Most investors are probably familiar with the “style box” approach to equity investing, but they may not know that the unintended consequences of this construct may result in underperformance.
In our view, style boxes have constrained managers and limited their opportunity set. Instead, PIMCO sees greater potential benefit to global portfolios in strategies that are unconstrained by a benchmark, and with managers who think about absolute return at least as much as they think about relative return. We believe that stepping outside the style box allows an active investment manager the opportunity to lower volatility and improve downside risk mitigation.
Inside the style box
Let’s quickly review how the style box was often constructed, simplifying to three basic steps.
First, the world was divided by region, style and market cap. Second, an emphasis was placed on the investor’s home market (in this example the U.S.), with international equity added for diversification. And third, managers were hired to fill a very specific and narrow role within the portfolio (see Figure 1).
For example, to fill the large cap value box, often several managers were hired – for manager diversification – and each manager typically was instructed to stay within that market segment. They were told not to drift down the market cap spectrum, as mid cap and small cap managers were playing those roles within the portfolio, and they were told not to drift over to growth, as that was the territory of other managers. And importantly, these managers were measured against a market cap and style specific benchmark, in this example the Russell 1000 Value index.
The attractiveness of the style box was twofold. In theory, it allowed investors to precisely control their equity exposures and to hire “specialist” managers who were expected to deliver strong performance in their specific area of focus. However, we believe this style box approach has failed investors.
Unintended consequences of the style box approach
We believe this approach resulted in too great a focus on returns relative to a very narrow index and led investors to have too short of an investment time horizon in which to evaluate their managers. By measuring managers against just a slice of the market, the cycles of style performance often dictated manager success.
Consider investor behavior in the technology bubble of 1999 to 2000. As tech stocks dominated the market, aggressive or momentum-oriented managers tended to outperform their more valuation-sensitive growth-at-a- reasonable-price (GARP) competitors. These GARP managers on average had good absolute returns that in many cases exceeded that of the broad market, but industry flows show that investors tended to hire the more aggressive growth managers to fill their growth boxes. As the bubble burst, though, the valuation discipline that kept many GARP managers largely out of tech stocks helped lessen, for many, dramatic losses. GARP managers appeared “smarter,” having navigated the tech bubble better, and as a result they eventually replaced the momentum players as the growth managers in many investment portfolios.
This dynamic of the timing of manager selection has repeated itself over time, as observed in a 10-year study of manager performance pre- and post-hiring and firing (Figure 2).
These charts suggest an element of performance chasing in the hiring and firing of equity managers. PIMCO believes that the cycles of style performance and the narrow benchmarks in the style box world encourages manager turnover and undermines long-term portfolio return potential.
Another drawback of the style box construct is that it may have been the catalyst for managers to “play the game” and become benchmark oriented. It seems many active managers recognized that assets and revenues were at risk if they deviated too far from the benchmark, and so they became closet indexers.
Research shows that managers have become less active over time (see Figure 3). According to Antti Petajisto of the NYU Stern School of Business, December 2010, thirty years ago, nearly all U.S. equity fund assets were with managers defined as “active” or “highly active,” based on their active share, or the percentage by which their holdings differ from the benchmark. The chart above shows the rise of index funds, which makes sense – we know there has been an increase in passive investing – but surprisingly, assets in closet indexers have grown at an even higher rate, such that closet indexers today represent approximately one-third of all U.S. equity fund assets! At the other end of the spectrum, assets in highly active managers, who tend to invest based on research conviction and do not have a benchmark orientation, have declined from 60% of overall fund assets to less than 20% today. In other words, it is easier to find managers that are benchmark constrained than managers that are highly active.
Even though many managers have become less active, many continue to charge active fees, and as a result, we believe, the style box framework basically became an expensive index strategy.
The dangers of benchmark hugging
The problem with a benchmark orientation is that it can cause managers to be reactive to the market. Consider that indexes frequently become distorted – think of the weighting of Japan in international equity indexes in the late 1980s, the weighting of technology in the U.S. equity and growth benchmarks in the late 1990s, and of financials in value benchmarks prior to the financial crisis in 2008.
The closet indexer defines risk primarily in benchmark terms, or tracking error, and therefore tends to follow the benchmark weights, i.e., buying more Japan equity as it becomes a larger part of the index, buying more technology and financials as they become larger parts of the index. While these managers are focused on minimizing benchmark risk, they may be creating significant absolute risk, and when these bubbles burst their investors learned the painful lesson that low tracking error does not necessarily mean low risk.
The performance problems associated with benchmark orientation and measuring managers relative to a narrow index can create a significant challenge for investors. A recent McKinsey & Company study (August 2011) cites a growing awareness that the structure and mindset that is entrenched in the investment management industry needs to change, that benchmarks create the wrong incentives, that manager focus should not be just on beating the index, and that fighting the relative investment mindset is a constant battle.
The case for global unconstrainedWhy global? Remember that traditionally equity portfolios tended to be built with a home-market focus, with international equity added as a diversifier. However, correlations between stocks globally have increased over the last decade and have stayed at this elevated level (see Figure 4).
