Defining Global Macro Quant Equity

In strictest terms, quantifiable information is that which investors can back test and assign a statistical significance level and then systematically implement.  A broader definition would include that which is quantifiable but cannot necessarily be assigned statistical significance or systematically implemented.  The broader definition is where I believe a new opportunity for global macro quant equity lies. 

Most of self-ascribed quant equity investors today use only the first set of quantifiable information, and mostly from the bottom up perspective.  My friend, Rodney Sullivan, and I have a new paper suggesting that quant investors should seriously consider applying a global macro approach that use the broader set of quantifiable information.  This paper can be found at the SSRN.  Both he and I have been keen to this approach for several years now.  I summarize this paper by way of an analogy. 

Before 2008, the quant ocean liners were faster the other ocean liners.  In particular, the pure quant ocean liners have the most systematic implementation and strictly rely on the quantifiable information that can be back tested with statistical significance.  They are the fastest of all.  These pure quant ocean liners are our focus. 

As a result of their success, the captains of the pure quant ocean liners hired many superb engineers and kept all of them under the deck tooling the ocean liners.  These captains made the rule that no one should waste any time above the deck.  This worked as long as the luck is on their side, i.e., the weather is great so that there are no icebergs, and all the undercurrents are flowing in the same direction as these ocean liners. 

From 2008, the weather starts to change.  Icebergs start to appear in the ocean and undercurrents start to turn against these ocean liners.  If these captains or his crews have been above the deck, they would have noticed this.  However, the under-the-deck rule prevents them from seeing that.  These ocean liners almost all hit the first iceberg, suffering significant damage. 

However, the captains of these ocean liners thought that this is just an anomaly and no iceberg will appear again.  So he insisted that his men mostly continue doing what they did, with the under-the-deck rule still strictly enforced.  What he and his crews did not know is that icebergs are popping up all around them this time, quite different from any difficult periods that they had before.  And the undercurrents are totally against them.  Soon enough, these ocean liners all hit icebergs again in 2009 and were severely damaged this time.  Undercurrents were initially against them strongly, pushing them backwards substantially in 2009, but then showed no direction in 2010.  With the engine sputtering and no clear direction of undercurrents, these ocean liners start to float with currents with no clear direction in 2010. 

Should the pure quant ocean liners do something drastic about it?  Just think about Euro.  It is a flawed system that may survive for a while, but its death is certain under large standard deviation events unless backed by a united fiscal authority and active efforts to bring down the culture and language barrier (so that labor and capital could flow freely within the Euro Zone).  The pure quant ocean liners are in the same situation. 

The captains of some of these ocean liners did start to order his crews to do something different, under the pressure of their passengers.  They gathered information about the history of the last two large iceberg accidents and maybe some smaller iceberg incidents in the past.  They want to use the information to predict the positions of any future icebergs.  This is a great improvement, and is part of what we advocate. 

However, trying to predict future icebergs purely with the limited information faces insurmountable challenges.  It is important to eliminate the under-the-deck rule so that the weather and current conditions can be taken into consideration in future navigation to avoid future icebergs even though those conditions cannot be back tested with an assigned significance level or systematically implemented.  After all, even the best dedicated surveyors of these conditions such as George Soros could not necessarily produce a statistically significant track record of forecasting (To understand why this is the case, please see my earlier post on global macro investing: http://xlpartners.blogspot.com/2010/08/what-is-important-in-global-macro_4346.html).  We also suggest that these captains should use all useful information instead of throwing some important information out. 

Further, these captains never had any idea about the issue of undercurrents. These captains do not fully understand that even if their ocean liners seems to be tightly constructed with everything based on statistical significance and systematically implemented, these ocean liners frequently have serious unintentional biases.  Some of these biases are persistent, e.g., small size, value, and momentum.  Other biases are incidental and last for a period and then disappear. 

When these unintentional biases are in the same direction as the undercurrents, like in the few years before 2008, they bring significant tail wind to the pure quant ocean liners.  When these unintentional biases are in the opposite direction as the undercurrents, the pure quant ocean liners face substantial head wind. 

The power of these unintentional biases is so large that many times most of the mileages covered by the pure quant ocean liners are due to this power.  Understand and be mindful of these unintentional biases is important to both alpha generation and risk management.  It is dangerous not to fully understand the unintentional bias while navigating the rough seas. 

For one example of unintentional biases, I would refer to one discussion by Jeremy Grantham in his most recent quarterly report (see the last two paragraphs on p5)

http://www.gmo.com/websitecontent/JGLetter_SummerEssays_2Q10.pdf

I have only read less than five reports of his ever due to the lack of time and their lack of short-term timing ability.  However, I find that all the reports are extremely insightful, especially for funds serving long-term institutional investors. 

A similar simple example is about small cap and value in different time periods.  As mentioned above, the pure quant ocean liners persistently have these biases.  In 2000, these segments are undervalued, which gives the pure quant ocean liners substantial tail wind in the last market down turn from 2000.  In 2008, the same segments are quite overvalued, which gives the pure quant ocean liners substantial head wind. 

So by global macro investing, we argue that the captains of these ocean liners need to add the top down quant models to systematically incorporate any global macro information that they can incorporate this way.  These captains also need to use quantifiable fundamental global macro information even if they cannot systematically implement it.  These reinvented ocean liners will be global macro quant equity ocean liners. 

Most of the issues discussed above are in three papers that I wrote in the past.  The first two papers are free for circulation if I am invited for presentation.  Feel free to contact me at xli5@hotmail.com if you are interested in these papers.  The last paper is only available for a handsome sum.  If you are interested in the last paper, you can learn about the first paper before you decide about the third paper. 

Li, Xi, 2007, Unintentional bets of pure quant investing, Working Paper, XL Partners, Inc.

Li, Xi, 2009, No religion in quant investing, Working Paper, XL Partners, Inc.

Li, Xi, 2008, Why does pure quant investing have unintentional bets, Working Paper, XL Partners, Inc.