Analysts' Cash Flow Forecasts and Accrual Mispricing

Analysts’Cash Flow Forecasts and Accrual Mispricing* SURESH RADHAKRISHNAN,University of Texas at Dallas

SHU-LING WU,National Taiwan University

1.Introduction

In a seminal paper,Sloan(1996)?nds that high accrual?rms experience negative future abnormal returns and that low accrual?rms experience positive future abnormal returns: the accrual anomaly.He provides evidence consistent with the notion that investors overes-timate the persistence of the accrual component of earnings when forming earnings expecta-tions.Hirshleifer and Teoh(2006)analytically show that investors’limited attention to earnings components contributes to accrual mispricing,an explanation that relies on inves-tors’irrationality.If the accrual anomaly is related to investors’irrationality,then the pro-vision of information on future accruals should help investors assess the implications of accruals on future earnings more appropriately.When analysts issue cash?ow and earnings forecasts for a?rm,they indirectly provide accrual forecasts as well.This leads us to exam-ine whether the incidence of a cash?ow forecast is associated with less accrual anomaly.

In recent years,?nancial analysts have begun to provide cash?ow forecasts in addi-tion to earnings forecasts,price targets,and stock recommendations.Call,Chen,and Tong(2009)?nd that analysts’earnings forecasts accompanied by cash?ow forecasts are more accurate and re?ect a better understanding of the implications of current period accruals and cash?ows for future earnings than those issued without cash?ow forecasts. This?nding suggests that the indirect accrual forecasts provided by analysts are of good quality,and thus,investors can better assess the implications of current period accruals on future earnings.In addition,in Hirshleifer and Teoh’s(2006)framework,the availability of information on accrual forecasts is likely to direct investors’attention toward accruals. When both the current period accruals and predicted next period accruals are available, investors can better gauge the persistence of current period accruals for?rms whose ana-lysts provide both earnings and cash?ow forecasts(hereafter referred to as earnings and cash?ow forecast?rms)than for?rms whose analysts provide earnings forecasts alone (hereafter referred to as earnings forecast?rms).Accordingly,we hypothesize that earn-ings and cash?ow forecast?rms have less accrual anomaly than do earnings forecast ?rms.1

The sample of?rms in this study contains33,241?rm-year observations spanning?s-cal years from1993to2009.Of the33,241?rm-year observations,all of which have earn-ings forecasts,9,977?rm-years have cash?ow forecasts as well.Prior studies on accrual mispricing show that accrual mispricing is mainly driven by abnormal accruals(Xie2001; Cheng and Thomas2006).Accordingly,to test the hypothesis,the one-year-ahead, *Accepted by Je?rey Callen.We thank Je?ery Callen and two anonymous reviewers for insightful comments and suggestions.We also acknowledge comments from Indranil Bardhan,Surya Janakiraman,Stanimir Markov and the participants of seminars at the University of Texas at Dallas,National Chengchi Univer-sity,and National Taiwan University.

1.McInnis and Collins(2011)?nd that accrual quality is better for earnings and cash?ow forecast?rms than

for earnings forecast?rms and could be an additional/alternative reason for less accrual mispricing for earnings and cash?ow forecast?rms than from earnings forecast?rms.

Contemporary Accounting Research Vol.31No.4(Winter2014)pp.1191–1219?CAAA

doi:10.1111/1911-3846.12060

1192Contemporary Accounting Research

size-adjusted returns are regressed on the cash?ow forecasts dummy,the decile-based rank of abnormal accruals,and the interaction between the cash?ow forecast dummy and decile-based rank of abnormal accruals.In addition to risk factors such as book-to-mar-ket,earnings-to-price,and beta,following Mashruwala,Rajgopal,and Shevlin(2006),we control for arbitrage potential and transaction costs using return variance,stock price, trading volume,and size.2We control for other factors that have been shown to explain accrual mispricing;speci?cally,the quality of earnings forecasts(see Call et al.2009)and investment patterns(Wu,Zhang,and Zhang2010).Furthermore,DeFond and Hung (2003)show that earnings and cash?ow forecast?rms are signi?cantly di?erent from earnings forecast?rms across various accounting-related?rm characteristics.As such,we use the Inverse Mills Ratio obtained from DeFond and Hung’s(2003)determinants model to deal with the potential self-selection issue.

We consider a cross-sectional design,where all?rms with analysts’earnings forecasts are included in the sample.In the cross-sectional design,the earnings forecast?rms provide the benchmark.We also consider a time-series design,where?rms without any cash?ow forecast during the sample period1993–2009are excluded;that is,in this design,?rms have at least one cash?ow forecast during the sample period.In the time-series design,the benchmark is the same?rms across years with and without cash?ow forecasts. This design potentially provides a better control for?xed di?erences in?rm characteristics across?rms.For both the cross-sectional and time-series designs,we?nd that accruals are mispriced;that is,the coe?cient on the decile-based rank of abnormal accruals is negative; and consistent with the hypothesis,the coe?cient on the interaction between cash?ow forecasts and the decile rank of abnormal accruals is positive.This shows that compared to earnings forecast?rms,earnings and cash?ow forecast?rms exhibit less abnormal accrual mispricing on average.We?nd that the results are driven by?rm-years belonging to the period1993–2001and not by?rm-years belonging to the period2002–2009.This result is consistent with Green et al.’s(2011)?nding that the accrual anomaly has disappeared in recent years.

Desai,Rajgopal,and Venkatachalam(2004)show that the cash?ow from the opera-tions anomaly subsumes the accrual https://www.sodocs.net/doc/c01867391.html,ing hedge portfolio abnormal returns, Call(2008)?nds that the cash?ow anomaly is mitigated for earnings and cash?ow fore-cast?rms.To provide a benchmark,we?rst consider the decile rank of cash?ows and interact the cash?ow forecast dummy with the decile rank of cash?ows,without includ-ing the accrual-related variables.We?nd that,as in Desai et al.(2004),Drake,Myers, and Myers(2009),and Barone and Magilke(2009),the coe?cient on the decile rank of cash?ows is positive;and similar to Call’s(2008)?ndings,the interaction of the decile rank of cash?ows and the cash?ow forecast dummy is negative.We then include the decile rank of cash?ows and the interaction of the cash?ow forecast dummy with the decile rank of cash?ows in the accruals model.While,in the cross-sectional design,the analysts’cash?ow forecast is associated with less accrual mispricing and cash from opera-tions mispricing;in the time-series design,the analysts’cash?ow forecast is associated with less accrual mispricing alone.

Using a di?erent research design,concurrent research by Mohanram(2014) documents similar evidence.Collectively,these?ndings show that earnings and cash?ow forecast?rms have less accrual mispricing and cash?ow mispricing compared to earnings forecast?rms.

These?ndings contribute to both the accrual anomaly and the analysts’cash?ow forecasts literatures.First,they provide evidence suggesting that the incidence of analysts’2.Earnings and cash?ow forecast?rms are signi?cantly di?erent in terms of risk,arbitrage potential,and

transaction costs.

CAR Vol.31No.4(Winter2014)

Cash Flow Forecasts and Accrual Mispricing1193 cash?ow forecasts is associated with less accrual anomaly and is consistent with Hirshleif-er and Teoh’s(2006)notion that investors’limited attention contributes to the accrual anomaly.It also adds to the prior literature suggesting that the availability of accrual information prior to the10-K?lings can direct investors’attention toward the accrual components of earnings and help investors appropriately assess the persistence of accrual information(Louis,Robinson,and Sbaraglia2008;Levi2008).

Second,this study adds to the growing literature on analysts’cash?ow forecasts. Following DeFond and Hung’s(2003)paper on analysts’cash?ow forecasts,subsequent studies have examined the quality of analysts’cash?ow forecasts(Givoly,Hayn,and Lehavy2009),the e?ect of analysts’cash?ow forecasts on?rms’accrual quality(McInnis and Collins2011),and whether analysts’earnings forecasts are more accurate when they also forecast operating cash?ows(Call et al.2009).This study adds to the literature con-cerning the consequences of the cash?ow forecasts by documenting the association of cash?ow forecasts with less accrual mispricing.

Finally,broadly speaking,our?ndings provide insights into the links between infor-mation environment,information intermediaries,and market anomalies.Prior studies show that analysts,as information intermediaries,help to improve the?rms’information environment(for example,see Asquith,Mikhail,and Au2005;Francis and So?er1997). Our?ndings add to this literature by documenting a consequence of analysts’information that helps to improve the information environment;more generally,analysts’information can help mitigate other market mispricing as well.

The remainder of the paper is organized as follows.Section2develops the hypothesis and the research design.Section3provides the empirical analysis and results.Section4 contains some concluding remarks.

2.Hypothesis development,variable de?nitions,and research design

Hypothesis development

Sloan(1996)documents the accrual anomaly:high(low)accrual?rms experience negative (positive)future abnormal returns.He suggests that investors are?xated on earnings and shows that the earnings expectation embedded in stock prices fails to fully re?ect the lower (higher)earnings persistence attributable to the accrual(cash?ow)component of earnings. LaFond(2005)and Pincus,Rajgopal,and Venkatachalam(2007)show that the accrual anomaly exists in other international stock markets as well.Lev and Nissim(2006)and Mashruwala et al.(2006)show that the accrual anomaly is not arbitraged away;partly due to arbitrage risk and transaction costs.However,Green et al.(2011)?nd that the accrual anomaly has decreased in the early/mid-2000s,potentially due to arbitrage by hedge funds targeting the accrual anomaly.3

Hirshleifer and Teoh(2006)provide an explanation for why investors?xate on earn-ings:investors’limited attention.They argue that attention requires e?ort;and accord-ingly,investors are selective in their analysis.Some investors with limited attention and processing power focus only on subsets of the publicly available information when valuing

a stock.If an investor with limited attention must select between a focus on earnings or

3.Several studies provide insights into the accrual anomaly.One stream of literature examines the component

of accruals that drive the accrual anomaly:in particular,it is shown that the accrual anomaly is driven by abnormal accruals(Xie2001),less reliable accruals(Richardson,Sloan,Soliman,and Tuna2005),and inventory changes(Thomas and Zhang2002;Chan,Chan,Jegadeesh,and Lakonishok2006).Another stream of literature provides insights into the na€?ve-investor hypothesis by examining whether sophisticated investors are able to appropriately assess the implications of accruals for?rm value.While Ali,Hwang,and Trombley(2000)and Bradshaw,Richardson,and Sloan(2001)?nd evidence inconsistent with the na€?ve-investor hypothesis,Collins,Gong,and Hribar(2003)and Lev and Nissim(2006)?nd support for the na€?ve-investor hypothesis.

CAR Vol.31No.4(Winter2014)

1194Contemporary Accounting Research

cash?ows,earnings is the better choice for?rm valuation because earnings are more informative than cash?ows(Dechow1994).When some investors do focus on earnings without distinguishing between accruals and cash?ows,they misprice?rms with high lev-els of accruals.Since the equilibrium stock price re?ects the average beliefs of investors who only focus on information in current period earnings and who focus on the accrual and cash?ow components of earnings,the stocks of?rms with extreme accruals are mis-priced.The correction of this mispricing results in the accrual anomaly.4 In addition to earnings forecasts,stock recommendations,and target stock prices,?nancial analysts have started to provide forecasts of cash?ow from operations.DeFond and Hung(2003)examine the determinants of analysts’cash?ow forecasts and?nd that analysts issue cash?ow forecasts for?rms with large absolute accruals,more heteroge-neous accounting choices relative to their industry peers,higher earnings volatility,high capital intensity,and poor?nancial health.They conclude from reading several full text analysts’reports that,“it appears that the cash?ow forecasts are not a trivial translation of predicted earnings,but rather the result of di?cult and costly information processing that involves the prediction of working capital and deferred taxes”(DeFond and Hung 2003,81).

When analysts issue both cash?ow forecasts and earnings forecasts,they indirectly provide accrual forecasts.Based on Hirshleifer and Teoh’s(2006)investors’limited atten-tion on earnings components arising from time constraints,when current period accruals and predicted next period accruals are available,investors can gauge the persistence of current period accruals for earnings and cash?ow forecast?rms with less e?ort.Further-more,Call et al.(2009)?nd that analysts’underweighting of the accrual and cash?ow com-ponents of earnings is less severe when analysts issue both earnings forecasts and cash?ow forecasts,suggesting that analysts better understand the persistence of the accrual and cash ?ow components of earnings when they issue earnings forecasts along with cash?ow forecasts.Thus,given that the cash?ow forecast information will help investors to assess the persistence of accruals more appropriately and with less e?ort,we expect less accrual mispricing for earnings and cash?ow forecast?rms than for earnings forecast?rms.