In addition, the lines between what it means to be a U.S company and a non-U.S. company are blurring, with over 30% of S&P 500 company revenues coming from outside the U.S., according to Goldman Sachs (2010). As the world has become more interconnected, and as diversification benefits fade, we believe the equity classifications of “domestic” and “international” have become outmoded. The implication is that as the world is evolving, managers who are not constrained to a specific region and have a global opportunity set may be better positioned to find the most attractive investment ideas regardless of company domicile.
Why unconstrained? Below we show a measure of performance of global managers vs. those that are regionally constrained and the performance of all cap managers vs. those that are market cap constrained (see Figure 5). The performance shown is the information ratio, or excess return over a benchmark, or alpha, divided by the standard deviation of that alpha. Information ratio is one of the risk-adjusted returns employed by consultants and institutions.
The performance suggests that investors should consider managers that are unconstrained by a benchmark. This creates another challenge – how do you know a manager is truly active?
The importance of active share
While there are a lot of different metrics that can be run to analyze an equity manager’s portfolio and performance, active share is likely the best objective measure of how active a manager is. The definition and calculation of active share are relatively simple: It is the percentage of a portfolio – considering actual portfolio holdings and weightings – that differs from its benchmark.
There are only four ways a portfolio can differ from the benchmark (see Figure 6):
- Own a benchmark company at an underweight (A)
- Own a benchmark company at an overweight (B)
- Do not own a benchmark company (C)
- Own a company not in the benchmark (D)
To calculate active share, you take the absolute value of the difference between portfolio holding and benchmark weights, sum all of these differences, and divide by two (to account for overweights and underweights). You will get an active share number between 0%, which indicates the portfolio perfectly replicates the benchmark, and 100%, which tells you that there is no overlap whatsoever between the portfolio and the index. A manager needs an active share of at least 60% to be considered active, with 80% active share defining “highly active” managers referenced earlier. (see “How Active is Your Manager? A New Measure that Predicts Performance,” by Martijin Cremers and Antti Petajisto of the International Center for Finance at the Yale School of Management, January 2007).
Let’s walk through an illustration of why active share is important. In this example, we have two hypothetical managers, one with 50% active share and one with 80% active share. We assume a benchmark return of 10%, manager fees of 100 basis points (bps, there are 100 basis points in 1%), and a net excess return target of 200 bps.
For the 50% active share manager, since half of the portfolio overlaps with the benchmark, that portion of the portfolio returns 10%, in line with the benchmark. This means that in order to hit the excess return target, the active portion of the portfolio has to work really hard and outperform by 600 bps (see Figure 7).
For the manager with 80% active share, we’ll assume the 20% of his portfolio that overlaps with the benchmark also delivers the benchmark return of 10%, which means that the active portion of the portfolio has to outperform by 375 bps to achieve the overall target excess return. Of course, 375 bps of outperformance is not easy – but it is 225 bps easier than 600 bps!
High active share, then, not only provides the potential for higher returns, but it increases the probability of achieving excess returns. Of course, high active share also leads to a greater opportunity to underperform the benchmark, so manager selection is key. We believe high active share should be top of mind when selecting active equity managers.
Evolving equity portfolio structures
Moving away from the style box raises the question: how should equity portfolios be structured? We see two solutions.
Core/satellite: If the style box construct has morphed into an expensive index strategy, then core/satellite addresses that problem by going passive at the core to obtain beta exposure more cheaply, particularly in more efficient markets. Active strategies with high alpha potential could then be added to complement this passive core. Satellite strategies typically include managers that are highly active, unconstrained, high-conviction and outcome-oriented, or in asset classes deemed less efficient, such as emerging markets.
Global unconstrained: The key change in this structure vs. the style box is that an investor would no longer make the distinction between domestic managers and international managers, and instead would hire several (for manager diversification) global, unconstrained managers. These global equity mandates would likely be complemented by allocations to active emerging markets. In addition, here global managers would be measured against broader benchmarks, such as MSCI World or ACWI. While managers still may have a style-oriented investment strategy – i.e., be value or growth investors – they will be held accountable to a broader benchmark and adhere to their discipline because they believe that discipline (investment process, research views) will allow them to outperform the broad market, not because they are playing a narrow, style-constrained role in a portfolio.
Restructuring an equity portfolio can be challenging. Part of the allure of the style-box approach is the feeling of control over equity allocations and exposures. Moving away from that comforting approach and toward unconstrained shifts some of that control to the managers. In addition, while high active share provides the potential for higher returns, it also creates the risk of significant underperformance. Because of this, our belief is that outside of the style box, manager selection and manager diversification become even more important. There must be in-depth due diligence on each manager’s investment philosophy, team, investment criteria and process to gain a clear understanding of how the manager may perform in various market environments and how they may complement other managers. Managers must demonstrate a research quality that allows them the potential to transform high active share into strong performance. We believe this combination of broader benchmarks and deeper due diligence will result in a better understanding of manager performance over shorter-term cycles, ultimately leading to longer investment time horizons and better returns.
At PIMCO, our active equity strategies are positioned for this evolving equity landscape. All of our strategies are global, unconstrained by the benchmark (seeking at least 80% active share), and consider downside risk mitigation as a critical part of the client experience. As investors move away from the style box quickly or more gradually, we will be there with equity solutions to help meet investor needs.