Active managers have exhibited skill in making asset-class decisions, our study shows.
Market volatility and low equity returns over the past decade have led many investors to seek new ways to invest. Tactical asset- allocation approaches, where changes are made to the portfolio allocation in response to market forecasts, hold a lot of appeal for many of these investors. The goal with a tactical strategy is to correctly time the market, or buy low and sell high, to achieve higher risk-adjusted performance than what an investor would achieve through a strategic asset-allocation, or buy-and-hold, approach. While the appeal of market timing is strong, the ability of portfolio managers to correctly time the market (the market-timing “intelligence” of the portfolio manager) is up for debate.
In this article, we will explore the ability of portfolio managers of actively managed balanced mutual funds to make “intelligent” decisions with respect to the historical asset- class exposures in the portfolio. We want to determine whether managers are able to add value by correctly timing portfolio decisions.
The Importance of Asset Allocation
In Brinson, Hood, and Beebower’s research paper, “Determinants of Portfolio Performance” (1986), perhaps the most well-known (and often miscited) paper on asset allocation, the authors concluded that the variance of a portfolio’s asset allocation, or “policy” return, explained 93.6% of the variation in the 91 large U.S. pension plans tested. Brinson, Singer, and Beebower (1991) confirmed the results in the original paper but found a slightly lower number, 91.5%. Additional research has confirmed the central theme of the findings regarding the importance of the beta decision (Ibbotson and Kaplan, 2000; Tokat et al., 2006).
Our study is similar in scope to a study published in the February/March 2012 issue of Morningstar Advisor by Jeffrey Ptak, titled “Tactical Funds Miss Their Chance”; however, there are some notable differences. First, Ptak defines a universe of funds that are truly tactical in nature based on the underlying philosophies of the portfolio managers. This test casts a much wider net than our study. Second, our paper focuses on the strategic decisions of the portfolio managers and uses indexes as proxies to estimate performance versus the actual returns. Third, a longer time period is considered, along with different measures of tactical performance to help better explain the results. Ptak’s piece is fundamentally a measure of the actual outcomes realized by investors, while this article is more of a theoretical exercise because we ignore the actual costs associated with trading, as well as the individual holdings themselves.
Tactical Allocation’s Promise
With the benefit of hindsight, the market seems predictable. Given the array of financial data and information now available, it is inevitable that mining through history will yield strategies that have outperformed historically to varying levels of statistical significance. Unfortunately, most of these strategies are more useful describing the market’s historical returns than predicting future returns. Despite the repeated failure of many of these approaches, tactical allocation is popular because the benefits from perfectly timing the market would be significant. Exhibit 1 displays the real growth1 of $1 from January 1926 to December 2011 for four different portfolios: 100% equity (represented by the return on the S&P 500); 100% bond (represented by the Ibbotson Intermediate Government Index); static 60%/40% equity/bond; and a “perfect-timing” portfolio. In the perfect-timing portfolio, we assume that for each month of the 1,032-month test period an investor successfully selected the asset class (bond or equity) at the beginning of the month with the higher return for that month. An initial $1 investment would grow to $21.88 billion if an investor perfectly timed the market, which works out to a 31.9% annualized geometric real return. Tactical strategies are tempting to investors.
How We Did the Study
To determine whether or not active managers have made intelligent tactical decisions, we performed an analysis based on aggregate data at the fund-holdings level, compiled by Morningstar every month. Morningstar breaks down the underlying holdings of a mutual fund into broad asset classes, to varying levels of precision. We used these holdings to construct a “synthetic” return of the asset-class-level decisions of the manager and compared it against a policy portfolio, which is simply the average of the monthly asset-class weights.
This approach to estimating the market-timing ability of fund managers is focused solely on their overall structural decisions. The actual returns of the fund are not considered because the underlying goal is to determine the timing ability of managers at the aggregate level, not whether or not managers can do so after fees. If structural changes in the asset classes did not add value (in this case, produce returns higher than the aggregate average of historical exposure to those asset classes), it could be inferred that the manager’s tactical changes do not add value. Even if these tactical choices do add value in this framework, it does not necessarily mean that tactical decisions will end up adding value for clients, because clients only realize the benefits of these decisions after paying the costs associated with the strategy. For the purposes of this analysis, these costs are assumed to be zero; therefore, our study is a test of whether balanced mutual fund managers have been able to exhibit skill versus determining if they are skilled enough to do so to cover their fees.