Is asset allocation dead?

The Wall Street Journal had an interesting article last Friday: Failure of a Fail-Safe Strategy Sends Investors Scrambling

It touches upon a very real issue that all the advisors I talk to are thinking about, yet no one has come up with a good answer to solve the dilemma: asset allocation doesn't work. In this latest financial crisis, all assets and even commodities were highly correlated, and thus there was no "safe" place to put your assets to mitigate the losses in the portfolio.

In the article, they spoke to Mohamed El-Erian, co-chief investment officer at Pimco, who noted the "breakdown of perceived relationships between asset classes." In our global economy, this is inevitable. Take a look at Boeing-what asset class do you put this firm in? Over 12 countries and 22 different companies provide the parts and labor to build a plane.

The theory of asset allocation is still a good one-diversify your (lowly correlated) assets to smooth the volatility of a portfolio. The problem is finding those lowly correlated assets. Do you diversify by sector? By frontier markets vs developed nations (since emerging markets no longer provide diversification)? Do you take down all geopolitical barriers and simply put the smartest people out there to try and find the best companies regardless of their size, location or industry? Or is it a combination of the above?

I believe that the traditional asset classes can stay in place-as long as there is overlay for sector, geography, political ramifications, etc. This will make asset allocation more complex, but not terribly so.

More importantly, investors need to remember what their assets are for--if they don't need the money right now, they should remember the long term picture. Six years after the 1929 market crash, investors made their money back. It may take longer this time, but in the end, it's a lot better than the "under the mattress plan". For investors who are very near or in retirement, it's another story. I'll post on that tomorrow.