What Place Does Fixed Income Have in a Portfolio?
In a recent article, a fellow SA contributor provided an interesting discussion of the merits of allocating some portion of an investment portfolio to fixed income assets (even for young investors). The primary support for his argument is, in essence, grounded in modern portfolio theory (MPT)-which is more or less a portfolio construction process that attempts to maximize a portfolio's expected return for a given level of risk.
The author goes on to explore ways to think about fixed income and the role it can play in a portfolio by considering the quintessential 60/40 portfolio. In this classic example a portfolio is constructed with an allocation of 60% to equities and 40% to fixed income. The author throws out a suggestion that the exact type of fixed income investment really depends on the "role fixed income is playing in your portfolio" and goes on to suggest that, "for aggressive investors young and old, there are some riskier sectors to be found in fixed income, such as high yield and emerging market bonds." While I respect the questions and arguments up for consideration by the author, I must respectfully suggest that there are a few important factors that have not been considered-factors which may potentially change an investors conclusions about what the exact role fixed income should play in a portfolio and how to best accomplish those objectives.
Why should investors consider incorporating fixed income investments into their portfolios? The quick and easy answer is diversification. But two vital follow up questions that investors need to answer include (1) What exactly am I diversifying and (2) Do all fixed income investments actually help a portfolio achieve asset class diversification?
Consider the classic 60/40 portfolio. I suspect a majority of the typical non-professional investor (and a potentially surprising amount of the professional investment crowd) would consider the classic 60/40 asset allocation mix diversified between stocks and bonds. In terms of dollar values they would be correct. But is the percentage of dollar value allocated to each asset class the most appropriate measure of diversification? I would argue it is far more important to consider the inherent risk exposure to each asset class. For the purposes of this article "risk exposure" is the percentage of portfolio volatility that is attributable to each asset class. Does the 60/40 asset mix actually provide diversification of risk exposures between each asset class?
In the figure below, I have taken monthly historical returns from January 2008 through August 2012 to construct a portfolio utilizing the classic 60/40 asset mix. I utilized the SPDR S&P 500 ETF (SPY) to represent the allocation to equities and the Barclays Aggregate Bond ETF (AGG) to represent the fixed income allocation. This combination pretty much follows the exact suggestions of the author in his article. By looking exclusively at allocation weights an investor would likely, and incorrectly, conclude that they have achieved sufficient diversification between asset classes. However, if that same investor were to consider the risk exposure to each asset class inherent in the underlying 60/40 portfolio, they may be surprised to find out that the portfolio is not even remotely diversified between equities and fixed income.
As can be seen from the graph above, the risk exposure inherent in the 60/40 portfolio is grossly attributable to equities at 95% versus just 5% of the portfolio risk attributable to fixed income. Most will agree, this is hardly a diversified portfolio. But, depending on which fixed income product an investor chooses, the statistical properties of the overall portfolio risk exposures can change dramatically.
For instance, rather than a 40% allocation to AGG, consider a 40% allocation to the iShares Investment Grade Corporate Bond ETF (LQD). Using the same time period as the graph above (January 2008-August 2012) the graph looks something like:
Incorporating LQD instead of AGG, as can be seen in the graph above, resulted in the portfolio's risk exposure attributable to corporate bonds (a.k.a. fixed income) increasing from 5% to 18%. This is probably still not what most people would consider diversified but it certainly has a more diversified risk exposure than the 60/40 portfolio using AGG as the fixed income instrument. It's somewhat amusing how far removed the classic 60/40 portfolio is from actually representing a true 60/40 portfolio in terms of asset class risk exposure. What is not remotely amusing is the vast amount of wealth managed by pension funds and other investment institutions that was destroyed by unintentionally constructing portfolios far more exposed to equities than anyone realized prior to the financial crises. As it turns out, and you will see in a moment, a more realistic 60/40 portfolio is actually something closer to a 40/60 portfolio.
By utilizing a combination of TLH, TLT, LQD, and SPY to construct a portfolio, I assembled a 60/40 portfolio that has more in common with a true 60/40 portfolio than simply the title.
As can be seen above, over the January 2008 through August 2012 time period it took an allocation of 60% to fixed income (mixed among government and corporate bond ETFs) and 40% to equities in order to achieve the true 60/40 risk exposure. While this relationship is dynamic, the underlying empirical fact is that the volatility inherent in equities overwhelms the volatility inherent in fixed income investments thus requiring a larger allocation to fixed income than many would realize in order to achieve greater diversification.
The primary focus of this article, so far, has been entirely upon the risk exposure between asset classes inherent in various portfolios. But how have all of these various portfolios performed over the time period in question (January 2008 - August 2012)? As you can see from the chart shown below the true 60/40 portfolio has outperformed all of the other portfolios.
Yes, as I am sure some critics will be quick to point out, there are some issues relating to data timing. If one were to pick the market bottom for the starting point and show a portfolio 100% allocated to equities, chances are it will have outperformed the true 60/40 portfolio over that same time period. But the entire point of diversification is downside protection. And, just in case it does not go without saying, there is certainly no guarantee that the exact weights of the portfolio I refer to as the true 60/40 will actually behave like a true 60/40 portfolio in the future-which is why a portfolio must be rebalanced overtime.
The previous SA contributor did a nice job setting up the conversation about the potential benefits that can be gained by adding some fixed income instruments to a portfolio, but hopefully I have shown that there is some additional information to consider with respect to how an investor can enhance their diversification between asset classes by utilizing a portfolio construction process that takes into account more than simple allocation weights.
Thanks for reading.