Scott Cameron, CFA
Multnomah Group, Inc. utilizes a rigorous, multi-step process to select investment managers for our clients. This process is grounded in the financial markets research of the past 50-plus years and utilizes proprietary research and reporting analytics to identify suitable managers in each asset class. The purpose of this paper is to explain our philosophy and process for investment manager selection.
Before I begin I want to highlight a few important caveats. First, this methodology is limited to the selection of core, single asset class investment strategies. The evaluation of multi-asset class portfolios (most importantly target date funds series) requires additional factors that are not covered in this white paper. For an explanation of the methodology for evaluating target date funds please see our white paper "Evaluating Target Date Fund Structure." Second, this methodology is targeted to the evaluation of actively managed investment strategies. Index, or passively-managed, investment strategies require their own set of analysis which is critically important but beyond the scope of this white paper.
Finally, the focus of this white paper is on the selection of investment strategies for participant-directed defined contribution retirement plans. These types of plans present a unique set of challenges compared to other portfolio types such as defined benefit plans, foundations, or endowments because portfolio construction is ultimately left to the individual participant. In this case, the investment decision needs to be made with recognition of how the investment choices might ultimately be utilized the individual participants in the plan.
The fundamental philosophy upon which this process is built is Modern Portfolio Theory (MPT) and market efficiency/equilibrium. We know that risk and return are highly correlated. To achieve greater investment returns, investors need to be willing to accept greater investment risk (typically defined as short-term volatility). We also know that different asset classes (i.e. bonds, large cap stocks, real estate, etc.), have different risk profiles, and therefore provide different risk premiums to investors willing to accept their inherent risk. Expected investment returns are based on an investor’s allocation to the various asset classes available in the marketplace. The primary determinant of a fund’s investment performance is going to be its asset class exposure.
With the advent of index funds in the 1970s it has become easy and inexpensive for any investor to capture asset class risk premiums through investment in an indexed or passively-managed investment product targeted towards a specific asset class. The most notable investment examples are the S&P 500 Index funds which provide investors with exposure to large cap U.S.-domiciled companies as represented by the S&P 500 Index. Because asset class exposure is so readily available through these types of funds at a very low cost, the only logical reason for placing money with an active investment manager is if we believe they can add value above the return of the asset class they are investing in.
By definition, market level returns are the returns for the average investor in the marketplace gross of any management expenses. With the growth of mutual funds and 401(k) plans, etc., professional investment average gross returns for professional investment managers are going to be market level returns. Factoring in the higher expenses associated with actively-managed products (i.e. research costs, trading costs, manager compensation, etc.), net of fees, the average professional investment manager will underperform the broad market.
Because the average investment manager cannot add value on top of the asset class returns, and because asset class returns are so readily available through index and passive investment products, we start with our default assumption being that we may index any given asset class. From here we look at the universe of actively managed investment products to determine if we can identify one of a minority of active managers that are likely to add value net of their fees.
Our investment selection process is as follows:
Define the Investment Universe
In order to make a recommendation for an investment manager we first need to analyze the options that are available. In order to analyze the options that are available we first have to define the universe of choices. The breadth of choices is almost overwhelming given the vast array of investment products available on the market today. Morningstar’s mutual fund database has over 10,000 distinct portfolios. Some plan types have a broader universe of investible assets and are able to utilize exchange traded funds (ETFs), collective investment trusts (CITs), and insurance company separate accounts. For purposes of this paper we will not differentiate between these products and collectively will refer to them as "funds."
We rely on third-party data providers to provide raw data, and we supplement their information with our proprietary in-house investment manager database. Combining these data sources into a single database allows us to identify strategies not captured in third-party sources alone. From this point, we are able to filter out products not appropriatefor institutional use, such as B and C share funds. This combined database is the starting point for our analysis.
Funds within the combined database are sorted into peer groups based on their Morningstar Category™. The Morningstar Category is utilized because of its broad adoption within the industry creates a common language for plan sponsors, consultants, and recordkeeping vendors. On occasion we will re-categorize a fund to another peer group that we believe more accurately represents the investment strategy but that is rare because it can create confusion among participants and plan sponsors that might reference that fund in other formats.
