In a prior life, Scott Stewart was an institutional fund manager.  But he changed career paths and, for the last seven years, has been a professor of Finance at Boston University.  His current research looks at the effectiveness of his former clients in allocating funds among institutional managers.  While many studies have looked at individual investors and the efficacy of their investment decisions, far less attention has been paid to whether the “big money” – assets managed by pension funds, endowments, and foundations (collectively referred to as plan sponsors) – does any better than the little guy.

Stewart and coauthors conducted three separate studies as part of their research.  The first sought to identify the reasons why plan sponsors hire and fire money managers.  The second looked at the key economic question – do plan sponsors gain or lose money when they shift assets between managers?  The third study looked at qualitative data to see whether the decision makers at plan sponsor firms understood the consequences of their decisions, and whether their professed investment policies were consistent with their actual investment decisions.

To answer the first two questions, Stewart utilized a database from a commercial vendor, Effron Associates.  The third question was explored utilizing a survey of 100 large public and corporate plan sponsors.  Effron’s data is widely used in the institutional space for manager selection.  It is free of survivorship bias and represents approximately $6.5 trillion of assets, two-thirds of which are in equity funds.  On average, approximately 8% of this asset pool is “in play” each year – representing the money shifted between managers.  The study utilized data from 1989-2000.

Why Do Plan Sponsors Shift Money among Managers?

Stewart began with the hypothesis that performance, either positive or negative, controls most investment decisions by plan sponsors.  He looked at performance over three separate time periods – one, three, and five years.  He also looked at performance relative to the S&P 500 and relative to the fund’s specific style.  For the style analysis, he used both the manager’s reported style and the style as determined by analyzing each fund’s historical returns, and found the results to be consistent with one another.  Lastly, he determined the level of style of each fund and evaluated whether this was used to shift assets.

One year total performance is very important, as is three and five year active performance.  A really bad one year record can be catastrophic for a manager, even if their three and five year records are strong.  Plan sponsors react quickly to poor one year results, putting managers on watch lists, not providing new money, and many times shifting assets away.  Conversely, managers with good one year and three records tend to get disproportionately higher asset inflows, even when they exhibit weak five year track records.   Commenting on this finding, Stewart said that “plan sponsors are neglecting their mandate when they fail to reward managers who exhibit superior long term performance.”

Performance relative to the S&P 500 is the key determinant of asset flows.  Performance relative to style benchmarks is equally important, suggesting plan sponsors understand manager style.

The level of “style extremeness” was less significant in predicting asset flows between managers.  Stewart noted that plan sponsors do not tend to look at this level of granularity when evaluating managers or, if they do, it does not play a significant role in their decision process.  The data show plan sponsors do not evaluate performance within a style box – e.g., they do not differentiate between deep value and relative value managers.  In 2007, deep value managers did poorly, because their investment strategies were amplified by the poor performance of value stocks in general.  Relative value managers did better.  But there was little adjustment for style extremeness.  Plan sponsors tended to treat a deep value manager who underperformed a value index but outperformed a deep value index the same as a relative manager who similarly underperformed the value index but also underperformed a relative value index.

As would be expected, managers with longer track records garnered more assets.  Managers that experienced high asset flows in the past attracted more flows in the future, but only to a point.  Once a fund gets to a certain size, it has difficulty attracting more assets.

Do Plan Sponsors Create or Destroy Value when Shifting Money among Managers?

Stewart’s data show convincingly that plan sponsors destroy value when shifting assets.  Managers fired by plan sponsors perform better than the managers hired and, as Stewart notes, “this is where the whole process breaks down.”  On average, over a one year period, the plan sponsors in the Effron database lost $20 billion through their allocation decisions and, over a five year period, this number ballooned to an astonishing $60 billion.  Managers most commonly fired outperformed those that were most commonly hired by 300 basis points over one year and by 100 basis points over five years.  Stewart was not surprised to see the value destruction that occurred in the first year, but expected (or at least hoped) these loses would be recovered over time.  They are not.  Stewart says that “the effort that plan sponsors are putting towards hiring and firing managers is not just a waste of time.  It is actually hurting them.  This is clearly disappointing and plan sponsors need to get on top of this.”

The graph below illustrates the performance differential:

The blue and green bars represent the difference in performance between funds in the top and bottom quintiles of cash inflows.  In the year prior to a cash inflow, the funds receiving the most inflows outperformed the group experiencing the least inflows by almost 400 basis points (the blue bar on the left).  But in the year following a cash inflow, the top quintile underperformed the bottom quintile by approximately 300 basis points.

Only a small portion of the value is lost through asset allocation (i.e., shifting assets between different asset classes), and this is primarily in year one, not over three or five year periods.  Shifting assets among different styles destroys value, but the greatest loss of value comes from manager selection, irrespective of asset class or style.  Stewart says that plan sponsors may be “cycling from one extreme to another within styles, leaving a path littered with value destruction.”

Do Plan Sponsors Understand the Consequences of their Actions?

The final piece of Stewart’s analysis involves the analysis of questionnaires completed by 100 plan sponsor firms.  These questionnaires were conducted in an academically rigorous fashion, for example asking the same question multiple times to insure the consistency of responses.  This methodology limited the number of questions that could be answered, but afforded a high degree of confidence to the results.

Responses to the questionnaires revealed that plan sponsors are unaware of the consequences of their decisions; they believe they are creating value (or at least not destroying it) through their manager selection process.   Stewart analyzed the responses by the level of experience of respondents (including criteria such as whether they had previously worked as portfolio managers) and found that those with more experience had a better appreciation for the performance dynamics of asset flows.  Respondents with CFA degrees also grasped the impact of asset flows to a greater extent than those with MBAs.

Stewart adds that, for plan sponsors, “the best experience comes from actually managing money and, in particular, to have been burned in the past.  In my opinion, you are not a true investment professional until you have been responsible for someone else’s money and lost some of it for them, hopefully temporarily, and this research supports that opinion.”

Implications for Financial Advisors

Stewart sees that institutional investors can be subject to the same performance chasing behavior as many individual investors.  Higher turnover among institutional accounts begets poorer performance, as it does among retail mutual funds.  His advice for plan sponsors applies equally to financial advisors:

• Don’t follow trends.  Build an investment process that seeks valuation and stick with it.
• Know the investment process of managers at a granular level.  For example, if you employ value managers, understand whether they are deep value or relative value managers.  Understand how managers’ portfolios are exposed to market dynamics.
• Before you fire a manager, make sure you are doing it at the right time and for the right reason, and not just before a turn in the cycle. 
• Evaluate your entire process, not just the managers you hire.  Put in place a process to insure, in hindsight, that this was the right decision.  Track the performance of the managers you fire.

Stewart recalled an incident from his days as an institutional manager.  He had strung together a solid track record of performance for one of his clients, and was confident he had a good relationship and good communication with the client.  But, as is inevitably the case, his performance hit a speed bump resulting in a poor year.  He was terminated by the client without the opportunity to explain his results.  He went on to have his best year the next year, resulting in the client leaving tens of millions of dollars on the table.  His hope is that plan sponsors improve their manager selection process and do not suffer from this same myopia that destroys value.

Display article as PDF for printing.

Would you like to send this article to a friend?