This is the third and final posting in a series on how advisors can take a fresh look at portfolio construction.
Part one covered how to get started on the process, while the second posting covered how to set goals with the client. Below, the conclusion.
Part 1: Portfolio Construction and Asset Class Suitability
This discussion assumes that advisors are interested in developing and implementing investment strategies based primarily on a method of investment analysis, portfolio design and performance evaluation expressed, in large part, by quantitatively evaluating risk (client and investment) and its relationship to investment return.
This methodology focuses attention on the overall composition of the portfolio (asset allocation) rather than the traditional method of analyzing and evaluating the individual components. The investment manager and client are, therefore, able to examine and design portfolios predicated on explicit risk-reward parameters and on the identification and quantification of portfolio objectives.
It is imperative that the advisor choose those assets classes that, when combined, will meet the investor’s growth objectives and risk profile. The importance of allocating capital based asset class assumptions was clearly demonstrated in the seminal Brinson, Beebower and Singer study of 91 large pension plans. The study sought to attribute the variation of total returns among the plans to three factors:
To some degree every investment program involves tactical and/or dynamic asset allocation. The example I often use is taking a sailboat from San Francisco to Hawaii.
If I set a course to get there in a straight line, I will never see the islands. Changes in the winds, currents and weather will all affect the heading of the boat. Consequently, we must always review our strategies and make whatever modifications (tactical allocation) are necessary to maintain the course to our objectives.
How Can You Determine Appropriate Investments?
Prior to designing the investor’s new portfolio, the advisor and the client should consider all assets that are acceptable to the investor. While scientific analysis is extremely important, there may be times when you, as the advisor, have certain knowledge or expectations of asset performance that warrant the inclusion or exclusion of some assets.
The classic example would be Japanese or Pacific Rim securities in 1988 when most advisors were under-weighting Pacific Rim securities in portfolios as they were considered substantially over-valued.
How Much Diversification Will Be Required?
It is important to note that diversification is not a function of how many assets or securities are included, but rather are they negatively correlated and do the selected assets balance the portfolio. For example, if you had an equity portfolio of 100 stocks, the portfolio is no more diversified and the systematic risk of the portfolio is not reduced any more than if you had only 20 stocks across industry lines. In other words, adding 80 more stocks to the portfolio did not reduce the risk of the portfolio.
In some cases you may find that although you have included many assets, an Optimization algorithm may only recommends some of those assets. Mathematically the relationships of the assets (returns, correlations and covariances) will support the recommendation. However, you may feel that additional “diversification” is necessary.
Remember, your knowledge, experience and understanding of your client’s needs are as important as any mathematical algorithm. It is also true that tools such as SCANalytics and the algorithms employed are not magical and therefore should only be relied upon as “mathematically” probable. There is no means by which these tools can know about current or possible economic events which might further influence the performance of any asset or assets without your introduction. Your experience, expertise and knowledge will be useful in guiding the composition of the portfolio.
It might be useful to establish a list of 15-20 generally recommended asset classes to start and add others as may be required for each individual investor.
Part 2: Designing the Proposed Portfolio
The process for developing portfolios is:
- Determine which asset types are appropriate (i.e. cash equivalents, Domestic Equities, International Bonds, Real Estate)
- Determine which asset classes are appropriate (i.e. Large Cap Growth, Small Cap Value, Domestic Bonds, Emerging Markets, Hedge Funds, REITs)
- Determine the estimated (historical or forecasted) rate of return and volatility assumptions
- Determine what, if any constraints are placed on Asset Types (i.e. maximum of 40% Equities, 25% International Bonds)
- Determine what, if any, constraints are placed on Asset Classes (i.e. 20% Large Cap Growth, 10% REITs, 5% Gold)
- Find securities that are highly correlated to each Asset Class
- Determine the probability of achieving the client’s objective
There are several ways to develop portfolio proposals.
A) Model Portfolios
Creating pre-set Model portfolios of assets can be a quick and efficient method of offering your expertise to clients. Based on the analysis of the client’s Risk Profile and the current portfolio, the advisor can expand or reduce the list of available assets and assign specific investment vehicle.
In some cases advisors might offer a complete portfolio (assets and securities), but it is not recommended prior to your initial evaluation and analysis. These Models might include one or more variations for each of the following:
Growth & Income
B) Forecasted Returns (Standard Deviations, Correlations)
In order to evaluate the performance characteristics of the recommended portfolio and comparison with the current portfolio, it would be appropriate to forecast expected returns for the near term (1-5 years). Clearly, your return assumptions will materially affect the probable performance.
B) Subjective Analysis of Alternative Investment Scenarios
In some cases advisors may simply be able to add or substitute asset classes iteratively based on alternative scenarios and manually modify allocations. The advisor would then run each scenario through the analysis procedure to determine if any of these portfolios will provide suitable returns within established risk parameters. While this is extremely subjective and time consuming, it may offer both the advisor and the client a better understanding of the various trade-offs suggested by each iteration.
D) Find Theoretically Optimal Portfolios
Optimization is the process of mathematically determining those combinations of investments (assets and/or securities) that will achieve the highest possible rate of return for any level of risk that an investor is willing to accept. In general, it is recommended that advisors optimize portfolios at the Asset Class level. Once this has been done, the advisor can find investment vehicles that are coordinated with the selected Asset classes and compliant with the investor’s risk and objectives.
When using historical data, optimizers may be unequivocally relied upon to provide the optimal portfolio in terms of return and “risk” for the specified time period. For example, using a time horizon of 1985 to the present, the optimal portfolio displayed would have been the “optimal” portfolio for that time period. If, however, you believe that performance over the next 1-5 years will be different, then you might want to use an historical time period that reflects your expectations, or forecast rates of return.
There may be important constraints that should be applied to asset holdings. If you believe that the optimization algorithm has allocated too much or too little to a particular asset, you should review the returns before setting constraints. It may well be that the asset in question has a rate of return far higher than you are comfortable with. Consequently, you should lower the return estimate and then re-optimize. Don’t arbitrarily constrain portfolio assets before optimizing the portfolio.
Optimize the portfolio and then iteratively set minimum or maximum levels for each asset (or you may set global minimums or maximums of some amount) until the portfolio meets your satisfaction. The reason for not pre-setting constraints is that there is no way the advisor can know precisely the mathematical relationship of each asset to each other asset in the portfolio. Consequently, pre-setting constraints may have the adverse effect of tilting the portfolio in the wrong direction.
Part 3: Find Appropriate Vehicles, Achieve Goals, Review Regularly
Once the portfolio asset allocation has been established the advisor will find those investment vehicles that match the performance characteristics of the asset classes (rerun, standard deviation, correlation to appropriate benchmarks).
Once the proposed portfolio has been designed it is now necessary to calculate and review cash flow analysis and/or Monte Carlo simulation to determine the probability of achieving the investor’s financial goals.
Various personal experiences and external factors can easily move the client’s risk profile in either direction requiring some modification of the portfolio and financial objectives. Changes in the dynamics of the financial markets may also require reevaluation of portfolio allocations and asset class participation.
All institutional, and most successful financial advisors, have an established method for understanding investor needs, reviewing existing investment portfolios, finding successful investment strategies, establishing strong client relationships and the work flow commensurate with that procedure.
This methodology not only increases confidence in the advisor’s ability to manage the investor’s investments, but virtually eliminates any potential liability concerns. Whether advisors follow the advice above or establish their own workflow procedures, they will realize substantial rewards by creating a clear, practical and repeatable method for managing client investment.