Conduct the Quantitative Analysis
Once the investment universe has been defined we assess all of the investment strategies using a proprietary quantitative scoring model. We start with the quantitative assessment because the sheer number of options that are available makes it difficult to conduct a thorough, in depth qualitative review on all of the strategies in a peer group. Our quantitative assessment is useful because of its ability to score a large population of investment strategies quickly, but it has serious limits and should not be relied on exclusively for selecting investment strategies.
Before I begin discussing our scoring model, I want to briefly discuss the limits of quantitative investment analysis. The biggest limitation of quantitative analysis is that relies on historical data. I’m sure most people are familiar with the disclaimer "Past performance is not an indicator of future success," as it appears in most investment literature. The reason that it is so commonplace is because it is true. Too often most quantitative analysis is purely a comparison of the historical performance a fund relative to its peers or benchmark. This type of analysis fails to account for differences in investment strategy, changes in portfolio managers or research personnel, and the dynamics of the market environment during the period analyzed. As a result of this limitation, investment selection processes that rely almost exclusively on quantitative analysis may result in simply chasing a hot fund rather than making a holistically prudent investment decision. Despite this, the ease with which quantitative assessments can be conducted makes them very popular in the marketplace, and a contributing factor in studies that have examined the underperformance of defined contribution plans for participants.
Our proprietary quantitative scoring model allows us to narrow the emphasis of our research and highlight key areas of interest or concern of which to focus our qualitative due diligence. A little later we will discuss the use of the scoring model results in the qualitative due diligence process.
Our proprietary scoring model is designed to analyze the key metrics we believe are relevant in evaluating an investment manager for suitability within a defined contribution retirement plan. Specifically we evaluate the following criteria:
Defined contribution retirement plans present a unique challenge from an investment perspective because we ultimately have no control over how the individual participants utilize the funds within the investment menu. We believe primarily that the focus for investors (whether individuals or large institutions) should be on determining a long-term, strategic asset allocation appropriate for their risk profile and investment objectives. Our perspective on this issue is not unique and allows our clients to leverage most of the investment education programs developed by service providers who seek to reinforce that view with participants.
In order for participants to create suitable strategic asset allocations, it is important that the building blocks offered are consistent with the participants’ expectations about those funds from which they are choosing. Put more simply, if a fund is advertised and communicated as a "large growth" fund we expect that it will invest in large growth stocks and generally behave as large growth stocks behave. This is important for most investors, but particularly necessary in defined contribution plans where most participants do not possess the same sophisticated tools to evaluate a manager’s "style" as institutional investors and consultants.
Evaluating a manager’s style purity is also a strong risk management technique. As asset class is a primary determinant of expected returns, managers that invest outside of their asset class regularly will exhibit variant risk to their benchmark. As a prime example, we saw managers that were supposed to be buying intermediate duration investment grade bonds, investing in more exotic bonds or lower credit quality bonds to generate higher returns. In a normal market environment this was an easy way to add "value." When the credit markets froze in 2008, those managers experienced the most significant losses.
In addition to analyzing where a manager is investing to see if it is consistent with our expectations, we also evaluate investment style for consistency. Style purity tells us where the manager invests on average over time while style consistency tells us how dependably they apply that discipline. Regardless of whether the manager plots as a true large growth manager, or more of a hybrid investing in large and small cap stocks, we want them to adhere to that style across market cycles.
The reasons for this are twofold. First, consistent with the point made with regard to style purity, we want participants to have faith that the managers they are using in their portfolio are investing as advertised. If they are jumping around into different asset classes to play the hot asset class or sector, they are not delivering the asset class exposure that participants are seeking.
The second reason for measuring style consistency is because there are significant risks related to attempting to add value through market timing. A manager that is trying to time the market by moving into and out of different asset classes is going to grade low in style consistency and raise a red flag for our qualitative analysis. Fundamentally we do not believe that market timing is a successful way to consistently add value through active management. Unfortunately, too many managers that try to time the market are reacting to what has happened rather than moving ahead of the market. The challenge in market timing is that managers may be correct on their fundamental analysis of the market, but that alone is insufficient in predicting the direction the market will move.
