Page 155 - 35-2

NTU Management Review Vol. 35 No. 2 Oct. 2025

We again conduct the analyses to examine whether the insignificant coefficient on
UK×POST is associated with the time-varying effect of the 2005 change in standards.
Our results reveal that the change’s effect in the UK is not constant throughout the post-
IFRS period. We do not identify a significant difference in forecast response time in the

early period following IFRS adoption (UK×YEAR_0506 = -2.078, t-statistic = -0.38;
UK×YEAR_0708 = 0.545, t-statistic = 0.13; UK×YEAR_0910 = -3.187, t-statistic = -0.80).
However, starting in 2011, we find that it takes less time for UK analysts to issue their

first forecast revisions (UK×YEAR_1112 = -7.923, t-statistic = -2.17; UK×YEAR_1314 =
-9.988, t-statistic = -2.22), and this reduction is economically meaningful. In 2011-2012,
both UK and US firms exhibit, on average, an increase in response time (YEAR_1112 =
9.343). However, this increase is 85% (-7.923/9.343) lower for UK analysts. Still, we note
that the magnitude of the coefficients and the significance level have increased in the last

four years of our sample period, suggesting that analysts spend less time issuing forecasts.
The results are similar to those in Table 3, which documents a time-varying effect of the
change in accounting standards on analyst forecast dispersion.

Finally, we examine the differential effect of the fair value model and the historical
cost model on forecast revision response time by checking the significance of the sum
of UK and the interaction term between UK and the vector of time-period indicators.
Panel B reveals that, aside from UK + UK×YEAR_1314, all sums are significant and
positive, providing additional evidence of analyst response time being longer among UK

firms reporting under the full fair value model than among US firms reporting under the
historical cost model.
Overall, we find that analysts spend more time issuing the first forecast revision

for UK firms reporting fair values, on average. However, the analysts’ forecast revision
response times are reduced when income statements and balance sheets are reported
under the same reporting model in the post-IFRS period (i.e., the full fair value model),
compared to the period in which balance sheets follow fair value accounting while income
statements follow the historical cost model (i.e., the partial fair value model). We note that

the reduction in forecast revision response time due to IFRS adoption is not immediate.
The delayed effect on forecast revision response time is consistent with the previously
presented findings on forecast dispersion (Section 5.1) as well as prior research indicating

that analysts improve over time, steadily learning how to use resources more effectively

147