Executive Summary
The foundation of investment education for CFP certificants is modern portfolio theory, which gives us tools to craft portfolios that effectively balance risk and return and reach the efficient frontier. Yet in his original paper, Markowitz himself acknowledged that the modern portfolio theory tool was simply designed to determine how to allocate a portfolio, given the expected returns, volatilities, and correlations of the available investments. Determining what those inputs should be, however, was left up to the person using the model. As a result, the risk of using modern portfolio theory - like any model - is that if poor inputs go into the model, poor results come out. Yet what happens when the inputs to modern portfolio theory are determined more proactively in response to an ever-changing investment environment? The asset allocation of the portfolio tactically shifts in response to varying inputs!
(Editor's Note: A version of this article originally appeared in the June 2012 issue of the Journal of Financial Planning in the Trends in Investing Special Report.)
The evolution of the industry for much of the past 60 years since Markowitz' seminal paper has been to assume that markets are at least "relatively" efficient and will follow their long-term trends, and as a result have used historical averages of return (mean), volatility (standard deviation), and correlation as inputs to determination an appropriate asset allocation. Yet the striking reality is that this methodology was never intended by the designer of the system itself; indeed, even in his original paper, Markowitz provided his own suggestions about how to apply his model, as follows:
To use [modern portfolio theory] in the selection of securities we must have procedures for finding reasonable [estimates of expected return and volatility]. These procedures, I believe, should combined statistical techniques and the judgment of practical men. My feeling is that the statistical computations should be used to arrive at a tentative set of [mean and volatility]. Judgment should then be used in increasing or decreasing some of these [mean and volatility inputs] on the basis of factors or nuances not taken into account by the formal computations...
...One suggestion as to tentative [mean and volatility] is to use the observed [mean and volatility] for some period of the past. I believe that better methods, which take into account more information, can be found."- Harry Markowitz, "Portfolio Selection", The Journal of Finance, March 1952.
Thus, for most of the past 6 decades, we have ignored Markowitz' own advice about how to apply his model to portfolio design and the selection of investments; while Markowitz recommended against using observed means and volatility of the past as inputs, planners have persisted nonetheless in using long-term historical averages as inputs and assumptions for portfolio design. Through the rise of financial planning in the 1980s and 1990s, though, it didn't much matter; the extended 18-year period with virtually no material adverse risk event - except for the "blip" of the crash of 1987 that recovered within a year - suggested that long-term returns worked just fine, as they led to a stocks-for-the-long-run portfolio that succeeded unimpeded for almost two decades. Until it didn't.
As discussed in the 2006 Journal of Financial Planning paper "Understanding Secular Bear Markets: Concerns and Strategies for Financial Planners" by Solow and Kitces, the year 2000 marked the onset of a so-called Secular Bear Market - a one or two decade time period where equities deliver significantly below average (and often, also more volatile) returns. The article predicted that the sustained environment of low returns would lead planners and their clients to question the traditional approach of designing portfolios based on a single, static long-term historical average input (which leads to a buy-and-hold portfolio), and instead would turn to different strategies, including more concentrated stock picking, sector rotation, alternative investments, and tactical asset allocation. In other words, stated more simply: planners would find that relying solely on long-term historical averages without applying any further judgment regarding the outlook for investments, as Markowitz himself warned 60 years ago, would become increasingly problematic.
The Growing Trend of Tactical
Although not widely discussed across the profession, the FPA's latest Trends in Investing study reveals that the rise of tactical asset allocation has quietly but steadily been underway, and in fact now constitutes the majority investing style. Although not all financial planners necessarily characterize themselves in this manner, the study revealed that a shocking 61% of planners stated that they "did recently (within the past 3 months) or are currently re-evaluating the asset allocation strategy [they] typically recommend/implement" which is essentially what it takes to be deemed "tactical" in some manner.
When further asked what factors are being re-evaluated to alter the asset allocation strategy, a whopping 84% of respondents indicated they are continually re-evaluating a variety of factors: 69% indicated following changes in the economic in general, 58% indicated they watch for changes in inflation, and another 58% monitor for changes in specific investments in the portfolio. Notably, only 14% indicated that they expected to make changes based on what historically would have been the most popular reasons to change an investment, such as changes in cost, lead manager, or other administrative aspects of the investment.