Expenses are one of the few factors that provide some predictive value for future performance. For all investors the only measure that matters is net of fees performance, but most people do not make the connection that higher fees create a higher hurdle for a manager to generate net of fee performance greater than the passive index. When excess performance is measured in basis points, any small difference in fees can have a significant impact on performance.
The most obvious fee data point that we look at is the expense ratio of the fund. This is the sum of all explicit fees charged by the investment manager for managing, marketing, and servicing the fund. While it is important that this is disclosed, it does not represent the total costs incurred by shareholders for investing in the fund.
Besides the explicit expense ratio, each fund incurs additional fees and costs for trading done in the portfolio. All portfolio managers incur brokerage commissions on the trades they make, which are paid out of portfolio assets. Additionally, the nature of all open markets display a spread between the bid and ask prices which represents another implicit discount on the securities sold and premium on the securities purchased. We refer to this cost as market impact. While difficult to quantify, research has been conducted that indicates that these implicit charges are in some cases equal to, or greater than, the explicit fees investors pay, depending on the frequency of trading and the liquidity of the securities the manager is trading. However, since it is never fully disclosed we utilize the turnover ratio as a proxy for the implicit costs borne by shareholders in the fund.
Manager tenure is evaluated for two primary reasons. First, and most simply, if a manager has a short tenure the historical track record of the fund is even less meaningful than it might otherwise be, because the decision-maker that created the strong track record is no longer the decision-maker making managing the portfolio going forward. Additionally, we favor long-term managers because of the stability it provides to the portfolio and the experience those managers possess.
We evaluate the diversification of the portfolio as a risk management tool for potential investors. Funds provide the benefit of security diversification within a single portfolio. Rather than owning a single stock, funds own a diversified basket of stocks that meet the portfolio manager’s investment criteria. In most cases, the resulting portfolios are sufficiently diversified to minimize the risk that a single security’s performance will have an outsized impact on the total portfolio. In this case, the manager is simply avoiding putting all of their eggs in one basket, or a handful of baskets. Managers that invest in fewer securities, or run more concentrated portfolios, are taking additional risks that may result in greater volatility, the potential for greater risk of loss, and deviations from the benchmark.
A sufficiently diversified portfolio manifests itself with breadth in the total number of securities in the portfolio and a lack of concentration of the portfolio’s assets in a handful of securities. As an example, we would generally view an equity portfolio with more than 60 stocks are sufficiently diversified, although if the top 5 stocks held 50% of the assets it would be lacking a measure of diversification, because those 5 stocks would have a significant impact on the total performance of the portfolio.
Ultimately, we are seeking to identify managers that have a unique skillset that enables them to outperform the market when controlled for the risk factors that impact performance. While not a perfect predictor of their future ability to do so, we would at least like to see that new managers have demonstrated that skill historically. In order to do this we evaluate the performance track record of the manager with a particular focus on identifying the portfolio of the total return that is attributable to the manager’s decision-making.
For any given return period, a manager’s return can be broken down into three main components: benchmark, market timing, and security selection. The benchmark return is straightforward; it is the return of the benchmark index against which we compare the manager. The return attributable to the benchmark is outside the manager’s control, and they should neither be compensated nor penalized for that return element. The other two factors are the responsibility of the manager and where the value should appear.
Market timing is the result of the manager making short-term changes in their investment style in an effort to outperform their benchmark index. As an example, a large growth manager benchmarked against the Russell 1000 Growth Index may decide that U.S. large growth stocks are unattractive relative to their European counterparts and decide to add European stocks to the portfolio. This short-term deviation from his long-term strategy is market timing and is a source of performance deviation from the benchmark (either good or bad). While this is a method of adding value, as discussed above in our evaluation of style consistency, we do not believe market timing is a strong method for adding value.
Security selection is the traditional domain of active investment managers that seek to add value by identifying the best securities within their targeted asset class. Using the above example of a large growth manager, the security selection component is the portion of the return that results from identifying the best stocks within their benchmark. It is our general preference to identify managers that have historically added value through their security selection decisions.
There are multiple methods for parsing a fund’s returns into these components. Holdings-based and returns-based methodologies both have their strengths and weaknesses, but most importantly both provide greater clarity into the relative merits of the historical track record.
