July 2001
Out With the Old,
In With the New
| "We cannot change the rules of the game; we cannot ascertain whether the game is fair." |
| -- David van Dantzig |
YEAR AGO WHEN WE FIRST CREATED Portfolio 4, we wanted to leave it unchanged for 12 months. The only turnover would be the result of corporate actions such as mergers and acquisitions. That gave it the opportunity to reveal its true performance untainted by reformulation.
Those twelve months have now passed and as we anticipated, the portfolio had very little turnover (8%). Unfortunately, it also had very poor performance. From its inception (July 1, 2000) through its rebalancing date (June 15, 2001) it was off a total of 26.73%, losing 19.19% year-to-date. Over the same periods, the benchmark S&P 500 was down 16.52% and 8.02%, respectively. This disparity was not supposed to happen.
|
|
 Portfolio 4 was designed to track the S&P 500. Based on our quantitative model, it contained the top stocks in each sector. Unfortunately the model wasn't able to anticipate 2000's abrupt shift from growth to value, making for a very tough year. |
What went wrong? Based on such a short period we can't speak with any certainty, but we can speculate. Given the way the underlying formulas for the portfolio were constructed (see The Starting Point), recent momentum plays an important role. Over the past 12 months, the market made a definitive switch from growth to value, so we'd submit P4 was unfortunate enough to be holding growth stocks when value was taking command.
This might also explain the remarkable turnover in the reformulation. As you'll see below, P4 has taken on a distinct value orientation. Before getting to that however, here's a brief review of how this portfolio is constructed:
The Theory
Like Portfolio 3, P4 was the result of regressing fundamental factors against past performance. For both we used seven different factors, measuring both value and growth characteristics. We then used the resulting equations and current fundamental values of the S&P 500 to pick the highest rated stocks for each portfolio.The main difference between the two is how we ran the regressions. Portfolio 3 is run against the entire index and simply looks for the 30 highest rated stocks regardless of weighting or sector. In its early formulations, it was heavily weighted towards tech, but as that sector faltered, it's become more diversified (more on that in September).
Portfolio 4 is a little more sophisticated. Rather than applying to all stocks, it runs the regression separately in each sector, making it more sensitive to the differing factors affecting each sector. This is a reasonable approach since you wouldn't expect the same fundamental factors to affect differing sectors equally. While Price-to-Book may be meaningful when comparing Financial stocks, it's pretty meaningless when you're looking at Tech companies. Similarly, Debt-to-Equity means different things for Basic Materials firms and Healthcare concerns.
To accommodate these differences, P4 has a different regression for each sector, based again on past performance. Also, unlike P3, P4 was initially weighted like the S&P 500. To accomplish this, we started with the highest rated stock in each sector and kept moving down the list until the market value of each sector roughly coincided with the sector weight in the S&P 500. We did this again when we reformulated it.
The Results
Prior to rebalancing, P4's weights remained remarkably close to those of the S&P 500. As tech stocks fell, its weighting came down accordingly. Healthcare and Financials moved in the opposite direction. As of June 15th, all sectors were within 2% of their counterparts in the index. We were pleased to see how well that worked and took it as indicating the stocks in the portfolio were fairly representative of the indexWhat was surprising was the degree of turnover we encountered. Of the 52 stocks on June 15th, 38 (73%) were replaced. Sure, the portfolio didn't perform well last year, but we didn't expect almost total turnover.
|
| |||||
|
When all was said and done, the portfolio went from 52 holdings down to 51. Sector weightings were rebalanced to the June 15th S&P 500, and the portfolio had a new look.
That new look was one of a value portfolio. So-called "old economy" companies such as Eaton, Dana, and K-Mart replaced sexier Solectron, Convergys, and Harley-Davidson. Even on the tech front, old value stalwarts Xerox and Novell replaced late-90s' favorites Cisco and Oracle. The only place where younger, more growth-oriented issues replaced more established companies was in the Utilities sector where AES and Enron were out and Calpine and Dynegy were in.
The Conclusions
Is this just another case of momentum tracking what worked in the recent past? If it is, it's a fundamental defect of our portfolio construction process, and will need to be addressed in future reformulations. But after only one year of evaluation, it's too early to tell if this really is an inherent flaw. Only time will tell.
To help facilitate measuring results on a year-to-year basis under the new formulation, we'll track P4's performance from the June 15th reformulation date as well as from inception and year-to-date. The latter numbers will continue to be posted weekly on our Home Page and results since reformulation will be updated in Historical Performance.
As before, we will not initiate any changes to the portfolio until its next scheduled reformulation in mid-June 2002. Given the portfolio's balance and the fact that it holds those stocks screening at the top of each sector, over the next few months we'd anticipate it closing the gap with the S&P 500, at least on a year-to-date basis.
It'll get even more interesting after the next two or three months, when the momentum factor may begin to wear off. As they say on TV, stay tuned.
Search this site! Just enter you key word or words:
Get current quotes or follow your own custom portfolio,
courtesy of E-Line Financials:
|
|||||||||||
| Homepage | Return to Top |
