When going passive, make sure it is for the right reasons

Arguments in favour of passive management have been made by a range of parties across both the academic and financial worlds. While there are grounds for favouring passive management over active management, the case for passive management within fixed income portfolios is not as clear cut as some have made it out to be. For this reason, it is useful for investors to take a closer look at some of the more debatable arguments in favour of passive management. The key point to note here is that whilst these arguments are not necessarily incorrect, it nonetheless pays to consider them in the appropriate context while also recognising their potential limitations.

Questionable arguments for going passive

1. Conceptual argument: zero-sum game

From a conceptual perspective, a commonly used argument against active management can be referred to as that of the zero-sum game. At its most basic, the notion of a zero-sum gain asserts that all the security holdings of all investors in a specific market must aggregate to form that particular market. 

As the holdings of all investors must aggregate to be the market, it follows that if one investor’s dollar outperforms the market over a particular time period, another investor’s dollar must underperform. The result is that the dollar-weighted performance of all investors equals the performance of the market. 

From this, it follows that the universe of active managers in aggregate cannot outperform the market they are operating in and therefore there is little point in actively managing portfolios. Given an appropriate definition of the market, such a conclusion would be technically correct. 

Yet a caveat must be added, which is that the zero-sum game is saying that active managers in aggregate cannot outperform the market, not that they cannot outperform a benchmark. 

There are several reasons for making this important distinction which are particularly relevant for fixed income markets.

Firstly, the zero-sum game only holds if the benchmark represents all the securities in an active manager’s investible universe, i.e. the benchmark does actually represent the entire physical opportunity set. Only then does one investor being overweight a security imply that another investor must be underweight the same security. In general, this assumption does not hold with fixed income benchmarks as they normally represent only a subset of the investible universe. In such a situation, the ability of managers to introduce ex-benchmark securities, i.e. securities which lie outside the benchmark, causes the zero-sum game argument to break down.

Secondly, even if a certain benchmark represented the entire investible universe, the zero-sum game argument only holds if all investors are trying to outperform the same benchmark. There are several reasons why this may not hold. 

1. Firstly, there may be classes of investors who are completely benchmark unaware, such as insurance companies and banks who are holding securities contained within the benchmark for liability or prudential/regulatory reasons. 
2. Secondly, there may be investors using entirely different benchmarks. This is relevant where a market may be characterised by material levels of non-domestic investors. In this case, the non-domestic investors may be using a materially different benchmark to the domestic investors. Indeed, if the country is a relatively small component of the global investors’ benchmark, their holdings in the market may be a relatively unrepresentative exposure even though they are actively managing global fixed income portfolios. In this situation, it is again possible for a subset of managers, such as domestic active managers, to outperform the benchmark even if the benchmark used by this subset of managers does represent the entire investible universe.

2. Statistical argument: historical performance of a manager universe

Simply put, the statistical argument is where a database of active manager returns is taken and compared over a set time period, generating the result that the average active manager has not outperformed the common benchmark. From such an analysis, it is then concluded that active management does not add value. Interpreted literally, such a conclusion is indeed correct, i.e. ‘given the sample of managers in the database and the set time period under examination the average manager hasn’t added value’ is an empirically correct statement. But let’s take a closer look at such a statement and highlight some of its limitations, keeping in mind that ‘there are lies, damned lies and statistics’.

The initial limitation to be aware of with respect to such comparisons is that when referring to the concept of the average active manager, it is important to ensure that any analysis considers the potential impact of size. It is common for a statistical analysis of active managers to treat all managers as effectively equal in size by simply averaging the performance results or displaying them within a distribution. Logically, it follows that the aggregate performance results should be asset weighted to derive a consistent result. Put another way, even the zero-sum game argument requires that ‘the dollar-weighted performance of all investors equals the performance of the market’. Accordingly, it is the number of dollars adding value which is the more important factor to quantify and measure; not necessarily the number of managers adding value.

The second limitation is that the time period under examination and the prevailing market conditions will also have an impact on the level of alpha generated by active managers. More generally, it would be anticipated that during market environments exhibiting high (low) risk premiums and high (low) dispersion in returns, the ability of active managers to generate alpha will be higher (lower). As market conditions will determine whether it is easier (harder) for skilled managers to add value, it is precisely the prevailing market conditions over which the analysis is taking place that needs to be accounted for when determining the capacity of active managers to add value.

3. Difficulty argument: picking a manager that will be successful is difficult

This argument against active management is a practical one which revolves around both (a) the dispersion in alpha between managers, and (b) the volatility of any one manager’s alpha over time. The argument goes that while some active managers may outperform at different points in time, it is just too difficult actually picking which managers will outperform over any timeframe. There is no denying that (a) identifying skilled managers is difficult, and (b) it is impossible for any active manager, no matter how skilled, to consistently add value over all timeframes. 

It is for these very reasons that one of the key benefits of using a multi-manager structure within the fixed income space is that it reduces the level of uncertainty associated with achieving a desired alpha target while not actually reducing the expected level of alpha to which the investor has access. 

This is not to say that with the benefit of hindsight, i.e. knowing what market conditions actually transpired, the investor may not have been able to do better by picking the ‘best’ manager. Just that such a choice is simply based on hindsight, whereas an investor needs to make decisions regarding the future in an environment of inherent uncertainty.

