1、 1 外文翻译 原文 Relationships Between Stock Prices And Accounting Information Material Source: http:/ Author: Scott Pirie and Malcolm Smith The term market based accounting research has been used by several authors, including Lev and Ohlson (1982) and Walker (1997), to describe the body of literature tha
2、t deals with relationships between market prices and accounting information.Most of the early empirical studies in this field focus on earnings and are usually concerned with the response coefficient that relates earnings to returns rather than the explanatory power of accounting information for sha
3、res prices or value. However, in a review of earnings research, Lev (1989) notes that earnings are generally found to have very low explanatory power, and he suggests that this casts doubt on the practical value of reported earnings. While there is evidence that earnings convey information that help
4、s form opinions about security prices, the earnings figures themselves have limited value because their relationship to prices is generally weak and unstable over time. Lev suggests this could be due to bias introduced by accounting rules, and he advocates a shift in research focus away from the inf
5、ormation content of accounting variables towards the rules that determine them.In another review of the relevant literature, Penman (1992) claims that much of the empirical work in market based accounting research has been misdirected in taking an informational perspective that assumes accounting va
6、riables only affect share prices if they provide new information. He advocates that there should be a return to fundamentals and a switch to a measurement perspective that views accounting numbers as useful determinants of asset values. Penman also maintains that traditional fundamental analysis and
7、 accounting practices have lacked the theoretical foundations required for rigorous economic analysis. However, he adds that the work of Ohlson has been a significant breakthrough in this respect.The question arises, however, of how well residual income models can explain variations in share prices.
8、 Several researchers have addressed this point, and some of the more significant studies are discussed in the following sections. The first group explores 2 the explanatory power of the models under different sets of accounting rules. The second focuses on identifying variables other than earnings a
9、nd book value that may form part of the other information in the theoretical framework. Finally, a number of studies are discussed that do not fall neatly into either of the previous two categories. With a view to examining the effects of international differences in accounting rules,Frankel and Lee
10、 (1998) explore relationships between share prices and accounting variables using data from 20 countries that include Australia, Japan, South Korea and Thailand. They use the general residual income model shown in equation together with reported book values and consensus earnings forecasts to calcul
11、ate an estimated value of the shares in individual companies. This value estimate is then included as an explanatory variable together with current book value and current earnings in a model that follows the version of the Ohlson model shown in equation. The explanatory power of the model is high, 8
12、8% for the US and 72% for the other countries combined.The study also finds that the value estimate based on consensus earnings forecasts is highly significant in all 20 countries, and that it consistently dominates current earnings and book value in explaining variations in share prices. Furthermor
13、e, the coefficients on the value estimates are relatively stable over time ranging from 0.67 in Norway to 2.56 in Italy suggesting some consistency in relationships between share prices and the value estimates under different accounting regimes. The authors use country-specific discount rates to cal
14、culate their value estimates, but conclude that most of the explanatory power of the estimate can be attributed to the use of consensus earnings forecasts rather than varying discount rates. They go on to suggest that the discounted residual income model could be an integral part of a broad solution
15、 to problems of international accounting diversity that goes beyond the harmonisation of standards. Their reasoning is that, in theory at least, value estimates based on this model do not depend on the particular accounting rules adopted in published financial statements.King and Langli (1998) exami
16、ne relationships between share prices and the two main accounting variables with data from Germany, Norway and the United Kingdom. They select these three countries because the accounting systems are considered quite different, particularly in their degree of conservatism and the extent to which the
17、 accounting rules deviate from the clean surplus relation. The authors estimate a model that links share prices to equity book value and current earnings in accordance with equation, and also estimate two restricted models that express share 3 prices as a function of either equity book value or earn
18、ings alone. They find that book value and earnings are both significantly related to share prices in all three countries, and that the two variables combined have explanatory power of about 70% in the United Kingdom, 60% in Norway and 40% in Germany. The authors conclude that these findings are cons
19、istent with the differences in the accounting systems of the three countries. They also find that the incremental and relative explanatory power of the two variables differ over time and between countries, with book value explaining more than earnings in Germany and Norway but less than earnings in
