Has Goals-Based Investing Ruined Modern Portfolio Theory (MPT)?
Modern or mean-variance portfolio theory (MPT) is an important financial concept. But it has little practical value for retail investors when it to comes to asset allocation.
Recently, goals-based investing has grown popular with both financial advisers and robo-advice tools. Financial advisers continue to apply an ‘”asset allocation overlay” check to ensure goals-based portfolios are not too risky when viewed through the MPT lens.
It’s time to develop a more practical risk-management measure.
It has taken me 20 years of wealth management industry to say this: while MPT is an elegant theory, it fails when used for retail investor client portfolios.
Let me explain why.
The Exclusion of Illiquid Assets
My clients were real people who liked real assets, especially a roof over their heads. So when I offered to rescue their haphazard collection of bank deposits, stocks, bonds, funds, and real estate by devising an “optimal” asset allocation, I would always be stumped as to whether to include their primary residence.
Since real property is a fairly common method to build wealth in a tax efficient way, I didn’t see the logic of excluding it. On the other hand, including it would swamp the asset allocation with “idiosyncratic residential property risk.”
I had no choice. I had to ignore pretty much all real estate due to its lumpy nature.
Often clients also had large amounts in bank-fixed deposits that they weren’t keen on investing in capital markets just because my efficient frontier calculation recommended it. Nor were they willing to sell any large direct-share holdings acquired as part of an inheritance or employee stock option plan. As a result, I would devise an optimal portfolio of mutual funds available on whatever platform I happened to use.
How much more efficient would the portfolio have been if I had access to private equity and hedge funds, or indeed, to those assets available to investors in other parts of the world? Does the difficulty of accessing an asset class render a portfolio made up of only immediately available asset classes inefficient?
The practical constraints of having to exclude certain real property assets, or not being able to include other assets currently unavailable, seemed to compromise the point of the exercise.
Difficulties in Forecasting, Risk Profiling, and Definitions
Often savvy clients sought advice from multiple advisers. They would question why the recommendations for the right or optimal asset allocation varied. Rightly so.
To the clients, it appeared as though the various advisers used different assumptions for long-term expected returns without necessarily having a reasonable basis for their views. Without a doubt, all advisers struggle with forecasting volatility and correlations. Most advisers use historical data as the basis for their opinions, but data histories vary. As clients focused on the discrepancies, financial advisers realized the difficulty of long-range forecasting.
The bigger issue, however, is the risk-profiling process. There aren’t any definitions of risk profiling, let alone any industry standards on how to conduct it. Firms regularly use a set of arbitrary questions to arrive at even more arbitrary risk profiles.
After some regulatory intervention, it became possible to compare risk profiles based on expected return and risk outcomes. Unfortunately, the return and risk assumptions varied, so the risk profiles weren’t strictly comparable — they still aren’t. And there are still no regulatory standards on risk profiling.
I left my financial adviser role to become a research analyst. I yearned for intellectual rigor. The more I researched, the more I realized that the MPT issue starts with the very definition of an asset class.
What constitutes an asset class? How different does the risk/return have to be for a subset of an asset class to be carved out into a separate asset class? For example, why do domestic equities qualify as a separate asset class to global equities? Should emerging markets and high-yield debt be carved out of global equities and fixed income, respectively?
The Rise of Behavioral Finance and Goals-Based Investing
I was delighted to discover behavioral finance in the late 1990s. Finally, here was evidence MPT wasn’t wrong, clients were just irrational. Their mental shortcuts — their anchoring, hindsight bias, and overconfidence — explained why client portfolios and returns were suboptimal.
Inspired, I started to educate investors about their mistakes in the hope they would realize their follies. That is, until I realized I had organized my own finances in three mental buckets just like the irrational investors in the behavioral finance textbooks. I had a lot of cash in my short-term bucket, real estate in my medium-term goals, and all of my superannuation into equities as a long-term prospect.
I was behaving pretty much like my clients.
Gradually, the investment industry evolved from correcting investor behavior to designing client-centered products and services. It started with “nudges” like better designed default options, auto-enrollment, and automatically saving future salary increases. Next we saw the advent of goals-based investing, creating separate portfolios for each client goal: buying a house, an education, or funding retirement.
But financial advisers (and their robotic counterparts) continue to calculate the weighted-average asset allocation of the various goals-based portfolios and force fit it into the assessed risk profile — which is still determined by arbitrary questions. Even Ashvin B. Chhabra’s brilliant paper on goals-based investing, “Beyond Markowitz,” advocates the MPT-based asset allocation overlay.
