This blog post was originally published on the MoneyGeek.ca blog by Jin Choi. The website no longer exists, but Jin has graciously allowed us to re-publish his research for the benefit of future investors forever.
Image credit:robuart/Shutterstock.com
A couple of months ago, I surveyed MoneyGeeks members to understand which new features they most wanted to see. Of all the options I made available, the members most often requested the ability to create their own optimized portfolios.
For those who are unfamiliar with the concept, portfolio optimization is a process which takes a number of investments and outputs a portfolio allocation that minimizes the portfolios risk without sacrificing potential returns. For example, if we specify that we want a portfolio that consists of Stocks A, B and C, portfolio optimization could spit out something like 20% in Stock A, 30% in Stock B, and 50% in Stock C, such that the risk of the portfolio as a whole is minimized. For a longer explanation, please read my previous article on this subject.
Today, Im happy to announce that such a feature is now available for MoneyGeeks paid members. If you are a member, you can access the tool using this link, or by going to the Members Only -> Model Portfolios menu, and then choosing the Create New option at the bottom of the page. Ive coded this tool myself, but the algorithm is based on the Black Litterman Model, which is the product of some famous employees of Goldman Sachs.
For the rest of this article, I shall explain how to use the new tool.
The first step to using the tool involves choosing the investments youd like to include in your portfolios. As of the current version, you can include either U.S. or Canadian stocks and Exchange Traded Funds (ETFs). You can think of ETFs as lower cost versions of mutual funds, and you can learn more about them in my free book.
In order to include an investment, you must specify the symbol, forecasted returns and the country that the stock or ETF trades in. For example, if you think that Apple stock will return 8% per year including dividends, you should specify AAPL for symbol, 8 for forecasted returns and U.S. for stock exchange.
You can specify up to 20 stocks and ETFs in your portfolio. The limit is necessary because portfolio optimization takes a lot of computing resources, and the more stocks and ETFs you have in the portfolio, the longer the optimization takes. Optimizing a set of 20 investments will generally take a few minutes.
Next, the Number of Portfolios setting asks you how many portfolios you want to create with the investments you mentioned. While all of the portfolios will contain the same investments, they will each have different risk and return characteristics. Furthermore, the first portfolio of the group will always be the most conservative, while the last portfolio will always be the most aggressive.
For example, lets say you choose to include stocks A and B in your portfolios, and that B is riskier and potentially higher returning than A. Then if you choose to build 3 portfolios, you may get something like 70% A, 30% B for the first portfolio, 50% A, 50% B for the second portfolio, and 30% A, 70% B for the third portfolio. Note that each of these portfolios will have been optimized to minimize risk.
Lastly, you will have to specify your conviction level. The conviction level basically asks how confident you are about your forecasted returns. For example, if you said that you think Apple will return 8% per year, are you dead certain that it will return about 8% per year? Or do you only have a mild gut feeling that Apple will return about that much?
If you specify a lower conviction level, the algorithm will tend to create more evenly allocated portfolios. For example, given stocks A,B,C, the algorithm may create a portfolio that allocates 25% to A, 35% to B and 40% to C. If you specify a higher conviction, on the other hand, the algorithm will create more unbalanced portfolios. For example, with the same stocks, the portfolio allocation may look something like 10% to A, 20% to B, and 70% to C. Let me explain why this happens.
Lets suppose you had an extremely strong conviction that Apple stock will go up by 8% per year. Also, lets say you have an equally strong conviction that Blackberry stock will go up by 6% per year, and you decide to include both in your portfolio.
As it turns out, Apple stock has historically been a safer stock than Blackberry, which means that Apple stock is assumed to generate higher returns at lower risk. If this is the case, there is no reason to include Blackberry in your portfolio, because its inclusion would only lower returns and heighten risk. In this case, our portfolio optimization algorithm will likely award almost all of its allocation to Apple stock, creating a very unbalanced portfolio.
But what if you only have a mild conviction that Apple will outperform Blackberry? In that case, the algorithm will give more allocation towards Blackberry, so that if youre wrong and Apple generates lower returns than Blackberry, your portfolio will perform better than if Blackberry were hardly featured. Diversification works as a protection against errors in judgment, and specifying lower conviction levels will give you more of that protection.
Once youve specified all of the information required on the page, you may click on the Create Optimized Portfolios. As Ive said, the process may last up to a few minutes, depending on the number of investments you included. While processing, our program retrieves the three year history of every investment, and analyzes the historical relationship between each of the investments to figure out the smartest way to diversify.
When the process is done, the site will take you to the page that shows your portfolios. This page will look like any of the other existing model portfolio pages, and tools like the risk analyzer and tax efficient portfolio allocator will work for your custom portfolios as well. You may modify the portfolios anytime by clicking on the Modify Portfolio button in the Overview tab. Youll also be able to track the performance of your custom portfolios in your own Performance page.
I hope that this new tool will prove useful for you. If you have any questions on how to use the tool, or if youd like to request additional features, please feel free to ask in the Q&A Forum.
This blog post was originally published on the MoneyGeek.ca blog by Jin Choi. The website no longer exists, but Jin has graciously allowed us to re-publish his research for the benefit of future investors forever.
