| |
 November 2006 Buy and Hold An Analysis of Portfolio 1 and 2's Inactivity
As a related lesson, quantitatively-oriented investors -- those who believe their investment decisions can be improved by analyzing statistical patterns in the market -- also believe they can learn from experience. They believe that while you can't control your portfolio's ultimate outcome, you can learn from understanding how that result came about. In other words, you may not always do well, but looking back, you should know why you did what you did.
By studying the performance of your portfolio's individual holdings and their correlation with one another and the market, you can gain a better understanding of how your portfolio works. Even more importantly, it can help guide your decisions about how to alter the mix to improve future results. In some instances, the best action may be no action at all. When markets are directionless and there's little leadership, trading of any sort may simply increase costs, reduce net returns, and add nothing positive. Understanding your portfolio's analytics and those of the market can reasonably lead you to this conclusion. That's exactly the approach we've taken with Portfolios 1 and 2 for a little over a year now. Was it the correct approach? Each model's portfolio analytics will provide an answer.
Background This approach usually renders a trading pattern that is distinctly different from our quantitative models. The latter select stocks on a purely quantitative (statistical) method and must be rebalanced on a periodic basis. As a result, turnover is typically higher than in P1 and P2. In contrast, P1 and P2 have no required periodic rebalancing or trading requirement. As long as their holdings continue to satisfy their original purchase criteria, they can remain in the portfolio. (P2 does have one additional requirement: In order to remain a small cap portfolio, its individual holdings must continue to be small caps, that is, have market capitalizations of $3 billion or less. When a stock's capitalization rises above this cutoff, it must be sold and replaced with a true small cap. While this increases potential turnover, in practice, it's only led to less than one transaction per year.) Unless there are compelling specific buying opportunities, there's very little reason to trade P1 and P2 when the market is stagnant or directionless. When this occurs over a prolonged period of time, they may take on the appearance of buy-and-hold portfolios. That's precisely what's happened over the past 22 months. Over that time, the S&P 500 has climbed 12% while the Russell 2000 has gained 13%. While that may look like a definite upward trend, it appears less so when put in context: The S&P has added 7% year-to-date through September 30 while the Russell 2000 has climbed 8%, much of which occurred since late July. In other words, the index's gains haven't been steady and the majority have come in the last two months. Given this environment, a buy-and-hold approach isn't unreasonable. The last fundamentally driven trades (i.e., those based on replacing shares with lower potential with those with better potential) occurred on December 2, 2004. Over that period, P1 stayed right with the S&P 500, both rising 12%. P2 subsequently experienced considerable turnover in September 2005, but that was necessary in order to remove stocks of companies that were no longer true small caps. At that point, three stocks were added and three replaced, resulting in a 30% turnover for the year. Since then, the Russell 2000 has risen 8% and P2 only 2%. In an effort to compare the buy-and-hold effect on both portfolios directly, we looked back over the past year (October 2005 - September 2006) as well as the year-to-date figures for the first nine months of 2006.
Portfolio 1 Corning, on the other hand, appeared to be a value. Following the demise of the tech bubble, shares fell hard as demand for fiber optic cable evaporated. As one of the leading producers of fiber optic cable, momentum investors fled from the stock as rapidly as they had jumped in in the late-1990s. Yet Corning once again demonstrated its ability to change with the times, becoming a major force in LCD panel production. With rising demand for flat-panel LCD displays for both TVs and computer monitors, we felt Corning was a compelling buy at December 2004's levels. As a single trade, it paid off well. From December 2, 2004 through September 30, 2006, Corning rose 107% while the S&P 500 was up a much more modest 12%. As mentioned earlier, P1 matched the S&P's 12% gain. Results were a little different over the past year, however. A look a the portfolio analytics (Chart 1) not only shows that to be true, but shows which elements were strong as well as weak. The first thing to notice about Chart 1 is that it isn't simply based on returns from individual equities. Although the blue bars do reflect returns from stock picks, they are relative returns within each sector of the benchmark index. Before looking at the specific results, let's take a few moments to discuss what that means and why it's informative.
Most bottom-up investors don't really care about sector analysis. All they're looking for are stocks with good potential regardless of sector or industry. In the purest form, a bottom-up portfolio may be comprised of stocks from only one sector if they're the ones with the greatest potential. There's nothing wrong with this per se, but it can lead to a concentrated portfolio subject to increased market risk due to lack of diversification. On the flipside, it can provide well above-average returns if that particular sector leads the market. Many bottom-up investors aren't this aggressive. Although they rely upon bottom-up research, they may still impose sector weighting limits to add diversification. Top-down investors almost always do this, at least to some extent. Although P1 and P2 don't have such a requirement, we always try to spread holdings across at least a handful of sectors and perhaps a formal limit may ultimately prove to be needed. Regardless, a sector analysis of returns is helpful for even the most hard-core bottom-up investor since it shows where returns originated; whether it was a winning sector concentration or astute stock picking. For example, consider the investor who concentrated all holdings in large cap tech stocks in the late 1990s. Without a doubt, such a portfolio would have handily bested the S&P 500 and most other indexes. The investor may have ended the century with the belief his or her returns were the result of stock picking prowess when in truth, the rising large cap tech sector lifted all shares. The results would have been dramatically different in the following three years when the tech sector collapsed taking all related stocks (e.g. Corning) with it. To a degree, it didn't matter what large cap techs were picked, they all fell. Most results were again sector dependent rather than stock specific. So it makes sense to look at a portfolio's return attribution to understand what was truly driving returns relative to the benchmark. With no trades in Portfolio 1 over the past year, the results shown in Chart 1 are quite informative. Reading backwards from right to left, you'll notice that neither stock picks nor sector weightings of consumer discretionary, consumer staples, materials, telecommunications, or utilities had much of an impact. In fact, they essentially balanced one another out.
