Is your portfolio entirely made up of small value stocks? You seem quite confident that they are the bee's knees.
I think Taylor was just making the point that SV is popular now because it has been doing well. When SV does poorly, it will be less popular.
Actually, what Taylor said was that the popularity (read: justification) for a small value tilt was based on recency and the last few years of outstanding returns.
My point was, the justification for tilting away from the market is actually based on over 8 decades of market research, a rational distribution of risk/return, and the prevailing asset pricing model in use today.
Not everyone will or even should tilt away from the market. I am fine with that. I think everyone, however, should at least consider it. Better to have reviewed the data and theories and decide against it than to say "its too complicated or complex", or, "i am intentially limiting my investment portfolio to X holdings". If you can figure out Int'l Investing and REIT investing, you can probably understand multifactor investing.
As my data showed, the tracking error from Int'l and REIT investing is probably greater than tilting within the dimensions of our domestic market, and the expected payoffs are much less.
As Michael has pointed out and I mentioned as well...all investment strategies will become unpopular when they "underperform". That is a lousy reason for excluding them from an investment portfolio.
If you embrace the theory/model that drives the expected equity risk premium (and thus an expected outperformance of US, Int'l, and EM Market portfolios), there is no reason to believe the same risk/return behaviors will not also generate positive excess returns to size and value dimensions globally
I have no particular insight as to the future premiums for small and value dimensions of the market any more than I do for the overall equity risk premium -- so my personal portfolio is highly diversified. As is is usually beneficial with most predicitions, I default to the naive forecast. Moderate equity, size, value, term and credit risks will probably be sufficient for future planning purposes. If one or more is way off (beta, for example), I would be much more comfortable not
concentrating my entire portfolio in that risk...
What prompted any comment from me in the first place is the assumption that US REITS are a CORE element of a diversified portfolio, yet size and value tilts are secondary (at best). There is credible research indicating that Int'l investing may be unnecessary if you target BETA exclusively (Sinquefield 1995), and only for those considering size or value dimensions of the Int'l markets are the higher investment costs justified.
As for my multifactor confidence, well...size and Value dimensions have brilliant explainatory powers in the context of a diversified portfoliio, and we have 80+ years of exchange traded data on their behavior through booms and busts (and info in over a dozen different countries over shorter periods). Size and Value dimensions can also be accessed in a more tax efficient manner, leaving valuable tax-deferred space for ST taxable fixed income. And, asset class or not, REITS contain industry specific risks (see 2007) that mirror most other sectors.
If one were to develop a REIT tracking variable (NARIET minus TSM), we would find that the premiums, persistence, and diversification benefits relative to a TSM portfolio are actually less than size or value tilts...yet the volatility of its premium is higher than the ERP, size, or value risks.
Furthermore, a REIT allocation introduces a portfolio to the sensitivity of the value premium (REIT minus TSM and Value minus Growth have an annual correlation of almost +0.7%, and poor
periods for Value minus Growth tend to coincide with very poor
relative periods for REITS). A logical question is whether or not you would prefer to take this "risk" in a 100 securitiy portfolio with industry specific risks, or a several thousand LV/MV/SV portfolio spanning over a dozen sectors?
Bottom line, there are as many reasons to develop a multifactor portfolio as their are a CAPM allocation. Completely ignoring this is the issue I have. Relegating certain factors in the FF5F Model to primary and secondary is not how I have interpreted the FF research. Adding REITs is not the diversification antidote some would have you believe. I understand that only using asset classes that Vanguard offers presents a bias for some, but its not a sufficient metric for building a long term portfolio for those of even modest net worths...