We consider abnormal accruals mispricing for two reasons.First,prior studies have shown that accrual mispricing is mainly driven by abnormal accruals(Xie2001;Cheng and Thomas2006).Second,noting that the total accruals consist of a normal and an abnormal component and the normal accruals are likely to be associated with the invest-ment-based accrual anomaly documented in Wu et al.(2010),the abnormal accruals are more suited for examining the investor?xation-based anomaly.Therefore,we focus our attention on abnormal accruals instead of total accruals.The above arguments are formally stated in the following hypothesis:

H YPOTHESIS.Analysts’cash?ow forecasts are associated with less abnormal accrual

mispricing,ceteris paribus.

We highlight a few noteworthy points about the hypothesis.The null/alternative to the hypothesis stated above occurs when either(a)the na€?ve or limited-attention investor assumption is not valid or(b)the cash?ow forecasts are of poor quality such that they do not provide information to investors to assess the persistence of current period accru-als.Wu et al.(2010)argue and show evidence that the accrual anomaly is consistent with a?rm’s optimal investment decision in response to discount rate changes.When the dis-count rate falls(rises),more(less)investment projects become pro?table.This increases (decreases)accruals,and future returns decrease(increase)on average because the lower 4.Ali and Gurun(2009)and Louis et al.(2008)provide indirect support for the limited-attention hypothesis.

CAR Vol.31No.4(Winter2014)

Cash Flow Forecasts and Accrual Mispricing1195 (higher)discount rate suggests lower(higher)expected returns going forward.As a result, accruals negatively predict future returns.This explanation does not depend on the na€?ve or limited-attention investor hypothesis.As such,if the accrual anomaly is not related to na€?ve or limited-attention investors,then the existence of cash?ow forecasts is not likely to be associated with less accrual anomaly.However,as argued by Wu et al.(2010),these two explanations—investors’rationality versus irrationality—are not mutually exclusive, and it is virtually impossible to distinguish the rational explanation from the irrational explanation of the accrual anomaly.

Givoly et al.(2009)examine the properties of analysts’cash?ow forecasts and?nd that cash?ow forecasts are less accurate than earnings forecasts.This?nding suggests that analysts’cash?ow forecasts are of low quality.By implication,it also suggests that the accrual forecast is of low quality.As such,the analysts’cash?ow forecasts may not help investors gauge the persistence of accruals,and thus the existence of cash?ow forecasts is not likely to be associated with less accrual anomaly.

Variable de?nitions and research design

Figure1shows the timeline of how we align the variables.We consider all year t+1earn-ings forecasts(EF t,t+1)and cash?ow forecasts(CFOF t,t+1)issued in year t,that is, between the earnings announcement dates of year tà1(EA tà1)and year t earnings(EA t). We also include earnings forecasts and cash?ow forecasts that are issued just after the earnings announcement in year t(EA t),but before the end of the?fth month after year t. A?rm-year is classi?ed as one with analysts’cash?ow forecasts,if analysts issued at least one EF t,t+1and CFOF t,t+1during the period between year tà1earnings announcement date(EA tà1)and the?fth month after year t;and a?rm-year without cash?ow forecasts if only EF t,t+1is issued during the period.

As in Bradshaw et al.(2001),annual returns are buy-and-hold returns,compounded annually beginning from the?fth month subsequent to the?scal year-end of year t using daily returns.The return accumulation period begins in the?fth month in order to ensure the availability of accruals information for forming accrual ranks.Size-adjusted returns SIZEADJR t+1are calculated by subtracting the returns of companion?rms in the same size-decile,where the size is the market value of equity at the beginning of the calendar year when the return accumulation begins.The breakpoints for the size-decile are identi-?ed using all?rms in the NYSE,AMEX and NASDAQ.As in Kraft,Leone,and Wasley (2007),when a?rm delists,we use the delisting return for the delisting day;if a?rm delists due to either a liquidation or a forced delisting by the exchange/SEC and the delisting return is missing in Center for Research in Security Prices(CRSP),then the delisting return is set atà100percent.We set the abnormal returns for the rest of the year for the delisted?rm to be zero.

Xie(2001)?nds that the accrual anomaly is driven by abnormal accruals.5Accord-ingly,we use abnormal accruals as the residual obtained from estimating Dechow et al.’s (1995)model in(1):

?TACC it=TA ità1 ?b1?1=TA ità1 tb2?eD REV itàD REC itT=TA ità1

tb3?PPE it=TA ità1 te it;e1Twhere subscript i refers to?rm i.TACC is the total accruals in year t(COMPUSTAT data item IBCàOANCF).TA ità1is total assets(COMPUSTAT data item AT)at the end of year tà1.D REV it is the change in sales from year tà1to year t.D REC it is the change in

5.We use total accruals instead of abnormal and normal accruals and?nd that the results are similar in our

main tests.We highlight the di?erence in the subsequent results when we control for cash?ow operations mispricing.

CAR Vol.31No.4(Winter2014)

1196Contemporary Accounting Research

net receivables from year tà1to year t.PPE it is the level of property,plant,and equip-ment at the end of year t.All continuous variables are winsorized annually at the one per-cent level in order to mitigate the in?uence of extreme observations on the results.The equation is estimated cross-sectionally for each year and each two-digit Standard Indus-trial Classi?cations(SIC)classi?cation with at least ten observations.The residuals from the above estimations are the abnormal accruals,ABNORMAL_ACC,and the predicted value is the normal accrual,NACC.Abnormal accruals(normal accruals)are ranked into deciles based on the prior year’s distribution of abnormal accruals(normal accruals)in order to avoid hindsight bias(Bernard and Thomas1989).6

Following Mashruwala et al.(2006),we control for other factors that are related to either future returns or that contribute to the existence of the accrual anomaly and esti-mate the following equation:

6.When these assignments are based on the rankings of accruals and abnormal accruals within the distribu-

tion for all?rms,including some that have not yet announced earnings for the year,there is a hindsight bias that makes the accrual-and abnormal accrual-based trading strategies not implementable.Assuming that investors rank accruals based on prior year’s accrual distribution therefore makes this trading strategy implementable.

CAR Vol.31No.4(Winter2014)

Cash Flow Forecasts and Accrual Mispricing1197 SIZEADJR itt1?b0tb0CFOF CFOF ittb1R ABACC ittb1CFOF CFOF it?R ABACC it

tb2R NACC ittb2CFOF CFOF it?R NACC it

tX it dtX it d A?R ABACC ittX it d N?R NACC it

tFixed-year effectte itt1;e2Twhere SIZEADJR it+1is the size-adjusted return for each?rm in year t+1.CFOF is a dummy variable equal to one for?rm-years with cash?ow forecasts,and zero otherwise. R_ABACC and R_NACC are the annual decile rank of abnormal and normal accruals, respectively,scaled to range fromà0.5to0.5(see,e.g.,Mashruwala et al.2006;Collins et al.2003;Drake et al.2009;Barth and Hutton2004).Scaled decile ranks helps to interpret the coe?cients as the annual size-adjusted returns for a zero-investment strategy with long positions in lowest accrual deciles and short positions in highest accrual deciles.

The vector X is the decile ranking of the following control variables scaled to be betweenà0.5and0.5:LNBM,EP,BETA,ARBRISK,PRICE,VOL,LNMV, EARNINGS FORECAST ERROR,INVESTMENT,and RISK FREE RATE(the pre?x “R”in front of the variable denotes that it is ranked).We also include the interactions of the scaled abnormal accruals and normal accruals decile ranks and the scaled decile ranks of these variables.We discuss and de?ne the control variables below.

LNBM is the natural logarithm of the book-to-market ratio for each?rm in each year.EP is the operating income after depreciation(COMPUSTAT item OIADP)scaled by the year-end market value of equity for each?rm in each year.BETA is the common stock beta obtained from CRSP for each?rm in each year,which follows the methods developed by Scholes and Williams(1977).7LNBM,EP and BETA are included in(2)to control for risk.

Mashruwala et al.(2006)?nd that arbitrage risk(ARBRISK)and transaction costs (PRICE,VOL,LNMV)prevent investors from exploiting the accrual anomaly.ARBRISK is the residual variance from a market model regression of a?rm’s daily return on the CRSP value-weighted market return over the year ending on the fourth month of the fol-lowing year.PRICE is the closing price at the end of the fourth month of following year. VOL is the daily closing stock price multiplied by daily shares traded averaged over a year ending on the fourth month of the following year.LNMV is the natural logarithm of the market value of equity at the end of year t for each?rm.

As discussed in the hypothesis development section,Wu et al.(2010)argue and?nd that investment patterns explain the negative relationship between current period accruals and future returns.Therefore,following Wu et al.(2010)we use the investment-to-assets ratio(INVESTMENT)and risk-free rates(RISK FREE RATE)as controls in(2); INVESTMENT is the annual change in gross property,plant,and equipment(COMPUSTAT data item PPEGT)plus the annual change in inventories(COMPUSTAT data item INVT) divided by the lagged book value of assets(COMPUSTAT data item AT).RISK FREE RATE is the monthly risk-free rates compounded over the?scal year.The monthly risk-free rates are the interest rates of the three-month Treasury Bill:secondary market rate 7.Speci?cally annual beta is calculated as follows:

b?f R teret i;t mret3tTàe1=nTeR t mret3tTg=R temret t mret3tTàe1=nTeR t mret tTeR t mret3tTwhere ret i,t=log of (1+return for security i on day t),mret t=log of(1+value-weighted market return on day t),mret3t= mret tà1+mret t+mret t+1(a three day moving average market window),n=number of observations for the year and the summations over t are over all days on which security i traded,beginning with the?rst trading day of the year and ending with the last trading day of the year.

CAR Vol.31No.4(Winter2014)

1198Contemporary Accounting Research

from the economic research data base at the Federal Reserve Bank at St.Louis(see,e.g., Welch and Goyal2008).

Call et al.(2009)show that analysts’earnings forecasts are more accurate when accompanied by cash?ow forecasts,which could suggest that the information environ-ment of earnings and cash?ow forecast?rms is better than that of earnings forecast?rms. This di?erence in the information environment could lead to lower accrual mispricing for earnings and cash?ow forecast?rms compared to that of earnings forecast?rms.To con-trol for this possibility,we include EARNINGS FORECAST ERROR as a control variable in(2);EARNINGS FORECAST ERROR is de?ned as the absolute value of the di?erence between consensus earnings forecasts and the company’s actual earnings de?ated by the absolute value of the company’s actual earnings.The consensus earnings forecast is the median of all individual one-year-ahead earnings forecasts outstanding immediately prior to the announcement of the current year’s earnings.If I/B/E/S does not have the actual earnings,the actual earnings are obtained from COMPUSTAT(annual data item IBC divided by the number of shares outstanding from I/B/E/S).

