Contact InformationJoleen Schultz
For press inquiries:
Lessons Learned from the Financial Crisis
An interview with Nobel prize-winner Harry Markowitzby Zach Jackson
In the early part of the 21st century, growth in unconventional financial activities such as over-the-counter trading, structured products and quantitative strategies were meant to increase market efficiency and cater to investor demand. Inherent in many of these activities were existing academic frameworks applied empirically.
Since the second half of 2007, however, we have witnessed unprecedented volatility and subsequent decline in capital markets from excessive leverage and contextual abuse of some of these frameworks. Now that the dust has begun to settle, focus is on recovery and understanding the new fundamentals of a different economic paradigm. To enable prudent growth, regulators, policy makers and central bankers have used extraordinary methods like purchasing troubled assets, slashing interest rates and enacting financial reform.
The following interview with Harry Markowitz is meant to be a compilation of reflections on the lessons learned from the financial crisis and the change in sources of diversification for mean-variance efficient portfolios. The interview helps combine views on the past and intuition about the future.
You wrote a book titled “Portfolio Selection,” which led to a Nobel Prize. Generally, can you tell us about its contents?
My 1959 book provides a method that recommends mean-variance efficient portfolios, which minimizes risk at various levels of return. In order to use this tool; one must supply the analysis with expected returns, variance and correlations for each security. Subsequently, an algorithm called the critical line, is applied to trace out an efficient frontier implied by inputs and constraints. That was my contribution.
What have you found to be the empirical applications of your framework?
It has been used in various ways. One application, as I expected, has been at the individual security level. But to a greater extent, in terms of money managed, it has been used with a top down approach – looking at asset classes instead of individual securities. The top down process is similar, in that you perform the mean-variance analysis at the asset class level. Implementation, however, is with products such as investment companies or exchange traded funds. Larger outfits like IBM or CALPERS can parcel out the various asset class mandates to specific money managers.
How does your framework help understand diversification?
The book, besides having an algorithm, has several theorems that help explain the relationships of securities, mean, standard deviation and covariances to the means and standard deviations of the portfolio. Specifically, in the “law of the average covariance,” the contribution of greater standard deviation may more than offset lower correlation. When looking at an equally weighted portfolio, the variance of the portfolio approaches the average covariance as the number of securities increases – it does not necessarily approach zero. If all the securities were uncorrelated, they would have zero covariance and in that case the variance of the portfolio would approach zero.
For a while it was popular to promote all sorts of exotic asset classes. Some of these salespeople represented that these assets were uncorrelated investment opportunities relative to other traditional asset classes. However, some of these were very volatile. So if you took their correlation and multiplied by standard deviation, you’d find that because they were high beta stocks they were adding more risk to the portfolio.
So the proportion of standard deviation contributed was not offset by the lower correlation.
Correct, and people were disappointed to find that many of these exotic asset classes didn’t diversify their portfolio.
What happened to portfolio theory in 2008?
People said that diversification failed because in 2008 all asset classes fell. Large cap stocks falling by 38 percent is a big drop, but it’s not an outlier empirically speaking. Not only was 2008 not an outlier, but it also wasn’t the worst year in history. 2008 was tied for the second worst year, just short of a 2 percent standard deviation move. If history were normally distributed, and I’m not saying that it is, you would expect the bad tail to occur 2.5 percent of the time or a move of this magnitude to occur once every 40 years. This is supported with empirical data.
During the same year emerging markets fell more, over 50 percent, whereas government bonds rose. Each of these respective asset classes has different betas. Emerging markets have higher betas and should indeed experience a larger relative decline.
Portfolio theory never promised high returns without risk. I spoke of the risk return trade-off. If you are higher on the frontier, you are accepting higher risk.
How do you believe practitioners misinterpreted your theory?
Since all practitioners are not created equal, I will make some distinctions.
There are some who stuck to a top down analysis, that Gary Brinson talked us into, which I described earlier. Those who practiced this over time were vindicated. Not in the sense that they protected clients from downside exposure, but ensured their clients knew, in advance, what their downside exposure was.
On the other hand, many people on the sales side advertised products that were supposed to be low risk. Many financial engineers also promoted products based on complex calculations that assumed continuous time models or frictionless trading. Practitioners also employed a high degree of leverage in some cases that was not well understood. Upon further examination, many of these products were in fact high risk, illiquid and leveraged. Portfolio theory cannot help with these distortions.
The last decade saw the growth of alternative asset classes such as hedge funds. These were designed to have high-risk adjusted returns and to maintain a lower correlation to traditional portfolios. Can you describe a process on how to better understand the attribution of returns and ways to include them in the top down approach?
In general, the use of a multi-factor model applied to historic performance of the fund can give you some intuition about where the fund is deriving returns. There are many multi-factor models already in existence, but selecting factors for a given a style of investment is important in gaining the right insight. Also, having done the proper due-diligence is necessary. After performing this process, one can sensibly build portfolios of hedge funds.
How do you see modern portfolio theory used in the future?
This method of analysis continues to empower millions of people who don’t have the same talents and resources as a Warren Buffet has. It has also helped to ensure that the right menu of investment companies is available in many retirement plans. These contributions combined with considerations for rebalancing and adjustments in expectation can be quite useful.
As we look to the future, prudence and common sense will have to accompany our endeavor to pioneer modern finance. Contributions, such as those of Dr. Markowitz, are well served in augmenting existing methods given proper context. Academics and professionals alike would agree that there is no substitute for continued diligence shaping the decision making process. This process should also not take existing frameworks for granted as it is this author’s opinion that empirical situations should always be approached with a prudent degree of skepticism.
Harry M. Markowitz, Adjunct Professor of Finance, Rady School of Management
Zach Jackson, CFA (Rady Full-Time MBA ’11) served as CFO of the Rady Student Board and vice president of the Finance & Investment Club. His professional experience includes a summer internship at Brandes Investment Partners preceded by a position analyzing investment opportunities and management of investment portfolios at a multi-strategy hedge fund.