Stock correlations, or the degree to which the prices of stocks move together, has been a hot topic in certain investment circles recently. In the summer and fall of 2010, correlations between stocks (as measured by the S&P 500 Index) spiked to 60%, levels never seen before. They have since dropped, closer to the 25 year average of 24% (see Figure 1). A similar pattern emerged among correlations between asset classes, which captured a lot of press given its implications on asset allocation and portfolio construction.
Figure 1: Correlation Cycles
Source: BofA Merrill Lynch US Quantitative Strategy
There are varied, well-documented reasons why correlations rise and fall, which we will discuss briefly later in the article. However, what is more interesting to us is whether or not this market phenomenon affects a manager’s performance versus an index. After all, as a provider of due diligence consulting services, it’s our job to understand the reasons behind performance results, good or bad. The market environment has a large degree of influence on how a manager performs, and in 2010 there was no shortage of “stock pickers” defending their approach in the face of rising correlations. The S&P 500 Index outperformed 70% of mutual funds in Morningstar’s Large Blend category last year.
Stock Correlation on the Rise
As you can see in Figure 2, the average level of stock correlation has been gradually increasing in recent years. A number of researchers (see footnote 1 below) have pinned the rise in correlations to a market that is fixated on broad macro-economic and geopolitical issues, in addition to the record use of index derivatives, ETFs, and high frequency trading.
Figure 2: Recent Rising Correlation Trend
Managers that rely on fundamental factors for stock selection, such as earnings revisions, for example, are most susceptible to swings in correlations. When macro events dominate stock movements and an underperforming manager tells you “the market is just not recognizing or focused on the right factors,” the evidence he often points to is rising stock correlations. The theory holds that in a high correlation environment, fundamentals can be overlooked and the largest portion of a stock’s price move may be caused by the market conditions. This makes it
more difficult for bottom-up managers to add value above and beyond the benchmark.
There is some evidence to back this argument, as illustrated by the following table. Over a 21-year period we measured the relative performance of domestic mutual funds versus their best fit benchmark (see 2 below) in both “high” and “low” correlation environments (see footnote 3 below). We found that when stock correlation was higher than normal, managers in aggregate performed worse during the quarter and the year that followed.
Table 1: Average Manager Relative Performance - Time Period: 1/1/1990 – 3/31/2011
When correlation is…
During the Quarter
Over the Next Year
Source: Prima Capital, Morningstar, BofA Merrill Lynch US Quantitative Strategy
However, the performance differential is not substantial considering how volatile a manager’s relative performance can be on a quarter to quarter basis. Also, there are many factors that influence whether managers as a collective group can outperform their benchmark.
One of the main factors is that not all mutual funds rely on a bottom-up fundamental investment approach. There are a significant percentage of managers who rely on factors such as price momentum and other technical indicators. There are also a number of managers who pride themselves on their macro-economic viewpoints and appreciate a high stock correlation environment because it is more suitable to their top-down strategy.
The Takeaways for Advisors
The issue of stock correlations and active management boils down to three key takeaways for advisors.
- First, if your bottom-up manager struggles during a high correlation environment and has held true to portfolio holdings, give them the benefit of the doubt until you can assess results when the market regains a focus on micro, fundamental issues. Conversely, if the manager changes stripes in the middle of a performance drought to become more macro-oriented, it may be time to move assets elsewhere since this is not the expertise that you sought when you hired the manager.
- Second, William Coaker of the University of California made valid arguments for why high correlations may persist in his 2010 IMCA Investments&Wealth Monitor article. Since this trend appears to be a headwind for fundamental managers, those who seek to deliver a reasonable degree of alpha need to take on bigger bets, increasing the argument for concentrated portfolios.
- Third, it is of utmost importance that advisors find the highest quality managers in expectation of more difficult alpha conditions for fundamental managers. Take the time to truly vet the manager’s credentials and investment process. Not only will this be a positive step in the direction of delivering investment quality on the behalf of your clients, you will have a better handle on the manager’s performance behavior in various market environments, including those featuring high stock correlations.
Author’s disclaimer: The views and opinions expressed are provided for general information only and do not constitute specific investment advice or recommendations from the author.
1) Citations: Marko Kolanovic, J.P. Morgan Global Equity Derivative & Delta One Strategy, 10/5/10. William Coaker, IMCA Investments and Wealth Monitor, 2010, Understanding the Recent Rise in Correlations. John Kemp, Rising Correlation and Computer Driven Trading, Reuters.com (July 13th 2010).
2)To determine the relative performance of domestic mutual funds we compared the returns of the following domestic Morningstar categories versus their corresponding benchmarks and averaged the results for each quarterly time period from 1/1/1990 – 3/31/2011.
Morningstar Category Index
US OE Large Blend S&P 500 TR
US OE Large Growth Russell Top 200 Growth TR USD
US OE Large Value Russell Top 200 Value TR USD
US OE Mid-Cap Blend Russell Mid Cap TR USD
US OE Mid-Cap Growth Russell Mid Cap Growth TR USD
US OE Mid-Cap Value Russell Mid Cap Value TR USD
US OE Small Blend Russell 2000 TR USD
US OE Small Growth Russell 2000 Growth TR USD
US OE Small Value Russell 2000 Value TR USD
3) The “high” and “low” designation was determined based on each quarterly time periods’ correlation in relation to the average over the 6/30/1986 to 4/31/2011 time period (24.4%).