Firms for which analysts choose to issue cash?ow forecasts may have certain features that would introduce a self-selection bias.In estimating(2),the analysts’choice of issuing cash?ow forecasts measured by CFOF could be driven by their private information which is not observable to the researcher.If these unobservable characteristics are correlated with abnormal accruals,then the coe?cient estimates would be biased(see Li and Prabha-la2007).In other words,the accrual anomaly could be correlated with certain features of the?rm that make the analysts choose to issue cash?ow forecasts.Heckman(1979)pro-vides a correction for such self-selection bias by estimating a selection model in the?rst stage for the analysts’choice of issuing cash?ow forecasts using the probit procedure and using the Inverse Mills Ratio(IMR)as a control variable in(2).The IMR is computed using the normal density and cumulative distribution functions of the predicted likelihood from the?rst-stage probit procedure.For the?rst stage,we use the selection model used in DeFond and Hung(2003).In particular,we estimate a probit model as in(3)below and obtain the IMR:

CFOF it?b0tb1MAGNITUDE OF ACCRUALS ittb2EARNINGS VOLATILITY it tb3ACCOUNTING CHOICE HETEROGENEITY ittb4ALTMAN Z it

tb5CAPITAL INTENSITY ittb6SIZE itte it;e3Twhere MAGNITUDE OF ACCRUALS is the absolute value of total accruals measured as net income before extraordinary items minus operating cash?ow divided by total assets.EARNINGS VOLATILITY is the coe?cient of variation of earnings measured over the sample period,calculated as the standard deviation of earnings over mean of earnings,where earnings is earnings per share before extraordinary items scaled by the beginning stock price.ACCOUNTING CHOICE HETEROGENEITY is the index rang-ing from0to1that captures the comparability of a?rm’s accounting choice with other ?rms in the same industry.For each of the following?ve accounting choices,(1)inven-tory valuation;(2)investment tax credit;(3)depreciation;(4)successful e?ort vs.full cost for companies with extraction activities;and(5)purchase vs.pooling,we assign a value of one to each?rm whose accounting choice di?ers from the most frequently cho-sen method in that?rm’s industry group.The score for each?rm is summed and then scaled by the number of accounting choices in the industry.ALTMAN Z is the Altman’s Z-score calculated as1.2(Net working capital/Total assets)+1.4(Retained earnings/Total assets)+ 3.3(Earnings before interest and taxes/Total Assets)+0.6(Market value of equity/Book value of liabilities)+1.0(Sales/Total assets).CAPITAL INTENSITY is the ratio of gross property,plant,and equipment divided by sales revenue in the year when CAR Vol.31No.4(Winter2014)

Cash Flow Forecasts and Accrual Mispricing1199 analysts made the cash?ow forecasts.SIZE is the natural log of market value of equity in the year when analysts made the cash?ow forecasts.We use the Inverse Mills Ratio (IMR)obtained from(3)as a regressor in(2)to mitigate problems of potential self-

CFOFTfor an earnings and cash?ow fore-selection.Speci?cally,IMR?/ed

CFOFT=Ued

CFOFTTfor an earnings forecast?rm,where

CFOFT=e1àUed

cast?rm and IMR?à/ed

/(á)andΦ(á)are normal density and cumulative distribution functions,respectively;and CFOF is the predicted value of(3)computed from the coe?cient estimates(see the d

Greene2003,759;Lennox,Francis,and Wang2012).

Lennox et al.(2012)discuss the importance of choosing the set of exogenous variables in (3)such that the association between the accrual mispricing and the exogenous variables is through IMR.8All of the regressors in(3);that is,MAGNITUDE OF ACCRUALS, EARNINGS VOLATILITY,ACCOUNTING CHOICE HETEROGENEITY,CAPITAL INTENSITY and ALTMAN Z are excluded in(2),and as such all of these variables are con-sidered exogenous for accrual mispricing.DeFond and Hung(2003)argue that each of these variables measure either the di?culty of interpreting earnings or the importance of cash?ows for the business model or both.Prior research on accrual mispricing have not identi?ed any of these variables as being associated with accrual mispricing.As such,we treat all of these variables as exogenous.Given that this argument for the exclusion restriction is based on prior empirical studies on accrual mispricing and not on any solid theory,we examine the variance in?ation factors in(2),and check the sensitivity of our results to this exclusion restriction.9

Equation(2)is estimated using ordinary least squares(OLS)with?xed-year e?ects and the standard errors are corrected using the Huber-White-Sandwich procedure by clus-tering the observations by industry.10Based on the accrual anomaly,we expect the coe?-cient on R_ABACC,b1to be negative because?rms with larger(smaller)accruals have larger(smaller)decile ranks and earn negative(positive)subsequent abnormal returns. Based on our hypothesis that compared to earnings forecast?rms,earnings and cash?ow forecasts?rms have less accrual mispricing,we expect the coe?cient on the interaction term CFOF it9R_ABACC it,b1CFOF,to be positive.The coe?cient b1+b1CFOF(b1)is the hedge return from the zero investment strategy for the earnings and cash?ow forecast ?rms(earnings forecast?rms).

3.Empirical analysis

Sample

The sample spans the period from?scal year1993to?scal year2009.11Analysts’cash ?ow and earnings forecasts are obtained from the I/B/E/S Detail History U.S.Edition database.Following prior research on the accrual anomaly such as Xie(2001),we delete

8.This is related to the weak instruments discussed in Larker and Rusticus(2010)and Chan,Chen,Janakir-

aman,and Radhakrishnan(2012).

9.In two separate sensitivity tests we consider only(a)ACCOUNTING CHOICE HETEROGENEITY,

CAPITAL INTENSITY and ALTMAN Z and(b)CAPITAL INTENSITY for the exclusion restriction.We discuss these along with the discussion of the results in Table2,panel A.

10.We do not cluster by year because in the subperiod analysis we have fewer than ten clusters.Peterson(2009)

shows that standard errors clustered over time produce unbiased estimates only when there are a su?cient number of clusters.Speci?cally,he?nds that ten clusters are insu?cient to produce unbiased standard errors.

11.The sample begins in1993since analysts’cash?ow forecasts data were not available until that year.An

implicit assumption is that the cash?ow forecasts reported to I/B/E/S are complete.If the cash?ow fore-casts are not reported or captured by I/B/E/S,especially in the initial years,we may not?nd any di?erence in accrual mispricing across earnings and cash?ow forecasts?rms and earnings forecast?rms.However,if I/B/E/S systematically omits to capture cash?ow forecasts of?rms with more accrual mispricing,our result could be driven by the I/B/E/S cash?ow forecast measurement error.

CAR Vol.31No.4(Winter2014)

1200Contemporary Accounting Research

?rms in the?nancial services industry(SIC codes6000to6999),American depository receipts(ADRs),real estate investment trusts(REITs),and units of bene?cial interest. Financial statement data are obtained from the CRSP–COMPUSTAT merged database, and returns,trading volume,and stock price data are obtained from CRSP.Following Kraft et al.(2007)and Ali and Gurun(2009),we delete?rm-years with stock prices (before-split)below$5to mitigate the in?uence of transactions costs driving the accrual anomaly.12Because large returns can cause misleading inferences in tests of mispricing,we delete observations with size-adjusted returns greater than225percent(Kraft et al.2007; Cheng and Thomas2006).

Table1,panel A provides the?rm-year distribution for the sample?rms used in the empirical tests.The sample includes9,977?rm-years with both cash?ow forecasts and earnings forecasts and23,264?rm-years with only earnings forecasts.The percentage of earnings and cash?ow forecast?rms has increased over the years:exceeding40percent of the sample in2002and reaching60percent in2009.Table1,panel B provides the mean and median of earnings forecast?rms and earnings and cash?ow forecast?https://www.sodocs.net/doc/c01867391.html,-pared to earnings forecast?rms,earnings and cash?ow forecast?rms have higher future size-adjusted abnormal returns,more negative total accruals,higher cash?ows,lower abnormal accruals,and lower normal accruals and,furthermore,are di?erent across the dimensions identi?ed in DeFond and Hung(2003).This shows the importance of control-ling for self-selection in(2).

Table1,panel C reports the Pearson(upper diagonal)and Spearman(lower diagonal) correlation among selected variables.Consistent with DeFond and Hung(2003),the provi-sion of analysts’cash?ow forecast is positively correlated with more heterogeneous accounting choices relative to their industry peers,poor?nancial health,and higher capital intensity.Consistent with prior research,future return(SIZEADJR t+1)is negatively corre-lated with total accruals and abnormal accruals and positively correlated with cash?ow from operations(Sloan1996;Cheng and Thomas2006;Xie2001).

Table1,panel D provides the results of estimating(3).We estimate(3)separately for the period before and after2002because Green et al.(2011)show that the accrual anom-aly ceases to exist starting roughly from2002.13The signs on the coe?cient estimates are in line with DeFond and Hung(2003).The area under the Relative Operating Characteris-tic(ROC)curve provides a measure of goodness-of-?t;it is0.8281and0.7953for the peri-ods1993–2001and2002–2009,respectively.This suggests that DeFond and Hung’s model is a reasonably good predictor of analysts’issuance of cash?ow forecasts.

Accrual mispricing across earnings forecast?rms and earnings and cash?ow forecast?rms Hedge return test

Table1,panel E provides the summary of a univariate hedge portfolio returns test for earnings and cash?ow forecast?rms and earnings forecast?rms based on the trading strategy of going long in?rms with the lowest abnormal accruals and short in?rms with the highest abnormal accruals.The left columns under the heading“Earnings and Cash Flow Forecast Firms”of panel E,Table1shows that the average hedge returns for earnings and cash?ow forecast?rms areà0.04(t-statistic=à0.40)andà0.02(t-sta-tistic=à0.37)for the1993–2001and2002–2009periods,respectively;this shows that for these?rms the hedge returns are not statistically di?erent from0.The right columns under the heading“Earnings Forecast Firms”of panel E,Table1shows that the 12.Other reasons for excluding?rm-years with stock prices less than US$5are to mitigate the impact of low-

priced stocks on the skewness of ex post returns and the e?ect of microstructure-induced return volatility (Loughran and Ritter1996).

13.We use the Inverse Mills Ratio computed by pooling both the periods,as well as separately for each year

and?nd similar results.

CAR Vol.31No.4(Winter2014)

T A B L E 1S a m p l e a n d d e s c r i p t i v e s t a t i s t i c s

P a n e l A :F i r m -y e a r d i s t r i b u t i o n o f s a m p l e

Y E A R E a r n i n g s a n d c a s h ?o w f o r e c a s t ?r m s

E a r n i n g s f o r e c a s t ?r m s

N u m b e r

%o f t o t a l

N u m b e r

%o f t o t a l

1993694%1,78896%19941427%1,87993%19951205%2,10795%19961838%2,12292%19972178%2,34392%19982029%2,01691%199945021%1,69079%200043824%1,41676%200122713%1,58787%200272645%87355%200392849%95651%20041,01355%82945%20051,08755%89245%20061,11156%85644%20071,07660%71240%200894963%54737%20091,03961%65139%T o t a l 9,977

30%

23,26470%

(T h e t a b l e i s c o n t i n u e d o n t h e n e x t p a g e .)

Cash Flow Forecasts and Accrual Mispricing

1201

CAR Vol.31No.4(Winter 2014)

P a n e l B :D e s c r i p t i v e s t a t i s t i c s

V a r i a b l e s

E a r n i n g s a n d c a s h ?o w f o r e c a s t ?r m s E a r n i n g s f o r e c a s t ?r m s

D i ?e r e n c e

M e a n

M e d i a n M e a n

M e d i a n t -s t a t s .

z -s t a t s .

S I Z E A D J R t +1

0.03à0.01à0.04à0.0812.89***16.91***T A C C à0.07à0.06à0.05à0.04à20.13***à20.49***T C F 0.110.100.070.0928.55***23.61***A B N O R M A L _A C C 0.010.000.010.01à1.97**à4.42***N O R M A L _A C C à0.08à0.06à0.06à0.05à16.52***à22.26***L N B M à0.93à0.87à0.91à0.83à2.45**à4.05***E P 0.080.080.080.085.26***2.98***B E T A 1.081.000.860.7831.78***32.91***L N M V 7.597.495.995.8989.30***80.86***A R B I T R A G E R I S K 0.000.000.000.00à46.96***à45.17***P R I C E 31.7527.1422.6717.8738.08***41.56***V O L U M E 47.9915.659.331.8340.81***85.07***I N V E S T M E N T 0.100.050.100.060.87à6.53***M A G N I T U D E O F A C C R U A L S 0.070.060.080.06à1.865.77***E A R N I N G S V O L A T I L I T Y 3.021.343.371.35à4.51***à0.18A C C O U N T I N G C H O I C E H E T E R O G E N E I T Y 0.050.000.040.008.88***16.58***A L T M A N Z 4.703.386.704.27à25.77***à28.38***C A P I T A L I N T E N S I T Y 1.310.660.770.4230.03***36.79***

N

9,9779,97723,26423,264(T h e t a b l e i s c o n t i n u e d o n t h e n e x t p a g e .)