Although not directly surveyed in the FPA study, another rising factor being used to alter investment allocations appears to be market valuation, on the backs of recent studies showing the value and effectiveness of the approach, such as "Improving Risk-Adjusted Returns Using Market-Valuation-Based Tactical Asset Allocation Strategies" by Solow, Kitces, and Locatelli in the December 2011 issue of the Journal of Financial Planning, and more recently "Withdrawal Rates, Savings ratings, and Valuation-Based Asset Allocation" by Pfau in the April 2012 issue, along with "Dynamic Asset Allocation and Safe Withdrawal Rates" published in The Kitces Report in April of 2009.
Notwithstanding the magnitude of this emerging trend towards more active management, it doesn't necessarily mean financial planners are becoming market-timing day traders. The average number of tactical asset allocation changes that planners made over the past 12 months was fewer than 2 adjustments, and approximately 95% of all tactical asset allocators made no more than 6-7 allocation changes over the span of an entire year, many of which may have been fairly modest trades relative to the size of the portfolio. In other words, planners appear to be recognizing that the outlook for investments doesn't change dramatically overnight; however, it does change over time, and can merit a series of ongoing changes and adjustments to recognize that.
Tactical Asset Allocation: An Extension of MPT
At a more basic level, though, the trend towards tactical asset allocation is simply an acknowledgement of the fact that it feels somewhat "odd" to craft portfolios using long-term historical averages that are clearly not reflective of the current environment, whether it's using a long-term bond return of 5% when investors today are lucky to get 2% on a 10-year government bond, or using a long-term historical equity risk premium of 7% despite the ongoing stream of research for the past decade suggesting that the equity risk premium of the future may be lower.
Consistent with the idea that financial planners are recognizing tactical asset allocation as an extension of modern portfolio theory and not an alternative to it, a mere 26% of financial planners answered in the Trends in Investing survey that they believe modern portfolio theory failed in 2008. For the rest, the answer was "no", modern portfolio theory is still intact, or at least "I don't know" - perhaps an acknowledgement that while MPT may still work, many of us lack the training in new and better ways to apply it. Nonetheless, that hasn't stopped the majority of planners adopting a process of making ongoing changes to their asset allocation based on the economic outlook and other similar factors.
Unfortunately, though, perhaps the greatest challenge for planners implementing tactical asset allocation is that we simply aren't trained to do so in our standard educational process. Some financial planning practices are responding to the challenge by investing in training, staff, and/or research to support a more tactical process. Others are responding by outsourcing to firms that can help; the Trends in Investing survey showed nearly 38% of advisors intend to outsource more investment management over the next 12 months, and 42% are already outsourcing more now than they were 3 years ago.
Regardless of how it is implemented, though, the trend towards tactical itself appears to have grown from a broad dissatisfaction amongst planners and their clients that the "lost decade" of equity returns has left many clients lagging their retirement goals. Even if diversified portfolios have eked out a positive return, it is still far behind the projections put forth when clients made their plans in the 1990s, forcing them to adjust by saving more, spending less, or working longer, to make up for the historical returns that never manifested. And as long as the secular bear market continues, the strategy will continue to be appealing. Ultimately, though, the sustainability of the tactical asset allocation trend will depend on it delivering effective results for clients.
So what do you think? Would you characterize yourself as a tactical trader? Is tactical asset allocation a short-term phenomenon, or here to stay? Is tactical asset allocation simply modern portfolio theory done right, or does it represent an entirely new investing approach?
Robert Henderson says
Michael,
Like the old “90%+ of investment performance is dictated by asset allocation” mantra, MPT is often misunderstood. It is important that investors and advisers understand the point you made about Markowitz’s assertion that allocations will likely not be static over time, but rather fluid, based on market and economic conditions.
I am a big follower of John Hussman and Ed Easterling (Crestmont Research). Both take different approaches to research, but provide a lot of the backdrop necessary to help make estimates of future potential returns.
I think the term itself (“Tactical Asset Allocation”) has also been misinterpreted by many to mean “market timing”. In fact, even standard buy-and-hold methods generally involve market timing by virtue of the fact that re-balancing represents one method of market timing.
jeremy mitchell says
I’d not characterize myself as a tactical trader, as I don’t think tactical trading is the same as tactical AA.
