Atlas – Our Portfolio Development Process

Our process at Atlas is different from most firms in the video accompanying this blog we were asked some specific questions about our process. This blog follows that format.

  • Steve, you speak at national conferences and write in publications aimed at certified financial planners to teach them about your portfolio management process.  Why are other certified financial planners interested in learning about what you do at Atlas and h ow is your approach different than what they have learned?

At Atlas, we use the most advanced mathematics and science related to the creation and design of our investment models. We have noted that few firms use this process which is based on a Nobel prize-winning theory called Modern Portfolio Theory. I am always surprised when I speak with other firms to find that they do not have a real process based on science. We also find that few firms and advisors have the extensive background in portfolio design that we have.

  • Why do you think more firms do not use MPT?

Prior to the 1950s mathematics was not used to create portfolios. In the 1950s a young Ph.D. candidate, Harry Markowitz, chose as his Ph.D. topic the application of mathematics and in particular statistics to creating portfolios, creating what is now known as Modern Portfolio Theory (MPT). MPT which led to a Nobel prize being awarded to its creator, Dr. Markowitz in 1990, has significant issues in its implementation in software. Using the software to “optimize” a portfolio leads to poor results because of the tendency for the software to be “error” maximizers. A fundamental part of using the theory is having a deep understanding of statistics and the limitations of the theory and the advanced statistics that have been developed over the last 20 years or so to help remedy the issues.

  • How does your and Laura’s background uniquely prepare you to understand the advanced statistics and mathematics associated with “fixing” MPT?

Laura has a Ph.D. in financial planning focused on investment and economic theory requiring extensive mathematical and statistical modeling skills to interpret and apply the body of knowledge., I have a bachelor’s degree in engineering did post-graduate work in physics and engineering and am a CFA. Laura and I are avid readers of the latest research from academia, and we have also helped create our own software and tools to both create and test the performance of our investment models. We think few other advisors have the broad experience and technical backgrounds in statistics and mathematics to do what we can do.

  • How does your advanced knowledge of MPT help you avoid some pitfalls others may face?

Understanding MPT leads to understanding what qualifies as an “asset class” and which asset classes should be in client portfolios. We see many advisors who confuse the definition and consider things like High Yield bonds which are junk bonds as an asset class. That example cuts across both mistakes, they are not an asset class and even if they were they are not appropriate for most client portfolios. Their performance becomes very “equity’ like during bear markets which makes no sense since they have none of the potential upside and growth of stock. We are also very careful how we create views on future returns. When MPT was developed Dr. Markowitz did not expect users to rely solely on history to project returns for the future. He expected advisors to impose their own “views” on future returns. That is a very difficult process and something we will discuss in our next video but in our experience, many advisors do not know how to create valid views and they create views that are too short, for example, a view that the stock market will return 8% a year for the next five years. If you look at the potential range of returns for a volatile asset class like that you can see an enormous range for five-year returns, trying to predict returns for such an asset class is dangerous since that return prediction dramatically affects how much would be allocated to that asset class in a model.

  • This sounds technical, as a client how does it benefit me?

Our goal throughout this rather complex process is to reduce the risk in your portfolio to as low a number as possible. For example, if we are targeting a 6-7% return, a number that we derive as necessary from a detailed analysis of your goals and objectives, and a number that will help you meet those goals, we want to take the risk (volatility) you experience achieving that return as low as we can. There is an effect called “risk-drag”; that is the difference between an average rate of return, for example, the rate of return you receive for one year vs. the compounded rate of return you receive by chaining together each year’s return. The average rate of return never matches the compounded rate of return because of the variation or volatility of returns each year. If we can reduce that volatility through the correct use of advanced science and math, then your compounded rate of return will more closely match the average and you will end up with more wealth through time.