The Hows, Whys, and Right and Wrong Way to Use Asset Allocation
The following is the fifth and final installment of a special five part series meant to be shared by professionals and non-professionals alike. This particular series covers
only one of the 7 Deadly Sins Every ERISA Fiduciary Must Avoid.)
Zombies have nothing on asset allocation. The current practice of asset allocation has been repeatedly shown to fail to yield the results it was intended to produce. We have, again, as others before us have, provided the empirical evidence that details how and why asset allocation fails. Asset allocation should have died a long time ago, but it hasn’t. Why not? Perhaps asset allocation isn’t what we think it is. Perhaps asset allocation is, and always has been, something more subtle, more instinctual, then we have imagined in these past few decades. Maybe there’s a reason to keep asset allocation around that has nothing to do with the way it’s thought of today, but harks back to an earlier era. We answer these and many more questions in our concluding segment to this series.
Recap of test results
We learned in Part III of this series that asset allocation doesn’t work for short-term periods. “It was never designed to do so,” says Paul Ruedi, CEO of Ruedi Wealth Management, Inc. in Champaign, Illinois, “it was never designed to do so as there is no optimal portfolio ahead of time.” He says mean variance optimization “is a silly way to invest…unless you like disappointment.”
The problem with short-term asset allocation is that it requires using short-term historic performance periods. These prove to yield inconsistent results. “Who on the planet earth would think they could design an optimal-short term outcome based portfolio?” asks Ruedi. “Any reasonable distribution of returns would suggest to anyone that mean variance optimization and short-term asset allocation (prediction based) are a fool’s errand.”
Unfortunately, the elegance of structured mathematical formulae makes asset allocation difficult to ignore. While the broader consensus appears to reject short-term asset allocation, there’s a reluctance to feel the same way about long-term asset allocation. “Asset Allocation is not perfect but it is the best we have,” says Jeffrey A Bogart, a Registered Investment Advisor with Sila Wealth Advisory in Mayfield Heights, Ohio.
But Part IV revealed the results of the long-term portfolio optimization test. They, too, showed the futility of asset allocation as is practiced by many. Unlike the common perception that a mixed portfolio reduces risk over the long term, the test showed a portfolio of 100% stocks offered not only the best range of performance results (based on highest, median, and lowest returns), but did so at virtually the same variance as other stock/bond mixes.
More surprisingly, rebalancing did not change the results. “There is no magic to rebalancing,” says Ruedi. “It isn’t going to increase returns other than by random chance. It isn’t going to deliver the holy grail of higher returns with less risk.”
All this is not news. Prominent researchers have for many years questioned, whether directly or indirectly, some of the fundamental axioms of using mean variance optimization to produce asset allocation “strategies.” (Indeed, one might go as far as to suggest combining the terms “asset allocation” and “strategy” is a non sequitur.) What does this leave us to conclude about asset allocation?
Asset Allocation is Dead!
As difficult as it is for many to admit, asset allocation, while a great common sense concept, fails to prove itself as popularly practiced. Portfolio optimization works wonders in spreadsheets and more complex computer programs, and it does have the merit of producing beautiful pie charts in vibrant living colors. But, unfortunately, as a prediction based tool, it’s no better than throwing dice (or, if you prefer, spinning the roulette wheel).
Nothing exposed the myth of asset allocation more than the opening decade of the new millennium. The promise of protecting the downside through asset diversification failed dramatically during the market extremes. “When bad things happen, all correlations go to one, so diversification does not work well, if at all,” says Joe Gordon, a partner at Gordon Asset Management, LLC in Durham, North Carolina. “In 2000 to 2002, bonds did OK and value investing, dead from 1995 to 1999, came alive as well, as valuations began to matter again. Small and mid-cap value did real well. In 2008, non-correlated managed futures (mostly available only to accredited investors then) returned over 20% and treasuries over 30% in Q4 2008.”
