on size and power strongly supports the use of the new variant instead of the traditional RESET.

This paper improves the test for deficit sustainability developed by Hatzinikolaou and Simos (2013, henceforth HS) by taking into account structural breaks when deriving critical values for the test. Using quarterly data on the US current-account deficit, 1947.1-2012.2, we find that taking into...