Denouncing Nobel-Prize winning MPT is heresy. Or is it? Wasn’t reducing risk its purpose? Aren’t there better ways to do that?
As a financial educator, I hope to influence the millions of people in middle-class India, and I am concerned about scale. If we can’t reach all investors through financial advisers, how can we teach them to fish — or invest — themselves? If I did just fine building my own portfolio with a few mental buckets, why can’t we teach investors to do the same?
I finally had the courage to admit what I had struggled with throughout my career: MPT was never practical enough for retail investors. The time has come for the wealth management industry to admit it as well .
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: Getty Images/MHJ
Agree. But what are your proposed alternatives?
Goals-based investing, like that proposed in the Beyond Markowitz paper, can be overlaid with qualitative/quantitative risk management.
MPT is over simplification of a complex problem. Using variance as a risk metric might make it easy to develop models but it has two major limitations, in my view: (i) Quantitatively speaking it does not reflect the concave utility (ii) Even bigger issue is that it assumes that rational investors would take decision purely based on mean-variance framework. The most investors don’t even understand mean-variance framework, obviously the theory is bound to fail.
*Thus, when most investors do not even understand mean-variance framework, obviously the theory wouldn’t hold good in practice.
Bless you for saying this!!! I have looked everywhere for an advisor who would include our house as part of our portfolio and who could calculate what it might mean to add a small rental property or two. So there is a niche where your talents are desperately needed…
Thank you Richard. I realised the house problem 20 years ago but didn’t have an alternative, not realising my 3 mental buckets were the solution. Yes I now use my experience to do investor education. Experimenting with a simple to explain concept – would love your feedback. Check out TheMoneyHans.com
While goals based investing may be more appealing to the retail public, how does one go about building the portfolios for each of the goals? Are they single asset classes? Fixed rate bonds? MPT provides a process for determining an appropriate asset allocation, given the tools one has to work with. Markowitz never said it was an exact science, nor that it would lead to the perfect portfolio. Assumptions matter, as do constraints, and asset classes. The true value of MPT is a disciplined process as opposed to guessing which asset classes to use in what amount. There is a reason MPT has survived for 60+ years and continues to be used by the largest asset managers in the world.
Retail clients can benefit from an understanding of the probabilities of outcomes provided by an MPT analysis. Where risk tolerance questionnaires can fail, ranges of outcomes and historical performance of an asset allocation can add tremendous value to a client determining an appropriate allocation.
Agreed, the utility of MPT for retail investors is for risk diversification rather than efficient frontiers. But wealth managers try to force-fit it with arbitrary assumptions.
All I am saying is we need to vaguely right than precisely wrong.
All assumptions in investing are arbitrary, whether you use MPT or not. Goals based investing can help address behavioral tendencies, but you still have to build the portfolios for each bucket. For example, you mention your long-term superannuation would be 100% equities. Would it be globally diversified or all domestic? Large cap? Small cap? Determining that would require arbitrary assumptions. One could certainly decide to weight by market cap, however if a more efficient solution was available by optimizing based on risk, return, and correlation, wouldn’t one want to do that?
That’s the point….as a retail investor I don’t care about ‘efficiency’ which is more about balancing volatility with return. If you are investing for the long term,volatility isn’t as relevant. I invested in Asian equities with an absolute oriented manager because I thought that gives me a good chance of a good return without having to check my portfolio every day or month or quarter. I may not have the most efficient portfolio but the point is I don’t care. And this is when I enjoy finance. Now imagine a person who is not into finance. What do they really care about?
I misunderstood then, I thought your article was regarding advising clients, not from the perspective of a retail investor. Regardless, I think retail investors do care about efficiency, they just aren’t aware of the opportunity. I’ve yet to explain to a client the benefit of maximizing return for a given level of risk and have them say that’s not appealing to them.
What people care about differs from person to person. Your example of investing in an absolute oriented Asian equities manager reflects that you are much more concerned about losing money than you are about the level of return. Someone who buys a levered equity ETF cares more about return. MPT doesn’t require checking a portfolio with any level of frequency. It’s just a process that provides guidance in decision making.
If our clients could embrace that when investing for the long term, volatility isn’t as relevant, we would all be better off. But again, as you alluded to, all investors have to deal with behavioral tendencies, which means volatility is always relevant.
