THE JOURNAL OF FINANCE*VOL. XLVII, NO. 2*JUNE 1992
EUGENE F. FAMA and KENNETH R. FRENCH*
Twoeasilymeasuredvariables,size and book-to-marketequity,combinetocapturethecross-sectionalvariationinaveragestock returns associated with market3,size, leverage, book-to-market equity, and earnings-price ratios. Moreover, whenthetests allow for variationin3that is unrelated tosize,the relation between market/3and average return is flat, even when3is the only explanatory variable.
THEASSET-PRICINGMODELOF Sharpe (1964),Lintner(1965),and Black(1972)
has long shaped the way academics and practitionersthinkabout averagereturns andrisk. Thecentral predictionof themodelisthat the marketportfolioofinvested wealthismean-varianceefficientinthe sense ofMarkowitz (1959).Theefficiency of the market portfolio implies that (a)expectedreturns on securitiesare a positive linearfunctionoftheir marketO3sthe slopeintheregression of a security'sreturn on the market'sreturn),and(b)marketO3suffice to describe the cross-section ofexpectedreturns.There are several empirical contradictions of the Sharpe-Lintner-Black(SLB)model. The mostprominentis the size effect ofBanz(1981).Hefindsthat marketequity,ME(astock'spricetimes sharesoutstanding),adds totheexplanationofthecross-section ofaveragereturnsprovided by
marketOs.Averagereturns on small(low ME)stocks are toohigh giventheirfestimates,andaveragereturnson large stocksaretoo low.
Another contradiction of the SLB modelis thepositiverelation betweenleverageandaveragereturndocumentedbyBhandari(1988).It isplausiblethat leverageisassociated with risk and expected return,but inthe SLBmodel, leverageriskshould becaptured bymarketS.Bhandarifinds,how-ever,thatleverage helps explainthecross-section of average stockreturnsintests that includesize(ME)as well asA.
Stattman (1980) and Rosenberg, Reid, and Lanstein (1985)findthat aver-agereturns onU.S. stocks arepositivelyrelated to the ratio of afirm'sbookvalueofcommonequity, BE,to itsmarket value,ME.Chan, Hamao, andLakonishok(1991)findthat book-to-market equity, BE/ME, also has a strongroleinexplaining the cross-section of average returns on Japanese stocks.
Graduate School ofBusiness, University of Chicago, 1101 East 58th Street, Chicago, IL60637. Weacknowledgethehelpful comments of David Booth, Nai-fu Chen, George Constan-tinides, Wayne Ferson, Edward George, Campbell Harvey, Josef Lakonishok, Rex Sinquefield,Rene Stulz,MarkZmijeweski,andananonymousreferee. This researchissupported bytheNational Science Foundation(Fama) and the Center for ResearchinSecurity Prices (French).