Risk allocation must replace asset allocation

The investment industry has made great strides in using technology to modernize the investment process. The growth of exchange-traded funds is testimony to the desire to increase liquidity and transparency, and lower transaction and management costs. Transaction fees are ever shrinking. Information technology puts market and investment information in managers' and investors' hands with lightning speed and laser accuracy.

But the most important part of the investment process — portfolio construction and management — is like using a rotary telephone! The asset allocation process is outdated, mired in 20th century thinking in a 21st century world, and — most importantly — fails to address investor needs.

Using risk allocation in place of asset allocation will change the way the investment industry thinks and operates proportionate to the way smartphones and networked devices have changed the way we communicate, gather and disseminate information. Risk knows the difference between T-bills and junk bonds, and sets the stage for all the enhancing apps that will not work in the rotary phone environment. We should protect our portfolios from significant risk of loss just the way we now protect ourselves from inclement weather, traffic jams and changes in tides — by using intelligently and coherently all the market information that is available at a moment's notice, rather than reading about the damage in the papers and on our account statements after the fact.

Risk allocation has several key advantages over asset allocation:

• Risk changes quickly, often daily — being unresponsive to it doesn't make sense. Traditional asset managers do not take the daily changes of risk into consideration. They instead focus on benchmark-relative performance, and wait until the end of each quarter to find out what happened to their portfolio and then rebalance. Would we wait 90 days to react to any other negative stimulus in our lives? Why allow asset managers to do so?

• Balanced allocations are no replacement for risk management. Traditional asset allocators use fixed allocations to bonds to control risk whereas risk managers use risk management to control risk.

Note the logic? Bonds don't control anything — they reduce returns too much when risk warrants more productive equity allocations, and do far too little to help when our portfolios really need it.

• Asset managers must rely on accurate forecasting, when evidence shows it cannot be done. According to Bloomberg, somewhere between 10% and 15% of economic and market forecasts come true. That means that seven out of eight times forecasts are wrong. That would be enough to get a student kicked out of school or an employee fired in any business except the investment industry. Perhaps John Kenneth Galbraith was right when he said: “The only function of economic forecasting is to make astrology look respectable.” Asset allocators stay up nights and weekends fretting over whether their allocation of assets to emerging markets should be 12.5% or 14%. The stress is palpable and unnecessary.

• Risk managers recognize that asset class relative returns are generally unpredictable, but risk and changes to risk are measurable – and if one has the wisdom to use it, significantly enhance returns over time. Rather than rely on impossible and unreliable forecasts, they rely instead on daily market observation and risk measurement. Risk managers play golf and tennis, and sail on weekends knowing that Monday might be treacherous, and they will be prepared to change market exposure on Tuesday if they need to get out of harm's way.

In Diagram 1, where the bubbles represent different investments, a diversified, low-volatility portfolio is poised for growth. In contrast, Diagram 2 demonstrates investments that are concentrated and highly volatile, signifying crisis.

Why do investment managers cling to their rotary phones while cell towers and high-speed networks surround them? Because managing risk daily is hard work. It requires a lot of computer power, data crunching, and experience to properly measure risk, changes in risk and the pace of those changes. Investors and advisers hold fast to old-school risk measures like option pricing, put/call parity, and the Chicago Board Options Exchange Market Volatility index, known as the VIX. They continue to invest infrastructure and human capital in fixed-income investments of little to no yield and significant duration risk.

Thought leaders will realize that adapting to the changing patterns of risk is not only the safest way to invest, but also the best way to take advantage of technology, new data management availability, and low trade-execution costs. Moreover, properly managing risk allows investors to avoid the zero interest QE environment that robs savers of any possibility of earning income through fixed-income investing.

Rotary phones and asset allocation will not make a comeback. Technology has made so many predictions in our lives less relevant or unnecessary. In times past, forecasting benefited greatly from the fact that the “unknowable” market factors would likely remain unknowable. When Harry Markowitz developed modern portfolio theory in the 1950s, investing in foreign countries offered excellent diversification that had little chance of losing value over time. Foreign markets were not correlated because investors in global markets were only concerned about local economies and news. News traveled parsimoniously and slowly. Peter Christoffersen at the Rotman School of Management at the University of Toronto has conducted extensive research on the changing patterns of market correlations over time. He points out that as recently as the mid-1990s there was a 0.21 correlation among the 26 developed equity markets. By the end of 2012, the information age dramatically increased the correlation to 0.81.

The sea change in global information flows makes it impossible for the future to emulate, or as Mark Twain put it, “rhyme” with the past. The further one goes back in history, the less relevant the market data become. Today, business is simply not conducted with the same information or in the same way. Equities have the widest range of volatility (the second derivative, or volatility of volatility) of securities in portfolios. While the return-forecasting business will remain a losing proposition, the daily measurement and impact forecast is the safer and more accurate route to success. The tools are now, for the first time, available. It is rare to have a solution that reduces stress and improves performance in investments the way that smart phones improve the quality of modern lifestyle.

Risk-allocated portfolios are constructed very differently from traditional asset-allocated models in that they are assembled with the expectation that they will be risk managed. For example, Diagram 2 shows the dramatic difference between portfolios that are allocated using asset percentages and those allocated by risk. Note the dramatic difference between the amount of space that fixed-income investments occupy in today's market and the relatively insignificant amount of market risk they bring to the portfolio. Also note that if the allocations to foreign markets are made in local currency, risk allocation is the only way to capture the potentially large amount of risk from changes in currency values.