20、the United Kingdom. The authors note however that, in this case, the results are not consistent with differences in conservatism and violations of the clean surplus relation embedded in the accounting rules. Finally, results with an extended model that includes realised earnings for the following fo
21、ur years as proxies for expected earnings show that these additional variables explain little of the variation in market prices not already explained by current book value and earnings.In another study of international accounting differences, Graham and King (2000) examine relationships between shar
22、e prices and accounting variables in Indonesia, South Korea, Malaysia, the Philippines, Taiwan and Thailand. Their regression model relates share prices to current book value and current residual income in accordance with equation (10) and they find that the coefficients on both these variables are
23、statistically significant in all six countries. They also find that the explanatory power of the model varies significantly between countries, ranging from a low of 24% in Taiwan to 55% in Thailand and 90% in the Philippines although, in this case, the sample size is relatively small. Their evidence
24、 supports the view that more strongly conservative accounting information is less value relevant. They also find that the incremental explanatory power of book value is higher than that of residual earnings in all six countries. Violations of the clean surplus relation such as revaluation of assets
25、and immediate write-offs of goodwill have the expected effect on the value relevance of book value but the effect on the value relevance of residual earnings is less clear. Ball, Kothari and Robin (2000) demonstrate the difficulties of classification in international comparisons, particularly for No
26、rwegian data, which subsequent authors have treated as misclassifications. In a study to explore the market valuation of research and development expenditure in the United Kingdom, Green, Stark and Thomas (1996) adopt a regression model based on the version of the Ohlson model shown in equation (10)
27、. However, they formulate it with the market-to-book premium (share price less 4 equity book value) as the dependent variable, and current residual income as the main explanatory variable. The authors then add current expenditure on research and development as a second explanatory variable that can
28、be considered as a part of other value relevant information variable in the theoretical model. They go on to include a number of control variables such as current advertising expenditure, market share, capital structure, and the variance of market returns as a measure of risk that can also be consid
29、ered as part of other value relevant information.The authors find that residual income has significant explanatory power for the marketto-book premium but the evidence on research and development expenditure is mixed.The other control variables have almost no effect on the explanatory power of the m
30、odel and have little effect on the inference tests.Rees (1997) adopts the version of the residual income model shown in equation (11) as his theoretical framework in a study that examines the impact of dividends, debt and investment expenditure on the market value of a large sample of industrial and
31、 commercial firms in the United Kingdom. He finds that the two summary accounting variables are highly significant in both the pooled and annual regressions, although the coefficients themselves vary over time. Rees then goes on to explore the effect of dividends, debt and capital investment. He doe
32、s this by decomposing earnings into dividends and retained earnings, by re-stating book value as total capital less total debt, and by including the annual expenditure on fixed assets as an additional explanatory variable. The results of the study show that earnings distributed as dividends have a l
33、arger impact on value than earnings retained within the firm and that, when dividends are included, the overall explanatory power of the model increases from 54% to 60%. There is also evidence that capital expenditure is positively related to market value, but there is inconsistent evidence on the v
34、alue relevance of debt. Many studies citing Ohlson (1995) as their theoretical framework include book value and earnings as explanatory variables in line with equation (11), but relatively few include dividends or the variable that represents other value relevant information. However, in a study wit
35、h United States data, Hand and Landsman (1998) include both dividends and net capital contributions in addition to the two main accounting variables and use this model to test different predictions that arise from two common assumptions about other value relevant information. Their model with indust
36、ry fixed effects explains about 80% of the variation in share prices. They also find that, if other value relevant information is set to zero as implied by empirical models that omit it, the sign of the coefficient on dividends is 5 reliably positive when the theory predicts it should be negative. A
37、lternatively, if realised earnings for the following year are included as a proxy for expected earnings on the assumption that the other information impacts future residual income through the linear information dynamics proposed by Ohlson, the sign of earnings coefficient is opposite to the theory.