But did asset allocation tell investors to place their assets in these specific classes in advance of these broader downturns? Short-term asset switches sound eerily like the dreaded “tactical” asset allocation, a.k.a., “market timing.” Sure, they’re up one year, but there down the next. Are we to expect disciplined long-term retirement savers to change their allocation with each change in the wind?
As we found out, this led to the infamous “sell-low/buy-high” behavior we saw during the depths of the recent market downturn. Ryan O’Donnell, Wealth Manager and Founding Partner of The O’Donnell Group located in Chico, California, says, “Look at investors who stayed the course during those downtimes, rebalanced when things got bad vs. those that tried to time the markets. Asset allocation doesn’t mean you won’t lose money; it means that you will essentially get the returns and variability for your given portfolio. It’s not sexy to say ‘when the market crashes, your portfolio is going to go down 30%,’ but that’s the truth.”
There’s a more inconvenient truth coming out of the realization that mean variance optimized asset allocation doesn’t work.
Bad news for Robo-Advisors and Target Date Funds.
“The problem is not with asset allocation, provided an intelligent allocation in the first place,” says Michael Prus, President of Scale Investment Group in Detroit, Michigan. “The problem is lack of understanding by both investors and many advisors that there is no magic silver bullet to all investment questions.”
Yet, if we’ve learned anything from behavioral studies, particularly as they apply to finance and economics, people – both professionals and the clients they serve – tend to look precisely for that magic bullet. They seek the ridiculously simple, easy-to-understand, easy-to-implement, decision. Consequently, we see the tidal surges of investment fads. Asset allocation was one such fad that has produced ripples of other fads based on its theology, including target date funds and, more recently, robo-advisors.
Therein lies the dire future for both of these popular products. Target date funds rely on a theory that an asset allocation should change as one ages. This may be true, but the trick is how you implement that change. If you’re relying on portfolio optimization techniques, then shareholders may be in for a nasty surprise at some point in the future. This problem is only exasperated in the case of robo-advisors, for they are predicated on the absence of any human supervision. Their dependence on computerized portfolio optimization to recommend asset allocation is exactly the misapplication of asset allocation we’ve been highlighting in this series of articles.
“Why in the heck would anyone think for a minute it would work?” says Ruedi, who describes mean variance optimization as “loading up on the stuff that has done the best over the past N periods. Good luck. Twenty years ago when mean variance optimization was just the buzz, I quickly figured it out as garbage in and garbage out. Anything you put in was essentially garbage. I can’t believe anyone is still using mean variance optimization.”
There’s just no other way to say it. Asset allocation is dead…
Long Live Asset Allocation!
In the immortal words of Otis Day (with the loyal support of his Knights, no doubt), “Now wai-a-ait a minute!” Remember, there is an intuitive sense to “asset allocation.” The problem with practicing asset allocation under the guise of mean variance optimization lies in its false promise, for there is no (and never was) causation. In other words, in most scientific theories, some action (or actions) directly cause some (sometimes equal and opposite) reaction. The dots of causation are clearly connected. No such connection exists with mean variance optimization and asset allocation (indeed, as in most stochastic models, there is a mistaken assumption regarding the independence of events). Still…
Something nags at the brain. We know statistical representations bear no causal relationship, but the notion of asset allocation cannot be denied. It’s not merely about diversification for the sake of diversification. There’s a greater meaning, a greater “there” there.
We know it isn’t the promise of reduced downside risk. We also know asset allocation isn’t the all-encompassing “93%” of anything when it comes to investing. Froilan Rellora, Chief Investment Officer at Catalina Asset Management in Mesa, Arizona, says, “Asset allocation is important when it comes to portfolio management, but investors shouldn’t rely on it to the point where 100% of their portfolio is invested in that manner.”
