Back to Forum

How to Avoid Making the Portfolio Assumptions that Harm Clients

Advisor Perspectives

By Scott MacKillop | June 24, 2016

Share on LinkedInShare on FacebookTweet about this on TwitterEmail this to someonePrint this page

A blueprint for building the "perfect little portfolio".

Related Resources

Link to Online Article

Download PDF of Article

Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.

My article, The Portfolio Management Assumptions That Harm Clients, generated a provocative conversation thread on APViewpoint following its publication in April.

One of the themes in the conversation was summed up by the venerable Bob Veres, who told me that he had read my article (my heart soared) and then quickly followed with, “you pointed out quite a few problems but didn’t really offer any solutions” (my heart sank).

This article is my attempt at redemption. I don’t want to be known as a person who stirs the pot and then walks away. And so it is that I offer up my perfect little portfolio (PLP). Showing how my firm has implemented the ideas in my original article will be more instructive than discussing those ideas in the abstract.

Introducing… (drum roll)… my PLP

My PLP is not perfect, nor is any portfolio. But my PLP has some very laudable characteristics that deal with the issues I raised in my first article as well as some other troublesome problems that I did not address. There are many other ways to address these issues. My PLP is only one of them.

I confess “my” PLP is not actually mine. It is the product of our investment committee, which includes seven members (five CFAs, a Ph.D., a CFP/CIMA/CAIA and a JD – for those of you who are counting, one of the CFAs is also a Ph.D. She is also a concert violinist.). Four members are independent outsiders, so we’ve had some spirited debates about how to build our PLP.

We manage portfolios at various risk levels, but for the purposes of this article, I will look at our “balanced” portfolio, which has a 60/40-type risk level. The portfolio has a “core-plus-satellite” structure. Let’s start by looking at the core. We will look at the satellites later.

The core consists of 36% Vanguard Total Stock Market ETF (VTI), 24% Vanguard Total International Stock ETF (VXUS), 24% iShares Core US Aggregate Bond ETF (AGG) and 16% Vanguard Total International Bond ETF (BNDX). The core can represent between 50% and 100% of the portfolio. In other words, we are not obligated to add any satellites but may do so at our discretion.

Having a starting point

In my previous article, I addressed what I called the “asset class selection problem.” This problem refers to the fact that many portfolios appear to have no logical starting point.

Instead, they are an amalgamation of asset classes that are thrown together without any logic, discipline or identifiable process. The operating assumption seems to be that if we keep throwing “diversifiers” into the mix, we will reduce volatility and somehow get a better result.

There are a few problems with this approach. First, asset classes that appear relatively uncorrelated become more correlated during periods of extreme market movement.

Second, some “diversifiers,” like managed futures, have no real long-term expected return. Including them permanently in a portfolio may provide a modest reduction in volatility but is almost certain to reduce long-term return. Clients cannot perceive these reductions in volatility, but they certainly will perceive the reduced terminal value of their account.

Third, there are some times when it is better to own one asset class than others. Forcing a portfolio to hold 12 to 15 asset classes at all times means you will be holding some of them when it is not wise to do so.

Finally, every asset class that is added to an account costs the client money. Transaction costs associated with buying, selling and rebalancing are a guaranteed drag on performance.

The core of our PLP is our logical starting point. It is designed to replicate the global market portfolio. Nobel Prize-winner Harry Markowitz developed the concept of the “optimal portfolio.” Nobel Prizewinner James Tobin developed the idea that there was really only one “super-efficient” optimal portfolio. Nobel Prize-winner Bill Sharpe identified Tobin’s super-efficient portfolio as the global market portfolio (GMP).

With all that Nobel-ity behind it, the GMP is a good starting point.

The GMP is the portfolio that represents how all of the investors in the world have invested their assets among the risky asset classes. Estimates of the composition of the GMP vary, but about 85% of it is represented by U.S. and international stocks and bonds. The remainder is represented by what we call “sliver positions” of less than 5% each. The sliver positions include assets classes like TIPS, private equity and high-yield debt.

We eliminate all the sliver positions and the remaining U.S. and international stocks and bonds form the core of our PLP. We eliminate the slivers because (1) each position in a portfolio costs a client money; (2) the expenses associated with owning the slivers is often higher than the expenses associated with owning the stocks and bonds; and (3) we can own the slivers when they are priced favorably and decline to own them when they are not.

The core of our PLP is implemented by holding the four ETFs mentioned above representing U.S. Page 2, ©2016 Advisor Perspectives, Inc. All rights reserved. stocks, international stocks, U.S. bonds and international bonds, respectively. This allows us to adjust the risk level of our PLP’s core up and down by altering the equity-to-fixed income ratio. It also allowed us to introduce a slight home-country bias. The average internal expense ratio for our PLP’s core is about .09%.

Our PLP’s core provides a client with exposure to over 18,000 securities across more than 40 countries, representing all major market sectors, styles and market caps. This simple structure provides a well-diversified foundation at a rock-bottom price.

Our PLP’s core addresses many of the problems I identified in my original article. Its structure and composition does not depend on the development of precise capital market assumptions. There is no temptation to tweak and twiddle based on unreliable optimizer inputs or outputs. It is not driven by datamined factors or period-specific research. It has firm theoretical underpinnings and is determined by the collective wisdom of all the investors in the world.

Rebalancing

In my original article, I pointed out a number of issues with current rebalancing practices. I won’t revisit them here but will refer readers to three excellent articles on the topic written by Michael Edesess (here and here) and Michael Kitces (here) for Advisor Perspectives. Vanguard also published a fine paper on the topic in July 2010 entitled Best Practices for Portfolio Rebalancing.

