数据与实证研究.ppt

单击此处编辑母版标题样式,单击此处编辑母版文本样式,第二级,第三级,第四级,第五级,*,单击此处编辑母版标题样式,*,单击此处编辑母版文本样式,第二级,第三级,第四级,第五级,数 据 与 实 证 研 究,高,宁,博 士,国,泰安信息技,术,有限公司常,务,副,总裁,西安交通大,学,教授,香港浸,会,大,学,商,学,院,Honorary Associate,香港浸,会,大,学,公司管制,与,金融政策研究中心,Research Fellow,什么是实证研究？,以事实、实际情况和收集到的数据为对象，通过分析、计算、实验、研究，解释和预测会计金融实务，回答,“,实际是什么,”,的问题,。,实证研究要求客观、准确、理性的描述现实,实证研究以解释现实为目的，认为存在就是事实,实证研究采用客观中立的立场,目前，在国际上，实证研究方法广泛的应用在经济、金融、会计等社会学科的研究中,实 证 研 究 的,发 展 与 趋 势,-,实证经济学,1953,弗里德曼,实证经济学方法论,发,展,历,程,-,实证会计学,1968 Ball,，,R.J.,P.Brown,An Empirical Evaluation of Accounting,Income,NumbersJournal,of Accounting Research,1986 Watts,，,Zimmerman,实证会计理论,趋,势,由于金融市场每天都产生海量的数据，这些数据又是从真实的交易,过程中产生的,这一特性使实证研究成为现代金融研究的主流话语,”,Ross,20,世纪,80,年代,Accounting Review,上实证性研究的论文占半数以上，有的年份还高达,81,。现在实证研究已成为会计，金融研究的主流。,推动实证研究,发展的因素,（,William Beaver),推动实证研究发展的因素,（,William Beaver,),财务和经济学的发展,1,证券市场在经济中的地位,2,政府对证券市场的积极监管，不断推出新的课题,3,机构投资者占股权比重的增大,4,计算机技术和数据库的发展,5,6,学术刊物受重视程度的增强,实 证 论 文 篇 数,实 证 论 文 篇 数,类型,1994,1995,1996,1997,1998,1999,2000,2001,2002,2003,2004,2005,2006,实证研 究论文,136,161,205,325,381,547,641,830,1153,1751,2482,3566,5043,经济类实证论文,82,72,115,168,188,236,276,335,432,697,1026,1399,1979,实 证 的 要 素,实 证 的 要 素,数据,：反映客观状况的数字材料。,模型,：刻画客观现象的数学形式。,假设,：对所研究问题的结果或状态的,一种预期。,检验,：利用数据使用统计学知识对假设的统计显著性作出判断。,推理,：基于知识和经验对假设检验结果进行推理。,结论,：利用假设检验的结果，通过合情的逻辑推理得出的结论，观点。,实证研究方法步骤,确立研究课题,实 证 研 究 方 法 步 骤,寻找相关理论,提出命题假设,设计研究方案,搜集事实数据,分析数据检验命题,得出研究结论,金 融 实 证 研 究,的 主 要 领 域,包括现代投资组合理论、资本资产定价理论、套利定价模型、期权定价模型、有效边界、资本市场线、证券市场线等。,包括资金成本传统理论、净利理论和营业净利理论、权衡理论和融资偏好次序等。,投资组合选择和资产定价,资金成本和资本结构理论,市场微观结构,研究交易价格发现过程与交易运作机制，包括价格发现的模型和市场结构与设计。,行为金融学,研究投资者的心理、个人特征等因素与其交易行为之间的关系，包括 个人信仰,(,过度自信、乐观主义、代表性、保守主义、确认偏误、定位、记忆偏误,),个人偏好,(,展望理论、模糊规避,),会 计 实 证 研 究,的 主 要 领 域,会计制度的选择,研究企业会计制度的选择与企业营运绩效之间的关系,盈余管理,研究企业管理当局借助会计政策的选择和会计估计的变更，寻求对自己有利结 果的行为及其影响,会计舞弊,研究公司采取伪造、掩饰的手法编造假账损害股东权益、影响投资者做出正确 投资决策的行为,财务预测,会计信息披露效应,财务困境,会计信息的价值相关性,研究如何根据财务活动的历史资料和现实情况对企业未来财务活动进行科学的预计和测算,研究上市公司会计信息披露与公司股票价格之间的关系,研究企业陷于财务困境的特征及影响因素，主要包括财务困境企业与非财务困境企业之间财务项目的分析,研究会计信息价值相关性对于会计准则制定、证券市场监管和投资者进行决策的作用,CSMAR,实 证,论 文 举 例,文章研究了中国上市公司盈余公告时间选择对股票交易量和未预期收益的影响。研究发现，与较晚月份公告盈余的公司相比，较早月份进行年度盈余公告的公司具有较强的股票交易量反应。文章认为愿意早些公告盈余的公司往往拥有利好的信息，并且这些较早的盈余公告含有更大的信息量,带来较大的交易量增幅和未预期收益；较晚公告盈余的公司则往往拥有利差的信息，而且更容易被市场预期,因而带来的交易量增幅和未预期收益也较小。,题 目,作 者,发表刊物,摘 要,Information Content and Timing of Earnings Announcements,陈工孟 高 宁 郑子云（香港理工大学）,Journal of,Business Finance and Accounting,January 2005,Vol,32,Iss,.1-2,Pg.65,95,数 据 样 本,以,1995,年至,2002,年间发行,A,股或同时发行,A,，,B,股，在时间区间内发表年度盈余公告的上市公司为研究样本。样本容量为,3802,。