Core Value Equity Review

by Bart McMurry & Robert Dombrower

The Impact of the Financial Crisis on Core Value Strategy

With the financial crisis continuing to pressure markets on a global scale it is useful to investigate and assess the overall impact of the crisis in terms of its effect on the Core Value strategy.  Thus, we depart slightly from our usual review and begin with a brief synopsis of our quantitative process in order to set the backdrop for our assessment.  The CVE philosophy is based on the belief that human bias is a detriment to portfolio performance.  At tops, euphoria is too high and at bottoms, pessimism too strong, resulting in inaction at key turning points in markets.  Therefore, what is critical for success as a value manager is a methodology to evaluate stocks that controls the human emotional part of the analytical process.  We achieve this by adhering to a multifactor modeling process for constructing our Core Value strategy.  This quantitative process is constructed to rank stocks based on factors that contain a high degree of predictive ability of relative investment attractiveness.  New information populates our model every five trading days ensuring that the model is very robust and reflective of current market conditions.

On a weekly basis, once our models have been updated, we aggregate the results according to broad economic sectors which yield a “signal” of relative attractiveness of the various benchmark sectors.  At this point, we have a quantitative, bias free evaluation of the sectors within the benchmarks as well as the individual names within the sectors.  We track this output and publish an internal graph of the results so we can evaluate the cyclicality of relative attractiveness of each sector over time.  This process serves as our framework for evaluating stocks and sectors within client benchmarks.  We then utilize additional construction techniques in order to build portfolios that meet performance goals.  These include benchmark weighting analysis, periodicity analysis and attribution systems that are used to confirm our performance expectations.  Throughout the entire process we are very strict in our adherence to client investment policy guidelines as we monitor holdings, weightings, and performance tracking error to ensure that portfolios are behaving as expected.  With the above description laid out, we can now relate our strategy within the financial sector to the background above for a clear historical understanding of our portfolio financial holdings and the weightings they have.

The following chart shows the long term historical signal output of the Core Value multifactor model for the financial sector, as it relates to the performance of the S&P financial sector.  As you can see, going back to 1999, the relative attractiveness of the financial sector relative to all other sectors has yielded a line that has cycled from strong buy to strong sell (as depicted by the green arrow for buy and red arrow for sell).  Corresponding to those buy / sell signals is the model output that issued timely and relatively accurate signals (also depicted by the green and red arrows) which were implemented accordingly by the CVE strategy over this time period.  Thus client portfolios have been either under or overweight the financial sector at these designated times by utilizing a quantitative process that is free of human bias and has proven over time to be a good indicator of investment performance.  This investment policy within the financial sector has been among the most important sources of alpha that the CVE process has had over the decade versus the benchmark.

S&P Financial Performance and CVE Financial Sector Signal

Benchmark Analysis Revisit:  The Financial Sector Then and Now  

Our clients generally know that the CVE modeling process undergoes rigorous scrutiny as signals reveal current and future strategic opportunities, but we also analyze the benchmarks themselves as a means of deriving important data that aid us in our understanding of model behavior.  Over the course of the last three years we have been sharing this data with our clients to inform them of the disconnects taking shape in the market which directly effect excess performance versus their benchmarks.  The following series of illustrations revisit some of the historical analysis that we included in previous quarterly reviews.  By taking a retrospective look at some of this data we can clearly understand some of the compelling data that supported our model downgrade of the financial sector and the reason behind our weighting scheme. 

Prior to our model downgrade of the financial sector in 2006, we had been sounding the alarm over value index concentrations that were steadily increasing within the areas of financials and energy.  Based on a ten year history of Russell Value sector data, we could see only one other time when sector concentration reached such a peak.  The following chart comes from our quarterly write-up of the fourth quarter, 2005 in which we wrote “we have been steadily positioning the CVE portfolio to take advantage of what we see as “overstretched” disparities in the market, such as index sector concentration within four sectors—Financials, Oils, Energy and Utilities.  We continue to view this phenomenon as a market anomaly that will ultimately revert to the mean.”   

Russell 1000 Value Sector Weightings – September 1995 through December 2005

Exactly one year later we presented an updated perspective of this concentration of sectors by illustrating that over a ten year period, two sectors accounted for over 50% of the Russell Value index.  Our main focus was to point out that this anomaly called into question the diversification of the index and its usefulness in measuring the performance of diversified portfolios.  We also reminded our clients that when these two sectors outperform all others, then by definition, more diversified portfolios would underperform this benchmark unless they had similar concentrations. 