38、Hand and Landsman conclude that neither of these assumptions cleanly fits the data and they suggest that an explanation for the positive relationship with dividends could be that, contrary to assumptions in the theoretical model, dividends provide information that is useful for predicting future res
39、idual income. Consistent with this explanation, they find that larger dividends are associated with higher future residual income, especially for firms that are currently making losses. 译文 股票价格之间的关系和会计信息 资料来源： http:/ 作者： 斯科特皮里和马尔科姆史密斯 基于市场需求的会计研究一直被许多作者所运用 ,包括 Lev， Ohlson (1982)和 Walker(1997)， 这些文献都
40、是描述会计信息与市场价格的关系。在这个领域中，大部分早期的实证研究主要集中于收入和响应系数所涉及的收益回报而不是股票价格对会计信息的解释力。然而，在对收入的研究中 Lev(1989)指出，收入通常被认为是很低的解释力。 他认为，这使人们对报告的收益的实用价值产生怀疑 。当有证据表明，收入传达的信息帮助对股票价格形成意见，但是因为这些数字价值有限，所以他们的关 系通常不稳定。 Lev 表明 这可能是由于对会计准则的偏见，因此他倡导其研究热点应从信息内容转变为变量规则来确定。在另一个回顾相关文献中， Penman(1992)声称大部分的会计实证研究是被信息化角度下，会计信息只影响股票价格所误导。他
41、主张应该从回归基本面的角度，认为会计数字是资产价值的决定因素。作者同样认为，传统的分析和会计实务是 理论基础 。不过，他补充说 , Ohlson 的工作已经在这方面取得重大突破。然而 , 对于如何利用剩余收益模型解释股价的变化 ，几位研究人员已经在着手解决这一点，有些比较重要的研究成果在接下来 的段落里进行了讨论。 第一组根据会计准则探讨不同设置模式的解释力。第二组根据收入与帐面价值以外的确定变量，可能构成在理论框架的其他部分信息 。 6 为了检验国际不同会计准则效果。 Frankel和 Lee(1998)运用来自 20 个国家，包括澳大利亚、日本、韩国、泰国的数据探讨股票价格和会计之间的关系
42、。他们用剩余收益模型和 帐面价值来计算 公司的股票价值 。然后，这个值估计为解释变量，包括同当前账面价值和当前收益。在模型的版本中运用了 Ohlson 模型方程。该模型的力量高，通过 88%美国和 72%的其他国家研究还发现， 在 价值估计盈 利预测基础上达成共识是非常重要的 。在 所有的 20 个国家中，它一直主宰目前 收益帐面价值解释股价的变化 。此外， 价值估计的系数都比较多 ，从挪威 0.67 到 2.56 意大利， 股票价格之间不同会计制度下价值估计的一致性时间 都保持稳定的关系 。作者用特定国家的贴现率计算价值估计，但是认为大部分估计的解释力可以归因于企业盈利一致预期而不是同享受折
43、扣优惠。他们继续建议 剩余收益折现模型可以解决一个广泛的国际会计多样性 ， 超越了统一标准。论据是，至少在理论上，价格估计在此模型上不依赖采用某种会计规则公布的财务报表。 King 和 Langli（ 1998） 探讨股票价格波动和会计数据之间两个主要变量的关系。他们数据来自德国，挪威和英国。选择这三个国家是由于会计系统被认为是完全不同的，特别是在保守主义和 在何种程度偏离了会计准则的净盈余 的 关联度 上 。 笔者估计 ， 股票价格相关链接的账面价值与当期收益按照方程估计为任何一方权益的账面价值或收益函数股价 。他们发现的账面价值和收入都显著相关，股票价格在这三个国家中，这两个变量的解释力结
