There are several things that investors do to protect their portfolios against risk. One significant way to protect one's portfolio is by diversifying. In short, this means an investor opts to include various types of securities and investments from different issuers and industries. The idea here is the same as the old adage “don’t put your eggs all in one basket.” When you are invested in many areas, if one fails, the rest will ensure the portfolio as a whole remains secure. This added security can be measured in the increased profits that a diversified portfolio tends to bring in when compared to an individual investment of the same size.
Diversification is a great strategy for anyone looking to reduce risk on their investment for the long term. The process of diversification includes investing in more than one type of asset. This means including bonds, shares, commodities, REITs, hybrids, and more in your portfolio.
- Investing in several different securities within each asset. A diversified portfolio spreads investments around in different securities of the same asset type meaning multiple bonds from different issuers, shares in several companies from different industries, etc.
- Investing in assets that are not significantly correlated to one another, or even negatively correlated. Negative correlation means that when some assets go in one direction the other assets go in the other direction. The magnitude of the move could be different but it ensures that when recession hits and many assets lose their value, the others could compensate. The idea here is to choose different asset classes and securities with different behavior toward systematic market risk to hedge against the impact of any negative conditions that could adversely affect your portfolio.
This final point is critical to keep in mind when composing a diversified portfolio. Without it, no matter how diversified your types of assets are they may be vulnerable to the same risk, and, therefore, your portfolio will react in unison. Therefore, it is key for investors to avoid choosing investments for their portfolios that are highly correlated. It is important to notice that within portfolio management practices there’s a distinction between naive diversification and effective diversification (also referred to as optimal diversification).
Naive and Optimal Diversification
The reason that diversification is usually a successful strategy is that separate assets do not always have their prices move together. Hence, a rather naive diversification can be beneficial (however, at worst, it can also be counterproductive). Naive diversification is a type of diversification strategy where an investor simply chooses different securities at random hoping that this will lower the risk of the portfolio due to the varied nature of the selected securities.
Naive diversification is simply not as sophisticated as diversification methods that use statistical modeling. However, when dictated by experience, careful examination of each security, and common sense, naive diversification is nonetheless a proven effective strategy for reducing portfolio risk.
Optimal diversification (also known as Markowitz diversification), on the other hand, takes a different approach to creating a diversified portfolio. Here, the focus is on finding assets whose correlation with one another is not perfectly positive. This helps to minimize risk in fewer securities which in turn can also help maximize return. With this approach, computers run complex models and algorithms in an attempt to find the ideal correlation between assets to minimize risk and maximize return.
As indicated above, both forms of diversification (naive and optimal diversification) can be effective, simply because diversification results when you spread your investable funds across different assets.
Naive diversification refers to the process of randomly selecting different assets for your portfolio without using any complex computation to decide which you choose. Despite its random nature, this is still an effective strategy to decrease risk based on the law of large numbers.
The Significance of Correlation
There is a “better” way to diversify—specifically, by examining the assets you intend to invest in, to find ones that don’t tend to move up or down in correlation with one another. By doing this, you can effectively lower the risk of your portfolio.
This works because of correlation—an important concept in statistics. Correlation is the measurement of the degree or extent to which two separate numeric values move together. Here, those values we are interested in are assets. The maximum amount of correlation possible is 100%, which is expressed as 1.0. When two assets have a correlation of 1.0, when one moves, the other always moves. Though the amount these assets move may be different, a correlation of 1.0 indicates they always move in the same direction together.
Conversely, when two assets move in opposite directions, their correlation is negative. If they always move 100% of the time in the opposite direction, this is considered -100% or -1. So when examining assets’ correlation, the closer to -1.0, the greater the effect of diversification.
The Bottom Line
Everyone is clear on this: investors must diversify their portfolios to protect against risk. Though it becomes less efficient to diversify under extreme conditions, typical market conditions will almost always mean a well-diversified portfolio can significantly reduce the risk that investors face. Therefore, it’s key to strive to continually improve or optimize your portfolio’s diversification to maximize the protection it offers your investments. This means performing due diligence to locate assets that don’t move in correlation with one another as opposed to simple, naive diversification.
On the other hand, the supposed benefits that complex mathematical diversification provides are relatively unclear. How to apply and operate such complex models is, even more, unclear for the average investor. Sure computerized models have the ability to appear convincing and impressive, but that does not mean they are any more accurate or insightful than simply being sensible. In the end, it is more important whether or not a model produces results than if it’s based on a highly complex algorithm.