Joint evaluation of the directional accuracy of corporate executives’ forecasts

Abstract

Although most research on directional analysis focused on either increase/decrease or acceleration/deceleration of variables of interest, we jointly evaluate the two-directional changes in the real GDP predicted by corporate executives. We applied an approach recently developed by Pesaran and Timmermann (2009) in addition to the tests extensively used in the literature. Besides determining that, the forecast is not useful in predicting an increase/decrease and acceleration/deceleration in real GDP, joint evaluation of the forecast illustrates the advantage of the new test. Further, we examine the two data vintages.

Keywords:

• directional analysis
• market-timing test
• forecast evaluation
• macroeconomic forecast

JEL Classification:

• C53
• E37
• E17

Notes

¹ An increase/decrease in inflation rate is equivalent to an acceleration/deceleration of the price level.

² Tsuchiya (2013a) is the only work that employs the new test.

³ Forecasts are available for all industries, including manufacturing and nonmanufacturing sectors. However, results of all industries are presented because all empirical results are equivalent.

⁴ See Ash et al. (1998) and Sinclair et al. (2010) for details. In particular, Sinclair et al. (2010) describe and remark on the tests statistics in the 4 × 4 case.

⁵ See Chou and Chu (2011) and Tsuchiya (2013b) for remarks.

⁶ For instance, see Croushore (2011) for a more detailed discussion.

⁷ In the 4 × 4 case for the FE test, hypergeometric distribution is used to directly calculate the probability of independence. Thus, only p-values are reported. Note that p-values in the test cannot be calculated because at least one column or row in the contingency tables has all zeros.