Variables used in the regression model Timeliness is tested with the Ball and Brown (1968) model:
NIit ¼ 0 þ 1Rit þ "
and conservatism with the Basu (1997) model:
N I i t ¼ 0 þ 1 R D i t þ 2 R i t þ 3 R i t R D i t þ " i t
Accounting income (NIt) is the net income before extraordinary items/preferred dividends. It represents income before extraordinary items and preferred and common dividends, but after operating and non-operating income and expense, reserves, income taxes, minority interests and equity in earnings. In our study, NIt is scaled by the year- end market capitalization of the firm for the period t 1 to become the accounting earnings yield.
Return (Rt) is represented by the total investment return, measured as:
MPt þ Dt þ MPt1 P
where Rt is the total investment return for year t, MPt the market price per share at end of year t, Dt dividends per share for year t and SDQ the special quarterly dividend per share.
Like Basu (1997), we use negative and positive annual investment return to proxy for bad news and good news. Here, RDt is a dummy variable equal to 1 if Rt < 0, 0 otherwise.
Recognizing that France has lower public disclosure standards, and that its capital markets are less mature and have lower liquidity, the validity of annual investment return as a proxy for economic income could be questioned. In defense of the proxy, we argue that even in countries alleged to have unusually large proportions of uninformed investors, information influences stock prices nevertheless through the trading behavior of informed investors. Also, we use annual stock returns, which are less sensitive than short-interval returns to issues of liquidity, transaction costs, and the precise timing for market incorporation of information (Ball et al., 2003).
Table II shows the descriptive statistics for the sample.
Variables relating to corporate characteristics Size. We use total market capitalization to measure the size of firms. Year by year, we classified our sample in order of market capitalization, then divided it into two sub- samples: large firms and small firms. We use a dummy variable where size equals 1 if large, 0 otherwise.
International financing. The existing literature quite often uses cross-border listing to measure the degree of internationalization in a firm’s financing structure. In our opinion, this proxy has lost relevance as the French market has progressively become more open to foreign investors and institutional investors have become ever more internationally mobile (both of these developments being facilitated by technological advances). Arguably, nowadays, a company that is listed in its own internationally important domestic market and also included in an internationally recognized stock
Timeliness and conservatism