Russell 1000 Value Sector Concentration: 10 Year Progression

These examples show that as early as 2005, we began providing quantitative research depicting the precarious position of finance as a sector in relation to its historical weighting.  We also took steps in 2006, to defend our position in the face of criticism of our model and its output as we instituted steps to underweight the sector to the maximum of our mathematical process and to hold it firm until the model changed.  Thus, the major source of our underperformance as managers in 2005-2006 became our major source of outperformance from then until now with the advent of the current credit crisis.

In evaluating the most recent effects of the crisis on our current strategy it is useful at this point to consider our weighting procedure of individual securities in order to better understand how the most recent volatility has impacted the overall strategy. According to our client mandates, we are given a benchmark with the understanding that the names within it are all free to be bought or sold in some weighting (subject to guideline limits) over time that will outperform the passive index with controlled risk, as measured by variance to the index.  Therefore to beat the index, the exercise becomes one of a sector weighting decision and ensuring that the names within the sector do not detract from performance.  We already reviewed how the model assesses sector attractiveness, but on an individual stock basis, the model provides us with a ranking system for each stock within the benchmark.   The distinction is that it does not establish buy or sell prices.   It merely quantitatively ranks the stocks against the range of other investable options.  The following exhibit illustrates an actual example of our model’s latest financial ranking (first 16 names).

We can see how the model ranks some of the names of the investable sector.  Notice that despite the price hardship that many of these names have endured year to date, they are at the top of our list of attractive candidates, and have been for most of the year.  Even names that have been recently transacted upon such as Washington Mutual (bought by JP Morgan) and AIG (80% of which is held by the U.S. Government), are still included in the benchmarks and thus continue to be rated by the model.  This is an important point because these names (among many others that have lost most of their value) prior to the crisis, comprised a significant weighting in the index, and by definition, were widely held by most benchmark sensitive plans.  The current credit crisis has brought to light the challenge of benchmarking because the sector weighting of financials was so high at the peak.  At 37%, in order to come close to the benchmark weight, with 5% as a limit on any individual holding, one had to hold generally about 12 names to get near the benchmark weight, thus increasing the probability that one or more of these troubled names were held.  But now let’s turn to overall client performance for the year to date period.  

S&P 500 Sectors and Overall Performance VS. Value Index and Core Value Equity

12/31/07 – 9/30/08

With financials down over 31% for the year to date, we have held some that have ridden out the volatility with small declines, and others that underwent sharp declines—an understandable positioning in a risk controlled portfolio.  Overall, however, our exposures to the benchmarks netted positive relative performance on a year to date basis as shown above.  We did not design the process for panics other than to point us to opportunities and we highlight the above examples to distinguish our general methodologies as entirely mathematical and quantitative in nature, especially as they relate to benchmark analysis. Furthermore, our portfolio analysis of returns shows that they rank well even against fundamental managers who purport to study these facts that are happening right now in great detail.  Our process bettered their results.  The firm of the well known analyst who allegedly set out the bearish case on the names on your list found itself to be holders of the same names despite her analysis.  How could a manager who relies on a process of fundamental research find themselves in this situation?  The answer lies in client benchmarking.

The Long Term Plan:  The End of 2008 and Beyond

The Core Value quantitative process was designed for the management of long term assets within the capital market under continuously shifting sector weightings.  It cannot be expected to predict the demise of 158 year old trading houses—no system or fundamental analyst can.  Its performance trail accomplishes its goal of outperforming a benchmark consistently so our clients can rest in the assuredness that their return assumptions are met in the long run.  The remainder of 2008 will undoubtedly be volatile as global markets await more stable footing in the wake of the deepening credit crisis.  Domestically, the change in administration will also play a role in shaping market sentiment, as history has shown, which in the past has played a salutary effect on markets.  Our strategy throughout the market turmoil has remained constant—to analyze our quantitative signals which are immune to the emotional swings of the market, and implement portfolio strategy accordingly.  In this regard the latest quarter has shown a slight improvement in our financial signal from a static bearish stasis read, as the sector works to consolidate into a leaner, more stable landscape.  We have been following this signal very closely with the knowledge that consolidation within a formerly manic sector is one of the tell tale signs of the healing process, and historically has served as a value opportunity.  As it applies to the financial sector, history also teaches us that these stocks do not stay down for very long.  Thus we have been methodically adding to our financial exposure as bearish sentiment reaches new peaks and many gauges of pessimism and fear (ie. The Volatilty Index, or VIX) reach all time highs.  Many of our older clients have experienced this type of strategic move with us in the past, which continues to be standard model implementation as we stoically forge ahead in our goal of achieving risk controlled excess portfolio returns.     

Thirteen Year Cumulative Growth of $100.00 in Style Benchmarks and Core Value Equity

(Period 09/30/95 – 09/30/08)