Chen Y.-T., Hsieh C.-S.Hsieh C.-S.Chen Y.-T.CHIH-SHENG HSIEHYI-TING CHEN2019-10-242019-10-24201014798409https://www2.scopus.com/inward/record.uri?eid=2-s2.0-77954373370&doi=10.1093%2fjjfinec%2fnbq016&partnerID=40&md5=bf587b6a887657961b6791f401a863d2https://scholars.lib.ntu.edu.tw/handle/123456789/427363Autoregressive conditional duration (ACD) models have been shown to be important for several applications in empirical finance. In this paper, we consider a set of generalized moment tests for the conditional mean specifications, the IIDness assumption of the error terms, and the distribution assumptions of the error terms in the context of ACD models. These generalized tests are also applicable to other multiplicative error models. We demonstrate that these tests are useful for unifying existing parametric tests, correcting the estimation effect ignored by some popular tests, and generating new tests for ACD models. Therefore, they can be applied to evaluate ACD models in a more complete way. This study also includes a Monte Carlo simulation and an empirical example to assess the performance of these tests.Autoregressive conditional duration model; Model evaluation; Moment test; Multiplicative error modeljournal article10.1093/jjfinec/nbq01636161777400