Pzena Investment Management commentary for the second quarter ended June 30, 2015.
Introduction
Passive investing strategies have become all the rage. Since 2000, Index strategies, together with their close cousins, Exchange Traded Funds, have gone from a little more than a tenth of U.S. Equity Mutual Funds assets under management (AUM) to just under one-third at the end of 2014. The numbers are even more dramatic for Global and International funds, with passive strategies going from under 3% in 2000 to 27% at the end of last year. Over that time span, passively managed AUM has grown at 15.3% per annum, more than triple the rate of actively managed AUM, with no sign of slowdown in recent years.
Given this undeniable trend, we ask the following question: is the rise of passive investing inevitable because it offers a superior value proposition? We examine the arguments most commonly cited in favor of passive strategies and look at the assumptions that underlie them. We conclude that passive investing is far from the panacea that many proponents claim it to be. Passive investing offers two good things – low fees and tax efficiency – and gross performance consistently close to a market index. A moment’s thought shows then that, on a net basis, passive strategies allow investors to systematically and consistently underperform the market throughout an investment cycle, the underperformance being precisely those fees and taxes. In contrast, active management strategies adhering to a consistent style in a disciplined fashion across the cycle offer investors a way to outperform an index over the totality of a cycle, despite higher fees and despite almost assuredly underperforming during meaningful portions of that cycle. Patient investors with long time horizons should seek to exploit the excess returns offered via disciplined active management, rather than the consistent below-market returns offered via passive strategies.
Pzena Investment Management – What is active versus passive investing?
The general consensus among professional investors – which we adopt here – uses the term passive to mean an investing strategy seeking to mimic the returns of a given market index or benchmark. We can then define active strategies to be those which seek to depart from a given index with the purposes of achieving some positive investment return over and above that of the index, commonly referred to as “alpha.” Smart Beta investment strategies, which occupy a middle ground between active and passive, construct portfolios to mimic alternative indices based on “factors” associated with outperformance (“factor premium”) over time (i.e., Value, Small Size, Momentum, Low Volatility, Dividend Yield and Quality). Rigorously executed Smart Beta strategies, like rigorous style-based active management, also offer the promise of market out-performance over the cycle.
Pzena Investment Management – Arguments for Passive Investing
We believe the following list captures most of the arguments advanced in favor of passive strategies:
- Passive strategies are transparent. Investors know what they are getting: a market return for a low fee.
- Passive strategies enable investors to customize exposures to a given set of asset classes or sectors at low cost.
- Passive strategies are much more tax-efficient than active strategies.
- Passive strategies present no performance disadvantage in comparison with active strategies. Few active managers out-perform the index in any given year, and those who do out-perform rarely show persistence in ensuing years.
Let’s consider each of these in turn.
Argument #1 seems to place a high premium on the value of transparency – passive investors will reliably get “the market” less fees and taxes. This is quite true – in fact, those low fees and taxes constitute precisely the consistent underperformance that passive strategies deliver. The question is whether such transparency is valuable enough to justify the underperformance. Highly stylized active management or rules-based Smart Beta strategies can also make claims about high transparency.
Argument #2 is widely suggested but wholly inconsistent with the idea of passive management. Nowadays, one can get sector indices, country indices, style indices, and a range of others. Armed with these, an investor can buy, say, a basket of country or regional indices and choose the allocation among them, electing to go “long China” or “long Energy.” Of course, in so doing, the investor is playing the role of the active manager, most likely an active manager adhering neither to a particular style nor to a disciplined process. This kind of active management – termed tactical asset allocation, which is really a special case of market timing – is virtually destined to underperform1 and is clearly at odds with the concept at the heart of adopting passive in the first place.
Argument #3 – the tax efficiency of passive over active strategies – is generally true. But the question is, how large is the difference? To examine the tax argument we compared a simple Large Cap Index Fund vs. an Active Value strategy similar to our own Large Cap Value Portfolio using identical dividend yields of 2% and annual price appreciation (net of fees) of 6%. The main difference between them is the turnover rate: 3% for the index fund, 45% for Active Value. Using a tax rate of 20% on dividends and 23% for capital gains and assuming full redemption after year 20, the index fund’s annual after-tax return was 6.78% versus 6.53% for Active Value, a difference of 25 basis points. Our conclusion is that the tax advantage is real but small – much smaller than most investors making this argument probably realize.
Argument #4 lies at the heart of the active-passive debate. Since the active manager operates with ingoing disadvantages in terms of tax (small, as noted) and fees, the whole case for active relies on the ability of an active manager to deliver enough alpha to counteract those two items. In making the anti-active case, passive proponents cite statistics showing that active strategies exhibit very low “batting averages” – that is, the fraction of actively managed funds beating benchmark in a given year is low, and out-performance tends to be non-persistent (Figure 1).
Yet there are several fundamental problems with the argument. First, “batting average” is a rather incomplete measure of success. An active manager can produce more down (i.e., below-benchmark) years than up years, but if the wins are large and the losses small, there is still outperformance over time. In other words, annualized return is a superior metric to batting average. Second, the argument rests on mutual fund data rather than institutional net returns data, thus adding a layer of fees within “active management” more representative of retail selling costs than manager fees. Third, the definition of active management in Figure 1 is very broad and in particular includes many funds with very low active share,2 funds that might be well-described as “closet indexers.” Since such funds are akin to passive strategies with higher fees, they tend to exhibit performance similar to passive strategies, i.e., highly consistent under-performance and thus very low batting average.
