The unprecedented seems to happen all too frequently in financial markets. Is there something wrong with the way financial advisors build their clients' portfolios? “Post” Modern Portfolio Theory says yes, and it's gaining currency.
Modern Portfolio Theory (MPT) has been the very bedrock of investment management and, more specifically, portfolio construction and asset allocation, for decades. To oversimplify, one might explain MPT to clients this way: It is literally a mathematical proof for the idea that you shouldn't put all of your investment eggs in one basket. According to MPT, a portfolio of non-correlated assets — distributed across the risk spectrum — can lower the overall risk of a portfolio. This idea was first articulated by Harry Markowitz in his doctoral thesis, published in 1952 in the Journal of Finance. As the late and noted financial historian Peter L. Bernstein wrote, “[Harry] Markowitz's work on portfolio selection was the foundation of all that followed in the theory of finance and of the Capital Asset Pricing Model [CAPM] in particular.” (See Glossary of Terms on page 38.)
Of course, MPT has been picked apart by legions of critics over the years. But suddenly, in the aftermath of the recent stock market debacle — in which nothing seemed to work at all — critics of MPT are gaining currency. Frank Sortino, founder of the Pension Research Institute and the Sortino ratio's namesake, is perhaps the most well-known supporter of what, in a play on the popular term coined for 20th century critical theory, is being called “Post Modern” Portfolio Theory (PMPT). Other well known critics of MPT include hedge fund managers Nassim Nicholas Taleb (author of The Black Swan) and Ray Dalio, CEO of Bridgewater Associates, one of the largest hedge funds in the world with roughly $80 billion in assets under management. The very foundation of modern asset allocation just doesn't work, they say.
The debate between believers in the two different approaches to portfolio construction centers around how they define risk, and how that risk influences returns. MPT models risk using standard deviation above and below expected returns (also called mean variance). PMPT models risk using only standard deviation below expected returns (semivariance). In other words, MPT assumes that there is such a thing as upside “risk,” whereas PMPT proponents believe that only downside risk matters to investors.
This difference seems to give PMPT modeling greater power to predict disasters. In fact, applying MPT's concept of standard deviation to the monthly returns of the S&P 500 indicates a monthly loss greater than 12.8 percent has nearly no chance of happening. But it has occurred 12 times since 1926, according to Brent Bentrim, founder of Carolopolis Fiduciary Counsel, an RIA in Charleston, S.C. PMPT, say supporters, allows for last year's upset because it measures asymmetrical return distributions.
James Breech, a financial advisor and founder of Toronto-based Cougar Global Investments, is one of PMPT's devoted followers. You've probably never heard of him, but after reading this, you may wish you had. Cougar Global — which serves mostly family offices, foundations and wealthy individuals and calls itself “a global tactical asset manager with downside risk management” (in other words, a long-only absolute return manager) — has just $200 million in assets under management. But Breech, who long ago adopted PMPT, has won his clients strong returns over the years. Last year, in particular, when the S&P 500 dropped 37 percent, his portfolios yielded audited returns of 2.8 percent.
How'd he avoid the mess that upturned nearly every asset class (financial and otherwise) over the last 18 months? Good asset allocation modeling? Yes. Sturdy economic analysis? Certainly. But underpinning his approach to the market, his economic analysis and his asset allocation modeling, Breech says, is his allegiance to this unconventional way of assessing portfolio risk: PMPT.
The Post Modern World
Neither Breech, nor the PMPT ideas that drive his practice, have exactly been magnets for assets. (Breech's $200 million in AUM is small potatoes in the wealth management world.) But that may soon change. A comparison of Cougar's returns with those of his peers yields pretty astonishing results. In March, international brokerage and consulting firm Brockhouse & Cooper audited Cougar's performance and compared it to a universe of 145 of the world's leading money managers with the same global investment mandate. Cougar ranked in the 1st percentile for every single one of the last eight years, with annualized returns of 11.9 percent. Breech doesn't try to beat the market, even if he has on occasion. The risk clients take is defined by how much money they will need to fund specific goals. No, he doesn't use historic risk/reward averages or historic return projections, either. That's a product of MPT and he says the MPT way of defining risk is why portfolios blew up in the last 18 months.
