Lec 2 - Share Price Anticipation of Earnings Flashcards
What did Lev find?
- Was reviewing empirical return-earnings in 68-69
- Found regressions of returns on earnings is usually low
- Put this down to deficiencies in the accoujnitng measurement system (usefulness of earnings)
What did Kothari find?
- Believed Lev’s found a premature conclusion because it failed to recognise the correlation between stock returns and earnings changes
- and how this related to how well an earnings change proxies for unanticipated component of the period’s earnings change
- and serves as a proxy for the revision in the market’s expectations of future periods’ earnings change
What happens if investors had more information than just historical earnings?
Investors’ expectations of next period’s earnings might differ from current earnings
- in this case - earnings change variable measures earnings surprise with error
What does the earnings surprise with error lead to?
An underestimation of the usefulness of earnings as measured in terms of ERC or R^2
Reported earnings is slow in reflecting value-relevant events, why?
Accounting conventions that are invovled with current historical cost regime:
- objectivity
- verifiability
- conservatism
Therefore it isn’t a good indication of future earnings
What is an Efficient Capital Market?
A market where we expect prices to respond as soon as the information reaches the market
What is the implication of reported earnings and the efficient capital markets?
Information affects reported earnings and share price in different periods - this is “Prices Leading Earnings” or “Share Price Anticipation of Earnings”
Research in market based accounting is based off this model, what do the componenets stand for:
Rt = return in period t
UXt = unexpected earnings in period t
Pt-1 = share price at end of period t-1
b = ERC (measures extent of a security’s abnormal return in response to unexpected component of reported earnings)
What are unexpected earnings?
- Difference between realised earnings and expected earnings
(Expectations can’t be observed by researcher, so expected earnings needs to be expressed in terms of observable variables)
In the case of value-relevant non-earnings information (when prices lead earnings), what is X(t-1)?
A noisy proxy for period t’s expected earnings
- X(t) - X(t-1) measures unexpected earnings for period t (with measurment error)
What does measurement error in dependent variables result in?
Lower explanatory power
What does measurement error in the independent variable result in?
Reduced explanatory power and leads to downward biased slope coefficients
What do Kothari and Sloan and US and Donnelly point out?
A way to improve statistical properties of return-earnings association is to increase the return window - resulting in higher ERC
Replacing the deflated earnings change variable with a deflated earnings variable results in higher ERC and R^2
How does inappropriateness in measuring earnings surprise when prices lead earnings affect statistical performance of the simultaneous return-earnings association?
- Reduced explanatory power of earnings for returns (lower R^2)
- Downward biased slope coefficient on the earnings variable (downward biased ERC)
What does the magnitude of ERC and R^2 in return earnings regresssions used for?
- Statements on usefulness of reported earnings in determining stock price revisions
- Lev argues that usefulness of reported earnings in explaining stock price revisions is rather limited