In fact, I do agree with Robert that TAA has been misinterpreted, and even misconstrued, by investment managers and our industry at large. TAA has become a catch all for all types of tactical investment and speculation. Markowitz’ comments are indeed instructive, in that he basically discloses that MVO and MPT are simply frameworks for thinking about asset allocation, not biblical instructions for investment practitioners. I perceive that his talks and writings have been quite instructive in that one might divine that Markowitz never thought his original work would gain such fervent, sometimes militant, following among practitioners.
In my view, “tactical trader” denotes high turnover strategies that are largely opportunistic and speculative in nature, while Tactical Investor or the connotation I perceive as being applicable to the term TAA, refers really to various methods of incorporating forward-looking quantitative or qualitative metrics to determining overal asset allocations, probably much more similar to how Markowitz really envisioned his ideas being implemented.
This being said, while correlation and standard deviation might be a useful framework for thinking about how assets can be combined in a portfolio, I think we (as an industry) focus, or have focused in the past, inordinately on the quantitative methods and MVO. David Swenson’s work is quite instructive, as he seems to be a very dedicated proponent of MVO, MPT, and of using forward looking assumptions…but these items are almost secondary to his risk management framework. In both of his books he spends a great portion of the texts evaluating asset classes for their fundamental risk exposures, such as liquidity, interest rate risk, default risk, manager risk, market risk, growth risk (GDP), etc. He then determines, for the purposes of whatever the investment objective is, which asset classes are suitable for inclusion in the portfolio, both from a risk exposure standpoint and from a hedging standpoint (i.e. US treasuries as a sole fixed income component for personal portfolios due to their lack of default risk and high liquidity in times of crisis).
Thus, Mr. Swensen practices some of these less quantitative, more qualitative methods for determining which assets are included in a portfolio from the get-go. Such a process, I would suggest, is the backbone of effective risk management in portfolio design. Some advisors and managers are so myopic in their pursuit of non-correlated assets, that they fail to take into account the fundamental attributes and risk factors of said asset classes and needlessly expose their clients to unnecessary and unproductive levels of risk.
Finally, though inclusion of certain assets in a long-term portfolio may take a more strategic bent, there are certain times when asset classes, which would be otherwise inappropriate, may be accretive on a temporary basis (i.e. high yield bonds in 2009-2011, etc). So called Global Macro TAA practitioners can be quite active in their trading, due to their high reliance on governmental and central bank activity and other economic news. I tend to be a big behavioral guy, so I tend to shy away from more active tactical managers unless they can demonstrate exceptional logical and computational rigor in their decision-making process.
That is all. 🙂
Rob Bennett says
Michael:
This is another super article. I am always encouraged when I see smart and good people addressing these fundamental questions, questions which I believe too many have been ignoring for too long.
It always rubs my fur the wrong way when people use the term “tactical asset allocation” to describe a decision to make changes in one’s asset allocation. My view is that changes that are tactical really are exercises in short-term timing and that the Buy-and-Holders are right that efforts in that direction won’t work.
Where are planners going to draw the line? Are they all going to make guesses as to what is going to happen to the economy in days ahead? Aren’t we going to end up with 20 different planners giving 20 different recommendations? Aren’t we going to end up with just personal opinion and with the idea that there is any sort of science to financial planning reduced to a joke?
My view is that what is needed is not acceptance of the need for tactical changes but acceptance of the need for strategic changes. Investors must change their allocations in response to big valuation shifts because, if they fail to do so, they are permitting their risk profiles to get wildly out of whack. Keeping your risk profile roughly constant is a strategic matter, not a tactical one.
That one form of change is justified because there is 140 years of historical data showing that it always increases returns and it always reduces risk. There is no such showing for tactical changes, so far as I know. I am suspect of the merit of making tactical changes while strongly favoring the idea of making strategic (valuation-related) changes.
Rob
Richard R says
Michael, keep pressing this. The financial industry has done a great injustice to Markowitz’s work as compliance departments, financial firms in general have applied this theory cowardly; in other words, when cycles change SO DO THE INPUTS. It’s best to leave it alone. Smaller firms can be more nimble-combining strategic allocations for secular cycles (as Ed Easterling identifies) along with a tactical overlay.
Thank you for bringing this info to the surface again. It’s anathema to dig too deep into Markowitz. The financial industry for the most part, doesn’t like what it discovers.
Thank you!!