“When Harry Markowitz presented the world with a mean variance analysis [the basis of MPT] for looking at risk/return, he was the first. No one had done anything like that,” says Frank Sortino, whose work in the 1980s and 1990s at the Pension Research Institute improved upon the Sharpe Ratio with a focus on downside risk measurement. (After all, upside volatility isn't necessarily a bad thing.) “The problem with the mean variance approach and Bill Sharpe's Capital Asset Pricing Model was that it assumed that every investor has the same objective. And that's just not true,” he says.
What is risk? Bernstein, investing legend and author of the book, Against the Gods: The Remarkable Story of Risk, said recently during a brief and philosophical interview about risk, posted in the McKinsey Quarterly, that it is essentially “that we don't know what's going to happen, good or bad.” In short, predicting the future is impossible — though both MPT and PMPT still try to do this with modeling. A major difference is that MPT assumes all investors have the same investment objective: to maximize the expected return for a given level of risk as measured by deviations around the mean. And so, for example, many retirement calculators suggest that 40 year olds who claim to have moderate risk tolerance plunk 40 percent of their assets in fixed income — which assumes these individuals, whether janitor or executive, will have exactly the same goals. Subscribers to PMPT say, conversely, that investors have different and often very specific goals. The focal point should not be the maximum return possible given a certain level of risk, but rather the rate of return that must be earned in order to accomplish these specific investment goals, such as retirement or paying for college tuition, with minimum risk. The risk, then, is defined as the possibility that the investor will be unable to accomplish the goal. As a result, returns below the target rate of return (“downside risk”) incur risk; returns above the target do not. With client portfolios suffering some of the largest losses in a generation, wouldn't everyone want a better handle on downside risk?
The funny thing is, this idea has been around for a long time. As Bill Sharpe wrote in his 1964 book, Capital Asset Prices: A Theory of Market Equilibrium under Considerations of Risk, Markowitz actually preferred downside risk or “semivariance” as a measure of investment risk. “Due to the complex calculations and the limited computational resources at his disposal, practical implementation of downside risk was impossible. He therefore compromised and stayed with variance,” Sharpe wrote. In short, Markowitz didn't have a powerful enough PC to do the math or make the models he thought would serve investors best.
When asked now whether PMPT and semivariance have since shown the simpler mean variance approach to be an inadequate risk measurement tool, Markowitz, who at 90 years old still teaches finance at the Rady School at UC San Diego, says no. “The mean variance approximation to expected utility is quite good. Going beyond that is not worth the extra fuss.” Translation: MPT is good enough, and it's easier to use. That's a tragic error, say Breech and other PMPT advocates.
Fighting a Monolith
Some PMPT proponents say PMPT is still an outlier mostly because they are fighting entrenched interests at marketing and technology firms that cater to the financial advice business. Frank Sortino, who has expanded his work on downside risk with a focus on determining upside potential, says that as long as these firms are making money with MPT, they have no incentive to change. (He has a new book coming out on the subject for institutional consultants and financial planners in November.) So even if Markowitz were to admit that a semivariance approach is “more plausible,” it's still a hard sell to an industry raised on mean variance. “The PMPT crowd has been marketing it like mad for years, but with little progress,” acknowledges Markowitz.
Brian Rom, the founder of Investment Technologies, a New York-based asset allocation software provider, says he gave up the fight against MPT after spending 10 years on the road attending debates, conferences and otherwise beating the pavement trying to convince nonbelievers that semivariance is a superior measure of risk for portfolio construction. Of course, he had his own financial incentives — selling his software. He founded Investment Technologies in 1986 and, with the help of Sortino, “introduced the first commercial applications of Post Modern Portfolio Theory, which uses downside risk measures for portfolio construction and performance analysis,” according to his company's website. It was Rom who created the Sortino Ratio. “Eventually, I gave up and just grabbed my little niche,” says Rom, who claims to have taken on MPT giants like Ibbotson and others. Still, today, more than 60 institutions use Investment Technologies' software worldwide, including Banc One, Deutsche Bank and CIBC. James Breech is also a client.