T A B L E 1(c o n t i n u e d )

1202

Contemporary Accounting Research

CAR Vol.31No.4(Winter 2014)

P a n e l C :P e a r s o n (u p p e r d i a g o n a l )a n d S p e a r m a n (l o w e r d i a g o n a l )c o r r e l a t i o n c o e ?c i e n t s f o r s e l e c t e d v a r i a b l e s f o r a s a m p l e o f 33,241?r m -y e a r s f r o m 1993–2009

(1)(2)

(3)

(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)

(1)C F O F 0.07

à0.10

0.14à0.01à0.10à0.010.030.170.45à0.210.220.290.00à0.01à0.020.05à0.12

0.19

(2)S I Z E A D J R t +10.09

à0.03

0.10à0.020.010.060.07à0.020.03à0.100.020.01à0.07à0.03à0.030.01

à0.04

0.00

(3)T A C C à0.11à0.04

à0.31

0.660.200.030.18à0.03à0.05à0.070.03à0.050.08à0.180.01à0.02

0.04

à0.15

(4)T C F 0.130.11

à0.46

à0.16à0.130.000.38à0.010.24à0.310.260.130.02à0.03à0.010.06

à0.06

à0.11

(5)A B N O R M A L _A C C à0.02à0.02

0.66

à0.27à0.550.010.11à0.03à0.01à0.030.02à0.010.11à0.060.00

0.01

à0.01

à0.03

(6)N O R M A L _A C C à0.12à0.01

0.30

à0.21à0.380.040.050.00à0.06à0.05à0.03à0.06à0.14à0.100.01

à0.04

0.04

à0.14

(7)L N B M à0.020.06

0.01

à0.180.000.010.32à0.14à0.29à0.10à0.31à0.18à0.07à0.11

0.06

0.05

à0.29

0.06

(8)E P 0.020.10

0.11

0.250.070.050.39à0.130.06à0.300.08à0.010.03à0.17

0.01

0.04

à0.25

à0.07

(9)B E T A 0.18à0.04

à0.02

à0.01à0.030.02à0.16à0.190.240.100.080.160.03

0.07

0.05

0.00

0.12

à0.04

(10)L N M V 0.440.08

à0.07

0.23à0.02à0.07à0.290.050.28à0.380.640.56à0.05

à0.10

à0.09

0.07

à0.06

0.10

(11)A R B I T R A G E R I S K à0.25à0.13

à0.02

à0.22à0.010.01à0.05à0.330.11à0.52à0.32à0.060.07

0.18

0.06

à0.03

0.23

à0.06

(12)P R I C E 0.230.08

0.03

0.320.03à0.02à0.310.130.080.68à0.540.360.02

à0.11

à0.09

0.07

à0.01

0.02

(13)V O L U M E 0.470.04

à0.07

0.19à0.04à0.05à0.31à0.070.430.87à0.240.57

0.00

0.00

à0.01

0.04

0.04

0.03

(14)I N V E S T M E N T à0.04à0.07

0.13

0.050.16à0.11à0.110.050.00à0.060.100.07

à0.03

0.02

0.01

0.10

0.06

0.16

(15)M A G N I T U D E O F

A C C R U A L S 0.03à0.02

à0.26

0.11à0.09à0.17à0.08à0.140.06à0.060.15

à0.11

0.00

0.01

0.05

0.00

à0.01

0.04

(16)E A R N I N G S V O L A T I L I T Y 0.00à0.12

à0.05

à0.13à0.02à0.010.11à0.150.16à0.200.30

à0.29

à0.06

à0.05

0.17

0.03

à0.01

à0.06

(17)A C C O U N T I N G C H O I C E

H E T E R O G E N E I T Y 0.090.02

à0.05

0.030.00à0.070.050.020.040.08

à0.04

0.06

0.06

0.01

0.11

0.12

à0.04

0.13

(18)A L T M A N Z à0.16à0.05

0.08

0.160.030.07à0.32à0.330.12à0.11

0.26

à0.01

0.01

0.12

0.01

à0.02

à0.02

à0.10

(19)C A P I T A L I N T E N S I T Y 0.200.03

à0.240.00à0.01à0.340.090.03à0.050.19

à0.22

0.10

0.09

0.09

0.09

à0.080.17à0.41

T A B L E 1(c o n t i n u e d )

(T h e t a b l e i s c o n t i n u e d o n t h e n e x t p a g e .)

Cash Flow Forecasts and Accrual Mispricing

1203

CAR Vol.31No.4(Winter 2014)

P a n e l D :P r o b i t m o d e l f o r d e t e r m i n a n t s o f a n a l y s t s ’c a s h ?o w f o r e c a s t s ,(3)

V a r i a b l e

P r e d i c t e d s i g n

1993–20012002–2009C o e ?c i e n t (z -s t a t i s t i c s )C o e ?c i e n t (z -s t a t i s t i c s )

I N T E R C E P T

?

à3.21***à3.49***(à48.87)(à47.99)M A G N I T U D E O F A C C R U A L S

+

1.22***0.43**(6.20)(

2.44)E A R N I N G S V O L A T I L I T Y

+

0.00**0.00**(2.51)(2.01)A C C O U N T I N G C H O I C E H E T E R O G E N E I T Y

+

1.27***0.89***(11.47)(5.74)A L T M A N Z

à

à0.04***à0.01***(à13.99)(à6.22)C A P I T A L I N T E N S I T Y

+

0.22***0.14***(27.43)(14.35)S I Z E

+

0.27***0.53***(32.21)(52.41)N 18,99614,245A r e a U n d e r R O C C u r v e 0.82810.7953P s e u d o -R 2

20.63%

21.43%

T A B L E 1(c o n t i n u e d )

(T h e t a b l e i s c o n t i n u e d o n t h e n e x t p a g e .)

1204

Contemporary Accounting Research

CAR Vol.31No.4(Winter 2014)

P a n e l E :A b n o r m a l a c c r u a l -b a s e d h e d g e p o r t f o l i o r e t u r n s

E a r n i n g s a n d C a s h

F l o w F o r e c a s t F i r m s

E a r n i n g s

F o r e c a s t F i r m s

A l l y e a r s (t -s t a t s .)

1993–2001(t -s t a t s .)2002–2009(t -s t a t s .)A l l y e a r s (t -s t a t s .)1993–2001(t -s t a t s .)2002–2009(t -s t a t s .)

H e d g e r e t u r n s

à0.03à0.04à0.020.04*0.11***à0.03(à0.61)(à0.40)(à0.37)(1.80)(4.63)(à0.82)#o f p o s i t i v e h e d g e r e t u r n s

8o u t o f 174o u t o f 94o u t o f 813o u t o f 179o u t o f 94o u t o f 8

N o t e s :

I n p a n e l B t h e d i ?e r e n c e c o l u m n i s t h e d i ?e r e n c e i n m e a n s (m e d i a n s ):“E a r n i n g s a n d c a s h ?o w f o r e c a s t ?r m s ”m i n u s “E a r n i n g s f o r e c a s t ?r m s .”I n p a n e l C ,b o l d n u m b e r s s u g g e s t t h a t t h e c o r r e l a t i o n s a r e s i g n i ?c a n t a t 10p e r c e n t l e v e l o r b e t t e r f o r a t w o -t a i l e d t e s t .I n p a n e l E ,t h e h e d g e r e t u r n s b a s e d o n a b n o r m a l a c c r u a l d e c i l e s a r e r e p o r t e d .T h e “A l l y e a r s ”c o l u m n s i n c l u d e ?r m -y e a r o b s e r v a t i o n s f r o m 1993t o 2009.T o i m p l e m e n t t h e h e d g e p o r t f o l i o t e s t ,w e s o r t ?r m s i n t o a b n o r m a l a c c r u a l d e c i l e s b a s e d o n t h e p r i o r y e a r ’s d i s t r i b u t i o n o f a b n o r m a l a c c r u a l s .W e t h e n c a l c u l a t e f u t u r e r e t u r n s f o r e a c h a b n o r m a l a c c r u a l d e c i l e f o r t h e y e a r f o l l o w i n g p o r t f o l i o f o r m a t i o n d a t e s .T h e h e d g e r e t u r n s r e p o r t t h e a v e r a g e s i z e -a d j u s t e d r e t u r n s f r o m i m p l e m e n t i n g t h e t r a d i n g s t r a t e g y l o n g i n ?r m s w i t h i n t h e l o w e s t a b n o r m a l a c c r u a l d e c i l e s a n d s h o r t i n ?r m s w i t h i n t h e h i g h e s t a b n o r m a l a c c r u a l d e c i c l e .T h e ***,**a n d *d e n o t e s i g n i ?c a n c e a t t h e 1p e r c e n t ,5p e r c e n t ,a n d 10p e r c e n t l e v e l s f o r a t w o -t a i l e d t e s t ,r e s p e c t i v e l y .A l l v a r i a b l e s a r e d e ?n e d i n t h e A p p e n d i x .

T A B L E 1(c o n t i n u e d )

Cash Flow Forecasts and Accrual Mispricing

1205

CAR Vol.31No.4(Winter 2014)

1206Contemporary Accounting Research

TABLE2

Analysts’cash?ow forecasts and accrual anomaly

Panel A:Return-abnormal accruals regressions,(2)

All years1993–20012002–2009

N=33,241N=18,996N=14,245

Coef.t-stat.VIF Coef.t-stat.VIF Coef.t-stat.VIF CFOF b0CFOF0.13*** 3.3315.010.10** 2.2114.280.14** 2.3813.62 R_ABACC b1à0.08***à4.81 2.07à0.11***à4.87 2.050.00à0.16 2.05 CFOF9R_ABACC b1CFOF0.07*** 3.09 2.210.13** 2.08 2.21à0.01à0.30 2.20 R_NACC b20.00à0.01 1.980.000.12 1.960.010.20 1.97 CFOF9R_NACC b2CFOFà0.04*à1.77 2.330.08 1.65 2.32à0.07à1.42 2.32 IMRà0.05**à2.2711.62à0.01à0.5411.27à0.06**à2.0611.01 Control variables Yes Yes Yes

Fixed-year e?ects Yes Yes Yes Adjusted-R2 2.56% 3.49% 1.56%

F-statistic(p-value)

0.31(0.58)0.11(0.74)0.40(0.53)

for b1+b1CFOF=0

3.43*(0.07)

4.20*(0.05) 3.08*(0.09)

F-statistic(p-value)

for b2+b2CFOF=0

Panel B:Return-abnormal accruals regressions for initiation and constant samples,(2)

Initiation Sample Constant Sample

N=20,521N=3,465

Coef.t-stat.VIF Coef.t-stat.VIF CFOF b0CFOF0.08** 2.4311.970.10 1.5614.93 R_ABACC b1à0.09***à3.66 3.32à0.15**à2.55 4.15 CFOF9R_ABACC b1CFOF0.09*** 3.03 4.110.15*** 2.85 4.00 R_NACC b20.03 1.33 3.40à0.06à1.05 4.36 CFOF9R_NACC b2CFOFà0.04à1.32 4.290.06 1.07 4.19 IMRà0.05**à2.3110.25à0.10***à3.0114.73 Control variables Yes Yes

Fixed-year e?ects Yes Yes

Adjusted-R2 2.41% 6.91%

0.01(0.93)0.02(0.88)

F-statistic(p-value)

for b1+b1CFOF=0

0.17(0.69)0.01(0.93)

F-statistic(p-value)

for b2+b2CFOF=0

(The table is continued on the next page.) CAR Vol.31No.4(Winter2014)

average hedge returns for earnings forecast ?rms are 0.11(t -statistic =4.63)and à0.03(t -statistic =à0.82)for the 1993–2001and 2002–2009periods,respectively;this shows that for these ?rms the hedge returns are positive for the period 1993–2001,which is consistent with the ?nding of Green et al.(2009)that the accrual anomaly ceases to exist after 2002.Collectively,these results provide preliminary evidence supporting the hypothesis.However,it is important to control for various risk,transaction cost,and economy-wide investment factors that prior research has shown to explain the accrual anomaly.

Regression based test

Table 2,panel A shows the results of estimating (2)for “All years,”1993–2001,and 2002–2009.14As mentioned earlier we do this because Green et al.(2011)show that the accrual anomaly ceases to exist starting roughly from 2002.For the sample containing all ?rm-years,that is,columns “All years”the coe?cient on R_ABACC ,b 1,is à0.08(t -statistic =à4.81)suggesting that for earnings forecast ?rms,the hedge strategy of a long (short)position in the largest negative (positive)abnormal accrual ?rms yields an annual size-adjusted buy-and-hold return of 8percent.The coe?cient on the interaction term CFOF 9R_ABACC ,b 1CFOF ,is 0.07(t -statistic =3.09),suggesting that for earnings and cash ?ow forecast ?rms,the hedge strategy of a long (short)position in the largest negative (posi-tive)abnormal accrual ?rms yields an annual abnormal return of à7percent less com-pared to earnings forecast ?rms.That is,for earnings and cash ?ow forecast ?rms,the accrual anomaly is less than for those without cash ?ow forecasts and supports the hypothesis.The sum of the coe?cient b 1+b 1CFOF indicates abnormal returns for the hedge strategy of a long (short)position in the largest negative (positive)abnormal accrual ?rms for earnings and cash ?ow forecast ?rms’returns.The F -statistic =0.31(p -value =0.58)for the test of whether b 1+b 1CFOF =0shows that the accrual anomaly is completely mitigated for earnings and cash ?ow forecast ?rms.These results are driven by the early

Notes:

SIZEADJR it t1?b 0tb 0CFOF CFOF it tb 1R ABACC it tb 1CFOF CFOF it ?R ABACC it

tb 2R NACC it tb 2CFOF CFOF it ?R NACC it tX it d tX it d A ?R ABACC it tX it d N ?R NACC it tFixed -yeareffect te it t1;

e2T

where the vector X includes the decile ranks of the following control variables scaled to be between

à0.5and 0.5:LNBM ,EP ,BETA ,ARBRISK ,PRICE ,VOL ,LNMV ,EARNINGS

FORECAST ERROR ,INVESTMENT ,and RISK FREE RATE (the pre?x “R ”in front of the variable denotes that it is ranked).IMR is not interacted with R_ABACC and R_NACC .The standard errors are clustered by industry.***,**,and *denote signi?cance 1percent,5

percent,and 10percent,two-tailed tests,respectively.All variable de?nitions are contained in the Appendix.

