22.12.2012 Views

Understanding earnings quality - MIT Sloan School of Management

Understanding earnings quality - MIT Sloan School of Management

Understanding earnings quality - MIT Sloan School of Management

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Exhibit 1: Summary <strong>of</strong> <strong>earnings</strong> <strong>quality</strong> proxies<br />

This exhibit lists the commonly used proxies for <strong>earnings</strong> <strong>quality</strong> and the most common specification(s) <strong>of</strong> the variable. The exact specification <strong>of</strong> the proxies<br />

can vary by study. For each, we summarize the theory for its use as a measure <strong>of</strong> <strong>quality</strong> and an abbreviated summary <strong>of</strong> the strengths and weaknesses <strong>of</strong> the<br />

proxy. Details are in the review in Section 3.<br />

Empirical proxy Theory Strengths and weakness<br />

Persistence<br />

Earningst+1=α+ βEarningst + εt<br />

β measures persistence.<br />

Smoothness<br />

σ(Earnings)/ σ(Cash flows)<br />

A lower ratio indicates more<br />

smoothing <strong>of</strong> the <strong>earnings</strong> stream<br />

relative to cash flows.<br />

Timely loss recognition (TLR)<br />

Earningst+1=α0+α1Dt+β0Rett<br />

+β1DtRett +εt<br />

where D = 1 if Ret < 0. Higher β1 is<br />

greater TLR.<br />

Benchmarks<br />

∗ Kinks in <strong>earnings</strong> distribution<br />

∗ Changes in <strong>earnings</strong> distribution<br />

∗ Kinks in forecast error distribution<br />

∗ String <strong>of</strong> positive <strong>earnings</strong> increases<br />

Firms with more persistent <strong>earnings</strong><br />

generate more accurate DCF-based<br />

equity valuations.<br />

Managers opportunistically smooth<br />

<strong>earnings</strong>. Therefore, greater<br />

smoothness is artificial relative to<br />

the fundamental process; or,<br />

smoothness reduces noisy variation<br />

in cash flows as a measure <strong>of</strong> the<br />

process.<br />

There is a demand for TLR to<br />

combat management’s natural<br />

optimism. TLR represents high<br />

<strong>quality</strong> <strong>earnings</strong>.<br />

Unusual clustering in <strong>earnings</strong><br />

distributions indicates <strong>earnings</strong><br />

management around targets.<br />

Observations at or slightly above<br />

targets have low <strong>quality</strong> <strong>earnings</strong>.<br />

172<br />

Pros: Fits well with a Graham and Dodd view <strong>of</strong> <strong>earnings</strong> as a summary<br />

metric <strong>of</strong> expected cash flows useful for equity valuation.<br />

Cons: Difficult to control for persistence <strong>of</strong> the fundamental <strong>earnings</strong> process,<br />

but this persistence is likely to be a large contributor to persistence <strong>of</strong> reported<br />

<strong>earnings</strong>. Thus, it is difficult to make statements about the effect <strong>of</strong><br />

measurement on persistence. Greater persistence may be achieved via<br />

opportunistic <strong>earnings</strong> management.<br />

Pros: In cross-country tests, measures <strong>of</strong> artificial smoothness appear to<br />

capture meaningful variation in <strong>earnings</strong> management.<br />

Cons: It is difficult to disentangle smoothness <strong>of</strong> reported <strong>earnings</strong> that<br />

reflects smoothness <strong>of</strong> the fundamental earning process from artificial<br />

smoothing.<br />

Pros: Aimed at disentangling the measurement <strong>of</strong> the process from the<br />

process itself.<br />

Cons: The net effect <strong>of</strong> TLR on <strong>earnings</strong> <strong>quality</strong> is unknown because TLR<br />

results in lower persistence during bad news periods than during good news<br />

periods (Basu, 1997). Both persistence and TLR affect the decision<br />

usefulness <strong>of</strong> <strong>earnings</strong>. TLR is a returns-based metric, see comments on<br />

ERCs.<br />

Pros: The measure is easy to calculate, the concept is intuitively appealing,<br />

and survey evidence suggests <strong>earnings</strong> management around targets.<br />

Cons: In addition to statistical validity issues, evidence that kinks represent<br />

opportunistic <strong>earnings</strong> management is mixed, with credible alternative<br />

explanations including non-accounting issues (e.g., taxes). It is difficult to<br />

distinguish firms that are at kinks by chance versus those that have<br />

manipulated their way into the benchmark bins.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!