Do you remember 50 years ago when the industry standard to determine a client's asset allocation was to subtract her age from 100 and then allocate accordingly to a basket of blue-chip stocks and bonds? Just look how far we have come; today, our industry has become increasingly reliant on complex asset allocation models that digest large amounts of asset class data and then generate a set of optimized target portfolios.
These portfolios use the concept of the "efficient frontier," theoretically maximizing returns for a given level of risk, or minimizing risk for a given level of return. After the efficient frontier is established, the next step for an advisor is to ascertain which portfolio is most appropriate for her clients given each one's time horizon, cash-flow needs and risk tolerance. In other words, the advisor must determine where on the risk/return spectrum each of her clients lies. Theoretically, the advisor is always choosing an efficient portfolio: It is just a question of where on the frontier it is appropriate for each client.
Ostensibly, it would seem that the complexity attached to these models means we have reached a new pinnacle in forward investment thinking. Accordingly, these models have become a very popular selling tool within the industry. However, lifting up the hood and studying the engine of these models reveals that the theory behind the efficient frontier is much sounder than its real-world application-for a few reasons:
The typical asset class inputs in a portfolio optimizer are the class's mean return, its standard deviation and its correlation to other classes. For various reasons, the efficient frontier in the unconstrained model often suggests portfolios that are unlikely to be implemented because they allocate too heavily to some of the more volatile asset classes like emerging markets and small-cap value. No advisor is going to recommend that a client invest 75% of a portfolio in emerging markets equity, even if it's what the model says is efficient. To combat that problem, advisors insert their own constraints into the models. For example, they may allocate no more than 10% to emerging markets equity or halt the small-cap allocation at half of what the large cap's is. As soon as advisors start tinkering with the model to create allocations that are more salable, the resulting efficient frontier ceases to be efficient and becomes subject to arbitrary, advisor-driven constraints.
The problem with feeding historical data to an optimizer is that, well, it's historical. Portfolio optimizers are not designed to predict future market movements, which is probably a good thing because they are very bad at it. Volatility and correlations vaulted much higher in late 2008 than what most models using historical data could have predicted or would have accounted for. In other words, investors in a model-generated target portfolio did not have a leg up on other similarly diversified, non-model-driven portfolios going into that year. Someone has yet to build an optimizer using such history that can actually help investors in times of stress. So a common-sense approach seems better, one that avoids unnecessary cost and complexity.
If we are not to rely on optimizing models, where do we turn for our target investment portfolios? First understand that in the world of asset allocation simplicity is not such a bad thing. It is important to be invested in different asset classes that are not highly correlated. An optimizing model that tells you exactly how much to invest in each asset class down to the hundredth of a percent is a waste of time, energy and resources. Instead, you should focus on the following:
Getting the Aggressive/Conservative Breakdown Correct
Appropriately determining the breakdown in your clients' portfolios between the more aggressive asset classes (for example, equities and real assets) and the more conservative ones (high-quality fixed income and cash) is the most important decision in the implementation process. You do not need a complex model for this exercise. You need to know your client and align her portfolio with her cash-flow needs, time horizon and risk tolerance. Once this breakdown is established, you can then determine what sub-asset classes to use within each bucket and in what capacity. It is important to allocate to all relevant sub-asset classes within each of the buckets, but the breakdown between your aggressive and conservative investments is the key.
You should spend most of your time ensuring that all aspects of your clients' financial lives (such as retirement planning, estate planning, income tax planning, etc.) are working in concert. Unfortunately, many advisors become absorbed with finding the hot manager or fixate on returns. A well-timed and thoroughly thought out planning strategy can add much more value to a client's bottom line than any one investment manager can. With the status of estate tax legislation in flux and the opportunities associated with estate planning changing by the minute, it is imperative that you are aware of these opportunities and can proactively present your clients with interesting ideas and strategies.
Our society often equates complexity with viability. Within the world of investment management, make sure not to be blinded by sexy asset allocation models that promise "efficient" portfolios when a simpler, more common sense approach might work just as well. There is no need to rely on something that is inherently flawed just to make yourself feel better that there is "science" behind your recommendations, and there is no sense in creating or paying for an optimizer to tell you what you want to see when you can implement something sensible without it. Once you have embraced this concept, spend your time and energy with clients on things you can control, like aligning the stock/bond breakdown with each of your client's goals. Your clients will feel the difference, and so will your bottom line.
Jarrett Solomon, CFP, CIMA, is a senior financial advisor with Connecticut Wealth Management LLC, a registered investment advisor firm.