Stephen C says
Michael, excellent article and good points. Of late there have been some excellent academic papers on a form of tactical asset allocation associated with the identification of “regimes.” I direct you in particular to an article in the FAJ (CFA Institute) by Kritzman, Page, and Turkington: Regime Shifts: Implications for Dynamic Strategies. I rejected the idea of tactical asset allocation based on guesses and subjective interpretation of data and therefore always avoided it. Now some strong mathematical techniques have been applied to create methodologies to assist in making tactical decisions. We are studying this here and intend to incorporate them in our portfolio management in the next few years. Unfortunately the mathematics are daunting and the analysis requires the use of high level software packages like MatLab or the R language probably placing its use out of the range of many advisors.
Kenneth Solow says
Michael, as you know this is my favorite subject. What is missing from MPT is any notion that valuation matters to subsequent asset class returns. MPT works when layered on top of the efficient markets hypothesis. However, once you introduce the idea of market inefficency, you find yourself on the slippery slope of answering the question, “what constitutes fair value?” Of course, Graham and Dodd tackled this subject long before Markowitz authored his famous paper. Our industry is hoping for a quantitative “magic bullet” that will give planners the “correct” inputs to MPT, or any other model, that will relieve advisors from the task of trying to understand WHY markets move. I’m afraid there is no such magic bullet, and we will have to acknowledge that investing is a craft, as well as a science. As such, some advisors will be better at investing than others. I realize this isn’t good news for an industry that uses MPT to pound out “scientific” portfolios located on an efficient frontier. MPT is a terrific model for financial planners who think of investing as just one of a suite of services they provide their clients. Unfortunately, I’m afraid those who are looking for an easy quant answer to the valuation problem, which fits neatly into their current paradigm of business structure, time management, and client service, are going to be disappointed.
Steve Smith says
Not sure where the idea came from that “what is missing from MPT is any notion that valuation matters to subsequent asset class returns.”
Reversion to the mean of securities prices has always been integral to MPT.
Michael Kitces says
Steve,
Reversion to the mean is different than valuation.
Valuation, and secular market cycles, suggests that the mean itself is not always the same, and instead changes over time in different environments.
– Michael
Actually the discussion should be rather about WHY planners
– aren’t aware of modelling illusion
– don’t look at century old worldwide benchmarks and up to 2hundred years old US ones
– don’t disclose to and discuss with their clients the SWOT of active or passive or mixed
– CMH (Bogle) either
– and whether any AA can really work w/o a deep understanding and client involvement… which is a clear condition for comprehensive & fiduciary planning service.
Not MPT but mainstream planning needs rethink.
Did you ever ask yourself this questions?
Or too busy with TAA decisions 🙂
RE:”Improving Risk-Adjusted Returns Using Market-Valuation-Based Tactical Asset Allocation Strategies”
Is there not issues with data mining in this study. The top and bottom decile P/E environments can not be known ex-ante.
Why no out-of-sample analysis?
Curious,
Eric
Eric,
Ultimately, P/E ratios define earnings and profitability of the company (as an inverse E/P ratio).
Although we didn’t frame it quite this way, we could have simply said “Buy stocks when their (earnings) yields are above 10%, and sell them when (earnings) yields are below 5%, simply because 5% is a lousy return for risky stocks and 10% is a decent return for risky stocks.”
The point is not at all about data-mining the results. The point is simply to set parameters on what a reasonable deal on stocks is, in ANY environment, and buy more when the yields are good and less when the yields are bad.
The thresholds we used in the study correspond almost precisely to this framework, as the buy signals were around P/E ratios of 10 and the sell signals were for P/E ratios north of 20.
I don’t believe an out-of-sample analysis is necessary to make the point that 5% yields are worse than 10% yields. It would be self-proving, as the yields speak for themselves.
But strictly speaking, since the peaks and troughs have been the same for almost every secular cycle for 150 years, we did some informal out-of-sample testing early on in our analysis and found the exact same buy and sell points every time, as people NEVER buy stocks for long when yields are below 5% and ALWAYS get a better return when yields are above 10% (by definition).
– Michael
Agree on buy equities when yields are >10%. Except this does not addresses the alternative, as the short term bonds yielding next to nothing, 10 yr is at 1.45%. How about the corporate balance sheet? Should not we give consideration to this?
Ram