Solving the Portfolio Puzzle

Dr. Frank A. Sortino, Director

Pension Research Institute

 

 

 

 

 

 

 

 

 

 

 

You can’t solve a puzzle without all the pieces, and if your equity portfolio doesn’t have a core…you are missing an important piece of the puzzle.  In the second quarter of 2005 the growth component of the S&P 500 was up 2.6%.  But the growth component of the Russell 200 was only up 3 tenths of a percent (See figure 1).  Conversely, the return on the value component of the S&P was only .1% while the Russell 200 was 2.1%.  This was due mostly to how they treat the core component of the market.

 

Figure 1

 

The S&P 500 ignores the core and classifies every stock as either growth or value.  Growth stocks behave differently than value stocks and there is no denying it.  There is also no denying there is something in between value and growth that behaves differently than either growth or value. About a third of the time core surprises by either outperforming or underperforming both value and growth. In other words, core tells you more than you would know by simply knowing value and growth.  That means it is an important diversification partner for value and growth.  Ron Surz, president of PPCA-inc., was the first to recognize this market segment and in 1992 called it core.  Morningstar also called it core in their iShares traded on the American Stock Exchange since 2004.

 

Russell acknowledges a core element but they put it in both their value and growth indexes. This creates a statistical problem called multicollinarity.  This statistical flaw results in indeterminate solutions and reduces their usefulness for returns based style analysis.  This is not an opinion, it is a statistical fact

Figure 2 shows the results of a study done at PRI and documents the importance of recognizing the core element of the stock market.

 

Figure 2

 

 

The Russell indexes can only explain 50% of what is going on in the core segment of the market.    

 

Implications for Performance Measurement

 

Suppose growth manager A is measured against the S&P 500 growth index and growth manager B is measured against the Russell growth index. Both managers earned 1%. Manager A measured against the S&P could have a negative alpha (1% - 2.6% = -1.6%) while manager B would have a positive alpha (1% - .3% = .7%) using the Russell value index.  Ironically, some consultants would be recommending terminating manager A while others would be recommending he be hired?  Even more bizarre, some consultants would claim they could port the positive alpha from manager A while others would claim they could port the positive alpha of B.   This absurd result is based on false assumptions about the validity of alpha and the failure to recognize the core component of the market place.

Assembling The Pieces of the Puzzle

 

There are over 6000 stocks in the U.S. equity market.  In our opinion, the Surz indexes are currently the most accurate description of this market (see Figure 3).

 

Figure 3

 

 

 

The S&P 500 over weights large cap, under weights mid cap and ignores small cap.  This makes it easy for a money manager to game the evaluation by just adding a little small cap. Notice the S&P is composed of 20% core stocks even though they do not formally recognize it.

 

The real added value

 

Alpha is supposed to measure the value a manager adds in excess of the return on the market.  According to a recent survey of institutional investors and consultants[1], 96% used the S&P 500 as a surrogate for the market.  The above example of managers A and B illustrates the error introduced by using alpha as a measure of added value.  What matters is the value the manager adds to your portfolio, not whether he beats the S&P 500. 

To really measure the added value of a manager to your portfolio you must recognize that no manager has a 100% pure style like mid cap growth.  By identifying the blend of styles that uniquely identify the manager one can truly find those managers who add value to the portfolio and port that added value without unwittingly changing the portfolio mix.

 

For example, many consultants consider Van Kampen mid cap growth fund to be invested solely in mid cap growth stocks.  In fact it is mostly mid cap core and only 28% mid cap growth.

 

 

Figure 4

 

 

To determine if this fund manager could really add value, one should calculate the return one could have earned by owning the percentage in each passive index bucket shown in Figure 4.  The return on this style blend should then be subtracted from the return earned by Van Kampen.  When we adjust for downside risk we call this Omega excess. Omega excess shows the value this manager added over buying passive indexes.  Our research indicates that almost 70% of managers in the first quartile in period repeated in the next period (see Omega articles at www.sortino.com).  We believe this is a better alternative to alpha.  If active managers cannot beat their style blend, then use passive indexes.

 

Is anyone doing this besides us?  Not that I know of.  Yet, the formulas we use have been published in financial journals and on our web site for years.

 

 

Implications for asset allocation

 

 The asset allocation decision is considered the most important investment decision by nearly all investment professionals.  Yet, the way that most consultants hire active managers guarantees they will change the asset mix so that they will not know what the final asset mix is?  Let me say that again.  Most consultants will change the most important determinant of the portfolio return in some indeterminate manner.

 

This is how it happens.  Let’s suppose a consultant proposes the asset mix shown in Figure 4.

 

Figure 4

 

 

 

The consultant believes that 10% should be allocated to mid cap growth and he hires Van Kampen to fill that box.  However, Van Kampen only has 28% in mid cap growth so he only ends up with 2.8% in mid cap growth instead of 10%.  What is more, he will now have more allocated to the other categories than he wanted.  By the time he hires 8 more managers to fill the 9 style boxes he won’t know what the final asset mix is.  If the pension fund did have the best asset allocation initially it is now suboptimal. Plan sponsors BEWARE!

The Solution

 

To hire active managers without changing the asset mix, draw from the pool of managers who beat their style blends and add as much value (omega excess) as possible. Managers should come into solution as style blends, filling the style buckets in the strategic asset allocation. Then, diversification and style neutrality can be attained by filling in voids with passive investment portfolios. In other words, go passive in the parts of the market where active managers do not show skill.  Figure 5 demonstrates how this could have been done in the past using the Fidelity platform of mutual funds and may not be applicable today.

 

Figure 5

 

 

 

How can anyone ignore this factual information?  Well, they have and they will.  Why?

There is no risk to consultants in doing what all the other consultants are doing…until somebody does it better and starts taking business away from them.  So, until plan sponsors demand innovation they will continue to get the old technology in a different wrapper.