Teresa Pereira Pinto
Assistant Lecturer, UTAD University, Portugal.
Carlos Machado-Santos
Associate Professor with Habilitation, UTAD University, Portugal.An Analysis of the Correlation between EVA® and
MVA®: The Case of a Portuguese Company Listed
on the Stock Market
Following the international trend, the management by value
practices start to proliferate in the Portuguese corporate context,
accentuating the Economic Value Added (EVA®) as the privileged
performance measurement in the control process of the taken strategic
options.
Being so, we start by reviewing the concept and the principles of a
management by value approach, counter measuring it to the traditional
corporate valuation models. We give special emphasis to EVA®,
referring a possible correlation between this measure and Market Value
Added (MVA®).
The main element of distinction between EVA® and traditional
valuation metrics points to the fact that EVA® incorporates not only
remunerated liabilities and financing costs of debt but all the invested
capital (also including market value of the firm's equity and the cost of
equity). Besides that, EVA® includes also a sort of adjustments that
minimize a set of distortions resulting from the accounting practice
adopted.
EVA® became a strongest indicator when considered together with
another one, also based on the value created by businesses: the MVA®.
These two indicators, in general, have a strong correlation, being MVA®
the present value of all the future expected EVA®. MVA® can also be
computed as the difference between the market value of the company
and the capital invested in it.
Using a case-study, we explore the use of EVA® in a corporate
group with marked presence in reference activity sectors, listed in
Lisbon’s Euronext. We examined, using regression models, the
incremental information of a set of performance measures from 2005 to
2009.
The empirical analysis allows us to measure the performance
associated with the creation of value for the capital holders.
Additionally, we analysed the MVA® performance and compared it to
the existing link between the latter and the former measures. We found
that the relation between EVA® e MVA® is, in this case, statistically
significant.