Page 147 - 35-2
P. 147

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




               forecast. Our analyses follow a difference-in-differences design. We include UK , POST ,
                                                                                             it
                                                                                      it
               and UK ×POST  as the test variables: UK  is an indicator variable that equals one if the
                                                     it
                      it
                             it
               company is a UK firm and equals zero if it is a US firm; POST  is an indicator variable that
                                                                     it
               equals one from 2005 if firm i has a December 31 fiscal year-end, equals one from 2006 if
               firm i has a non-December 31 fiscal year-end, and equals zero otherwise. The coefficient
               on UK  (β ) captures the effect of the partial fair value model (i.e., unrealized fair value
                        1
                     it
               gains/losses do not pass through net income) on the properties of analysts’ earnings
               forecasts relative to that of the historical cost model under US GAAP. The coefficient on
               POST  (β ) captures concurrent temporal effects on US firms that coincide with UK IFRS
                       2
                    it
               adoption. The coefficient on the UK ×POST  interaction term (β ) captures the incremental
                                                     it
                                              it
                                                                       3
               effect of the full fair value model (i.e., unrealized fair value gains/losses pass through net
               income) under IFRS relative to that of the historical cost model under US GAAP.
                   Equation (1) includes a series of firm-level control variables that are likely to affect
               the properties of analyst forecasts. ROA  is profitability, measured as net income scaled
                                                   it
               by total assets at the end of year t-1. SIZE , referring to firm size, is the logarithm of
                                                      it
               the market value of equity at the end of year t-1. LEV  is the leverage ratio, measured
                                                                 it
               as total long-term debt divided by total assets at the end of year t-1. BM  is the book-to-
                                                                                it
               market ratio at the end of year t-1. OTHER_A  and OTHER_L  are the percentage of non-
                                                       it
                                                                      it
               investment property assets and non-debt liabilities in total assets, respectively, measured at
               the end of year t-1. EPS_CHANGE  is the change in EPS, measured as the absolute value
                                              it
               of the difference in EPS between years t-1 and t divided by share price at the end of year t-1.
               GROWTH  is the sales growth rate between years t-2 and t-1. REIT  is an indicator variable
                        it
                                                                         it
               that equals one if firm i is classified as a real estate investment trust and zero otherwise.
               DIVERSIFIED  is an indicator variable that equals one if firm i has a diversified portfolio
                            it
               of investment properties and zero otherwise. STD_RET  is the standard deviation of daily
                                                                it
               stock returns for firm i in year t-1. PINT  is the percentage of institutional ownership in
                                                   it
               year t-1. FOLLOW , referring to analyst coverage, is the number of analysts issuing an
                                it
               EPS forecast for firm i in year t. HORIZON  is the log of 1 plus the average length of time
                                                     it
               between the forecasting date and the earnings announcement date for firm i in year t.
                   Next, we examine whether the effect of the full fair value model relative to the
               partial fair value model and the historical cost model varies over time. To capture time-

               varying effects, we replace POST in Equation (1) with a vector of time-period indicators:


                                                     139
   142   143   144   145   146   147   148   149   150   151   152