Based on this research and some of our own, we rebalance our PLP’s core once a year. But we reserve the right to rebalance more or less frequently if we have a strong conviction that departing from our one-year default is appropriate. As a default, annual rebalancing is better than more frequent rebalancing, which is guaranteed to generate higher transaction costs without necessarily improving performance.

Building on the core

Our PLP’s core is a pretty nifty portfolio. A reasonable person might leave this little-engine-that-could well enough alone. But we can add incremental value over the long term through the selective addition of satellites.

Satellites are actively managed mutual funds, index funds or ETFs designed to provide exposure to asset classes that are not represented in the core or to overweight those that are. Satellites can represent up to 50% of the portfolio.

This brings me to the next harmful portfolio management practice that our PLP addresses. The debate between the passive- and the active-management communities has reached new levels of silliness at the expense of our clients. Both sides are more interested in being right than in doing what’s best for clients.

Rather than approaching the active versus passive question as if we were choosing a religion, why not approach it more practically? Index funds and ETFs are cheap and can be tax-efficient. This gives them a built-in performance advantage. There are strategies and exposures that cannot easily be accessed through passive products. Active products make sense here.

And then there is the question of skill. Let’s start off by agreeing that many, even most, active managers don’t consistently add value after fees. Let’s also agree that the mythical “average manager” is an unskilled amateur whose investment products should be avoided at all costs. That still leaves hundreds of managers from which to choose.

I am fully aware that those who want me to provide detailed solutions will be disappointed by my lack of guidance on how to select satellites to complement the PLP’s core. That is a more complicated discussion, and the number of possible good solutions is quite large. The key is being highly selective, patient and adding to the core only when you have a high level of conviction that you can add value.

When in doubt, do nothing.

Our PLP has three satellites representing approximately 17% of the portfolio. They are SPDR Barclays Intermediate Corporate Bond ETF (ITR – 6%), Vanguard Market Neutral Fund (VMNFX – 4%) and PRIMECAP Odyssey Stock Fund (POSKX – 7%). The internal expense ratio of our PLP is .14% with the satellites. I won’t explain here the role that each satellite serves in the portfolio. What matters is that we have a high conviction about each one, but we are not obligated to own any of them.

The courage to be simple

Albert Einstein once said, “Everything should be made as simple as possible, but no simpler.” I’m pretty sure he wasn’t talking about portfolios, but our PLP embodies his thinking.

The financial services world has fallen in love with complexity, and this is not a good thing. A good deal of it is designed to increase fees and/or inject an air of mystery into the equation that no mere mortal could parse through. Clients (and some advisors) are bedazzled and confused by it all. Transparency and confidence suffer.

Our PLP is simple on the surface, but it took a tremendous amount of work to arrive at this solution. We had to hack through a jungle of traditional and stale thinking to arrive at a portfolio structure that accomplishes the following:

  • It has a logical starting point
  • It is well diversified
  • It is highly liquid It has low internal expenses
  • It has few positions, which minimizes transaction costs
  • The rebalancing strategy minimizes transaction costs
  • The rebalancing strategy is thoughtful, not mechanical
  • We are not obligated to own anything except the core
  • We can add satellites when we see opportunity and otherwise do nothing
  • It combines active and passive management, so clients can benefit from both
  • Clients can understand it

Because of its focus on low costs and minimizing transaction costs, our PLP has a performance headstart at the beginning of each year. Strategies that hold more positions and trade more – even ETFbased strategies – have to play catch-up. That is hard to do on a consistent basis.

A broader definition of “portfolio”

Normally, the discussion about our PLP would end here, but our firm has a broader view of what a portfolio is than the traditional asset management firm.

Our fee schedule is also a feature of the portfolio. The fees we charge affect the performance of our portfolios every bit as much as the internal expense ratio of our core or the performance of the satellites we choose.

Because we care about the overall net return to our clients, we have given as much thought to our fee schedule as we have to the construction of our PLP. In our view, the traditional percentage-of-AUM fee schedule doesn’t make much sense. In dollar terms, our costs do not increase as the size of the accounts we manage increases. Therefore, there is no justification for charging as though they do.

We charge a low basis-point fee for smaller accounts and cap the fee once it reaches a certain size. So for clients of advisors who work with us directly, the fee is .50% of AUM, and the fee is capped at $1,500 annually. Once an account reaches $300,000, the fee never increases. A $1 million account pays the same as a $300,000 account. Essentially, it becomes a flat fee.

Successfully getting a client to their financial goals is partly a portfolio management problem and partly a behavioral problem. Traditionally, asset management firms focus on the portfolios and leave the behavioral part of the equation to financial advisors. We have taken a broader view of our responsibility.

Victor Hugo once said, “An idea expressed is a responsibility assumed.” We know that our portfolios only represent two wheels on the bus. The client is unlikely to make it to their goal without help on the behavioral side. So we have developed videos and written material to help clients with the other two wheels on their bus. We’ll continue to create more in the future.

This educational material does not sell our products or our firm. It is designed to reinforce the messages that advisors are already giving to their clients. We have made it available through our website to anyone who wants to use it. Our hope is that, together with our PLP, we can increase the chances that clients will reach their long-term goals.

Yes, I stirred the pot and walked away. But I have tried to make up for that by providing some ideas about how to address the issues I raised in my first article. Let the conversation begin! I’ll be waiting for you on APViewpoint.

Scott MacKillop is CEO of First Ascent Asset Management, a Denver-based firm that provides investment management services to financial advisors and their clients. He is a 40-year veteran of the financial services industry. He can be reached at scott@firstascentam.com.