We look at the historical track record of a fund because it gives us something concrete to evaluate in trying to determine the likelihood that a manager is going to perform well in the future. As mentioned above, we try to differentiate between the market-driven sources of return and the value added (or subtracted) by the manager. While this is helpful in understanding the drivers of historical performance, even this analysis is not perfect. We also want to see how consistently the manager has been able to add value to see if it was a singular phenomenon or is more consistently present in different environments.
As an example, a manager might have generated positive excess returns from good security selection decisions. But, was this the result of one good call on a single security, or the result of multiple good decisions over a few years? To evaluate this consistency we utilize the metric called "batting average." This metric compares the number of months that a fund generates excess returns relative to a benchmark to the total number of periods in the evaluation window. If the fund had positive excess returns in 6 out of 10 months its batting average would be 0.600. To carry this analogy further, our preference is for single hitters rather than a home run hitter that may experience significant droughts between periods of strong relative performance.
Risk is a difficult concept to define in any discussion on investments. There are thousands of articles, white papers, and research papers on differing forms of risk. Risk can be defined in a myriad of ways to suit different purposes. However, for participant directed defined contribution plans the simple, but important measure of standard deviation is still relevant and informative. Standard deviation measures the volatility of a fund’s monthly returns.
Aside from evaluating the standard deviation of a fund, risk is incorporated into the scoring model in other ways as highlighted above. Style purity and consistency along with diversification are all measures of risk in some way. The goal is to combine these measures to create a more holistic view of the risks associated with an investment strategy.
Evaluating funds against all of the criteria discussed above culminates in a total score for each fund. We compare the total scores to the peer group by ranking funds from 1 to 100. Some funds score well and look like attractive candidates for inclusion in the investment lineup, and other funds score poorly. The scores are used to narrow the field of potential candidates for selection and to red flag issues for due diligence among funds that we already utilize.
Complete the Qualitative Due Diligence
The quantitative assessment is the starting point for the qualitative due diligence process. Funds that score well are researched and evaluated further. Here we seek to understand "why" a particular fund scores well and whether we think they are likely to be successful in the future. We cannot buy past returns so we are trying to use all of the information available to us to make the best-informed decision about future performance.
The biggest challenge in identifying attractive investment managers is that there is no template or standard against which to evaluate managers. By definition a successful investment manager is one that is unique from its peers and the test for us to identify the positively unique managers from the rest.
Our process begins with the firm. Investment management firms come in all sizes and stripes. There are large stand-alone investment managers, subsidiaries of insurance companies or banks, and small boutique managers. In each category we can find successful firms and strategies. So what differentiates the good ones from the bad? As with nearly any entity it comes down to the stability and the culture of the organization. Both go hand in hand as a firm lacking in stability is unlikely to have a strong, positive culture.
The process starts with understanding the ownership structure of the firm. Is it a small boutique with the equity concentrated between a few key shareholders? If so, what is the transition plan? If the firm is part of a larger organization, how does it fit strategically within the parent’s business plans? If the firm is publicly-traded, how does the pressure of outside shareholders impact decision-making in the firm? Each model of ownership has its strengths and weaknesses that must be evaluated.
Stability comes from ownership structure but also from the leadership and people of an organization. To evaluate the leadership and people we have to look at employee turnover and compensation structure. Firms with high employee turnover in the investment ranks or in their leadership, create concern about leadership and culture. High turnover creates uncertainty among existing employees, makes it challenging to create a positive investment culture, and diminishes the intellectual capital of the firm. At the same time, organizations that have low levels of employee turnover may also indicate a culture that is not performance driven. Trying to balance the two issues is difficult. It is necessary to understand what is creating the turnover. Is it related to changes in leadership, compensation structure, retirement, or a combination of factors?
Too often lost in the evaluation of investment products, is that ultimately we are hiring people, not the products or brand names. Any fund might have a commendable track record but what we seek to evaluate are the people making the decisions. We touched on this a little bit in the quantitative analysis, where we can evaluate the portfolio manager’s tenure but that is just touching the service. We do not always have the benefit of selecting managers that have been with a single fund for years, and even if we do, the portfolio manager is frequently supported by multiple research analysts or other portfolio managers that also have an impact on the decisions that are made.
