UNDERSTANDING THE DYNAMICS OF PASSIVE INVESTING

A BROADER SET OF PERFORMANCE FACTORS

Many believe that passive funds are a nearly-invisible part of an actively-managed portfolio. These innocuous, low-cost samples of market exposure are often used for these reasons:

• Temporary investments
• Replacements for poorly-performing funds
• Market exposure in sectors where investors believe that “you can’t beat the market.”

In all these circumstances, investors view passive funds as pure market exposure in the segments where they are used. As such, they contribute nothing to excess return, and by inference they contribute nothing to active risk (i.e., tracking error.) If anything, the assumption is that active return and risk decline as passive exposure is added.

None of these statements is true.

We use a real-life case study to demonstrate the role of passive investments in “hybrid” portfolios that contain both passive and actively-managed funds. This involves taking a fresh look at the role that passive investments play. Our approach splits the portfolio into two parts: one active, the other passive. We will demonstrate that the passive “portfolio” makes its own contributions to both active return and risk. Equally important, we also show how and why portfolios with substantial passive exposure can perform as well – and sometimes better – than all-active portfolios. 

HYBRIDS: EQUAL TOTAL RETURNS AND BETTER ACTIVE RETURNS

We had created a set of 24 optimized “fund teams” which minimized tracking error while delivering its specific alpha target. Our “alpha” was a volatility-adjusted excess return relative to the four equity vs bond strategies (i.e., 50-50, 60-40, 70-30 and 80-20) and our six alpha targets went from 50 bps to 175 bps in 25 bps increments. We then re-optimized these portfolios with the addition of passive investments, maintaining the same allocation and alpha target while minimizing tracking error. Our first chart summarizes our results. We see that the two sets of portfolios (all-active vs a hybrid mix of passive and actively-managed funds) have equivalent results on a total return basis. Another surprising result is the extent of passive funds that were used: on average (across strategies) we saw passive exposures between 18% and 69% (which we describe more generally as 1-in-5 dollars to as much as 2-out-of-3 dollars in passive exposure.)

This clearly demonstrates that passive funds are not quite so “passive” after all, since they can be combined with the right active funds so that they replicate all-active results. These passive funds do more than simply supply market exposure and market returns. They appear to increase the efficiency and effectiveness of the active funds that are used. We see that ths is partly because the passive funds allow us to select a different set of active funds, and also because of the passive funds’ interaction with both the active funds and the overall portfolio.

Our “all-active” portfolios clearly demonstrate the direct relationship between each allocation’s volatility and the tracking error delivered by its funds. Both total volatility and active risk increase as our equity exposure increases. This produces a decline in the Information Ratio (alpha vs tracking error) as our strategies shift more toward equity.

The hybrid portfolios are superior to the all-active portfolios in two regards:

• They are more consistent in their active results. The alpha-vs-tracking error of our hybrid solutions is close to being a single line. The Information Ratio across the four strategies are almost identical.
• The hybrid portfolio tracking error is substantially lower than for the all-active portfolios. This is especially true at the lower alpha targets. It is obvious that more active product is needed at higher alpha targets.

A NEW LOOK AT HYBRID PORTFOLIOS

The traditional view of performance attribution is to view the portfolios in two parts: market exposure and an active return. The market exposure is represented by the return of a benchmark or index, and the active return is calculated as the portfolio return minus the benchmark return.

We suggest breaking the portfolio into two parts: a passive portfolio and an active portfolio. Instead of “benchmark plus active return” we use “passive portfolio plus active portfolio.” Each of our components contributes to the portfolio’s total return and volatility risk, and also to its excess return and tracking error risk.

Our novel approach allows us to identify the true contributions of the passive portfolio, while revealing the dual contributions of the active portfolio: allocation return and true selection return.

IN-DEPTH REVIEW OF A HYBRID PORTFOLIO

Our sample hybrid portfolio uses a 60-40 equity-bond allocation at an alpha target of 100 bps.

We begin with a look at the geometric total return vs volatility of the portfolio and its benchmark. The portfolio return of 7.27% and volatility of 10.21% outperformed its benchmark (6.34% return with 10.33% volatility) on both an absolute and a risk-adjusted basis. Its simple excess return is 93 bps but given its lower volatility, it’s alpha (i.e., risk-adjusted excess return) is 99 basis points.