The Initiation Sample excludes ?rms without any cash ?ow forecast from 1993to 2009and the year

prior to the initial cash ?ow forecast year.The Constant Sample excludes ?rm-years without cash ?ow forecasts after initiation from the Initiation Sample.

TABLE 2(continued)14.

The coe?cient estimates on the control variables are not shown for brevity.The complete table is available from the authors.

Cash Flow Forecasts and Accrual Mispricing

1207

CAR Vol.31No.4(Winter 2014)

1208Contemporary Accounting Research

years of the sample,1993–2001,which is consistent with Green et al.’s(2011)?nding that the accrual anomaly ceases to exist after2002.Overall,the results suggest that the pres-ence of analysts’cash?ow forecasts is associated with less accrual mispricing.15 The coe?cient estimates on the interaction term CFOF9R_NACC,b2CFOF,areà0.04 (t-statistic=à1.77),0.08(t-statistic=1.65)andà0.07(t-statistic=à1.42)for“All years,”1993–2001,and2002–2009periods,respectively;while the coe?cients on R_NACC are not (or only weakly)statistically di?erent from0.This suggests that for our sample for earnings and cash?ow forecast?rms,hedge return trading strategies based on normal accruals could yield abnormal returns,albeit weakly.One possible explanation for this result is based on the investment-based economic rationale proposed by Wu et al.(2010);this rationale is likely to be correlated with normal accruals.Even though we have controlled for the invest-ment-based economic rationale,these controls may not be adequate.Also,the hedge return trading strategy needs to be reversed for the periods1993–2001and2002–2009,which is indicative of normal accruals proxying for other potentially omitted factors.

The coe?cient estimates on IMR are statistically signi?cant in the“All years”and 2002–2009models,suggesting that the unobservable factors a?ecting the analysts’choice of issuing cash?ow forecasts are related to the abnormal returns.The VIF of IMR and CFOF are high enough that the IMR not being signi?cant for the1993–2001period is not indica-tive of self-selection not being an issue(Lennox et al.2012).However,in unreported analy-sis the results on our test variables are not a?ected whether IMR is included or excluded in (2).This shows that self-selection does not introduce a bias for abnormal accruals itself.

As discussed earlier we consider all of the regressors in(3)as exogenous for accrual mispricing—this exclusion restriction is based on prior empirical studies on accrual mispricing.However,it is plausible that the MAGNITUDE OF ACCRUALS and EARNINGS VOLATILITY are correlated with accrual mispricing because both of these are likely to be a?ected by the accruals.Thus,in(2)in addition to all of the control variables,we include the MAGNITUDE OF ACCRUALS and EARNINGS VOLATIL-ITY as additional regressors.The results are summarized as follows:For the sample con-taining all?rm-years the coe?cient on R_ABACC,b1,isà0.07(t-statistic=à4.15);and the coe?cient on the interaction term CFOF9R_ABACC,b1CFOF,is0.07(t-statistic= 3.00);and the F-statistic=0.09(p-value=0.76)for the test of whether b1+b1CFOF=0. These are qualitatively similar to the tests reported in Table2,panel A.

In an additional sensitivity test,we only consider CAPITAL INTENSITY for the exclusion restriction,based on the correlation table reported in Table1,panel C. ACCOUNTING CHOICE HETREOGENEITY is likely to be correlated with the accrual mispricing,because accounting choices map into accruals;and ALTMAN Z is likely to be correlated with accrual mispricing because it is likely to be correlated with performance. Thus,in(2)in addition to all of the control variables,we include the MAGNITUDE OF ACCRUALS,EARNINGS VOLATILITY,ACCOUNTING HETEROGENEITY and ALTMAN Z as additional regressors.The results are summarized as follows:For the sample containing all?rm-years the coe?cient on R_ABACC,b1,isà0.09(t-statistic =à4.71);and the coe?cient on the interaction term CFOF9R_ABACC,b1CFOF,is0.07 (t-statistic= 3.21);and the F-statistic=0.46(p-value=0.50)for the test of whether b1+b1CFOF=0.These are qualitatively similar to the tests reported in Table2,panel A. Accrual mispricing around analysts’cash?ow forecast initiation

Accrual mispricing is likely to be associated with certain?rm-speci?c characteristics.In the research design employed in Table2,panel A,we consider?rm-years without analysts’cash?ow forecasts as the benchmark group.As such,the sample contains earnings forecast

15.We obtain similar results when total accruals,instead of abnormal accruals,are used.

CAR Vol.31No.4(Winter2014)

Cash Flow Forecasts and Accrual Mispricing1209?rms for the entire sample period.Even though we have controlled for self-selection,risk factors,transaction costs,and economic rationale in(2),it is still possible that the accrual mispricing is driven by those earnings forecast?rms.Accordingly,we examine two subs-amples where each?rm is its own benchmark.First,we remove?rms without any cash ?ow forecasts during the entire sample period.In addition,we delete the year prior to the initiation year because the one-year-ahead,size-adjusted returns,which are the dependent variables,will track the year of initiation.In this sample,each?rm is its own benchmark for years without analysts’cash?ow forecasts;and this process ensures that there are ?rm-years before the?rst cash?ow forecast that also form part of the benchmark.16We refer to this sample as the“Initiation Sample.”

Second,we consider a subsample of the“Initiation Sample,”where we restrict the ?rm-years without cash?ow forecast to two years prior to the initiation year and two years after the initiation year.As with the“Initiation Sample,”we drop the year prior to the initiation year.For the years after initiation,we only include those?rm-years with analysts’cash?ow forecasts.This subsample ensures that changes in?rm characteristics over time with a long period of pre-initiation do not drive the results,and neither do the ?rm-years after initiation but without cash?ow forecasts.We refer to this subsample as the“Constant Sample.”It is important to note that even though this design helps to con-trol for latent?rm characteristics that are sticky over time,the results in this design could be driven by the alternative explanation that the accrual anomaly was arbitraged away by investors after2002due to arbitragers taking advantage of the accrual anomaly(see Green et al.2011).

Table2,panel B provides the results of estimating equation(2)for the Initiation and Constant subsamples.There are7,991(1,155)?rm-years in the Initiation(Constant)sam-ple in the pre-initiation period,that is,with CFOF=0;and12,530(2,310)?rm-years in the Initiation(Constant)sample in the post-initiation period.17For the Initiation sample, the coe?cient on R_ABACC,b1,isà0.09(t-statistic=à3.66),suggesting that the hedge strategy based on extreme abnormal accruals would have been pro?table on average for these same earnings and cash?ow forecast?rms,before such forecasts were initiated.The coe?cient on the interaction term,CFOF9R_ABACC,b1CFOF,is0.09(t-statistic=3.03), suggests that after cash?ow forecast initiation,the hedge strategy of a long(short)posi-tion in the largest negative(positive)abnormal accrual?rms yields an additional annual abnormal return ofà9percent compared to earnings forecast?rms.We?nd that the F-statistic=0.01(p-value=0.93)for testing whether b1+b1CFOF=0,which shows that the accrual anomaly is mitigated completely.The results are qualitatively similar for the Constant sample as well.

Accrual mispricing across earnings forecast?rms and earnings and cash?ow forecast?rms, controlling for cash?ow from operations anomaly

Desai et al.(2004)examine whether the accrual anomaly is related to the value-glamour anomaly,where one of the value-glamour measures is the cash?ow from operations to price(CFO).They?nd that the accrual anomaly is subsumed by the CFO anomaly.As such,more recent studies that examine the accrual anomaly also consider the CFO anom-aly(see,e.g.,Drake et al.2009;Barone and Magilke2009;Cheng and Thomas2006)

16.For?rms where the cash?ow forecasts are initiated in the?rst three years of the sample period,we go

back as far as to1989to get the?rm-years without cash?ow forecasts.

17.In the post-initiation period,CFOF=1in the Constant sample only when there is a cash?ow forecast.

However,for the Initiation sample,CFOF=1whether or not a cash?ow forecast is recorded in the I/B/E/S database.We do this because,given the notion of investors’attention,even one cue to help inves-tors direct their attention to cash?ows could be su?cient.We include all the years after initiation with only CFOF=1and obtain similar results.

CAR Vol.31No.4(Winter2014)

1210Contemporary Accounting Research

because the accrual mispricing could be a manifestation of cash?ow mispricing.Further-more,Call(2008)uses hedge returns and shows that cash?ow from operations mispricing is mitigated by analysts’cash?ow forecasts.

Combining the insights from these studies,our results could simply be driven by the cash?ow from operations mispricing,and not the accrual mispricing.However,Desai et al.(2004)state,“If the reader views the traditionally used C/P measure,and not the expanded CFO/P measure introduced here,as the de?nitive value-glamour variable,then our evidence does not support Beaver’s(2002)conjecture that the accrual anomaly is the glamour stock phenomenon in disguise.Rather,this reader would view CFO/P as a com-prehensive mega proxy that subsumes the role of accruals and the traditional value-glamour variables in predicting future returns.”(Desai et al.2004,357)Under this view,if the accrual anomaly is subsumed by the cash?ow from operations anomaly,it is still the accruals that are associated in part with predictable future returns.

Similar to Desai et al.(2004)we obtain cash?ow from operations from COMPUSTAT (item OANCF)and scale it by average total asset and independently sort cash?ow from oper-ations into decile ranks(R_CFO)and scale them to be betweenà0.5and0.5.To control for the cash?ow anomaly,we augment(2)by including R_CFO and its interaction with CFOF.

In Table3,panel A,column Model1,we estimate(2)without the accruals variables, that is,only with cash?ow from operations and the interaction between cash?ow from operations and CFOF to provide a benchmark.We?nd that the coe?cient on R_CFO, b3,is0.11(t-statistic=6.04)and the coe?cient on the interaction term CFOF9R_CFO, b3CFOF,isà0.06(t-statistic=à2.13).These results are similar to those of Call(2008),

suggesting that cash?ow forecasts are helpful in mitigating the cash?ow anomaly by using a di?erent method and controlling for various other factors.

In Table3,panel A,column Model2,All Years,we estimate(2)by adding the R_CFO and R_CFO9CFOF.We?nd that controlling for cash?ow from operations mis-pricing,the accrual mispricing also exists,especially before2001,and both cash?ow and accrual mispricings are less for earnings and cash?ow forecast?rms.

In Table3,panel B,we estimate(2)by adding the R_CFO and R_CFO9CFOF as in Table2,panel B for the Initiation and Constant subsamples.The results are qualita-tively similar to those discussed with Table3,panel A.18Overall,the results show that both accrual mispricing and cash?ow from operations mispricings are less for earnings and cash?ow forecast?rms than for earnings forecast?rms,which is consistent with the hypothesis that relates cash?ow forecasts to accrual mispricing.19

Announcement and non-announcement period size-adjusted portfolio stock returns in the year following portfolio formation

The tests in Tables2and3are based on annual returns.However,if the mispricing is due to accruals,then the hedge returns should be driven by the accounting events;that is, earnings announcements.For this purpose,similar to Sloan(1996)we disaggregate the annual abnormal returns into the announcement and non-announcement period abnormal returns.Speci?cally,we aggregate abnormal returns across the announcement periods of the four quarterly earnings’numbers for the year after each accrual ranking year.The non-announcement period is the period beginning the day after the announcement of the ?rst quarter’s earnings and ending three trading days prior to the announcement of the

18.We estimate the models with cash?ow from operations controls for1989–2001and2002–2009separately

and?nd that the results are driven by the initial period as in Green et al.(2011).

19.The results when cash?ow from operations ranks are controlled for are statistically much weaker when

total accruals instead of abnormal accruals are used.