Philip Hensler, CEO and chairman of DWS Investments Distributors, the retail asset management arm of Deutsche Bank Asset Management, says the current situation for investors amounts to a “Bermuda Triangle of asset allocation — lower return expectations, higher volatility and increasing correlation of asset classes.” These three things, he says, negatively affect the efficient frontier, defined as a set of portfolios with expected return greater than any other with the same or lesser risk, and lesser risk than any other with the same or greater return. Hensler is marketing structured products as a possible solution for investors. Apparently, some advisors see the merits: DWS sold $800 million worth of the complex products in 2008, more than half to the RIA community. Hensler, an avid reader of behavioral finance theorists and mathematicians like Daniel Kahneman and Henri Poincaré, says the past is the past. “We're not saying MPT is wrong. The trouble is in the input variables. We've been using long-term averages for inputs. Historic averages have no predictive power at all.”
Indeed the concept of average expected returns, a central assumption in MPT, has taken a huge whack in the wake of the worst bear market in years, a bear so savage that it wiped out 12 years of equity returns in 16 months. Until now, who would have imagined the following could be true: Between 1969 and 2009, investing in 20-year Treasury bonds yielded better returns than investing in the S&P 500, according to research published in May by Rob Arnott, CEO of Research Affiliates, a pioneer in fundamental indexing who sub-advises for firms, such as PIMCO. So much for the idea that over the long term, greater risk means greater reward, huh?
The Unprecedented Happens All The Time
Nassim Nicholas Taleb, the author of The Black Swan, has been railing against the orthodoxy of both efficient market theory and MPT for some time. In an October 23, 2007 article for the Financial Times, he wrote: “In 1990, William Sharpe and Harry Markowitz won the [Nobel] prize three years after the stock market crash of 1987, an event that, if anything, completely demolished the laureates' ideas on portfolio construction. Further, the crash of 1987 was no exception: The great mathematical scientist Benoit Mandelbrot showed in the 1960s that these wild variations play a cumulative role in markets — they are ‘unexpected’ only by the fools of economic theories.” He concluded: “We learn from crisis to crisis that MPT has the empirical and scientific validity of astrology (without the aesthetics).” Taleb criticizes the establishment for underestimating the chances of the unexpected; his critics call him a “catastrophist,” one who always plans for a crisis, making him look a genius when one eventually comes to pass.
James Breech was in the same Wharton graduating class with Taleb in 1983 and, even though they didn't know each other then, they share a similar attitude to the events of last year. Breech has only contempt for people who speak of recent events as statistical anomalies. He also shares Taleb's dislike of MPT. But they appear to share something else, too — a perceived eccentricity that makes it difficult for the mainstream to appreciate or understand them. Now that Breech can show audited performance on the portfolios of his U.S. clients, it's getting easier to reach a wide audience, he says. But the next hurdle is getting them to grasp his approach. “Selling my concept has been hard,” he admits. “Eighty percent of my clients are entrepreneurs and other types who have always thought differently or taken a different path.” (Cougar's investment minimum is $250,000. Average account size is $1.4 million.) A focus on downside risk management is eccentric enough for all but the most conservative, but the other key element of Breech's strategy is a proprietary economic scenario analysis he conducts that is based on Mordecai Kurz's so-called Rational Beliefs Theory. (See glossary at left.) But Breech doesn't talk downside risk management or Rational Beliefs to clients anymore, even with more sophisticated investors like family offices, because it's too complex. “The emphasis is on avoiding bear markets,” he says.
In the simplest of terms, this is how Breech puts theory into practice: Every month, he assigns probabilities to five different economic scenarios based on one-year consensus forecasts for key economic variables. Then, taking a semivariance approach to risk, and using a “three parameter log normal distribution for each asset class,” he calculates expected returns for each asset class, and the correlations between them, under each scenario. This probability-weighted risk/return data is entered into Rom's software to create portfolios (using 13 ETFs) optimized to achieve the client's minimum acceptable return (MAR) goal. Most importantly, Breech says, he takes the least amount of downside risk possible to achieve MAR. So when Breech's January 2008 model showed there was a 10 percent chance that client portfolios would experience a loss greater than 6 percent, he moved nearly every last penny out of equities. What little he left in emerging markets equities he also pulled out in the following months. “We don't like to lose clients money,” he says.
Is there room for PMPT in the mainstream? Tom Idzorek, chief investment officer at Ibbotson Associates, thinks elements of MPT and the “gazillion” downside risk models being churned out in academia are cross pollinating — and that's a good thing. “I think we'll continue to see innovation, more advanced techniques for measuring risk come along,” says Idzorek. “Markowitz made a huge leap — we're making incremental improvements at this point.”