What is it about asset allocation that refuses to go quietly into the night? (Dare we say it “Shouts” out?) Suggesting asset allocation doesn’t work “is like suggesting a hammer doesn’t work because the user of the hammer was using the wrong end of the hammer,” says Ruedi. “How hard is asset allocation to understand in the end? Let’s see, the more money I invest in assets that historically earn 10% a year vs. asset classes that have historically earned 5%, the higher the expected return of my portfolio is. I suggest folks begin to use the correct end of the hammer.”
Ah, if it were only that simple.
The Dilemma of Timing
Ronald Surz, President, PPCA Inc in San Clemente, California, says, “The observation that ‘all-stocks wins in the long run’ has been Professor Jeremy Siegel’s position, especially in his book Stocks for the Long Run. It’s like Dirty Harry’s question: ‘Do you feel lucky today punk?’ We each go through the ‘risk zone’ only once. Losses suffered during the 5-10 years preceding and following retirement can dramatically reduce our standard of living and our ability to have enough money for our lifetime.”
Therein lies the rub. For all our tests in Parts III and IV, we have been shackled by the same nagging frailty of all experiments in the realm of finance (in particular, portfolio and investment theory): we have been burdened by the albatross of the “average” return. Over long periods, returns do regress towards a mean. But as the periods get shorter, the variability of returns increases. Here’s an example of the variability of what Part IV concluded was the best asset allocation, a 100% stock portfolio:
How Return Variance Decreases as Time Interval Increases
The graph clearly shows the significant decrease in the variability of returns as we move from 1-year reporting periods to 40-year reporting periods. Of interest, the median return falls between 10% and 11% for nearly all time intervals. This shows the fallacy of putting your faith in the “average” return. Of note, as we approach reporting periods of 30-years or more, the highest and lowest returns converge on the mean. This phenomenon has come to be known as “time diversification.” This mean, in investments as with many other things in life, time heals all wounds.
But what of time periods of less than 30-years, when the variance in returns starts to become more meaningful? Could this be the vindication for asset allocation? Is it possible for asset allocation to dampen this volatility in such a way that makes the 100% stock allocation less attractive? The following table, albeit a bit difficult to read, compares three different asset allocations: The extremes of 100% stocks and 100% bonds, as well as the traditional 60% stocks/40% bonds asset allocation.
Performance Comparison of Three Different Asset Allocations Over a Range of Time Intervals
The table contains the highest, median, and lowest return for reporting periods from 1 to 40 years. It’s color coded to indicate the best (green) and worst (red) performing asset allocation for each of those three categories in every reporting period. We immediately see the 100% stock portfolio has the best highest return and the best median return in every single reporting period. It’s the lowest return category where things begin to get a bit dicey for the 100% stock portfolio. It does perform the best for all reporting periods of 23 years or more, but the other two allocations capture all the remaining reporting periods. Furthermore, the 100% stock allocation possesses the worst lowest performance for all reporting periods of 14 years and less. In terms of the “best” downside for the periods of 22 years of less, the 100% bond portfolio wins for periods of 6 years and less, while the traditional 60/40 asset allocation does the best for periods between 7 and 22 years.
Here we begin to see why asset allocation has that intuitive feel we’ve been talking about. As the investor comes closer to the date he needs to start spending his investment, the 100% stock portfolio, while continuing to offer the chance for the greatest and most consistently highest returns, also begins to take on an element of higher downside risk. If asset protection is the primary goal, then this chart suggests a shift to the 60/40 allocation and, ultimately, a shift into the 100% bond portfolio.
Again, if it were only that easy.
Practical Uses for Retirement Savers and Fiduciaries
Robert R. Johnson, President and CEO, The American College of Financial Services located in Bryn Mawr, Pennsylvania, says, “There is no free lunch in terms of protecting investors against adverse market moves. One can certainly purchase put options on a broad index to protect against falling markets. But, that kind of portfolio insurance is expensive. A portfolio that fully participates in the upside of markets while not declining in market corrections is akin to a unicorn.”
Perhaps the answer lies in the pre-Modern Portfolio Theory application of asset allocation.
