Friday, 28 April 2017
Last updated 16 hours ago
Nov 16 2010 | 8:20am ET
Alpha is dead? Not according to Sebastian Ceria, CEO and founder of Axioma, which provides portfolio optimization tools to top asset management firms. Before founding Axioma, Ceria served as a professor at Columbia Business School, where his students voted him "best core teacher.”
In a Q&A with FINalternatives, Ceria weighs in on what he calls ‘provocative’ statements by the media, the challenges of portfolio diversification, and why factor-based risk management is important.
FINalternatives: A Barclay’s analyst noted that correlations across stocks hit all time highs this year, prompting a reporter to declare: “Alpha is dead.” An article in September in the Wall Street Journal echoed this sentiment, declaring that “Stock picking is a dead art form.” Will stock pickers ever make money again?
Ceria: The statement should not be blown out of proportion, it is provocative, but it may be more showmanship than reality. I would rather say alpha is in the intensive care unit, but the patient will live. However, I can appreciate the sentiment. The correlation of equity returns in equity markets has been steadily rising and experienced two sharp peaks, one in late 2008 during the financial crisis, the other in the summer of 2010 during the European debt crisis. These were extreme periods for the markets and it is reasonable to anticipate that they may work to fundamentally alter the way markets behave.
Do you think the argument is nonsense?
I don’t summarily dismiss it by any means at it raises important points for investors to consider.
Logically, either stock correlations will remain at crisis-like, historically high values and create elusive alpha and stock picking opportunities for investors; or they will return to historically “typical” values, which may well be accompanied by a post-crisis-like “more normal” market (a sigh of relief from investors).
Based on our analysis, we do believe the market is in a state of high correlations relative to pre-2007, but this is not to declare alpha or stock picking dead, just playing hard to get. So, I do agree with the fundamental assertion that we are in a new reality for stocks. But I prefer to think that alpha and stock picking opportunities still remain.
So we are in a whole new stock market and it is permanent?
Yes, but it’s been evolving for a while, actually. Independent of the two correlation spikes, equity correlations have been steadily rising in the U.S. since 1995 and world-wide side 2003 or 2004. From a longer historical perspective, the recent correlation spikes have merely punctuated a longer term trend of steadily increasing asset-asset correlations. In the U.S., equity correlations have been increasing steadily for the last 15 years. In fact, the post-crisis “lows” in colorations are as high as or higher than most correlations prior to 2008 and 2006. So the relief rally of today would be seen as a highly-correlated and frustrating market in, say 2004.
There are, of course, many possible explanations of the rising correlations circulating in the investment community, and each explanation provides a different perspective on the current state of the market. On the one hand, some observers have noted that many quantitative portfolio managers appear to have increasingly used and traded with the same underlying quantitative factors. As a result, their reactions to market events tend to be similar and their subsequent “crowded” trading lead to increased correlation. This story has been circulating since at least July 2007 when the Quant world experienced a “Quant Scare.”
On the other hand, most people suspect that the emergence of the ETF market has driven a lot of the increased equity correlation. When someone trades an ETF, all the equities underlying the ETF are traded in an extremely correlated manner, and it is easy to believe that this is a significant part of the rising correlations. For Global Equities, ADRs are having a similar effect. This can be seen explicitly in the difference between global equity correlations measured in U.S. dollars vs. local currencies.
So what has really been happening? New investment opportunities and tools such as ETFs have altered the way in which the market works, and, as a result, old insights need to be updated. This is not a new story on Wall Street. Just take a look at how Graham and Dodd’s original data and ideas have evolved over 80 years.
How should investors play it?
In our view, the main takeaway is that diversifying your investments is more challenging than it used to be because underlying market forces affect equity returns more than they used to. Whereas in the past, an equity portfolio could be acceptably well diversified by holding enough different equity names, which probably is no longer be true. Similarly, whereas 10 years ago a mutual fund position may have represented a diversified set of underlying positions, nowadays, the diversification and risk reduction obtained by holding a mutual fund is probably less than it used to be. Just ask anyone who had their 401K in an SP500 mutual fund in the fall of 2008 how diversified that investment turned out to be.
How is this going to affect the investment world? We expect managers to become more focused on risk management than they used to be. For the quants, this is not a significant change in the way they do business. However, fundamental managers who managed risk using qualitative approaches such as rules of thumbs (“Hold at least 40 names”) will need to learn how to use sophisticated quantitative risk models to better ensure that their holdings are as diversified as they should be. The old rules of thumb may no longer work as well as they did in the past.
What is factor-based risk management?
What high equity correlations imply is that factors such as market beta, size, and momentum are more important than ever as these factors explain an ever larger fraction of equity returns. Factor-based risk management involves using a risk model that explicitly models how those factors interact. By explicitly managing the exposure of a portfolio to different factors, a portfolio manager can ensure that his holdings are as diversified as he needs them to be.
Where do you see the market in five years?
We believe that equity correlations will remain high, as they will continue to be driven by new investment opportunities such as EFTs. In other words, we do not think we will experience a correlation bubble and a return to the investment environment of three or five years ago.
We expect that many if not most of the new ETF investment opportunities will be driven by factors (momentum, gold, etc.). Previous ETFs covered a fairly narrow range of factors – size, value, and growth. We expect the new ETFs to cover a much broader and diversified range of investments. We expect managers to diversify across factors, and the new investment opportunities will reflect that trend.
We also expect to see fundamental portfolio managers becoming increasingly quantitative and sophisticated about risk management and portfolio construction.