Risk budgeting is the closest practice to risk-allocated portfolios in the market today: CIOs select markets they believe have potential for profit over the next year and then calculate the current levels of volatility and correlation to evaluate the best potential “fit” in current market conditions.

The main drawback to risk budgeting, however, is that there is no dynamic way to protect assets from market turmoil or insure the initial portfolio will remain stable over time. Risk-allocated portfolios combine the best mathematical advantages of modern portfolio theory with the use of the latest technology to accurately define the desired mix of investment returns. Another significant advantage of risk-allocated and managed portfolios is that they do not require fixed-income exposure to control risk and, in a zero-interest-rate environment as we have now, bonds have no return potential, provide insufficient coupon to cover inflation, and have significant downside risk. When fixed-income markets resume to price levels that pay reasonable coupons for credit risk, bonds can be considered for their value as a cash flow producing investment. Until such time, investors should avoid fixed-income exposure.

Since the initial allocations to a risk-allocated portfolio will be modified as markets change over time, daily risk measurement of volatility, correlation, return and pace of change sharply reduce the stress of having to predict accurately. Risk allocations may be set as absolute percentages, or as limits for exposure to each asset or investment class. Once the limits are set, the manager determines the amounts of cash necessary to fund each class. The “best fit” is then measured as a function of the absolute levels of risks associated with each asset class, the expected potential return and the correlations between the investment classes.

The best way to calculate the strength of a risk-allocated portfolio is to consider the relationship between the total risk of all investment components in the portfolio and the correlated (diversified) risk of the portfolio. A high concentration of portfolio risk serves as a warning that there are common risk factors embedded in the selected asset classes that might not have been expected in the selection process. Thoughtful consideration may result in the selection of different or additional asset classes to improve diversification. A low concentration of portfolio risk signals broad portfolio diversification and high dependence on future low correlation. In some cases, the volatility of correlation can be a higher source of risk and return than the volatility of the investments themselves.

Initial allocations may be set using a metric that measures the marginal utility of adding and subtracting allocations to the risk-adjusted returns of each of the asset classes of the portfolio. The marginal utility of each return is then calculated to determine the impact of each investment class on the portfolio as a whole. Those classes that have the best impact on the risk-adjusted performance of the portfolio receive the highest allocations and those with less favorable impact are given less allocation. When the portfolio manager/CIO is comfortable with the market exposures, the next step is to fund the risks selected with the necessary amounts of cash to invest in each asset class to achieve the desired portfolio mix.

Typical allocations for an asset-based portfolio are assigned as a percentage of total assets and left to the mercy if the markets. Risk-allocated portfolios recognize that the factors that will determine success or failure of each asset class are largely unknowable and equally unpredictable. Because risk-allocated portfolios are subject to daily quantitative analysis, they do not require painful and usually inaccurate forecasts.

It is important to note that such quantitative analysis must be robust and sufficiently practical to signal change whenever necessary. For example, should any investment in the portfolio become unacceptably risky by virtue of high volatility accompanied by high correlation to the overall portfolio, the allocation must be reduced. The best expected outcome is determined by assigning a maximum and minimum range of risk for each asset class and running an analysis to determine the relative marginal contribution of each asset to the risk-adjusted return of the portfolio. Running the utility function daily enables the portfolio to adapt quickly to changing market conditions.

The outcome of the risk-allocated portfolio construction process is a dynamic portfolio with performance potential that benefits from the elimination of low-returning fixed income, daily MPT analysis to readjust allocations to stay in tune with multiple market behavior and the early detection and elimination of poorly performing asset classes. Daily monitoring and management may reduce exposures and the use of revised allocations to find new sources of returns that support the risk-adjusted portfolio performance.

The takeaway is clear: Investors want to participate in growing economies and rising markets. They should not waste time trying to determine what the market should or should not care about. Investors should get out of the evaluation/forecasting business and let the risk levels in the markets tell them what troubles or comforts the millions of investors who buy and sell billions worth of securities every day. The information stored in price behavior speaks volumes about investor perception. The relationship between greed and fear is widely debated; markets tell us when they are confident or scared.

While forecasting return is impossible, measuring risk can signal the best time to be invested in potentially risky assets and when to return to the sidelines. The valuable insights are hidden in full view for those willing to mine the data thoughtfully and frequently.

A robust risk management system that can accurately measure volatility, correlation, and pace of change on a daily basis is the hard work required to mine the market data. The effort is rewarded in a smoother portfolio glide-path and superior performance.

Risk-allocation technology is much more than a new money management strategy. Using risk as a proxy for investment exposure eliminates the need for many of the investment industry's longstanding, if not very successful, practices. Risk allocation:

• recognizes that global and local investment is in constant flux and static assumptions are useless;


• eliminates the asset allocation process;


• eliminates the need to use low-return/high-risk investments to water down expected volatility;


• reduces forecasting risk to the selection of investible asset classes — not the allocations to them;


• protects investments pro-actively when market behavior threatens;


• properly places investor goals/needs ahead of “benchmark” performance; and


• provides a healthy way to permit money managers to focus on their clients' needs.

We are bombarded with the burden of knowing so much more about world events faster than ever before. The stress of our investment decisions, however, can be softened by the same technology that makes our investments more unreliable. Perhaps there is some justice after all.

William G. Ferrell is managing partner and chief investment officer of Ferrell Capital Management LLC, Greenwich, Conn.