The Pension Plan Approach
Let’s begin with pension plans. In the earliest days, pension plans focused almost exclusively on bonds. This reflects the tradition of trusts in general which, in some states, were prohibited from investing in stocks well into the twentieth century. Following the securities industry overhaul in the 1930s, with the help of finance professors like Ben Graham, stocks became attractive long-term alternatives.
Following World War II, when the retirement plan industry blossomed, pension plans and profit sharing plans were managed in a “balanced” fashion. We often refer to this as the 60/40 split referenced earlier, but the actual allocation was determined by the specific demographics of the employees within the plan. A plan that had a higher than average percentage of younger workers would find its stock weighting increased. A plan that had a higher than average percentage of older workers would find its bond percentage increased. In either case, these plans all exhibited some variation of this 60/40 split. How did this come to be?
Irene Burke, Principal and Consulting Actuary, Burke Group, Rochester, New York, says, “Section 404(a)(1)(C) of ERISA states that fiduciaries of defined benefit pension plans have a duty to diversify investments. This is a responsibility under the prudent person rule, which would prevent them from investing ‘an unreasonably large proportion’ of assets in a single security or a single type of security. Hence, the separate allocation of assets to fixed income investments and equities. However, a plan may invest in a balanced mutual fund or a combination of fixed income and equity funds.”
If this sounds like the usual “diversification” justification for asset allocation, then you’ve been paying attention. Pension plans in particular, however, had a more urgent need for bonds. They need a dependable source of income to pay out to the plan’s retirees. “Often plans structure a bond portfolio such that future interest and maturity values will cover the expected benefit payments for a specific duration,” says Burke. “This ensures that there will be enough liquid assets to pay benefits when due.”
While this might work well for asset pooling vehicles like retirement plans, utilizing this strategy on an individual basis could present some very practical problems. Burke says, “To set up a bond portfolio to generate sufficient interest streams to cover future income needs, a bond portfolio needs to be very large. Let’s say you wanted to generate $50,000 in annual income. Assuming a 5% interest rate, you would need $1,000,000 in long term bonds.”
Given today’s yield curve, you’d need far more than that million dollars. If you aren’t worried about future inflation, you can buy 30-year treasuries (yielding 3%), but you’d need $1,700,000. If you care about inflation and you decide to build a laddered portfolio of 10-year treasuries (yielding 2.25%), you’d need $2,225,000.
Maybe there’s another way that more likely in reach of the typical retiree.
“Assigned-Asset” or “Bucket” Asset Allocation
Asset allocation does have a purpose, albeit very limited and only relevant for certain styles of investing. You may have heard of “bucket” or “assigned-asset” investing. This is where you separate your vast portfolio into “buckets” for each of your life’s goals and assign those assets to those life’s goals. Different life goals may require different types of investments, and this can be represented by different asset classes. Prus says, “The alternative is to accept that even a great strategy like asset allocation doesn’t always work all of the time. A better approach is to make sure your allocation is appropriate for you in a worst case scenario situation.”
This, by the way, is the idea behind a pension plan investing in bonds for the purposes of generating income. There’s a need for that income (the pension liability), so bonds are purchased to address that goal. On the other hand, current workers will likely receive a greater payout, so the pension plan needs to grow in order to pay out those future liabilities. For this goal, the pension seeks assets with superior long-term growth potential. That would be stocks.
Even in the case of pension plans, as the current retirees begin to deplete the bond holdings, new bonds must be bought by selling stocks. This same idea can be used by individuals. Ruedi says, “Rebalancing is a common sense venture. You create a financial plan based on your goals. You then back into a portfolio that historically would have gotten you there with room to spare. You don’t take on any more risk that necessary to do so. If the equity allocation rises above the needed allocation, duh, reduce equities. I guess you could call that rebalancing.”
What does the father of modern portfolio management have to say about this?
Ben Graham Redux?
