PURPOSE : To investigate the extent to which different prices within the bid-ask spread
are used for fair value measurements and evaluate the potential consequences thereof.
DESIGN/METHODOLOGY/APPROACH : The paper investigates different level one fair value
measurements of exchange-traded funds’ equity investments. Using descriptive
methods, it compares actual and stated fair value measurement policies. In addition,
comparative value-relevance of these measurements is investigated in regression
FINDINGS : Most fair value measurements are based on closing prices, but stated
accounting policies and actual measurements frequently differ. Results also show that
the bid-close spread of underlying investments is value-relevant in determining the bidclose
spreads of ETFs themselves.
RESEARCH LIMITATIONS/IMPLICATIONS : Findings are specific to unleveraged ETFs, the sample country and sample period used and only apply to investments in listed equities.
Conclusions from this study may assist in predicting market perceptions of the risk of
listed equity portfolios.
PRACTICAL IMPLICATIONS : This paper sheds light on the practical impact of the recent
change in fair value measurement guidance.
ORIGINALITY/VALUE : This study provides evidence on the size of the bid-ask spread of
actual investment portfolios and its potential impact. It shows that bid-close spreads of
underlying investments are used to price the bid-close spreads of ETFs themselves and
that stated and actual accounting policies often differ. Findings imply that standardsetters
might be influenced by actual accounting practices.