,年份,样本数,月（）,月（）,月（）,4,月（）,1995,265,6(2.26),9(3.40),81(30.57),169(63.77),1996,294,1(0.34),6(2.04),33(11.22),254(86.40),1997,350,4(1.14),10(2.86),52(14.86),284(81.14),1998,590,4(0.68),45(7.63),269(45.59),272(46.10),1999,350,8(2.28),9(2.57),87(24.86),246(70.29),2000,531,45(8.47),50(9.42),188(35.41),248(46.70),2001,663,13(1.96),108(16.29),299(45.10),243(36.65),2002,759,15(1.98),84(11.07),277(36.50),383(50.45),Total,3802,96(2.52),321(8.45),1286(33.82),2099(55.21),CSMAR,总 体,样 本 容 量,CSMAR,总 体 样 本 容 量,文 献 回 顾,和 假 设,为什么选交易量而不是价格,Bamber,Barron and,Stober,(1997)suggest that trading volume is related,to the magnitude of the disagreement among investors about a firms,earnings.,Kim and,Verrecchia,(,1991a)argue that price changes reflect the average,change in the aggregate markets average beliefs,while trading volume,is,the sum of all individual investors trades,which also depends on the,prevailing information asymmetry level before disclosure.They suggest,that although all investors have equal access to public pre-disclosure,information,they acquire private pre-disclosure information with,different degrees of precision.,为什么选交易量而不是价格,Atiase,and,Bamber,(1994)andKross,et al.(1994)suggest that trading volume is an,increasing function of the degree of divergent pre-disclosure expectations,。,Bamber,and,Cheon,(1995)argue that the reason for different reactions is that price,reactions reflect the average belief revision,while trading volume arises when,individual investors make differential belief revisions.,更 进 一 步 的 分 析,Kim and,Verrecchia,(1994)suggest that there may be more information asymmetry,at the time of an announcement than in a non-announcement period.This is because,earnings announcements provide information that allows certain traders to makejudgements,about a firms performance that are superior to the,judgements,of other,traders.,Lobo and,Tung,(1997)find that the trading volume around quarterly earnings,announcements is related to the level of pre-disclosure information asymmetry.For,firms with a high level of pre-disclosure information asymmetry,the trading volume,is low prior to and after the announcement,but high during the announcement.,更 进 一 步 的 分 析,Bamber(1986)employs the divergence of earnings forecasts from analysts forecasts as a proxy for information asymmetry.She finds that the higher the information asymmetry,the greater the abnormal volume reaction.,In this study,we first use unexpected earnings as a control variable for information asymmetry.,Earlier announcements should generate a greater surprise in the market because it is more difficult to predict,earlier announcements than later announcements.Chambers and Penman(1984)argue that longer reporting,lags provide the opportunity for more of the reports information to be supplied by other sources,either,through search activity by investors,through other voluntary disclosures by firms,or through predictions that,are supplied in the