Ben Graham was a true pioneer of modern investing. Called the “Father of Securities Analysis” and the “Dean of Wall Street,” he influenced many successful portfolio managers, most notably Warren Buffet. “Asset allocation” wasn’t nearly the term it is today when Ben Graham advocated for maintaining a 50% stock/50% bond asset allocation. He did permit overweighting one way or another, but never holding less than 25% of either class. His basic idea was this, “The sound reason for increasing the percentage in common stocks would be the appearance of ‘bargain price’ levels created in a protracted bear market. Conversely, sound procedure would call for reducing the common-stock component below 50% when in the judgment of the investor the market level has become dangerously high.”1
Graham didn’t see investing in asset classes merely for the sake of investing in asset classes. He was a numbers guy, and he developed a numerical relationship between treasuries (which gave him a “risk free rate of return” number) and stocks (which he valued based in part on that risk free rate of return number). For him, it was all about balancing safety with growth.
“Graham insisted on having a margin of safety in his investments and therefore paid close attention to valuation,” says David Larrabee, Director, CFA Institute, Charlottesville, Virginia, says. “As a result, asset allocations were driven by his security selection. Age-based or target-date asset allocation strategies that are popular today typically don’t consider valuation levels. The advantage of Graham’s approach is that it allows a skilled portfolio manager the flexibility to ‘buy low and sell high,’ and deliver superior performance over time.”
But is this practice really “asset allocation”? Yale Emeritus Professor Roger Ibbotson, Chairman and CIO at Zebra Capital and Founder of Ibbotson Association, says, “Going for the bottom-up is very different than going top-down. Value is certainly a piece of asset allocation. In the overall perspective, it’s difficult to do asset allocation from the bottom-up. You’re going to end up with assets that might not give you the diversification you need, Warren Buffet aside.”
Ben Graham didn’t practice diversification merely to say he was diversified. He only cared about buying stocks for an amount less than their value. This casual attitude towards diversification might be questioned in today’s world of asset allocation, but Graham’s performance record (said to exceed 20% per year) and that of his most prominent disciple (Warren Buffet) make his philosophy and practice difficult to ignore.
A Look at Things to Come
Asset allocation deserves neither the credit nor the blame it has been assigned. Neither should it be idolized nor should it be relegated to the trash heap of history. To paraphrase Jessica Rabbit, asset allocation is not bad, it’s just practiced that way. The fault lies not with the idea of asset allocation, for it was an art form long before mean variance optimization came into vogue. In that sense, asset allocation is merely a victim. “Asset allocation is not a hoax,” says Bogart. “[It] is based on the Noble winning prize research of Dr. Harry Markowitz’s Modern Portfolio Theory.”
And that may be a clue of things to come…
Part I: 7 Deadly Sins Every ERISA Fiduciary Must Avoid: The 5th Deadly Sin – Asset Allocation
Part II: How’d an Innocent Fiduciary Like You End Up Asset Allocating?
Part III: Asset Allocation’s Greatest Failure: Short-Term Investing
Part IV: Why Asset Allocation Doesn’t Matter In The Long Run
Part V: The Hows, Whys, and Right and Wrong Way to Use Asset Allocation
Are you interested in discovering more about issues confronting 401k fiduciaries? If you buy Mr. Carosa’s book 401(k) Fiduciary Solutions, you’ll have at your fingertips a valuable reference covering the wide spectrum of How-To’s (including information on the new wave of plan designs) every 401k plan sponsor and service provider wants and needs to know. Alternatively, would you like to help plan participants create better savings strategies? You can buy Mr. Carosa’s latest book Hey! What’s My Number? How to Improve the Odds You Will Retire in Comfort right now at your favorite on-line or neighborhood book store.
Mr. Carosa is available for keynote speaking engagements, especially in venues located in the Northeast, MidAtantic and Midwestern regions of the United States and in the Toronto region of Canada.
1Graham, Benjamin, The Intelligent Investor, Harper & Row, 4th Edition, 1985