In 1993 Eugene Fama, often referred to as the “father of mod­ern port­fo­lio the­ory,” along with his col­league Ken French of Dart­mouth, pub­lished “Com­mon risk fac­tors in the returns on stocks and bonds.” By test­ing his­tor­i­cal stock and bond prices going back to 1926, they con­cluded that smaller com­pa­nies and cheaper stocks out­per­formed the mar­ket, albeit with extra risk. Port­fo­lios that have an excess of small com­pany stocks and ‘cheap’ value stocks (mea­sured by a high ratio of book value to mar­ket value) sig­nif­i­cantly out­per­form less con­cen­trated port­fo­lios. These find­ings have been widely accepted by both aca­d­e­mics and pro­fes­sional investors. Such effects have been observed for nearly a cen­tury and in mar­kets around the world, but do not occur in every time period. You can see the small and value effect for the US mar­ket in the chart on the right.

Notice the dra­matic increase in annu­al­ized return when mov­ing from growth to neu­tral to value stocks and as stocks decrease from large to small. While this visual rep­re­sen­ta­tion shows the small/value effect, a math­e­mat­i­cal model cre­ated by Fama and French pre­dicts — with very high accu­racy — port­fo­lio per­for­mance based on the small and value ‘weight­ing’ of a port­fo­lio. Aca­d­e­mics con­sider this new model a sig­nif­i­cant improve­ment over an ear­lier Nobel Prize win­ning model (the Cap­i­tal Asset Pric­ing Model).

3 factor has ana­lyzed a num­ber of indus­try and aca­d­e­mic experts’ port­fo­lios. Some of the allo­ca­tions we reviewed were designed with a ‘tile’ (or over­weight) toward small-value. Oth­ers’ allo­ca­tions, for exam­ple, Malkiel and Swensen, allo­ca­tions are described in broad brush­strokes — just the high­est level asset classes, such as US, Devel­op­ment, REIT, etc. – with no small-value over­weight­ing.. It is not clear where these two acclaimed experts sit with regard to “tilted” port­fo­lios, but it might be inter­est­ing to explore what would hap­pen if we take their over­all, top level allo­ca­tions and over­weight them with small and value indexes:

Better Returns Through Research

Sign Up for Shop Talk Newsletter

When we over-weighted the Swensen and Malkiel port­fo­lios (“tilted” them with small and value equi­ties), for each of the time peri­ods (1, 3, 5 and 10 years) their per­for­mance increased over the time frame given. Does this mean you should “tilt” your port­fo­lio? As we dis­cussed above, not nec­es­sar­ily; it depends on your will­ing­ness to accept risk. Nonethe­less, this was an inter­est­ing experiment.

Bob Lux­en­berg, March 2010

The views and infor­ma­tion con­tained within this arti­cle are pro­vided for infor­ma­tional pur­poses only and are not meant as invest­ment advice. They rep­re­sent the author’s cur­rent good-faith views at pub­li­ca­tion time and are sub­ject to change with­out notice. As with any strat­egy, it is impor­tant for an investor to fully under­stand the strat­egy prior to invest­ing; seek­ing advice from a pro­fes­sional advi­sor can be a good place to start.

The infor­ma­tion that is pro­vided on herein has been com­piled to the best of our abil­ity. How­ever, the authors makes no war­ranty of any kind, expressed or implied, and will not be held respon­si­ble, or liable for errors, or omis­sions result­ing in any loss or dam­age caused or alleged to be caused, directly, or indi­rectly, by infor­ma­tion con­tained in 3 factor’s publications.

Links to other web sites are pro­vided as gen­eral infor­ma­tion sources for the reader. 3 factor has not for­mally eval­u­ated the infor­ma­tion pro­vided via these sites and inclu­sion of these links does not con­sti­tute an endorse­ment of any orga­ni­za­tion. The links pro­vided are main­tained by their respec­tive orga­ni­za­tions and they are solely respon­si­ble for their con­tent. Trade­marks are the prop­erty of their respec­tive owners.

3 factor has con­ducted exten­sive analy­sis con­cern­ing port­fo­lio per­for­mance. See “impor­tant dis­clo­sure” for details and dis­claimers regard­ing our state­ments con­cern­ing per­for­mance, and the var­i­ous assump­tions we have made in our analy­sis.