Rather than thinking in growth or income terms on the investment side, the investor must separate out expense streams and allocate a portion of the portfolio to more closely track these expenses over time. The largest non-discretionary expenses incurred by most households are for housing, taxes, energy and food. To manage the risk that these expenses will increase, investors can pick assets that more closely track them than a diversified investment portfolio would. In addition, if an investor anticipates other future liabilities and estimates when they might occur, he can create an investment allocation that better minimizes the risk of potential shortfalls. He could, on the other hand, well come up short if he simply estimates expected asset class returns, correlations and annual standard deviations.
There is one other advantage of this approach: Investors who don't use it might otherwise be sacrificing return by keeping the overall risk level of their portfolios too low. They aren't taking the proper time horizon into account and they could be mistakenly misallocating assets in excess of their future non-discretionary needs. They are more likely to assail their overall risk if they tackle the individual components of risk separately, allocating portions of their portfolio to retirement income, education savings, living expenses and estate planning. Not only will they be managing their risk better, but they will also be better able to use what assets remain, maximizing their long-term non-essential discretionary spending for things such as philanthropy, bequests, collecting or lifestyle. If a client separates these assets from particular savings goals, he will be more likely to invest them in more illiquid long-term investments with a higher expected return.
Managing Risk
To use this risk framework, an investor must first insure risks of uncertain but improbable events such as premature death and disability, catastrophic medical costs or property loss. Next, he should segregate out and invest specifically for future liabilities with unknown costs: for example, retirement income, education costs and living expenses. Since the risk associated with these unknown costs is primarily related to inflation (not necessarily CPI inflation), the diversified portfolio for each will have to contain asset classes with a higher expected return than inflation. Additionally, the shorter the time horizon, the more heavily weighted the portfolio should be to assets that directly track the inflation of the costs specific to the liability.
Take household expenses. It is now possible to use commodity ETFs that track the same futures used by corporate risk managers to hedge against future household price increases. Similarly, if an individual doesn't own real estate but he might have future housing needs, he can find liquid real estate investments that will reduce his exposure to future housing inflation without giving him the liquidity risk of holding actual real estate.
For principal protection against inflation as measured by CPI, an investor may turn to Treasury Inflation-Protected Securities, but he might also consider ETFs or assets with a lower tracking error than TIPS-for example, gold for general dollar risk or international currencies and bonds for specific non-dollar exposures.
Then there are school expenses. Since higher-education costs are rising faster than the returns of all the asset classes, the only effective way an investor can eliminate the inflation risk for them is by using a prepaid tuition plans. Unfortunately, these limit the student's choice of college or university (just one of their drawbacks).
After an individual identifies and isolates specific future liabilities and their associated risks, he can manage the remainder of his assets with more traditional mean variance portfolio theory. But even here, it helps us to categorize our risks as we diversify the portfolio.
Some assets are illiquid. Some are vulnerable to tax changes in a charged political environment. Others depend on the asset manager's diligence and integrity. It may help an investor to hold international assets, for example, whether they are outperforming or not, simply because it reduces exposure to the U.S. dollar and domestic political changes. Alternatives carry their own risks. Many come burdened with lockup periods. There is also a higher risk of fraud and inaccurate marks and the risk that the manager will drift from his stated objectives.
This kind of asset matching requires investors to ponder different scenarios. It's easy now to look back at the housing crisis, for instance, yet difficult for investors to model scenarios for similar cataclysms in the future. For example, the risk of home ownership is not just the potential downside of the residential real estate market, but the possible liquidity squeeze if there is a mortgage involved and the principal exceeds the possible sale price. It is easier to imagine severe downturns in the real estate market now than it was a few years ago. But there are still other downside risks considered unlikely now that would require an investor to tap into his liquidity at an inopportune moment.
Insurance can cover some of these risks. But investors have to put thought into more complex financial situations and see how the risks and risk mitigation strategies are interconnected. The liquidity and duration of a portfolio should be determined not just by an asset allocation model and expected expenses, but by careful consideration of all that could go wrong and what decisions would have to be made if they did.
Scenario modeling means asking a client to imagine what would happen if she lost her income, her stock options sank and her home dropped 30% in value. That kind of question will give her a much better indication of her risk tolerance than a Monte Carlo analysis of her proposed asset allocation.