Look inside the wrapping
The European mutual fund market has now grown to e2.6trn with a choice of more than 20,000 funds. Supporting this growth is an increasing trend towards institutional investing in mutual funds. Depending on individual countries, the institutional market now makes up 30–60% of the total mutual fund market. In spite of this trend, the question of whether mutual funds are appropriate for institutions is still sometimes asked – implying that mutual funds are inherently different from separate accounts and have an implied bias in their risk structure or their investment horizon.
In a professional investment management environment this is the wrong way to look at the question, focusing, as it does, on the legal wrapper of a portfolio instead of the contents of the portfolio. To determine the appropriateness of a portfolio for an investor, whether a commingled vehicle or a separate account, one must address four fundamental issues:
q the benchmark of the portfolio;
q the targeted tracking error and expected outperformance;
q the investment universe and process, and
q the legal structure of the portfolio.
The benchmark determines the fundamental aim and character of the investment. The main issue in this regard, given the recent proliferation of benchmarks covering the same sectors, is to ensure that the characteristics of the benchmark are appropriate to the aims and risk tolerance of the investor in terms of volatility and concentration. Investors may sometimes not be aware of the implied sector and risk choices they are making by selecting “traditional benchmarks”.
Based on the selected benchmark, the targeted tracking error must be selected. The target tracking error will determine the universe of expected returns relative to the benchmark. In more technical terms, it determines the shape of the expected return distribution of the portfolio including shortfall risk and, given return expectations for the benchmark, the risk of absolute negative returns. Indeed, one can think of the entire portfolio management process as nothing but the construction of different shapes of expected return distributions, from narrow distributions for passive management to wider distributions for active management (Figure 1).
The tracking error target will determine the expected risk level of the portfolio itself. It is also intimately linked to the expected level of outperformance, with higher levels of targeted outperformance typically requiring a higher level of expected tracking error. Lastly, it will impact the contribution that the portfolio will make to total portfolio risk, which may be higher or lower, depending on the correlation of the elements of the overall portfolio .
The investment universe determines by what means the investment manager may seek to add value and, by implication, tracking error to the portfolio through deviations from the benchmark.
The same numerical tracking error may be achieved in different ways by over- or underweighting securities contained within the benchmark or by adding non-benchmark securities to the portfolio, by adding cash or by taking currency risk relative to the benchmark. Let us consider, for example, a US dollar fixed-income portfolio with a US government bond benchmark. Tracking error may be achieved by taking solely duration risk against the benchmark, – by investing exclusively in US government bonds with, however, a shorter or longer duration than the benchmark, depending on the manager’s interest rate expectation. This investment style remains within the investment universe of the benchmark. It results in concentrated tracking error since one single bet (rising or falling rates) determines the investment return. An alternative strategy would be also to invest in corporate bonds and/or foreign currency bonds to diversify the sources of tracking error while still remaining inside the same limits for the amount of overall tracking error for the portfolio.
The choice of investment universe has therefore a significant impact on the structure and diversification of the tracking error within the overall tracking error budget (Figure 2).
Once these decisions are taken, the portfolio and its expected future development are essentially determined. The legal structure, be it separate account or commingled vehicle, may then be thought of essentially as a wrapper inside which the portfolio is managed. This wrapper, provided it does not imply restrictions incompatible with the decisions taken previously, should have no impact on the portfolio itself. A global asset manager would typically manage a number of identical portfolios with the same benchmark, tracking error and universe characteristics within different wrappers.
These wrappers could include separate accounts, branded UCITS as well as UCITS or unit trusts under a distributor’s brand. The difference in the wrappers would, however, be based solely on investor-specific preferences regarding the legal structure and not on differences in the portfolios themselves.
Based on this analysis, it is clear why commingled vehicles are attractive to institutional investors. Provided that the portfolio wrapped inside the commingled vehicle is appropriate for the investor in terms of benchmark, tracking error and universe, the decision to use a commingled vehicle may be taken based on a number of features:
q convenience, including the ability to outsource. Unlike a separate account with a large number of securities, the mutual fund has one single price line and one booking entry per year, since the entire booking process for payments and trades is handled within the fund. For institutional investors with limited resources, outsourcing the booking of hundreds of trades and coupon or dividend payments is attractive. The same convenience argument applies to the ease of account opening and transactions which typically can be handled by fax with standard documents. Lastly, the mutual fund management includes the selection and monitoring of the custodian, a process that can be difficult and onerous for the trustees for investment in more specialised markets.
q costs coupled with flexibility for fees and total expense ratios. The large size of commingled vehicles typically reduces trading and transaction costs. Where commingled vehicles have higher stated investment management fees or higher total expense ratios – due to, for example, the costs of the yearly audit and annual report which are typically included in the fund expenses – than the equivalent separate account, the European regulatory environment allows flexibility in pricing. Unlike the US, where regulations prevent managers from differentiating between investors in the same share class in terms of fees and expenses, European regulation allows flexible, tailor-made rebates. These rebates can be used for institutional investors to adapt the total costs of investing in a commingled vehicle to the amount invested and make the cost comparable to equivalent separate account costs.
Commingled vehicles are not always suitable, however. Institutions may have specialised needs in terms of portfolio structuring based on internal guidelines or the structure of their remaining portfolio. Since there are fixed costs associated with the setting up of commingled vehicles and their cost advantage depends on volume, there may be situations where a manager does not have a commingled vehicle available for specific needs and where the creation of such a vehicle may not be efficient for a limited investment volume. Additionally, in certain specific situations, tax issues may make the use of commingled vehicles inappropriate for an investor.
The increased investment in commingled vehicles by institutions is therefore no coincidence. Provided that the underlying portfolios are professionally structured and are adapted to the needs of the institution in terms of benchmark, tracking error and universe, the convenience advantages of commingled vehicles may prove a powerful incentive to make use of the fee and expense flexibility available under European regulatory systems to take optimal advantage of these vehicles.
Adam Lessing is executive director at Goldman Sachs Investment Management in London