Since the portfolio matches the asset allocation and the style characteristics of its benchmark, there are no strategic or tactical allocation effects at work. All return and risk differences are due to the active effect resulting from the interactions of the alpha streams within the “team of funds.” We have previously shown how a negative correlation between the alpha stream and the total portfolio total return subtracts a percentage of the portfolio’s tracking error from portfolio volatility.

Our chart of performance shows the total return for both the passive and the active funds that make up the portfolio. Both passive and active portfolios outperformed the benchmark, with the passive funds taking substantially greater volatility risk, while the active funds took less risk.We also isolated the return on the market exposures of the set of active funds (“Active Allocation Return.”)

What is the source of the Passive portfolio’s greater volatility risk? The answer is obvious: it has a more aggressive asset allocation than the benchmark. And since the sum of the market exposures of the active and passive strategies is equal to the exposure of the benchmark, we know that the Active portfolio follows a more conservative allocation than the benchmark. We show the return and risk for this active allocation, using benchmark returns. This creates an “allocation line” that connects the active and passive asset allocation returns to the benchmark. This allocation line helps us to unbundle the drivers of performance.

INITIAL INSIGHTS

We now see that the passive funds are not simply an invisible contribution to the market exposure of the portfolio. In our example, they had an effect on both the volatility and the return of the portfolio. Our first “surprising insight” is that our passive portfolio contributed to the portfolio’s excess return through its “allocation effect” (this is essentially the traditional allocation effect in most performance attribution analyses – but our analysis also looks at contribution to active risk!)

We also see the ”true selection effect” as the difference between the return of the active funds and their “true benchmark,” which is their active allocation return. The active portfolio’s true selection effect is significantly higher than the simplistic difference between the portfolio and the benchmark. In the All-Active portfolio, the active excess return was 94 bps, while the Hybrid portfolio’s true selection effect (relative to the active allocation) was 155 basis points.

It becomes clear that the presence of the passive portfolio contributes to the creation of a “selection portfolio” with a stronger active effect. As noted earlier, using passive funds selectively allows us to focus on the “best of the best” active funds - in the context of building the best “active team” - without the constraints of the asset allocation. Since constraints generally degrade solutions, we employ passive funds to bypass the constraints that the asset allocation could have on fund selection. This shows that using passive funds is an important part of the process of creating greater active efficiency.

UNDERLYING ALLOCATION EFFECTS

The asset allocation of the passive portfolio is driven by the fund selection process.

In optimizing the active return component of the portfolio, we focus on selecting those funds in terms of their individual performance as well as their contribution to the active team’s results. This process is rooted in the concept of “alpha diversification” and its ability to equalize each fund’s contributions to the active risk and active return of the portfolio. This is the method underlying our goal of minimizing tracking error at a given allocation and alpha level. When funds are not delivering this desired result, then passive funds may be chosen. This ensures that only the highest active contributors are included - this is key to creating a superior active team.

The selection of passive funds occurs at the most granular level in the portfolio’s allocation. We summarize these results up to higher levels of the asset allocation. There are no limits placed on the asset allocation of either the active or passive sub-portfolio; we only require that the combined allocations match the benchmark.

Our first chart shows the passive portfolio allocation to be close to an 80-20 mix of equity and bonds, while the active portfolio is nearly a 50-50 split between these asset classes. This explains the higher return and volatility of the passive portfolio. However, this does not explain why the equity allocation is so high. For that, we must dig deeper.

Initially, we see that the higher allocation to equity in the passive portfolio was driven by an aversion to foreign equity funds. Conversly, the strong emphasis on active fixed income was driven by the attractive opportunities in high quality bond funds. This was to be expected, given the aggressive “core-plus” strategies available on the fund platform. Both of these allocations reflect the attractiveness of the active opportunities as the main driver of the asset allocation in both the active and passive sub-portfolios.

Digging a bit deeper in the asset allocation hierarchy, we see the distribution of asset segments that drive the higher-level asset allocation effects. It may seem counter-intuitive that the supposedly inefficient “alpha opportunity” sectors are overweighted in the passive portfolio. These include emerging markets equity and high yield bonds. Also surprising is the active overweight of the “plain vanilla” segments that supposedly offer little alpha opportunity, led by high-quality bonds.

The answer to this puzzle is primarily found in the specific funds available on the fund platform. Those funds that don’t perform well individually, or which fail to diversify their tracking error in the context of the team of funds, will not be included in the team of active funds. We are building the best team; we are not running a beauty contest of individual funds.