We look specifically for investment teams that are stable, experienced in executing the investment strategy for the fund (even if not in this specific fund), and have clearly articulated roles and responsibilities for each person on the team. Understanding everyone involved in the portfolio and their roles/responsibilities enables us to understand how decisions are made about which securities end up in the portfolio. In some cases that is left to a single portfolio manager. In other instances it might be a team. If it is a team, do they reach consensus? Does each team manager control a slice of the pie, or does one person have override authority?
Understanding the structure of the team also enables us to understand how investment research is completed and decisions are ultimately made. Research can be completed by portfolio managers or research analysts. If the portfolio manager is supported by a team of analysts, are they dedicated to this strategy, or are they a centralized resource that the manager can access? Are the research responsibilities assigned based on sector or industry, or are analysts expected to be generalists? Again, there is not a single model that works better than every other model, so we are trying to evaluate whether the model that is employed makes sense for the organization implementing it, and enhances the investment strategy in a meaningful way.
After understanding the firm and its people, it is critical to understand the investment strategy that we are evaluating. This is really where the "why" comes into play. It begins by understanding the portfolio manager’s philosophy on managing money. Every manager has a certain way that they view the markets, and a belief that they can add value. That should translate directly into the investment strategy that they have developed. Again, there is not a single way of doing things. What we are looking for are investment managers that are able to articulate a well-defined process for evaluating securities and making decisions. The process should be consistent with their investment philosophy and should incorporate the key differentiators that they believe add value.
The investment strategy needs to also be reasonably grounded in fundamental research. In the extreme, a portfolio manager that says he or she likes companies that begin with the letter "A" has an investment strategy that might have generated great returns but is not likely to be a strong candidate for selection. Finally, the investment strategy needs to be consistent with the historical track record of the investment strategy. If the manager says that he or she only invests in U.S. stocks, but the performance attribution shows significant international exposure that should be a red flag that something is not right.
In discussing the investment strategy with a portfolio manager, we focus on the "buy" and "sell" disciplines. The buy discipline is the process for determining which securities to buy. It begins with the universe of stocks that the manager may purchase and narrows that list down to the few securities that make it into the portfolio. It should include the steps that the manager uses, the key criteria that they evaluate, and ultimately who retains decision-making authority over the portfolio. The sell discipline is the process for determining how securities that are already in the portfolio are sold. Every good idea has a shelf life and the portfolio manager should be able to clearly articulate the parameters that they use for deciding when to sell an existing position.
As part of the due diligence process, we are trying to reach some conclusions about the investment managers approach to fund shareholders. Key issues are their willingness to close large or rapidly growing funds, their inclination to limit fast traders, their limitations on using soft-dollars, and their philosophy towards fees. Managers should be able to articulate their capacity for managing assets, and should design compensation systems that reward investment personnel for strong risk-adjusted performance over complete market cycles, rather than for growth in asset base.
All managers have limitations on the size of the portfolio they can manage effectively before the portfolio size affects their performance. This is especially true for managers targeting less liquid or illiquid asset classes such as small cap stocks or emerging market stocks, although it is even true in large markets like U.S. large cap stocks. Our expectation is that the managers have an understanding of that limit, and are proactively willing to place limitations on asset growth as they approach or exceed that limit, so that they protect existing shareholders. This obviously impacts a manager’s revenue as they shut down the flow of assets into their most popular products.
Based on our analysis of the quantitative and qualitative factors discussed above we ultimately narrow down our recommendations to a couple of managers we deem suitable for inclusion in a client’s investment lineup. Our final recommendation is made based on our understanding the client’s objectives, the structure of their investment menu, and the other investment strategies that they already utilize within the investment menu. This recommendation is made to the client’s Investment Committee, who provides final approval of our recommendation.
† Information herein is provided for general informational purposes and not intended to be completely comprehensive regarding the particular subject matter. Multnomah Group does not represent, guarantee, or provide any warranties (express or implied) regarding the completeness, accuracy, or currency of information or its suitability for any particular purpose. Receipt of information does not create an adviser-client relationship between Multnomah Group and you. Neither Multnomah Group nor our advisory affiliates provide tax or legal advice or opinions. You should consult with your own tax or legal adviser for advice about your specific situation.