Wab,
I take your point about market cap being "The Market" but the Splitter approach is looking for a way to dampen volatility (for a given return) or increase yield (for a given volatility) by being in minimally-correlated asset classes. The theory is that a balanced blend of these assets will in fact be the optimum allocation.
Now to quibble with the cap-weighted 'Lumper's' premise a bit more: We've moved our definition of "The Market" from a purely US-based stocks and bonds to an international one -- that is good. But if we are really trying to own "All Assets in the proportion of their value" then we'd need to go a lot further than publicly traded stocks and bonds (although perhaps not all the way to the lemonade stand!).
In this way, we would pick up the vast amount of value in real esate, private companies, commodities, precious metals and so forth.
When we do this, we'd be well on our way to the Coffee-House or Bernstein "Gap" portfolios, though the equity proportions of those portfolios might still be overweighted to small and value.
So your point about overweighting Small and Value as being a bet that we are 'smarter than the market' is right on target -- at this point, I am going on the basis of the Portfolio Optimizer studies which show how these portfolios have behaved over time -- the virtues of their low correlations -- and that has given me the confidence to deviate from pure market weights. The optimizer studies are based on the long run behaviors of the asset classes interacting in a portfolio, and that an asset class is not always represented by a common cap-weight index, nor is the final mix necessarily the cap-weight of the traded securities market.
I'm comfortable with it; it makes sense to me, but as you point out it is different from the Lumper's Efficient Market Theory approach of following the standard cap-weighted indexes. Thus this should serve as a warning label to any Splitters: the implications need to be understood and accepted before one should become a 'splitter'.
ESRBob
I take your point about market cap being "The Market" but the Splitter approach is looking for a way to dampen volatility (for a given return) or increase yield (for a given volatility) by being in minimally-correlated asset classes. The theory is that a balanced blend of these assets will in fact be the optimum allocation.
Now to quibble with the cap-weighted 'Lumper's' premise a bit more: We've moved our definition of "The Market" from a purely US-based stocks and bonds to an international one -- that is good. But if we are really trying to own "All Assets in the proportion of their value" then we'd need to go a lot further than publicly traded stocks and bonds (although perhaps not all the way to the lemonade stand!).
In this way, we would pick up the vast amount of value in real esate, private companies, commodities, precious metals and so forth.
When we do this, we'd be well on our way to the Coffee-House or Bernstein "Gap" portfolios, though the equity proportions of those portfolios might still be overweighted to small and value.
So your point about overweighting Small and Value as being a bet that we are 'smarter than the market' is right on target -- at this point, I am going on the basis of the Portfolio Optimizer studies which show how these portfolios have behaved over time -- the virtues of their low correlations -- and that has given me the confidence to deviate from pure market weights. The optimizer studies are based on the long run behaviors of the asset classes interacting in a portfolio, and that an asset class is not always represented by a common cap-weight index, nor is the final mix necessarily the cap-weight of the traded securities market.
I'm comfortable with it; it makes sense to me, but as you point out it is different from the Lumper's Efficient Market Theory approach of following the standard cap-weighted indexes. Thus this should serve as a warning label to any Splitters: the implications need to be understood and accepted before one should become a 'splitter'.
ESRBob