Belonging to the “best of the best” active team is not a reasonable expectation for any fund that simply earns a higher return than its benchmark. The first potential disqualifier is too much tracking error. The second (and more important) disqualifier is an excess return pattern than is too highly correlated to the other funds. Such a fund has an alpha that is simply not worth the risk that it contributes to the overall portfolio. As such, it will be bypassed to give a place to a superior fund. If no better fund is available, then that place in the asset allocation goes to a passive fund. The result is a higher-peforming alpha team.

Our most detailed asset allocation breakout shows the location of passive vs active exposure by investment style within the equity segment. The passive equity portfolio was two-thirds invested in global growth, while the active portfolio was less than 40% in the growth style. These allocations were primarily driven by the lack of attractiveness in the US large cap and non-US developed sectors.

SOURCES OF RETURN FOR HYBRID PORTFOLIOS

The traditional view of total return performance analysis is a useful starting point. Here is our “benchmark plus active” explanation of the contributors to the total return and risk of the portfolio:

We see that the active process subtracted 11 bps of total return volatility. Since its correlation to the portfolio’s total return was negative, its tracking error lowered total return volatility. Our calculation of tracking error multiplied by correlation of excess return to the portfolio’s total return was: 52 bps x -0.211 = -11 bps. Clearly, our hybrid approach increased the efficiency of the portfolio’s active performance by increasing return while lowering volatility.

Critical insight: the “active process” includes the decision to use passive funds!

We now adapt our contribution to return analysis to evaluate the effects that the active and passive sub-portfolios had on overall performance. It is clear that the passive portfolio was a significant contributor to the portfolio’s total retun and risk.

PERFORMANCE ATTRIBUTION FOR HYBRID PORTFOLIOS

Our next task is to explain 93 bps of excess return and 52 bps of tracking error. First, we acknowledge that this Information Ratio of 1.78 is substantially higher than the 1.21 earned by its all-active alternative. Clearly, the availability of passive options paved the way to identifying a set of complementary funds that delivered the target alpha with lower active risk.

This active risk decomposition uses each component’s weighting, tracking error and correlation of excess returns vs the portfolio’s total excess returns. This approach to risk assessment flows through our holistic investment process, from asset allocation to portfolio construction and through to our “risk aware” performance evaluation process.

We begin with our return data, followed by the data for our risk attribution process:

We see 39 bps of excess return contributed as an “allocation effect” by the aggressive exposures of the passive portfolio. This passive portfolio allocation effect simultaneously subtracted 28 bps of active risk. Clearly, the passive portfolio contributed significantly to the active return of the portfolio!

The active portfolio contributed 54 bps of excess return as a “selection effect.” That amount is correct, but it is incompletely attributed if we fail to account for the corresponding negative allocation effect (-39 bps) from the active portfolio’s conservative asset allocation. This adjustment produces a result of a 93 bps contribution from the active selection effect.

We noted earlier that the “true selection” excess return of the active portfolio was 155 bps, when compared to its asset allocation: (7.5% minus 5.94 percent.) The inclusion of passive funds helped to create this superior active return earned by the other 60% of the portfolio. Essentially, the “hybrid” process creates the highest-returning active portfolio, while using some of its assets for passive investments in those spaces which don’t contain enough of the “best of the best” active funds.

CASE STUDY SUMMARY

We replicated the total return performance of an optimized, all-active portfolio using a hybrid portfolio of 40% passive and 60% active funds. Active performance was greatly improved, as measured by an Information Ratio of 1.78 vs 1.21 in the corresponding all-active portfolio.

Our passive component added value by lowering both overall volatility and tracking error, while also adding to active return through its aggressive asset allocation. The active component was more opportunistic than its all-active counterpart portfolio, given its greater flexibility in selecting the most complementary set of funds, without the negative influence of asset allocation constraints. This positive effect more than offset any of the presumed “active drag” from the passive portfolio.

ROLE OF PASSIVE INVESTMENTS AT A GLANCE

• Passive investments can play a substantial role in an active portfolio that seeks to maximize its efficiency.
• Including passive funds is part of the active process; this is driven by the platform of funds.
• Passive funds eliminate asset allocation constraints that may limit the creation of the best “active fund team”
• Clients appreciate “using active only where it works best” while reducing active costs.

.

.

.

Written in partnership with Stephen Campisi