photo of toy robot

Automated Retirement Investing
I spent several hours yesterday (more than I intended) reading about so-called robo-advisers. These are relatively new firms plus some established ones that implement and manage portfolios via computer algorithms. They replace human financial portfolio managers at much lower cost. These new firms include Wealthfront and Betterment. Wealthfront is the pioneer, or at least one the earliest to offer algorithm driven portfolio management services. The approach is based upon a solid foundation of asset allocation, Modern Portfolio Theory and the use of ETF’s to keeping costs as low as possible. Wealthfront puts it this way: “By implementing a completely software-based solution, informed by world-class financial expertise, Wealthfront is able to deliver its automated investment management service at much lower cost. We democratize access to high-quality financial advice.”

The term robo-adviser, to my mind, conveys the wrong notion. Robo-advisers are not some mechanical mindless action. Think of these advisers as more like Dr. Spock in Star Trek, emotionless relentless rationality. These firms offer automated, algorithm-driven portfolio management. For the retirement investor, they take the emotion out of managing a portfolio. The emergence of these firms extends the roll of technology and the internet, which began with low-cost on-line investment and brokerage firms, notably Charles Schwab. They really challenge the present model of human managers for your retirement portfolio.
I have been aware of Wealthfront in particular since one of its founders is Burton Malkiel. He is the author of “A Random Walk Down Wall Street”, first published in 1973 and now in its tenth edition (at least). He is as close to a rock star as you can get in the index investing world. A series of articles on Market Watch initiated my interest. The foundation of the robo-adviser approach is asset allocation and I am critically interested in that topic. These advisers target the younger generations that really need help in saving for retirement. But that does not mean that investors that have already retired cannot take advantage as well.
The first article points out that robo-advisers can offer sound advice. Incidentally, it also circumscribes the limits of the algorithmic approach. Robo-advisers provide portfolio management only. But this is the basis upon which most financial management and wealth managers charge for their services. Robo-advisers cannot provide financial planning in a broader sense, such as advice about mortgages, college savings, insurance, annuities, how to handle an inheritance and other important areas. So the human advisers still perform an important role. I see this role as similar to lawyers. We have, on occasion, a specific problem that requires specialized expertise. We should be prepared to pay for this. But for everyday management of your retirement portfolio robo-advisers represent perhaps the best option.
Consider the basic premise that you cannot beat the market. Your returns over time are generally proportional to the amount of risk you accept. Asset allocation serves to mitigate risk. As noted in a previous post the research described in a book, Global Asset Allocation, by Mebane Faber of Cambria Investment Management, concluded that asset allocations proposed by several financial luminaries generated closely grouped performance results over time. He reported that the keys to success in building your retirement assets are minimizing your investing costs and maintaining your asset allocation. Robo-advisers reduce the emotional element in the investing equation. And they operate at a much lower cost than conventional portfolio managers. By the way, if you want an assessment of the damage to your retirement investments induced by emotion, see the post by Daniel Kahneman on the Wealthfront website about the illusion of stick picking skill: (Kahneman won the Noble memorial prize for economics in 2002.)
The second article in the Market Watch series is one that hypes up the situation, kind of John Henry versus the steam driven machine. In the end it gives the algorithmic approach its due and points out there are indeed circumstances such as an early retirement buyout that require a human intervention. It is clear to me though that for building and maintaining your retirement portfolio in the most cost-effective and efficient manner the robo-advisers have the edge.
The third article is an extended one that provides a good comparison of four automated advisers and four human advisers. It compares the portfolio compositions and the fees. The various asset allocations and portfolios presented provide a really helpful resource and it will worth your while to spend some time reviewing the results of the comparison.
One big issue for those of us that are retired and have a retirement portfolio is the cost of switching to a robo-adviser. If you had to sell a portfolio with substantial unrealized capital gains the cost could be prohibitively high. I investigated this briefly on the Wealthfront website. I will say that they are very sensitive to tax issues and I believe in this regard would be more attractive to a retired investor. The size of the problem of course depends upon the make-up of your current retirement portfolio. It looks like you can transfer a tax-deferred account quite easily without any serious penalties.
I actually signed up at Betterment for a 30 day free trial. I’m not sure yet what I really get to try unless I commit funds. But I did get a recommended portfolio allocation—48% stocks: 52% bonds. Betterment listed twelve asset classes and allocated dollars to eleven of them. It did not differentiate between a taxable allocation and a non-taxable allocation. Wealthfront on the other hand pays a great deal of attention to taxes and tax harvesting is one of the key elements of their approach. It recommended two allocations, one taxable and one non-taxable over eleven asset classes. A key difference in the two allocations is that Wealthfront included real estate and natural resources, while Betterment did not. I spent more time on the Wealthfront website. They provide a very helpful description of their asset class choices along with a detailed explanation of their methodology which is worth reading whether you use their service or not.
The Wealthfront methodology white paper offers estimates of expected returns and standard deviations and correlations for the selected asset classes. Expected returns are real returns, that is they subtract the impact of inflation. Wealthfront’s expected returns really caught my attention. They are real returns so might appear low to most of us accustomed to thinking in nominal rates of return. I think they are actually historically low and reflect the “new normal”. For most bonds their real returns are negative. Now wonder both Wealthfront and Betterment recommended what I consider high stock allocation for someone my age with limited income. For bonds they both proposed a very high allocation to municipal bonds. This surprised me and is not something that had occurred to me before.
What I can say at this point, especially if you are recently retired, is that your retirement assets will have to last you for a long time. You will need to squeeze every nickel out of these assets. Especially if you are not a do-it-yourself retirement portfolio manager you should seriously investigate these algorithm-driven portfolio management firms. The great obstacle most likely will be fear of the new, resistance to change and the obfuscations of your current adviser. At the very least your adviser should unequivocally demonstrate the value they offer you.