Another criticism of asset allocation is that it ignores contemporary market valuations, a complaint integrally related to the historical data issue. Although admittedly there is no guarantee of accuracy, if a practitioner does not use historical projections but rather develops his or her own input for the optimization process using forward-looking expectations, the assumptions will factor in market valuations. In fact, even after the practitioner incorporates valuation-based, forward-looking expectations in the analytical process, he will still need to then use judgment that takes into account other factors, as Markowitz suggests.

The third criticism of MPT is that allocations are determined at the beginning of the investment process and never changed, except when they are rebalanced. Again, this criticism should be leveled at those practitioners setting their policies in stone. There is nothing in the literature or in practice to suggest that a policy allocation should not be revisited and revised when and if forward-looking market expectations change. I can't speak for other planners; however, it is our practice to review our assumptions at least annually, and our "strategic" allocations do in fact vary over time as a result of changes in our world view.

The bottom line is that practitioners may develop allocation models based solely on projections of historical data, but MPT does not. Practitioners may also ignore valuations; MPT does not. And practitioners may design allocation models and set them in stone; again, MPT does not.

The Alternatives
I only know of three alternatives to MPT-market timing, absolute return and tactical allocation.

As for market timing, I can make this argument short and sweet. Name the ten most successful market timers of all time. How about the top five? The top one? I agree that if an advisor could consistently predict which markets would be up and which down, he would have to be pretty foolish to diversify. Why on earth would I invest in stocks if I knew the market was headed down? Had I posed my challenge back in the '80s, many would have pointed to Joe Granville.

What, you haven't heard of Joe Granville? Until the late '80s, he was the market guru. Like many gurus, he had absolute confidence in his crystal ball. According to Robert Shiller in the book Irrational Exuberance, Granville was quoted by Time magazine as saying, "I don't think that I will ever make a serious mistake in the stock market for the rest of my life," and he predicted that he would win the Nobel economics prize. In 1981, when he was grossing $6 million a year for his newsletter advice, his two-word "sell everything" warning to his subscribers triggered a massive market sell-off with a record number of shares trading around the globe.

Just before the 1987 crash, he again warned of a market disaster. He was obviously correct on that call and his picture was on the cover of major magazines and papers around the world.

Maybe you haven't heard of him because his crystal ball, like others, had flaws. A few years ago, the Hulbert Financial Digest reported that The Granville Market Letter "is at the bottom of the rankings for performance over the past 25 years-having produced average losses of more than 20% per year on an annualized basis."

Until someone can name at least a few successful long-time market timers, I remain a skeptic and will continue to "bet" on some form of diversification.

Another alternative to modern portfolio theory is absolute return. If an advisor can manage to develop an entire portfolio of investments that provide an absolute return high enough to accomplish the investor's long-term goals, he might reasonably forgo worrying about MPT, allocation and diversification. For those who cannot locate the appropriate investments for such a portfolio, incorporating absolute return strategies into a strategic or tactical model may allow them to have the best of both worlds.