44、合了约 70%在英国， 60%在挪威与 40%在德国。 由此 的结论是 ， 这些发现与这三个国家的会计制度的差异是一致的。 他们还发现，解释力和相对 增量随时间不同而不同，以及国家之间账面价值超过收益使得在德国和挪威少于收入在英国。然而，该研究报告中作者还指出，在这种情况下的 结果没有差异与保守的净盈余违反会计规则嵌入关系是一致的 。最后，延长模型 其中包括作为代理实现盈利 于 以下四年预期收益表明 ， 这些额外的变数解释的变异甚少 。市场价格没有解释当前的账面价值和收入显示国际会计的差异。 Graham 和 King(2000)探讨股票价格和会计变量在印尼，南非、韩国、马来西亚、菲律宾、台湾
45、和泰国的关系。他们用回归模型对当前股价与帐面价值剩余收入推算，发现系数对这两个变量在 统计上是显著相关的。模型的解释力对国与国之间是明显不同，从 24%台湾 55%泰国到 90%菲律宾。虽然在这种情况下，样本的规模是相对比较小。他们认为，更保守的会计信息与价值不大相关。还发现最小的国家 账面价值的增量解释能力比在所有六个国家的剩余收益高 。 净盈余关系的侵犯，如资产和商誉重估取舍上有价值相关性 ，但账面价值的影响与收入剩余的价值相关尚不清楚。 Ball，Kothari 和 Robin（ 2000） 表明 ，国际 分类比较困难，特别是挪威的数据， 这些数据随后的作家 是 作为错误分类处理 的。
46、一项 研究 ，探讨在英国绿色市场开发支 出的估值 （ 1996 年）采用回归模型7 Ohlson 模型版本 。然而，他们用制定的费用作为因变量，当前的剩余收益为主要表达变量。之后，研究人员加入当前研究与发展经费的投入作为第二表达变量，可以被视为一部分变量值相关的理论模型。他们继续包括一系列的控制变量如当前广告费、市场占有率、资本结构、市场回报的方差的风险作为一种方法，也可以考虑作为其他相关信息的研究发现 剩余收益有显着溢价解释的权力 ，但在研究和发展 开支 的证据 对 其他的控制变量几乎没有解释力。影响极小推理测试 Rees(1997)采用的版本中显示的剩余收益模型方程作为 其理论框架，研究股
47、息、债务和投资分红在对 英国的工业和商业公司的市场价值 的 影响 。他发现这两个会计变量的集中回归显现着，但是系数随着时间的流逝有所不同。 Rees然后远期探讨股息、债务和资本投资的作用。 通过分解 ， 重新陈述为总资产价值减去总债务包括固定资产作为额外的解释变数 对 每年的开支 ， 红利 并且 保留盈利 。研究结果表明 盈利作为股息分派 的 价值大于企业内部留存收益的影响 。而且当股息计算在内 ， 整体由 54 至 60 的模型解释力增加 。 也有证据表明 ,资本开支呈正相关，但关于市场价值与债务价值不一致相关。许多研究引用Ohlson(1995)作为他们的理论框架包括账面价值和收入解释变量
48、，不过很少会包括股息或变量代表其他值相关信息。然而， 对 美国的数 据 Hand 和 Landsman（ 1998） 研究 两个主要的会计变量包括红利和网络出资，使用这个模型来验证不同的预测。表明，两种常见相关信息 与产业固定效应模型解释了约 80 股票价格的变化 。他们也发现，如果其他值相关的信息的设置为所经验模型就忽略。标志的系数股利是可靠积极的。反过来说，如果意识到下一年的收入是代理预期的盈利，假设未来影响的其他信息 以 Ohlson 收益系数为标志 通过剩余收入 线性资讯动态 所 提 出 ， Hand 和 Landsman 认为 ，这两种假设没有干净地拟合数据，他们建议解释良好股息可能与假设条件提供信息有关，理论模型能有效预测将来的剩余收益。 与此相一致的解释 ， 他们发现 较大的红利与未来剩余收益较高 。特别是公司 ， 因为它们目前正在 遭受 损失。