4. Portfolio bias argument: most active managers add value by just overweighting credit

The portfolio bias argument is that most of the additional return from active managers is from the assumption of higher levels of credit risk compared to the benchmark. Therefore, active managers are just charging fees for the assumption of more risk. 

Such an assessment is fair for a material portion of the active fixed income manager universe, but only up to a point. While history does indicate that biases toward credit premiums are the major source of value add for a material proportion of active fixed income managers, it is unclear why this is necessarily a bad thing for investors. As the benchmark is simply a subset of the investible universe, it does not follow that the level or composition of the credit exposure contained within the benchmark is optimal in the first place. If adding a higher allocation to appropriately-managed credit exposures results in a higher return for investors over the cycle then why shouldn’t investors pay to incorporate such an allocation within their fixed income portfolios? The key point here is the concept of appropriately-managed credit exposures. 

Simply adding credit risk without fully understanding the additional risks being assumed is inadvisable. Investors should therefore ensure that a manager has the resources to suitably manage such exposures in response to changing market conditions. In this context, it needs to be highlighted that risk is a multi-dimensional concept and that it is the level of absolute risk assumed which is relevant for the investor, not necessarily the benchmark-relative level of risk.

5. Risk argument: the appropriate definition of risk is measured relative to the benchmark

This brings us to the point that it is implicit in passive management that the relevant measure of risk, typically measured as tracking error, is risk-relative to the benchmark. This highlights a key point, which is that passive investors are, by implication, indifferent to the overall risk characteristics of the selected benchmark. This can lead to some perverse results with respect to risk management, especially where a benchmark is relatively concentrated; namely that by focusing on tracking error as the relevant risk characteristic an investor may be increasing other forms of risk. 

To highlight this point, it is again worth noting that the exposures of fixed income benchmarks will tend to be biased toward the largest issuers of debt. As a result, traditional benchmarks may not necessarily represent the best risk/return profile for an investor. In such a situation, active managers attempting to increase issuer diversification may be reducing the default risk of the portfolio but are incorrectly perceived to be increasing risk as the tracking error may rise. This highlights how a focus on tracking error can have the perverse effect of exposing the investor to other risks, especially where a benchmark may be relatively concentrated with its characteristics changing materially over time. 

Put another way, by minimising tracking error, an investor may actually be inadvertently increasing the total risk profile of their fixed income portfolio.

6. Non-discretionary argument: passive is better as it removes the manager’s discretion

It is often viewed that passive managers are simply replicating the holdings in the benchmark and hence have removed the scope for manager discretion to impact outcomes. Often this is not the case when dealing with fixed income portfolios due to transaction costs. One of the key factors a passive manager needs to manage is the cost of trading in securities. For this reason, passive managers often aim to match the broader characteristics of a benchmark while excluding less liquid securities. This is normally done via a process referred to as stratified sampling, which may or may not include a judgmental overlay applied by the passive manager. The result of this process is that the passive manager’s holdings will not physically replicate the benchmark but instead statistically replicate the overall characteristics of the benchmark.

In the same vein, it’s worth noting that passive managers may actually be using active tilts or strategies such as stock lending which aim to enhance the returns of a passive portfolio to assist in managing or recouping transaction costs. This means there may be an active/discretionary decision-making element to passive portfolio management which, while not apparent under normal conditions, can still impact relative performance under more extreme market conditions. So it is important for investors to recognise that passive management of fixed income portfolios is not always as passive as it may first appear. Therefore, when it comes to investors choosing between active and passive investing, it isn’t so much a question of not allowing manager discretion, but rather how much manager discretion is allowed.

The right reasons for having passive management within a portfolio

When determining whether or not to use active or passive management within a portfolio, it is not simply a case of stating categorically that one is better than the other. There are a range of factors which investors need to consider when determining which type of strategy is best for them.

Firstly, there may be markets where the benchmark does capture the investible universe and the exposures are so concentrated that it is debatable whether active management is worth the time and effort. Such a situation may occur when there are relatively few securities or the level of return dispersion between securities is so low that the ability to add value though active management is significantly reduced. In this case, it may be better for the investor to simply assume the market beta though passive management.

Secondly, there could be an expectation on the part of the investor that market conditions going forward will not be conducive to active management. In this case, it may not make sense for an investor to pay the extra fees involved in hiring active managers. Care, however, needs to be taken here as this decision is an active bet which is predicated on an expectation regarding future market conditions which may prove to be either right or wrong.

Thirdly, the investor may be fee constrained. 

When faced with fee constraints, it is important for the investor to appropriately allocate their risk budget to active management within those markets which will provide them with the most ‘bang for their buck’, i.e. the highest alpha potential for each unit of active risk assumed. 

In such a situation, adopting passive strategies in markets with lower alpha expectations is an appropriate means of allocating a constrained fee budget to maximise the investor’s expected returns.

Much has been said about the perceived benefits of passive management. Some of those arguments are highlighted in this paper along with the caveats investors should keep in mind when considering them. There is no disputing that active management is difficult and that in an uncertain and changing financial environment there is no silver bullet to adding value to portfolios. However, investors need to be mindful that when adopting passive strategies, they are doing so for the right reasons. Failure to do so may result in investors assuming unintended outcomes/risks which may adversely impact the total returns realised in the longer term.