CAR Vol.31No.4(Winter2014)

pos机刷卡手续费标准

pos机刷卡手续费标准 目前国内刷卡费率与行业分类挂钩，餐娱类的刷卡手续费率最高，为 1.25%;百货等一般商户为0.78%;超市、加油站等为0.38%;医院、教育等公益类则是零费率。 根据新规，对发卡行服务费实行不再区分商户类别。也就是说，商户行业分类定价取消，总体上大幅降低了刷卡的费率水平。 新政显示，发卡银行服务费费率水平降低为借记卡交易不超过交易金额的0.35%，贷记卡交易不超过0.45%。 银行卡清算机构收取的网络服务费费率水平降低为不超过交易金额的0.065%，由发卡机构、收单机构各承担50%。从类别上看，餐饮类企业刷卡手续费支出可降低53%63%，百货等行业商户可降低23%39%。新政实行后，医院、教育等公益类刷卡仍为零费率。 信用卡大额消费成本高 不过，信用卡刷卡手续费取消封顶后，一旦遇到大额消费，商户费用成本必将增加，这部分成本由谁买单。 此前信用卡交易虽然费率高，但单笔交易有封顶，如果不封顶，持卡人刷信用卡消费10万元，单笔手续费要500多元。 记者注意到，最近几天，不少汽车4S店通过朋友圈开始营销：信用卡改革，9月6日前信用卡购车，省千元手续费。 南京一家汽车4S店负责人告诉记者：目前已经有一部分4S店明确，客户负担多出的刷卡成本。对比看，9月6日前信用卡购车，刷卡手续费是0元。但是，

9月6日后信用卡购车，如果刷10万元，手续费600元;20万元则是1200元;50万元则需要支付3000元成本。 事实上，免手续费的概念早已深入人心，突然出现数百元刷卡付费成本，不少持卡人很难接受。不过，上述负责人指出，虽然成本增加了，但是，上不封顶提高了套现成本，可以防止一些非正常的套现客户。上述负责人指出：有些车商没有获得厂家授权，就无法贷款。但是，他们会刷卡套现支付车款，等于是变相贷款。 pos机刷卡手续费标准 新版手续费施行在即，银联5月正式开始对全国市场存量的商户进行重新入网的施行工作，原本MCC的几大类都被改名字了，MCC几大类别的概述变化，分为：标准类，优惠类，减免类，特殊计费类(或取消) 原一般类、餐娱类改称标准类：1.25%/1.28%-80封顶/0.78%/0.78%-26封顶 2017年9月6日实行的新版刷卡手续费取消了以上两类商户行业分类定价，对房产、汽车、批发行业不再实行贷记卡手续费封顶计费，并对标准类实行借贷分离收费，即 借记卡最低费率：发卡行0.35%(单笔13元封顶)+收单服务费(市场调节价) 贷记卡最低费率：发卡行0.45%+收单服务费(市场调节价) 按收单方需要有0.15以上%的运营成本计算，新版费率收单机构的成本大约为：借记卡费率：0.35%+0.15%=0.5%(单笔20元左右封顶) 贷记卡费率：0.45%+0.15%=0.6% 以上是成本，想要赚钱估计还要加0.1%左右。也就是标准收费商户刷信用卡的手续费在0.7%左右或以上!

POS机费率怎么算

我们都知道使用pos机刷卡消费或者转账是十分方便的，对于手续费也是有相关的规定，pos机手续费是根据行业的标准计算的，每个行业费率不同，因此相关的费率计算不同。 对于不同的POS类型还有行业不同相关的费率是不同的，具体的费率以及算法如下： 一、传统出小票POS机： 1、民生类：包括超市家电等，刷卡手续费是0.38%； 2、一般类：包括酒店宾馆等，刷卡手续费是1.25%； 3、批发类：包括家具批发家电批发等，手续费是26-80封顶。 4、标准类：信用卡刷卡0.55%~0.6%；储蓄卡刷卡0.5%，20元封顶。以刷卡一万为例，信用卡手续费55元~60元，储蓄卡手续费20元； 5、是优惠类：比加油站，手续费通常为信用卡0.38%；储蓄率0.4%，18元封顶； 6、减免类，比如公立医院、学校，手续费为0 费率0.38%是指刷卡手续费按照百分之0.38计算，比如刷100块钱手续费

是3角8分，刷1000块钱手续费是3元8角，以此类推。所以，刷卡费率等于刷卡金额乘以所属行业费率。 二、智能POS机 智能POS刷卡手续费和传统出小票POS的手续费是一样的。智能POS机和传统POS机相比，最大的优势就是支持主流二维码付款，收款更加便捷。二维码收款手续费一般为0.38%，交易一万，手续费38元。 三、手机POS机 手机POS机，顾名思义，就是通过手机蓝牙连接POS机刷卡消费的。一般手机POS机手续费不同支付公司会有一点区别，主流机器都是0.68%+3元/笔、0.69%+3/笔。手机POS机一般用于个人使用。 详细上述的内容可以帮助到您。如需了解更多请关注后续文章。投资有风险，加盟需谨慎。

比较PageRank算法和HITS算法的优缺点

题目：请比较PageRank算法和HITS算法的优缺点，除此之外，请再介绍2种用于搜索引擎检索结果的排序算法，并举例说明。 答： 1998年，Sergey Brin和Lawrence Page[1]提出了PageRank算法。该算法基于“从许多优质的网页链接过来的网页，必定还是优质网页”的回归关系，来判定网页的重要性。该算法认为从网页A导向网页B的链接可以看作是页面A对页面B的支持投票，根据这个投票数来判断页面的重要性。当然，不仅仅只看投票数，还要对投票的页面进行重要性分析，越是重要的页面所投票的评价也就越高。根据这样的分析，得到了高评价的重要页面会被给予较高的PageRank值，在检索结果内的名次也会提高。PageRank是基于对“使用复杂的算法而得到的链接构造”的分析，从而得出的各网页本身的特性。 HITS 算法是由康奈尔大学( Cornell University ) 的JonKleinberg 博士于1998 年首先提出。Kleinberg认为既然搜索是开始于用户的检索提问，那么每个页面的重要性也就依赖于用户的检索提问。他将用户检索提问分为如下三种：特指主题检索提问（specific queries，也称窄主题检索提问）、泛指主题检索提问（Broad-topic queries，也称宽主题检索提问）和相似网页检索提问（Similar-page queries）。HITS 算法专注于改善泛指主题检索的结果。 Kleinberg将网页（或网站）分为两类，即hubs和authorities，而且每个页面也有两个级别，即hubs（中心级别）和authorities（权威级别）。Authorities 是具有较高价值的网页，依赖于指向它的页面；hubs为指向较多authorities的网页，依赖于它指向的页面。HITS算法的目标就是通过迭代计算得到针对某个检索提问的排名最高的authority的网页。 通常HITS算法是作用在一定范围的，例如一个以程序开发为主题的网页，指向另一个以程序开发为主题的网页，则另一个网页的重要性就可能比较高，但是指向另一个购物类的网页则不一定。在限定范围之后根据网页的出度和入度建立一个矩阵，通过矩阵的迭代运算和定义收敛的阈值不断对两个向量authority 和hub值进行更新直至收敛。 从上面的分析可见，PageRank算法和HITS算法都是基于链接分析的搜索引擎排序算法，并且在算法中两者都利用了特征向量作为理论基础和收敛性依据。

pos机刷信用卡手续费怎么算-信用卡刷pos机手续费收取标准

pos机刷信用卡手续费怎么算|信用卡刷pos机手续费收取标准 【--出国祝福语】 如果你是消费者，pos机刷卡是不需要手续费的，pos 机刷信用卡手续费是扣商家的钱。 在全国各地所有POS机的商户，pos机刷卡消费都不收手续费，只要不出国，都不收手续费。无论是任何银行的任何卡。POS终端不向持卡人收取手续费。有收取的情况可拨打银联客服根据POS单的商户及商编投诉。 不同性质的商户，pos机刷卡时商家承担的手续费比例是不同的，这个主要看当时跟POS提供单位的签约情况来确定，通常是按照比例扣。

POS机刷信用卡卡手续费一般分为5类： 第一类商户含餐饮、宾馆、娱乐、珠宝金饰、工艺美术品类（一般扣率为2%-2.5%） 第二类商户含房地产、汽车销售、批发类（一般扣率为1%，可申请单比交易封顶） 第三类商户含航空售票、加油、超市类（一般扣率为0.5%-1%） 第四类商户含公立医院、公立学校（一般扣率视地区不同各有差异） 第五类商户含一般类（一般扣率为1%-3%）

如果你是消费者，pos机刷卡是不需要手续费的，pos 机刷信用卡手续费是扣商家的钱。 在全国各地所有POS机的商户，pos机刷卡消费都不收手续费，只要不出国，都不收手续费。无论是任何银行的任何卡。POS终端不向持卡人收取手续费。有收取的情况可拨打银联客服根据POS单的商户及商编投诉。 不同性质的商户，pos机刷卡时商家承担的手续费比例是不同的，这个主要看当时跟POS提供单位的签约情况来确定，通常是按照比例扣。 POS机刷信用卡卡手续费一般分为5类： 第一类商户含餐饮、宾馆、娱乐、珠宝金饰、工艺美术品类（一般扣率为2%-2.5%）

2017年各银行POS机刷卡手续费收取标准

2017年各银行POS机刷卡手续费收取标准 提交需求，马上获得5家保险公司报价银行卡刷pos机的时候会根据一定费率收取刷卡手续费，那么2017年各银行pos机刷卡手续费收取标准是多少呢?2016年9月6日开始，新的银行pos机手续费收取标准开始实施，对餐饮、百货行业来说，pos机刷卡手续费可降低2—4成，不过信用卡封顶优惠政策也取消了，这意味着消费者可能将承担更多的pos机手续费。下文将为您介绍各银行pos机刷卡手续费收取标准。各银行pos机刷卡手续费收取标准2016年9月6 日银行pos机手续费新收取标准落地后，各银行也调整了手续费收取标准，下表是我国几大银行pos机刷卡手续费收取标准。各行业pos机刷卡手续费收取标准提癌症后，餐饮、宾馆、娱乐、珠宝金饰、工艺美术品、房产汽车类一律按照1.25%的费率收取pos机刷卡手续费，批发、百货、中介、培训、景区门票等则按照0.78%收取pos机刷卡手续费，加油、超市类、交通运输售票、水电气缴费、政府类、便民类pos机刷卡按照0.38%的费率收取手续费。此外，公立医院、公立学校、慈善类刷卡不收取pos机刷卡手续费。下表是各行业pos机刷卡手续费收取标准。 大类细类商户类别码（MCC)适用范围手续费率（％）备注餐娱类餐饮、宾馆、娱乐、珠宝金饰、工艺

美术品类5094贵重珠宝、首饰，钟表零售1.25 5811包办伙食、宴会承包商1.25原手续费率0.8％,2013年2月25日停用。5812就餐场所和餐馆（包括快餐）1.25不包括营业面积100 平方米（含）以下的餐饮5813饮酒场所－酒吧、夜总会、茶馆、咖啡馆1.25 5932古玩店-销售、维修和修复服务1.25 5937古玩复制店1.25 5944银器商店1.25 5950玻璃器皿和水晶饰品店1.25 5970工艺美术商店1.25 5971艺术商和画廊1.25 7012分时使用的别墅或度假用房1.25 7011住宿服务1.25宾馆酒店餐饮部分可再独立申请5812或5813编号7032运动和娱乐露营地1.25 7631手表、钟表和首饰维修店1.25 7033活动房车及露营场所1.25 7829电影和录像创作发行1.25 7911歌舞厅、KTV1.25 7922戏剧制片（不含电影）、演出和票务1.25 7297洗浴、按摩服务1.25原手续费率0.8％,2013年2月25日停用。7298美容、SPA1.25原手续费率0.8％,2013年2月25日停用。7929未列入其他代码的乐队、文艺表演1.25

pos机的费率怎么计算

pos机的费率怎么计算 POS系统基本原理是先将商品资料创建于计算机文件内，透过计算机收银机联机架构，商品上之条码能透过收银设备上光学读取设备直接读入后（或由键盘直接输入代号）马上可以显示商品信息（单价，部门，折扣...）加速收银速度与正确性。每笔商品销售明细资料（售价，部门，时段，客层）自动记录下来，再由联机架构传回计算机。经由计算机计算处理即能生成各种销售统计分析信息当为经营管理依据。 POS机是通过读卡器读取银行卡上的持卡人磁条信息，由POS操作人员输入交易金额，持卡人输入个人识别信息（即密码），POS把这些信息通过银联中心，上送发卡银行系统，完成联机交易，给出成功与否的信息，并打印相应的票据。POS的应用实现了信用卡、借记卡等银行卡的联机消费，保证了交易的安全、快捷和准确，避免了手工查询黑名单和压单等繁杂劳动，提高了工作效率。 磁条卡模块的设计要求满足三磁道磁卡的需要，即此模块要能阅读1/2、2/3、1/2/3磁道的磁卡。 内部分析 通讯接口电路通常由RS232接口，PINPAD接口，IRDA接口和

