An "asset class" is defined as a specific category of related finan­cial instruments, such as stocks, bonds or cash. These three catego­ries are often considered the primary asset classes, although people often include real estate as well.

Institu­tions and other financial profes­sionals advocate the need for portfolios to be properly diversified across multiple asset classes in order to di­versify risk. Decades of academic research support the view that the se­lection of each asset class in a portfolio is far more important than the selection of the individual stocks or other posi­tions.

Why Diversifying Across Asset Classes Can Actually Be Risky

However, there's a single and significant flaw in their research findings. It's the same fault inherent in asset classes in gen­eral. Asset classes are intentionally self-limiting, and their use is incapable of creating true portfolio diversification for two primary reasons:

  • By classifying asset classes as being long-only in re­lated markets, the various asset classes expose peo­ple to the same "re­turn drivers." As a result, there is a risk that the failure of a single return driver will negatively impact multiple asset classes.
  • Limiting your diversification opportunities to asset classes eliminates numerous trading strategies that, be­cause they are powered by entirely separate re­turn drivers, can pro­vide tremendous diversification value to your portfolio.

When looking at a portfolio diversi­fied across asset classes, it is clear that most of those asset classes are dependent on the same return drivers or baseline conditions for producing their returns. This unnecessarily exposes the portfo­lio to "event" risk, meaning that a single event, if it is the wrong one, can negatively affect the entire portfolio.

The figure below displays the performance of the components of a portfolio based on conventional wisdom during the bear market of 2007-2009. This portfolio includes 10 asset classes "diversified" across stocks, bonds and real estate, in both the United States and internationally. According to conventional wisdom, this portfolio is considered highly diversified. Of the 10 asset classes, however, only U.S. and international bonds avoided losses. All other asset classes declined sharply. A portfolio allocated equally to each asset class declined by more than 40 percent over a 16-month period. It is obvious that conventional investment wisdom failed.

Portfolios constructed around asset classes are unnecessarily risky, and taking on unnecessary risk is the equivalent of gambling with your money.

Diversification Across Return Drivers and Trading Strategies

True portfolio diversification can only be achieved by diversifying across return drivers and trading strategies, not asset classes. This process starts with first identifying and understanding the necessary baseline conditions and return drivers that underlie the performance of each trading strategy.

A "trad­ing strategy" is made up of two components: a system that exploits a return driver and a market that is best suited to capture the returns promised by the return driver.

Return Driver (system) + Market = Trading Strategy

Trading strategies are then combined to create a balanced and diversified investment portfolio.

The asset classes contained in the portfolios of most investors are simply a restricted subset of the potentially hundreds (or more) of combinations of return drivers and markets avail­able to be incorporated into a portfolio.

For example, the U.S. equity asset class is actually the strategy comprised of the system of buying into long positions in the market "U.S. equities."

How Risk Is Defined by Return Drivers

Risk is typically defined in portfolio theory as the standard deviation (volatility) of returns. Without an understanding of the return drivers underlying the strategies within a portfolio, however, this typical definition of risk is inadequate. Sound, rational return drivers are the key to any successful trading strategy. Risk can only be determined by an understanding and evaluation of these return drivers. History is full of examples of seemingly low-risk investments (characterized by consistent monthly returns with low volatility) that suddenly became worthless because the investment was not based on a sound return driver.

Therefore, risk is not determined by the volatility of returns. In fact, highly volatile trading strategies, if based on a sound, logical return driver, can be a "safe" contributor to a portfolio.

So, if the volatility of returns is not an acceptable definition of risk, what is a more appropriate definition? The answer is drawdowns. Drawdowns are the greatest impediment to high returns and the true measure of risk. It is far easier to lose money than it is to recover from those losses. For example, to recover from an 80 percent drawdown requires four times the effort compared to a 50 percent drawdown. A risk management plan must specifically address the destructive power of drawdowns.

Volatility and Correlation

Although volatility is not the true measure of risk (since volatility does not adequately describe the underlying return driver), volatility still makes a significant contribution to the power of portfolio diversification.

Portfolio performance measurement ratios, such as the Sharpe ratio, are typically expressed as the relationship between risk-adjusted returns and portfolio risk. Therefore, when volatility decreases, the Sharpe ratio increases.

Once we have developed trading strategies based on sound return drivers, true portfolio diversification then involves the process of combining seemingly riskier individual positions into a safer diversified portfolio. How does the combination of positions improve the performance of a portfolio? The answer is based on correlation, which is defined as "a statistical measure of how two securities move in relation to each other."

How is this helpful? If returns are negatively correlated with each other, when one return stream is losing, another return stream is likely winning. Therefore, diversification reduces overall volatility. The volatility of the combined return stream is lower than the volatility of the individual return streams.

This diversification of return streams reduces portfolio volatility and since volatility is the denominator of the Sharpe ratio, as volatility decreases, the Sharpe Ratio increases. In fact, in the extreme case of perfect negative correlation, the Sharpe ratio goes to infinity! Therefore, the goal of true portfolio diversification is to combine strategies (based on sound return drivers), which are non-correlated or (even better) negatively correlated.

The Bottom Line

Most investors are taught to build a portfolio based on asset classes (usually limited to stocks, bonds and possibly real estate) and to hold these positions for the long term. This approach is not only risky, but it is the equivalent of gambling. Portfolios must include strategies based on sound, logical return drivers. The most consistent and persistent investment returns across a variety of market environments are best achieved by combining multiple uncorrelated trading strategies each designed to profit from a logical, distinct "return driver" into a truly diversified investment portfolio.

True portfolio diversification provides the highest returns over time. A truly diversified portfolio will provide you with greater returns and less risk than a portfolio diversified only across conventional asset classes. In addition, the predictability of future performance can be increased by expanding the number of diverse return drivers employed in a portfolio.

The benefits of portfolio diversification are real and pro­vide serious tangible results. By sim­ply learning how to identify additional return drivers, you will be able to shift from gambling your portfolio on only one of them to be­coming an "investor" by diversifying across many of them.