Thursday, September 3, 2009

Low Quality Rally

Many Advisors and fund managers have been complaining about how difficult it has been to beat their stock market benchmarks since the March 9th bottom. It turns out that, for active managers with a rigorous process for selecting the highest quality stocks, it has been quantifiably frustrating. This post will shed some light on this topic, to the delight of disciplined stock-pickers everywhere.

Over long periods (> 2 years) 'high-quality' stocks outperform 'low-quality' stocks by a substantial margin. A stock's 'quality' in this sense is defined by its relative standing among factors such as earnings growth and momentum, earnings stability, debt-to-equity, trading liquidity, return-on-equity, credit rating, analyst earnings estimate revisions, price level and price momentum, etc. However, since March 9th, an investor who stuck to a discipline of choosing only 'high-quality' stocks for portfolios would have substantially underperformed North American benchmarks.

It may seem intuitive that stocks that dropped the most during the bear market (worst momentum, smallest market cap at the bottom, lowest price at the bottom) have shown some of the greatest returns off the bottom. This intuition is faulty however, for several reasons. First, those who own the stocks that dropped the most - owned the stocks that dropped the most. Very few investors were able to time the purchase of these cataclysmic names, like Citigroup, Bank of America, and AIG so that they avoided the bloodbath before experiencing the ecstasy of rebirth. Even the bravest value investors like Bill Miller, whose Legg Mason Value Trust had one of the strongest 10-year track records of any mutual fund until late 2007, saw his fund value drop from $85 to $22, a drop of 74% (!!) before seeing his fund rebound by almost 100% since March 9th. His fund is still down over 50% from its peak.

Another reason why it makes little sense to try to buy the stocks that have dropped the most during bear markets is that these stocks usually underperform over the duration of the subsequent bull market. The 2000 - 2003 bear market is a great example. The stocks that went down the most in the bear market - JDS Uniphase, Nortel Networks, Mindspring, Cisco, etc. were very weak performers from 2003 - 2007. The strongest performers throughout the 2000 - 2003 bear market - energy, materials, and gold companies - were enormous winners during the following bull market. This is the rule, not the exception.

Myles Zyblock, Chief Institutional Strategist for RBC Capital Markets wrote the following in his September 1st note to clients:

"North American stock-specific leadership, regardless of sector membership, has been characterized by a type of constituency that most analysts refer to as “low quality”. This has been a rather unnerving shift for managers who follow a discipline that emphasizes particular attributes such as earnings quality or profitability. Many mandates have not allowed managers to enter this low-quality style box, to the detriment of relative performance. Strict adherence to a process more often than not leads to investment success, but the market can and often does move against a winning long-term strategy for brief time periods. Keep in mind that low quality cycles, even in up markets, tend to be relatively short-lived and our best guess is that there are only another 1-2 quarters of life left in this one.

One way to get a sense of what the low quality rally has been all about is to view the performance of stocks since the March low sorted by a few simple metrics that isolate the impact of quality factors including size/liquidity, valuations and profitability. Based on this methodology, it’s pretty clear that the smaller, less liquid, relatively inexpensive and more fundamentally broken companies have been the big outperformers over the past five months. " (See Table 1.)

Table 1. S&P 500 Returns Since March 9th By Decile Rank

Source: RBC

The next chart shows the performance of North American stocks, with S&P performance on the left, and TSX performance on the right. The grey bars represent the performance of highest quality stocks in each of the four major investment styles - Momentum, Predictability, Growth, and Value - while the blue bars show the performance of the lowest quality stocks. All performance is from the March 9th bottom.

Source: RBC

High quality U.S. stocks underperformed low-quality stocks in every category, though the outperformance in the Value category is minor. This makes sense, as at the bottom many of the low quality stocks were priced well below book value, on the assumption that many would be liquidated as non-viable businesses. Of course, without unprecedented government intervention, many of the worst performing stocks would have collapsed, and these numbers would look quite different. 

In Canada, high-quality outperformed low-quality Value stocks quite substantially. Canadian Value managers have had the most prospective stock-picking environment in generations, while growth and momentum managers have dramatically underperformed.

Overall, the above charts show that rapidly declining companies with highly unpredictable financials and significant earnings contraction as of March 9th very significantly outperformed stocks with predictable financials, strong earnings growth, and which were already in positive price trends.

History shows that this dynamic is not that unusual in frequency, but that it doesn't last for very long. According to RBS research, low-quality stocks have outperformed high-quality stocks during 8 periods in the past 30 years, and this performance advantage tends to persist for about 6 to 9 months. So low quality stocks may continue to dominate for another three months or so.

To further illustrate the point of low-quality dominance, I have included two tables below describing the performance of several style models from Canada's Computerized Portfolio Management Services (CPMS). The top table shows U.S. models, and the bottom table shows Canadian models. 

Each style model (Asset Value, Earnings Value, Earnings Momentum, etc.) tracks a portfolio of stocks chosen using different high-powered factors, such as earnings momentum, earnings predicability, analyst estimate revisions, price-to-book value, price momentum, etc. The 'Dangerous' model in each table tracks the performance of a portfolio of stocks which the system suggests would make good short candidates. These are the lowest quality stocks, with poor earning growth and quality, negative analyst estimate revisions, high debt levels relative to equity, etc. 

Note that over the past year the 'Dangerous' models have outperformed every model except the Canadian 'Earnings Value' model. This makes sense, as the Dangerous portfolio closely approximates a low-quality value portfolio, while the 'Earnings Value' model replicates a high-quality Value portfolio.

The Momentum portfolio in Canada, and the Earnings and Price Momentum portfolios in the U.S. have demonstrated the best performance since inception (1985 and 1993 for the Canadian and U.S. models respectively) by a wide margin (see far right column). However, these models have been terrible under-performers since the beginning of this year.

Source: CPMS

Source: CPMS

Fortunately, we can be confident that the low-quality rally won't last forever. If this rally provides us with another leg up, disciplined adherents to high-quality stock picking strategies will almost certainly get their day in the sun.

No comments:

Post a Comment