RS485等接口电路组成。RS232接口通常为POS程序下载口，PINPAD接口通常为主机和密码键盘的接口，IRDA接口通常为手机和座机的红外通讯接口。接口信号通常都是由一个发送信号、一个接收信号和电源信号组成。 MODEM板由中央处理模块、存储器模块、MODEM模块、电话线接口组成。首先，POS会先检测/RING和/PHONE信号，以确定电话线上的电压是否可以使用，交换机返回可以拔号音，POS拔号，发送灯闪动，开始拔号，由通讯协议确定交换机和POS之间的信号握手确认等，之后才开始POS的数据交换，信号通过MODEM 电路收发信号；完成后挂断，结束该过程。

pagerank算法实验报告

PageRank算法实验报告 一、算法介绍 PageRank是Google专有的算法，用于衡量特定网页相对于搜索引擎索引中的其他网页而言的重要程度。它由Larry Page 和Sergey Brin在20世纪90年代后期发明。PageRank实现了将链接价值概念作为排名因素。 PageRank的核心思想有2点： 1.如果一个网页被很多其他网页链接到的话说明这个网页比较重要，也就是pagerank值会相对较高； 2.如果一个pagerank值很高的网页链接到一个其他的网页，那么被链接到的网页的pagerank值会相应地因此而提高。 若页面表示有向图的顶点，有向边表示链接，w(i,j)=1表示页面i存在指向页面j的超链接，否则w(i,j)=0。如果页面A存在指向其他页面的超链接，就将A 的PageRank的份额平均地分给其所指向的所有页面，一次类推。虽然PageRank 会一直传递，但总的来说PageRank的计算是收敛的。 实际应用中可以采用幂法来计算PageRank，假如总共有m个页面，计算如公式所示： r=A*x 其中A=d*P+（1-d)*(e*e'/m) r表示当前迭代后的PageRank，它是一个m行的列向量,x是所有页面的PageRank初始值。 P由有向图的邻接矩阵变化而来，P'为邻接矩阵的每个元素除以每行元素之和得到。 e是m行的元素都为1的列向量。 二、算法代码实现

三、心得体会 在完成算法的过程中，我有以下几点体会： 1、在动手实现的过程中，先将算法的思想和思路理解清楚，对于后续动手实现 有很大帮助。 2、在实现之前，对于每步要做什么要有概念，然后对于不会实现的部分代码先 查找相应的用法，在进行整体编写。 3、在实现算法后，在寻找数据验证算法的过程中比较困难。作为初学者，对于 数据量大的数据的处理存在难度，但数据量的数据很难寻找，所以难以进行实例分析。

pos机的费率怎么计算

刷卡服务的费用也因行业而异。一般行业为1％，餐饮/娱乐/珠宝/票务为2％，大型仓库超市/飞机票行业为0.5％，汽车/房地产为50元/笔，一些批发商品为20元/笔。 刷卡服务费按商家类型区分 百货商店约占1％，餐馆和酒店约占2％。批发，汽车和房地产按笔收费（从10至50不等）。慈善机构可能不收取服务费。如果您使用外国卡，银行将从商家处收取服务费。 异地刷卡服务费 在其他地方（包括香港，澳门和海外），民生卡和工银卡（所有银联卡）均不收取服务费。对于带有签证徽标的银行卡，香港将收取刷卡的服务费，称为货币转换费。因此，如果卡是双币种，则必须在国外使用银联卡，并且不收取服务费。 0.38是POS卡服务费的0.38％。扣除100元后，将扣除0.38元的服务费，扣除后实际收到的金额为99.62元。 根据国家发改委的通知，自2016年6月9日起，银行卡刷卡服务费将正式调整。

根据调整后的计划，卡的总体费率降低了53％至63％。 根据新准则，从2016年6月9日起，餐饮类别的税率从1.25％下降至0.6％。 百货公司刷卡服务费从0.78％下降到0.6％，而超市刷卡服务费从0.5％下降到0.38％。 公益仍然是0％。 房地产和汽车行业参照餐饮和娱乐业的收费标准，并且提高了单一费率。 扩展数据： 银行系统有两种类型的手续费： 1，贷款以外的利息统称为手续费。例如：信用卡分期还款产生的利息和逾期信用卡产生的利息； 2，在银行做生意的成本。

1.中国工商银行：卡手续费为5元，年费为10元/年，小于300元的小账户管理费为3元/季度，在同一个城市免费取款服务费，同城同业取款4元/笔；其他地方的信用卡存款免收手续费，取现金金额的1％，最低1元/笔，最高50元/笔 2，中国农业银行：卡手续费为5元，年费为10元/年，小于300元的小账户管理费为3元/季度，同一城市的银行间取款为2元/笔交易（广东省为4元）；收取信用卡取款金额的1％，最低为1元/笔，其他银行加收2元同业手续费 3）中国建设银行：卡手续费为5元，年费为10元/年，不足400元，小账户管理费为3元/季度，同城同业取款4元/交易+ 1％提款金额 4，中国银行：卡手续费为5元，年费为10元/年，不收取小额账户管理费，同城同业取款为4元/笔 5，交通银行：卡手续费5元，年费10元/年，小于500元的小账户管理费3元/季度，同一城市同业取款2元/交易

PageRank算法的核心思想

如何理解网页和网页之间的关系，特别是怎么从这些关系中提取网页中除文字以外的其他特性。这部分的一些核心算法曾是提高搜索引擎质量的重要推进力量。另外，我们这周要分享的算法也适用于其他能够把信息用结点与结点关系来表达的信息网络。 今天，我们先看一看用图来表达网页与网页之间的关系，并且计算网页重要性的经典算法：PageRank。 PageRank 的简要历史 时至今日，谢尔盖·布林（Sergey Brin）和拉里·佩奇（Larry Page）作为Google 这一雄厚科技帝国的创始人，已经耳熟能详。但在1995 年，他们两人还都是在斯坦福大学计算机系苦读的博士生。那个年代，互联网方兴未艾。雅虎作为信息时代的第一代巨人诞生了，布林和佩奇都希望能够创立属于自己的搜索引擎。1998 年夏天，两个人都暂时离开斯坦福大学的博士生项目，转而全职投入到Google 的研发工作中。他们把整个项目的一个总结发表在了1998 年的万维网国际会议上（WWW7，the seventh international conference on World Wide Web）（见参考文献[1]）。这是PageRank 算法的第一次完整表述。 PageRank 一经提出就在学术界引起了很大反响，各类变形以及对PageRank 的各种解释和分析层出不穷。在这之后很长的一段时间里，PageRank 几乎成了网页链接分析的代名词。给你推荐一篇参考文献[2]，作为进一步深入了解的阅读资料。

PageRank 的基本原理 我在这里先介绍一下PageRank 的最基本形式，这也是布林和佩奇最早发表PageRank 时的思路。 首先，我们来看一下每一个网页的周边结构。每一个网页都有一个“输出链接”（Outlink）的集合。这里，输出链接指的是从当前网页出发所指向的其他页面。比如，从页面A 有一个链接到页面B。那么B 就是A 的输出链接。根据这个定义，可以同样定义“输入链接”（Inlink），指的就是指向当前页面的其他页面。比如，页面C 指向页面A，那么C 就是A 的输入链接。 有了输入链接和输出链接的概念后，下面我们来定义一个页面的PageRank。我们假定每一个页面都有一个值，叫作PageRank，来衡量这个页面的重要程度。这个值是这么定义的，当前页面I 的PageRank 值，是I 的所有输入链接PageRank 值的加权和。 那么，权重是多少呢？对于I 的某一个输入链接J，假设其有N 个输出链接，那么这个权重就是N 分之一。也就是说，J 把自己的PageRank 的N 分之一分给I。从这个意义上来看，I 的PageRank，就是其所有输入链接把他们自身的PageRank 按照他们各自输出链接的比例分配给I。谁的输出链接多，谁分配的就少一些；反之，谁的输出链接少，谁分配的就多一些。这是一个非常形象直观的定义。

pos机费率一般是多少

现在，使用pos机进行刷卡支付的人群越来越多，因此商家为了满足消费者需求，纷纷办理大pos机设备，然而对于很多商户来说，并不了解具体的费率是多少。 从费改后，刷卡手续费不再像原先细分的区分商户类别，不再对房产汽车、批发行业实行贷记卡手续费封顶计费，同时对标准类实行借贷分离收费：借记卡最低费率：发卡行0.35%+收单服务费(市场调节价); 贷记卡最低费率：发卡行0.45%+收单服务费(市场调节价)。 按收单方需要有0.15%以上的运营成本计算，新版费率收单机构的成本约为： 借记卡费率：0.35%+0.15%=0.5%; 贷记卡费率：0.45%+0.15%=0.6%。 以上是成本，正常运营需加0.05%左右，标准类商户刷信用卡极速到账的手续费在0.6%-0.65%之间是合理的。

标准类属于正常商户，刷卡均有积分，根据表中各方收取的手续费比例，加上公司的运营成本等因素，有积分的商户综合一般在0.60%费率基础之上。根据不同需求选择POS 很多用户都喜欢用低费率的POS，觉得手续费低可以省很多手续费，实际上，自己使用什么样的POS要根据自身的刷卡需求来定。 有的卡友刷卡要求有积分，以完成刷卡任务免除年费;有的卡友想获取积分参加活动换取优惠礼品，有的则是想提高自己的消费品质和信用卡额度。对于此类用户，建议选用0.60%左右的费率POS，刷卡交易有积分，发卡行获得一定利益了才愿意帮你的卡提额，避免出现降额甚至封卡的情况。 对POS的使用需求较少的卡友来说，每个月时不时地用现金消费，那就对费率没有太大要求，可选择快捷、扫码等方式，只要资金安全到账，加上平时逛商场、吃饭、交水电费等，既是正常现象，又丰富了账单，反正没人要求你需在有积分的商户那里消费。

pos机的费率怎么计算

pos机的费率怎么计算 依据我国银联和人行规定，商户使用银联卡POS机需要支付如下费用： 一般商户类：0.78%（是指一般零售商店，服装店，理发店，小餐馆等等绝多数商户） 餐饮娱乐类：1.25% （是指洗浴按摩，酒店旅馆，珠宝黄金等等）民生行业类：0.38% （是指大型卖场超市，加油站，票务等等） 封顶费率：26至80元（是指批发市场，房产汽车销售等等） 上述收费依据来源为中国人民银行2012（263）号文件，自2013年2月25日执行。 POS(Pointofsales)的中文意思是"销售点"，全称为销售点情报管理系统，是一种配有条码或OCR码技术终端阅读器，有现金或易货额度出纳功能。其主要任务是对商品与媒体交易提供数据服务和管理功能，并进行非现金结算。 POS是一种多功能终端，把它安装在信用卡的特约商户和受理网点中与计算机联成网络，就能实现电子资金自动转账，它具有支持消费、预授权、余额查询和转帐等功能，使用起来安全、快捷、可靠。大宗交易中基本经营情报难以获取，导入POS系统主要是解决零售业信息管理盲点。连锁分店管理信息系统中的重要组成部分。

POS系统基本原理是先将商品资料创建于计算机文件内，透过计算机收银机联机架构，商品上之条码能透过收银设备上光学读取设备直接读入后(或由键盘直接输入代号)马上可以显示商品信息(单价，部门，折扣...)加速收银速度与正确性。每笔商品销售明细资料(售价，部门，时段，客层)自动记录下来，再由联机架构传回计算机。经由计算机计算处理即能生成各种销售统计分析信息当为经营管理依据。 POS机是通过读卡器读取银行卡上的持卡人磁条信息，由POS操作人员输入交易金额，持卡人输入个人识别信息(即密码)，POS把这些信息通过银联中心，上送发卡银行系统，完成联机交易，给出成功与否的信息，并打印相应的票据。POS的应用实现了信用卡、借记卡等银行卡的联机消费，保证了交易的安全、快捷和准确，避免了手工查询黑名单和压单等繁杂劳动，提高了工作效率。 磁条卡模块的设计要求满足三磁道磁卡的需要，即此模块要能阅读1/2、2/3、1/2/3磁道的磁卡。