earnings releases of earlier reporting firms.,Haw et al.,(1999)study the Chinese stock market and find that firms with good news publicize their annual reports earlier than those with bad news,and loss-making firms are the last to release their annual reports.They define the reporting lag as the number of days from the fiscal year-end to the report announcement date.,Earlier announcements should generate a greater surprise in the market because it is,more difficult to predict earlier announcements than later announcements.Chambers and Penman(1984)argue that longer reporting lags provide the opportunity for more of the reports information to be supplied by other sources,either through search activity by,investors,through other voluntary disclosures by firms,or through predictions that are,supplied in the earnings releases of earlier reporting firms.,Haw et al.,(1999)study the Chinese stock market and find that firms with good news,publicize their annual reports earlier than those with bad news,and loss-making firms,are the last to release their annual reports.They define the reporting lag as the number of days from the fiscal year-end to the report announcement date.,更 进 一 步 的 分 析,1.First,normally due to potential insider,trading and information leakage,it is,possible that the market reaction starts,long before the actual announcements.,Consequently,we employ-20,2,and-20,-3 to capture the possible pre-,event reaction.,2.Second,in the relatively efficient market,announcement,effects should not exist in the,long event w,i,ndow.Therefore,we use four,short symmetrical event windows to capture,announcement effects.They are-1,+1,-2,+2,-5,+5,and-7,+7.,时 间 窗 口 的 确 定,-20,2,-20,-3,-1,+1,-2,+2,-5,+5,-7,+7,共,6,个,250 trading days from day 280 to day 31.,A time gap between the end of the estimation window and the beginning of the event window(i.e.from day 30 to day 21)is employed to avoid using unusual price or volume data(due to information,leaka-ge,)for model estimation.,比 较 期 间（,beta,期 间）,To focus our analysis on the number of tradable days,we define the reporting lag as,the number of working days from the fiscal year-end to the annual release date.,1.a continuous variable,Announcement Timing Index(ATI),to proxy the reporting lag,which is,defined as ATI=n/N,where n is the nth working day from January 1 on which,the earnings announcement is made.N is the total number of working days in the period,from January 1 to April 30 in,the event year.,三个不同的时间变量,(TEA),定义,三个不同的时间变量,(TEA),定义,2.the unexpected ATI(UATI),a proxy for the unexpected reporting lag,is defined as the difference between the actual and expected ATI(the expected ATI of the current year should be the same