MPT fails because we don’t have an appropriate risk measure. In the real world there is no risk/reward trade-off over a wide rank of scenarios. The variance/covariance matrix is not stable. Asking people to come up with their risk tolerance without a sound way to measure risk is senseless. I think Mr. Mehrotra points this out well.
I agree that understanding variability in returns is useful for investors, but the profession has a lot of work to do in coming up with new ways to define risk. MPT is fatally flawed.
Thanks Catherine. Yes, we need to come up with a more intuitive and subjective way to help retail investors spread their risks. Perhaps something like Bridgwater’s All Weather Fund. I am exploring a factor-based approach that can be explained easily. Happy to hear other suggestions.
ps – It’s a Ms, not Mr. I know my name sounds Swiss German but it’s Sanskrit for swan.
As Howard Marks puts it, for most investors the only real measure of risk is the permanent loss of capital. That occurs for two reasons: investing in something that can become worthless, or an emotional response at an inopportune time, preventing mean reversion to occur for a given allocation. MPT is not fatally flawed, but mischaracterized. It is not designed to prevent or eliminate losses! Even the comment that the variance/covariance matrix is unstable may be a reference to 2008, when correlations theoretically went to 1 (for a brief period of time). However an investor who owned stocks AND bonds lost less money than an investor who only owned stocks, proving that MPT works. An investor’s appropriate risk measure is substantially dependent upon their time horizon. And even then, not necessarily their investing time horizon, but their emotional time horizon. An investor who has the ability to monitor their portfolio less frequently will find they can tolerate a lot more risk.
The fatal flaw is in characterizing MPT as a volatility elimination tool. Earning return requires a willingness to bear risk. MPT provides a defined methodology to determine what level of risk / return tradeoff an investor can tolerate.
See my paper with Markowitz, where we show that you can map goals based investing to MPT.
http://journals.cambridge.org/repo_A772rEdS
Very nice article, Hansi! Kudos for the fluid language and comprehensive coverage…
I use a combination for my clients. Risk profile based asset allocation, and within the asset class using factor-investing to optimize security selection and more importantly to avoid behavioral follies…
As a risk measure, clients feel more comfortable with Max Drawdown or Downside volatility, which I believe are more suitable to retail investore.
Nehal Joshipura, PhD
Isn’t this ‘both /and’ as in quantum physics? Ashvin Chhabra’s wealth allocation framework and MPT are like ‘being a particle and a wave’ – as confounding as that is for an advisor. Hansi, you’ve nailed it, and so did Ashvin Chhabra; you both advocate a more complete approach. MPT continues to offer an excellent metric for measuring the advisor, e.g., Wharton professor Marston’s alpha star uses the capital market line to illustrate to a client how they’re doing on a risk adjusted basis. Clients intuitively grasp this, just as they love the logic of Chhabra’s ‘buckets.’
MPT is still valuable if max. drawdown is taken as risk measure and the fact that correlations of most traditional assets can go up to 1 in crises taken into account. To avoid such desastrous losses, which retail investors fear most, but still provide good long-term returns, it is necessary to combine assets with high return and correspondingly high volatility but zero to negative correlation, particularly in crises. Besides equities and real estate this necessitates the allocation of alternative assets such as trend following managed futures, which are becoming more and more available for retail investors.
Thanks very for this post, Hansi–it’s an enlightening and a refreshing read. All the fancy underpinnings of MPT don’t mean much to the retail investor. I believe, also, that it’s of use to advisors ensure some kind of balance between risk and return even while their portfolio constructions are goal-driven.
Thanks for your feedback, Eunice.
Yes, goals-based investing also require for a risk/return framework. I am trying to tackle it by grouping goals into various risk/return buckets. So there are essentials, for which a low risk/return is fine. The long term goals can be invested in diversified equities. The aspirational goals have to be concentrated bets, hence high risk/return. The trick is to figure if and how to determine asset allocation.
Hello,
I tried accessing the article mentioned in the blog, but it does not work anymore.
http://journals.cambridge.org/repo_A772rEdS
Can someone help?
Thank you!
Max
But even in the Chhabra article, MPT is really doing the heavy lifting. The third mental bucket is arguably tacked-on and not relevant to every investor and the first two buckets can be considered jointly in a MPT framework with constraints on downside.
Thanks very for this post, Hansi–it’s an enlightening and a refreshing read. I believe, that it’s of use to advisors ensure some kind of balance between risk and return even while their portfolio constructions are goal-driven.