2020年一清POS机手续费如何收取

手续费是大家在使用POS机时候需要进行支付的费用，而且是每个POS机刷卡机使用者必须进行支付的费用，本次就分享商用pos机刷卡机手续费的相关介绍，希望对大家使用商用POS 机刷卡机有所帮助。 各大支付机构的商用POS机的手续费是不尽相同的，也就是说大家在使用的时候手续费会根据商用POS机品牌的不同而有所不同。 下面与大家分享常见的商用POS机费率： 付临门POS机在刷信用卡时的费率是0.6%，正好是国家规定的费率收取标准。 汇付天下的POS机刷信用卡的费率分为两种情况，其中使用APP自主进件的贷记卡费率则为0.58%。平台商户进件的费率会根据情况在0.55%到0.65%之间进行调整。 嘉联立刷POS机的刷卡费率则为0.6%+3，而国家自九六费改后费率规定为0.6%。除此之外，嘉联立刷的POS机在使用的时候无押金不冻结，而且办理嘉联支付的VIP后，POS机的费

率为0.5%+3。 瑞银信POS机采用的是一机三费率的模式，具体为小额付款的费率为0.49%费率并且带积分模式。刷卡时可以自由选择切换；大额付款模式是35元封顶模式即35元手续费，手续费单笔最低收取0.1元；超级付款模式是1%费率模式即刷卡金额的1%。 当然值得一提的是如果大家在使用POS机刷信用卡的时候到账模式是秒到的还需要额外缴纳3元的手续费。但是无论机构对于POS机费率的规定为多少，其中都有0.515%部分是缴纳给发卡行和银联的，所以大家不要为了节省这笔费率而做出一些不会规定的事情。 江苏勇斌电子支付技术有限公司是一家专业从事pos机办理的服务商，公司秉承“以人为本、科学管理、真诚服务”的经营理念，以开发市场、创造市场、服务市场为战略目标，为不同行业的企业、公司以及个人办理对私以及对公POS机办理业务。

pos机的费率怎么计算

pos机: POS的中文意思是“销售点”，全称为销售点情报管理系统，是一种配有条码或OCR码技术终端阅读器，有现金或易货额度出纳功能。 POS是一种多功能终端，把它安装在信用卡的特约商户和受理网点中与计算机联成网络，就能实现电子资金自动转账，它具有支持消费、预授权、余额查询和转帐等功能，使用起来安全、快捷、可靠。大宗交易中基本经营情报难以获取，导入POS系统主要是解决零售业信息管理盲点。连锁分店管理信息系统中的重要组成部分。 原理解析 pos机基本原理 pos机基本原理 POS系统基本原理是先将商品资料创建于计算机文件内，透过计算机收银机联机架构，商品上之条码能透过收银设备上光学读取设备直接读入后（或由键盘直接输入代号）马上可以显示商品信息（单价，部门，折扣...）加速收银速度与正确性。每笔商品销售明细资料（售价，部门，时段，客层）自动记录下来，再由联机架构传回计算机。经由计算机计算处理即能生成各种销售统计分析信息当为经营管理依据。 POS机是通过读卡器读取银行卡上的持卡人磁条信息，由POS 操作人员输入交易金额，持卡人输入个人识别信息（即密码），POS 把这些信息通过银联中心，上送发卡银行系统，完成联机交易，给出

成功与否的信息，并打印相应的票据。POS的应用实现了信用卡、借记卡等银行卡的联机消费，保证了交易的安全、快捷和准确，避免了手工查询黑名单和压单等繁杂劳动，提高了工作效率。 磁条卡模块的设计要求满足三磁道磁卡的需要，即此模块要能阅读1/2、2/3、1/2/3磁道的磁卡。 通讯接口电路通常由RS232接口，PINPAD接口，IRDA接口和RS485等接口电路组成。RS232接口通常为POS程序下载口，PINPAD接口通常为主机和密码键盘的接口，IRDA接口通常为手机和座机的红外通讯接口。接口信号通常都是由一个发送信号、一个接收信号和电源信号组成。 MODEM板由中央处理模块、存储器模块、MODEM模块、电话线接口组成。首先，POS会先检测/RING和/PHONE信号，以确定电话线上的电压是否可以使用，交换机返回可以拔号音，POS拔号，发送灯闪动，开始拔号，由通讯协议确定交换机和POS之间的信号握手确认等，之后才开始POS的数据交换，信号通过MODEM 电路收发信号；完成后挂断，结束该过程。 pos机的费率怎么计算: 刷卡手续费是按行业不同点位也是不同的，一般的行业都是1%，餐饮/娱乐/珠宝/票务都是2%，大型仓储超市/机票行业0.5%，汽车/房地产50元/笔，一些批发类20元/笔。

2018pos机刷卡手续费标准

---------------------------------------------------------------范文最新推荐------------------------------------------------------ 2018pos机刷卡手续费标准 目前国内刷卡费率与行业分类挂钩，餐娱类的刷卡手续费率最高，为1.25%;百货等一般商户为0.78%;超市、加油站等为0.38%;医院、教育等公益类则是零费率。 根据新规，对发卡行服务费实行不再区分商户类别。也就是说，商户行业分类定价取消，总体上大幅降低了刷卡的费率水平。 新政显示，发卡银行服务费费率水平降低为借记卡交易不超过交易金额的0.35%，贷记卡交易不超过0.45%。 银行卡清算机构收取的网络服务费费率水平降低为不超过交易金额的0.065%，由发卡机构、收单机构各承担50%。从类别上看，餐饮类企业刷卡手续费支出可降低53%63%，百货等行业商户可降低23%39%。新政实行后，医院、教育等公益类刷卡仍为零费率。 信用卡大额消费成本高 不过，信用卡刷卡手续费取消封顶后，一旦遇到大额消费，商户费用成本必将增加，这部分成本由谁买单。 此前信用卡交易虽然费率高，但单笔交易有封顶，如果不封顶，持卡人刷信用卡消费10万元，单笔手续费要500多元。 记者注意到，最近几天，不少汽车4S店通过朋友圈开始营销：信用卡改革，9月6日前信用卡购车，省千元手续费。 南京一家汽车4S店负责人告诉记者：目前已经有一部分4S店明确， 1 / 24

客户负担多出的刷卡成本。对比看，9月6日前信用卡购车，刷卡手续费是0元。但是，9月6日后信用卡购车，如果刷10万元，手续费600元;20万元则是1200元;50万元则需要支付3000元成本。 事实上，免手续费的概念早已深入人心，突然出现数百元刷卡付费成本，不少持卡人很难接受。不过，上述负责人指出，虽然成本增加了，但是，上不封顶提高了套现成本，可以防止一些非正常的套现客户。上述负责人指出：有些车商没有获得厂家授权，就无法贷款。但是，他们会刷卡套现支付车款，等于是变相贷款。 pos机刷卡手续费标准 新版手续费施行在即，银联5月正式开始对全国市场存量的商户进行重新入网的施行工作，原本MCC的几大类都被改名字了，MCC几大类别的概述变化，分为：标准类，优惠类，减免类，特殊计费类(或取消) 原一般类、餐娱类改称标准类：1.25%/1.28%-80封顶/0.78%/0.78%-26封顶 2018年9月6日实行的新版刷卡手续费取消了以上两类商户行业分类定价，对房产、汽车、批发行业不再实行贷记卡手续费封顶计费，并对标准类实行借贷分离收费，即 借记卡最低费率：发卡行0.35%(单笔13元封顶)+收单服务费(市场调节价) 贷记卡最低费率：发卡行0.45%+收单服务费(市场调节价) 按收单方需要有0.15以上%的运营成本计算，新版费率收单机构的

pos机的费率怎么计算

pos机的费率怎么计算 使用标准费率0.6%以上较为合理，低于0.6%的反而要进行辨识。 一、传统POS机：0.6%以上（市场上的流行费率） 二、手机POS机：0.6%以上（0.6%以下100%商户有问题） 有很多同行问在96费改政策新的标准后我们的费率到底多少，为什么商户看电视报纸说“国家降费率0.35%、0.45%了你们却收我们0.65%、0.7%”。 那么，刷卡手续费调整为多少?银联刷卡手续费标准怎么算?新的手续费收费有哪些标准?接下来跟你一一道来。 先说说以前，在费改以前，国内刷卡费率与行业分类挂钩，餐娱类的刷卡手续费率最高，为1.25%;百货等一般商户为0.78%;超市、加油站等为0.38%;医院、教育等公益类则是零费率。 根据96费改后的要求，对发卡行服务费实行不再区分商户类别。也就是说，商户行业分类定价取消，总体上大幅降低了刷卡的费率水平。 新政显示，发卡银行服务费费率水平降低为借记卡交易不超过交易金

额的0.35%，贷记卡交易不超过0.45%。但是商户朋友们你别就看定了，得往下看，银联和收单公司的还没算上呢。 银行卡清算机构收取的网络服务费费率水平降低为不超过交易金额的0.065%，由发卡机构、收单机构各承担50%。从类别上看，餐饮类企业刷卡手续费支出可降低53%—63%，百货等行业商户可降低23%—39%。新政实行后，医院、教育等公益类刷卡仍为零费率。 POS机刷卡手续费，信用卡大额消费成本高。 不过，信用卡刷卡手续费取消封顶后，一旦遇到大额消费，商户费用成本必将增加，这部分成本由谁买单。 此前信用卡交易虽然费率高，但单笔交易有封顶，如果不封顶，持卡人刷信用卡消费10万元，单笔手续费要500多元。 一家汽车4S店负责人告诉我：“目前已经有一部分4S店明确，客户负担多出的刷卡成本。”对比看，9月6日前信用卡购车，刷卡手续费是0元。但是，9月6日后信用卡购车，如果刷10万元，手续费600元;20万元则是1200元;50万元则需要支付3000元成本。 事实上，免手续费的概念早已深入人心，突然出现数百元刷卡付费成

pos机的费率怎么计算

刷卡手续费： 商店接受客户刷卡后，需支付百分之二至三的手续费给银行和信用卡中心。称为刷卡手续费。2012年11月央行下发《中国人民银行关于切实做好银行卡刷卡手续费标准调整实施工作的通知》称，此次手续费下调仅涉及境内银行卡的消费交易，2003年施行的商户刷卡手续费规定自2013年2月25日起同时废止，此次刷卡费率总体下调幅度在23%至24%。 特约商店接受客户刷卡后，需支付百分之二至三的手续费给银行和信用卡中心。有些厂商为节省成本，会要求持卡人另外支付手续费，因为刷卡就必需开立发票，使商店无法逃税。 根据现行的《中国银联入网机构银行卡跨行交易收益分配办法》，银行卡收单业务的结算手续费全部由商户承担，但不同行业所实行的费率不同，费率标准从0.5%到4%不等。一般来说，零售业的刷卡手续费率在0.8%-1%，超市是0.5%，餐饮业为2%。 商家转嫁： 案例举证 2010年08月29日上午，市民陈先生反映，他在大利嘉城刷卡购物时被商家收取手续费。对此，银联工作人员指出，商家的做法违反了相关合同协议，消费者可向银联公司举报。 陈先生告诉记者，他去大利嘉城买相机，没带现金只带了银行卡，老板说如果刷卡需要他支付20块钱的手续费。平时他在永辉超市、

沃尔玛以及国美电器等商场超市也是刷卡购物，可商家从没向他收取任何手续费。 随后，记者在大利嘉城调查了15家安装了POS刷卡机的商家。其中，14家商家表示遇到消费者刷卡消费时，会要求持卡人交手续费。“手续费不是商家要的，而是直接从卡里转给了银行。”大利嘉城一家专营笔记本电脑的郑老板称，对消费者来说，刷卡购物方便快捷，还有积分累积等优惠，但对商家来说，消费者每使用POS机刷卡消费一次，商家需支付给银行消费数额1%的手续费。因此，大利嘉城大部分商家都会向消费者收取1%~2%的刷卡手续费，将POS 机的交易费用和成本“转嫁”到消费者头上. 银联解答 “商家把刷卡手续费转嫁给消费者，这是一种违规行为。”中国银联客服工作人员告诉记者，在全国范围内，银联卡刷卡消费，不分异地本地都不向持卡人收取任何手续费。刷卡消费后结算方是银行和商家，按照当初安装刷卡机时的相关协议，应是商家向银行支付一定比例的手续费。因此，如果刷卡消费时被要求加收手续费，持卡人有权拒绝，并可向中国银联客服热线投诉.