Interestingly, the energy sector weighting added .9% to the return relative to the index, yet the model held no energy shares. That's right, by completely ignoring this sector, performance was improved relative to the index. The decision not to hold stocks in a specific sector can be just as important as which stocks to hold in a sector.. Why would this be the case in this particular example? If you recall, energy prices and energy shares spiked following the devastation of Hurricane Katrina. This was in the beginning of the measurement period. Since then, energy shares have gyrated, falling in the final months of 2005 and again in August - September 2006 as well as rising in the first half of 2006. The end result was essentially a wash -- arguably a good reason to engage a buy-and-hold approach over the total period. But at the other extreme, that same buy-and-hold strategy didn't work so well in the tech sector. Unfortunately, P1 is heavily weighted there, with IBM, Microsoft, and Intel in addition to Corning. This overall sector weighting lost 2.1% relative to the index but the four stocks themselves actually added .1%. This was a case of having fairly good stocks in a lousy sector. The industrial sector weighting was comparable to the index but the stock selection (General Dynamics, +20% over the period) added a relative 1.1%. The exact opposite happened in the financials sector. Contributions were split in the healthcare sector. Here the two components (AmerisourceBergen and Johnson & Johnson) outperformed the index but the sector weighting detracted. That was a case of picking good stocks in a poor sector. Overall, the buy-and-hold approach to P1 hasn't worked too well in the past year although it did manage to keep up with the benchmark over the past 22 months. Of course that isn't much of a victory with a model whose goal is to beat the index. On the other hand, one of the fastest ways to destroy long-term returns is to over-trade in a listless market. Short-term moves are often exaggerated and can easily whipsaw an inexperienced investor. Just this past summer panicked investors sold in the mid-May to June period when the market sold off, locking in losses and missing the subsequent upturn. In that context, P1's recent buy-and-hold strategy may have served it well after all.
Portfolio 2
Chart 2 gives the same portfolio analytics for P2 as Chart 1 showed for P1. Again telecom, consumer staples, energy, and utilities had little impact on the overall performance relative to the benchmark index, the Russell 2000. Consumer discretionary (home of PF Chang's China Bistro, -23% and Corinthian Colleges, -19%), healthcare, and industrials, all lost ground to the index. Notice that the industrial sector weighting added (+.3%) although the stocks detracted. That was a case of picking poor stocks in a good sector. The biggest plusses relative to the index came from stock picks in the financial and materials sector. Private Bancorp (+33%) added to relative performance although it was in a relatively poor performing sector. The biggest gain was provided by OM Group (+118%) in the materials sector. Before concluding that the buy-and-hold strategy failed P2, it's worth looking a little closer at OM Group. That may seem like a strange consideration because who wouldn't want to hold a stock on a 118% tear. Truth is, many investors didn't.
You see OM was added to the model on September 13, 2004. It was one of the three stocks replacing others being removed because they had become mid caps. Shortly after its addition, OM shocked investors when the company announced it would be unable to file its quarterly accounting reports in a timely manner. Shares plummeted to $12.85 -- 44% below the initial purchase price. On October 1, 2005, the beginning of our measurement period, shares traded at $20.13 on their way down to the $12.85 low. Investors were dumping the stock. Small caps as a group were still leading the market and by retaining OM, P2 was falling behind. One might think this was a good time to sell and move on to an alternative -- almost any alternative -- that would do better. Instead, we held on. All the research leading up to OM's purchase indicated it was a solid company with steady to rising earnings. The only thing that changed was an accounting snafu that led to the delayed filing. No one was charged with corruption, the company wasn't facing any lawsuits a la Merck, and its market was holding up. Active traders were leaving in droves, but did this make sense for long-term investors? We decided it didn't. Perhaps the most compelling reason to hold on was the fact that most of the small cap alternatives were already expensive and becoming more so. Ultimately, they fell back toward fair value in the May-June selloff, but OM continued to recover. By September 30, 2006, it was trading at $43.94, more than twice its September 2005 purchase price. By the way, it subsequently traded up into the upper-$50s and may finally be overvalued. The point here is that it's hard to say the buy-and-hold approach failed P2. Yes, there have been some sharp losses -- OM was one less than a year ago. Small caps are volatile by nature and anyone holding on for any period of time will be subject to their ups and downs. Selling one volatile security in favor of another equally volatile one is often a recipe for selling low and buying high. If nothing else, P2 illustrates how broad sector diversification can help smooth the effects of volatile short-term markets. This is true even for strictly bottom-up models like P1 and P2. So did the recent buy-and-hold strategy work for P1 and P2? The jury's out. Would we take the same approach in similar market conditions? Absolutely. Search this site! Just enter you key word or words: Get current quotes or follow your own custom portfolio,
courtesy of E-Line Financials:
|
||||||||||||||||||||||||||||||||||||
NE OF THE FIRST THINGS YOU PROBABLY learned as an investor is that you can't control the market. No matter how well thought out your investment strategy or security picks may be, there's no guarantee that the market will conform to your expectations. Well reasoned analysis can help your odds, but it's still no guarantee.