Examining other data on active management presents a more optimistic picture than Figure 1. Drawn from Morningstar data on large-cap actively managed mutual funds and eVestment data on large-cap institutional active managers, Figure 2 shows that the median annualized institutional net return over the past twenty years has exceeded the S&P 500 by 60 basis points, and the median mutual fund return by 105 basis points. Investor behavior, however, appears to be heavily influenced by the most recent five year period in which median returns for active managers lagged the benchmark, overshadowing the strong evidence for outperformance of active managers over the prior 15 years. Although it is not possible to predict future performance, we believe the data support the case that there are institutional active managers with a demonstrated ability to harvest excess return over the long-term.
Now, Figure 2 has one primary flaw: both the retail and institutional data are subject to survivorship bias. Specifically, the returns comprising the sample set exclude non-surviving funds and managers – and these almost surely are below-average. Morningstar has estimated the magnitude of survivor bias to be 80 basis points; we are not aware of comparable estimates for institutional data, but intuitively we would expect the figure to be lower, probably significantly so. Adjusted for this bias, then, it is probably then correct to say that the median institutional net return approximates the benchmark. But a little thought suggests that there is still scope for outperformance, given that many managers do much better than median. Alpha is seemingly available for those willing and able to conduct the necessary due diligence.
In addition, it is critically important to note that both the Morningstar and eVestment include many active managers with low active share. This type of active management – “closet indexing” – achieves investment outcomes akin to high-fee passive strategies. When such funds/managers are excluded from the sample, results change dramatically. A recent paper authored by Invesco suggests that if active management is defined to be only those funds having active share above 60, then 61% of active managers have delivered out-performance over a period spanning five market cycles (Wendler and Peckham, 2015). This study is also subject to survivorship bias issues. However, academic work from Yale (Cremers and Petajisto, 2009) based on datasets correcting for survivorship bias suggests a similar conclusion. In the latter paper, the authors show that funds exhibiting both high active share and high tracking error achieve (net) positive alpha.
Pzena Investment Management – Discipline: The Real Issue
How do we make sense of all this? The key is to recognize the role of discipline in the investment process and to acknowledge that outperformance and consistency are two investing concepts that do not generally go together. Consider the schematic shown in Figure 3, which divides the universe of strategies along two dimensions –alpha and volatility of alpha (i.e., the predictability of relative performance). It thus divides into four quadrants described as follows:
Quadrant I (“Madoff Land”): positive alpha with high alpha consistency. This is investing nirvana, whimsically termed “Madoff Land” in the figure to emphasize the point of its unattainability. Few, if any, (honest) strategies inhabit this quadrant.
Quadrant II (“Alpha Generators”): positive alpha with low or modestly low alpha consistency. This quadrant includes active management winners employing a disciplined style aimed at exploiting factor premiums. Also in this quadrant are “patient” Smart Beta strategies, i.e., factor-based index strategies that keep relatively constant factor weights over time and thus avoid the risks of dynamic weights and “style drift.”
Quadrant III (“Value Destroyers”): negative alpha with low alpha consistency. This quadrant is every investor’s worst nightmare – cumulatively poor performance with highly idiosyncratic variation. Here reside “unskilled” (really better described as “undisciplined”) active strategies, including the prominent specific case of market timing strategies described in the previous section.
Quadrant IV (“Indexers”): negative alpha with high consistency. This quadrant is the land of consistent underperformance, the province of passive investing as well as their more expensive (and worse) brethren, the closet indexers. We believe Figure 3 is the way to reconcile the pessimistic numbers presented in Figure 1 with the extensive body of literature that asserts the existence of positive-premium investing factors like Value. Passive investing (just below the horizontal line in the chart) produces a simple result: performance which consistently and predictably trails the market index by the sum of fees and taxes. Further below the horizontal line live a host of active strategies that are demonstrably inferior to passive investing – closet indexers in Quadrant IV, and active managers employing no disciplined framework (including market timers) in Quadrant III. Above the horizontal line in Quadrant II lives a smaller group of active managers offering positive alpha over the cycle accompanied by the virtual certainty of extensive periods of underperformance. One such group of managers inhabits the Value Investing universe. Elsewhere in our work on cycles, we have documented this phenomenon of periods of both outperformance and underperformance for this style. Periods of underperformance have occurred frequently and have sometimes been of considerable duration. Notwithstanding those facts, patient investors following such a strategy through a full cycle have been rewarded with above-market returns. Arguably, those premium returns would not exist were it not for the fact that they are earned neither consistently nor smoothly. There is no free lunch.
Conclusion
The current trend away from active management towards indexing is undeniable, but is far from inevitable. While a number of forces are at work, the rationale at the heart of passive management is an investor preference for a consistent and predictable level of underperformance relative to a cap-weighted index which approximates “the market.” A growing number of investors appear to prefer such a path to the more variable but ultimately more rewarding one offered by disciplined active strategies. And the recent poor relative performance of the latter group has no doubt reinforced this investor preference.
As we have seen, the concept of alpha (i.e., factor premium) exists. Over the long run, we believe the rewards accruing to the consistent application of disciplined active frameworks will endure and are well worth seeking out. In the end, such is opportunity for the patient long-term investor.
DISCLOSURES
Past performance is no guarantee of future results. The historical returns of the specific portfolio securities mentioned in this commentary are not necessarily indicative of their future performance or the performance of any of our current or future investment strategies. The investment return and principal value of an investment will fluctuate over time.
The specific portfolio securities discussed in this commentary were selected for inclusion based on their ability to help you understand our investment process. They do not represent all of the securities purchased, sold or recommended for our client accounts during any particular period, and it should not be assumed that investments in such securities were, or will be, profitable.