as the ATI of the previous year),UATI=,ATIt,ATIt-1.,3.The final TEA is a dummy variable,called MAD,with a value of 1 for March and April announcements and 0 otherwise.,Null Hypothesis:Firms with earlier and later earnings announcements should receive similar abnormal market reaction.,简 单 的 假 设,Alternative Hypothesis:Firms with earlier earnings announcements should receive a higher abnormal,maket,reaction.,主 要 模 型,主 要 模 型,异常交易量的决定因素多变量回归模型,CATV(CAR)=,0,+,1,UEA(UERW,UEGM)+,2,SIZE+,3,POWN+,4,TEA,(UATI,ATI,MAD)+,5,EXCH+,i,YEARi-5+,j,INDj-12+,18,FOR,CATV,POWN,UEA,EXCH,IND,SIZE,TEA,YEAR,FOR,CAR,累积异常交易量,累积异常收益率,未预期盈余的绝对值,人民币计价的总资产的自然,流通股所占百分比,盈余公告时间,交易所哑变量,公告年的哑元变量,行业哑变量,外资股的哑变量,Abnormal Trading Volume around Earnings Announcement by bi-monthly sample,January and February(#,Obs,=417),March and April(#,Obs,=3385),Day,ATV,t-value,ATV,t-value,-7,0.0015,1.64,0.0007,1.63,-6,0.0024,2.39*,0.0010,2.12*,-5,0.0024,2.25*,0.0009,1.86,-4,0.0044,3.40*,0.0007,1.61,-3,0.0045,3.59*,0.0011,2.38*,-2,0.0055,4.41*,0.0010,2.05*,-1,0.0092,6.25*,0.0019,3.78*,0,0.0134,7.87*,0.0071,12.02*,+1,0.0129,7.63*,0.0071,12.24*,+2,0.0091,5.62*,0.0036,6.95*,+3,0.0055,4.05*,0.0018,3.61*,+4,0.0032,2.64*,0.0008,1.67,+5,0.0030,1.86,0.0006,1.31,+6,0.0018,1.44,0.0009,1.89,+7,0.0020,1.58,0.0010,2.02*,Interval,CATV,z-value,CATV,z-value,-20,2,0.0841a,13.32*,0.0380a,15.67*,-20,-3,0.0340b,7.57*,0.0173b,8.98*,-7,7,0.0808c,14.62*,0.0302c,14.76*,-5,5,0.0731d,14.94*,0.0266d,14.92*,-2,2,0.0501e,14.21*,0.0207e,16.57*,-1,1,0.0355f,12.56*,0.0161f,16.19*,Abnormal Trading Volume around Earnings Announcement by bi-monthly sample,Most of the ATVs for all monthly samples are significant,which indicates that the announcements do provide information to the market.,The magnitudes of the ATVs and,CATVs,for the January and February sample are much greater than those for the March and April sample.,CATV3,CATV5,CATV11,CATV15,CATV18,CATV23,Lowest 40%,of,ATI,Sample,0.0253,0.0337,0.0478,0.0545,0.0265,0.0602,Highest 40%,of ATI,Sample,0.0141,0.0182,0.0229,0.0258,0.0479,0.0298,Difference in,Mean CATV,0.0112cd,0.0155cd,0.0249cd,0.0287ab,-0.0033,0.0123,Panel,A,:Between the Lowest 40%of the ATI Sample,and Highest,40%of the ATI Sample,CATV3,CATV5,CATV11,CATV15,CATV18,CATV23,Positive UATI,Sample,0.0106,0.0132,0.0160,0.0166,0.0110,0.0242,Negative UATI,Sample,0.0290,0.0413,0.0631,0.0755,0.0820,0.0407,Difference in,Mean CATV,-0.0184cd,-0.0281cd,-0.0471cd,-0.0589cd,-0.0297a,-0.0578cd,Panel,B:Between the Positive UATI Sample,and Negative UATI Sample,T,he lowest 40%,of ATI samples demonstrates a significantly greater volume reaction than those of the highest 40%of ATI samples.,The negative UATI samples demonstrate a significantly greater