Page 182 - 35-1

The Impact of Economic Policy Uncertainty on Non-GAAP Earnings’ Quality

evidence also indicates that firms with more SPIs in their financial statements are more
inclined to voluntarily disclose the non-GAAP earnings figures, which are calculated by
excluding SPIs from GAAP earnings figures (Lougee and Marquardt, 2004).
In summary, we anticipate that EPU will incrementally promote voluntary disclosures

of non-GAAP earnings by firms with SPI recognition. Hence, our first hypothesis is as
follows:
H1: The positive relation between the likelihood of voluntarily disclosing non-GAAP

earnings and SPI recognition is incrementally stronger with increasing EPU.

Nevertheless, based on the inference of opportunistic motives, under the
circumstances of intense information asymmetry, managers may intentionally disclose
non-GAAP earnings information, misleading investors into overvaluing firms for their

private interests by exploiting information asymmetry (Doyle et al., 2003). Therefore,
we elucidate the rationale for the voluntary disclosure of non-GAAP earnings amidst
elevated EPU. Because non-GAAP earnings are calculated by excluding certain items

from GAAP earnings, we utilize the exclusion quality to represent the non-GAAP earnings
quality. We infer that if the motive is informative, the excluded items should be more
transitory or nonrecurring, indicating higher exclusion quality. Conversely, if the motive
is opportunistic, the excluded items should be more recurring, indicating lower exclusion
quality. Given that previous studies mostly support the informative motive (Nagar et al.,

2019; El Ghoul et al., 2021; Jiang et al., 2022), we expect a positive relationship between
exclusion quality and EPU for firms with voluntary non-GAAP disclosures. Consequently,
our second hypothesis is as follows:

H2: The relation between exclusion quality and EPU is positive for firms that voluntarily
disclose non-GAAP earnings.

3. Research Design

Our sample predominantly comprises US listed-firms’ data from 2003 to 2017 at
the firm-quarter level, as indicated by previous studies (e.g., Bentley, Christensen, Gee,
and Whipple, 2018). Non-GAAP EPS and the EPU index data are derived from datasets

constructed and extended by Bentley et al. (2018) and Baker et al. (2016), respectively.

174