Ownership Structure and Accounting Method Choice: A Study of European Real Estate Companies

Abstract

Companies can under IAS 40 Investment Properties choose between the fair value and the cost models. The fair value model arguably results in more relevant information but is also more costly to use. Based on studies suggesting that financial reports are a more important medium for communication with investors if ownership is dispersed, we hypothesize that the use of the fair value model is positively associated with ownership dispersion. We study European Real Estate firms and find support for this prediction. We also find a positive association between trade of shares and ownership dispersion, supporting the view that financial statements are less important if ownership concentration is high. Finally, we examine whether the choice depends on the identity of large owners. Companies with a financial company as the largest owner are somewhat more likely to choose the fair value model. Overall, the results indicate that accounting rules facilitating optional accounting policies have benefits.

Keywords:

• accounting method choice
• IAS 40
• ownership structure
• investment property
• fair value

JEL classification:

• G32
• M41

Acknowledgement

Previous versions have benefited from comments made by Jörgen Hellström, Benita Gullqvist, AFAR workshop in Vaasa (Finland), the Fekis conference in Umeå (Sweden) and participants at the EAA conference held in Paris in May 2013.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 To mention a couple of examples, LIFO was abolished from IAS 2 Inventories in 2003 implying that companies after this only can choose between the FIFO and weighted average cost formula. And in 2009 the optional capitalization of borrowing costs relating to the acquisition or construction of a long-term asset in IAS 23 Borrowing Costs was abolished. Currently, companies have to capitalize qualifying borrowing costs.

2 We use the ado file for two-way clustering in Stata written by Guan and Petersen. The file is available at: http://www.kellogg.northwestern.edu/faculty/petersen/htm/papers/se/se_programming.htm (retrieved January 2015).

3 Recent research (e.g., Laeven & Levine, 2008) stresses the role of multiple large shareholders. Based on this literature we predict that we attempted to include an indicator variable taking the value one if there was a second large owner controlling 10% or more of the shares. However, we found no significant association between the existence of a second large owner and the choice between the cost and fair value models. These analyses were based on 254 observations for fair value model users and 43 observations for cost model users.

4 Corporate investors and governmental bodies are likely to have different incentives and we attempted to analyze corporate investors and governmental bodies separately in regression 2. Probably as a consequence of the small number of observations, the coefficients of corporate investors as well as governmental bodies were then insignificant. However, the coefficient estimates suggest that both categories are less likely to use the fair value model than financial owners but the coefficient of corporate investors was smaller indicating that in particular corporate investors contribute to the negative association reported in Table 3.

5 The ownership type variables are mostly insignificant when we include country controls in the regressions. The decision not to include country variables in the reported regressions is based on the assumption that ownership type is exogenous and that the inclusion of country indicator variables capture part of the ownership effect. However, we cannot conclusively rule out that country factors rather than ownership type drive the reported associations.

6 We replaced OWNERTYPE_other with two indicator variables for corporate investors and governmental bodies. The results show that especially companies owned by a corporate investors are less traded at the stock market.