volume reaction than those of the positive UATI samples.,earlier announcements provide more information content to the market than later announcements do.,CATV3,CATV5,CATV11,CATV15,Intercept,0.1590,0.2480,0.4760,0.7070,(4.12)*,(4.28)*,(4.40)*,(5.16)*,UERW,0.0005,0.0010,0.0018,0.0023,(2.42)*,(3.19)*,(3.25)*,(3.21)*,SIZE,-0.00680.0110,-0.0228,-0.0341,(-3.31)*,(-3.57)*,(-3.96)*,(-4.67)*,POWN,-0.0052,-0.0105,-0.0085,-0.0362,(-0.43),(-0.57),(-0.25),(-0.83),UATI,-0.0282,-0.0384,-0.0568,-0.0596,(-3.47)*,(-3.14)*,(-2.48)*,(-2.06)*,EXCH,0.0082,0.0159,0.0336,0.0392,(2.37)*,(3.05)*,(3.45)*,(3.19)*,YEAR2,-0.0410,-0.0623,-0.1090,-01420,(-5.69)*,(-5.74)*,(-5.36)*,(-5.53)*,YEAR,3,0.0117,0.0107,0.0078,0.0057,(1.72),(1.04),(0.41),(0.24),YEAR,4,-0.0596,-0.0941,-0.1800,-0.2580,(-5.17)*,(-5.42)*,(-5.55)*,(-6.28)*,Results of Regression Model for CATV,YEAR5,-0.0397,-0.0604,-0.1050,-0.1550,(-3.52)*,(-3.56)*,(-3.31)*,(-3.88)*,YEAR6,-0.0599,-0.0893,-0.1560,-0.2180,(-5.59)*,(-5.54)*,(-5.16)*,(-5.71)*,YEAR7,-0.0592,-0.0877,-0.1590,-0.2230,(-5.73)*,(-5.63)*,(-5.46)*,(-6.07)*,IND,1,0.0549,0.0753,0.1540,0.2190,(2.84)*,(2.59)*,(2.83)*,(3.19)*,IND2,0.0000,-0.0019,0.0010,-0.0011,(-0.01),(-0.20),(0.06),(-0.04),IND3,0.0010,0.0023,-0.0027,-00008,(0.14),(0.20),(-0.12),(-0.03),IND4,0.0039,0.0046,0.0034,-0.0002,(0.79),(0.61),(0.24),(-0.01),IND5,0.0060,0.0100,0.0175,0.0203,(1.13),(1.25),(1.18),(1.08),FOR,-0.0048,-0.0048,-0.0022,-0.0028,0.0550,0.0510,0.0430,0.0440,R2adj,(-0.88),(-0.58),(-0.14),(-0.14),F,9.8060,9.1480,7.8220,8.0040,p-value,0.0000,0.0000,0.0000,0.0000,Results of Regression Model for CATV,The earlier the announcement by one company relative to other companies,and the earlier the announcement of the company relative to its time of,disclosure of the previous year,the greater the abnormal trading volume.,Greater unexpected earnings(UERW),smaller firm size(SIZE),and,Shanghai stocks(EXCH)also lead to greater volume reactions.,模型中各变量所用数据在,CSMAR,中的位置,变量名,变量含义,所用数据项目,所用数据在,CSMAR,中的位置,CATV,累积异常交易量,日个股交易量,日市场交易总股数,交易数据库日数据库,日个股回报率文件,CAR,累积异常收益率,日个股回报率,日市场回报率,同上,UEA,未预期盈余的绝对值,每股盈余,财务指标库股东,获利能力文件,SIZE,人民币计价的总 资产的自然对数,总资产,财务库资产负债表文件,POWN,流通股所占百分比,流通股数，总股数,交易数据库股本变动文件,TEA,盈余公告时间,公告时间,年中报公布日期 数据库日历文件,IND,行业哑变量,行业代码,交易数据库公司文件,/,一级市场研究库,FOR,外资股的哑变量,外资股代码,一级市场研究库 公司基本情况表,模型中各变量所用数据在,CSMAR,中的位置,ChinaSSRN,中文社,会科学研究网,(,www.ChinaSSRN.com,),ChinaSSRN,中文社会科学研究网,(,www.ChinaSSRN.com,),由国泰安公司与北京大学国际会计与财务研究中心联合创建。,是高等院校等教育机构的研究者（包括老师与学生,),，社会研究人员，政府机,构，学术组织等进行论文投稿、学术交流与资料搜索的专业平台。,作为一个资源发布与共享平台，,包括电子论文库、论文投稿、期,刊专区、会议专区、咨信推荐、,信息公告、文章讨论区、个人公,文包和用户服务区等九个版块。,ChinaSSRN,学术委员会义务联名向知名学术杂志推荐优秀投稿论文,定期举办年度优秀论文评选，收录成册。,受邀参加,ChinaSSRN,年度学术盛会。,选材广泛，内容丰富,专业组织、严格评选,高端路线、注重品牌,简洁页面、强大搜索引擎,便利、个性化服务,会员权益,特 色,THANK YOU!,