Wednesday, July 19, 2017

A case study on the weakness of the sole fiduciary model

A case study on the weakness of the sole fiduciary model


Treasurer Dale Folwell came to office with a few clearly stated promises, including a commitment to reduce the public pension’s fees by at least $100 million.[1]  As I’ve previously stated, fee reduction is a worthwhile goal.  Regrettably the Treasurer has gone about achieving his goal in a way that has cost the public pension more than the projected savings.  In two previous blog posts I have criticized North Carolina’s Treasurer over the way he went about allocating the proceeds from terminating thirteen equity managers.[2] 



The State Treasurer deserves credit for being transparent with the press and public about his decision-making process.  The Treasurer conducts a monthly conference call with the press[3] called “Ask me anything” where he explains his policy and decisions. He also seems willing to share documents with the press. 

However, his transparency allows us to see the damage that can be caused by a shift from one sole fiduciary to the next.  Without the continuity provided by a board of trustees, the priorities of a new treasurer can too easily create upheaval in an investment program.  As long time readers know, I was a critic of the prior Treasurer’s foray into alternative investments and the soaring fees associated with that initiative.  I have characterized my own involvement in beginning to hire hedge funds in 2002-2003 for the pension plan as a “failed experiment.”  Thus, I recognize that it is productive for the new Treasurer to examine and make improvements to the pension’s investment programs.  However, the series of memos furnished to the News & Observer show that Treasurer Folwell’s fee reduction initiative is running roughshod over policies meant to keep the pension focused on long-term returns and sound risk control. 

In the June edition of “Ask me anything” the Treasurer indicated that the equity managers had been fired because their performance has lagged their peers and benchmarks over extended periods of time.[4]  Among the thirteen managers terminated by the Treasurer, some definitely deserved to lose their mandates based on performance.  Treasurer Cowell had already identified three of the top candidates to be fired before Treasurer Folwell came into office in January.  According to a memo from the CIO to the State Treasurer, $1.1 billion in assets representing $4.7 million in fees had already been slated for termination before the Treasurer launched his fee-cutting initiative.[5],[6]

Other memos reveal managers with strong investment records terminated because they charge relatively high fees.  While the staff was writing memos and relying on consulting data to justify terminating Hotchkis & Wiley’s Large Cap Value and Sand’s Large Cap Growth portfolios, Kevin SigRist was simultaneously recommending that they be retained in the state’s define contribution (DC) plans (401(K) and 457 also known as the Supplemental Retirement Plan).[7]  While the pension plan’s consultant Callan was providing a rationale for firing Hotchkis and Sands, the DC plan’s consultant, Mercer, was singing the praises of the same managers.[8]  Sand’s even offered to cut their fee by 10%.[9]  Both managers have beaten their benchmarks (net of fees) over long periods, so they’ve demonstrated some ability to add value in the nine [Hotchkis] and eleven [Sand] years they managed money for the pension.[10]  Time Square Mid Cap Focus falls into the same category.  It was eliminated in the pension plan, but retained in the DC plan.[11] 

There’s an investment lesson in all of this.  On those rare occasions when a money manager has skill and can deliver performance, higher than average fees may be warranted.  The goal of any investment program is to keep fees low and only reward the limited number of managers who can add value.

As part of the fee cutting initiative, the Treasurer terminated three fund-of-fund programs.  In these programs, the pension hired managers, who in turn, tried to identify small and new money managers.  Thus these programs have two levels of fees and are usually too small to add meaningful value for the pension.  I criticized the state for creating the program in 2013[12] and understand why these mandates were eliminated as part of the fee cutting initiative.

What I see in the memos and analyses that accompanied the manager terminations is a hodgepodge of rationales designed to justify the Treasurer’s promise to cut fees.  In some cases, there are sound reasons for terminating a manager.  In other cases, the termination isn’t well grounded.  I hope this will not be with Treasurer’s modus operandi when he tackles the far more complex issue of restructuring and reforming the pension’s hedge fund, real estate, and private equity exposure, where the real performance, risk and fee issues reside.  While these areas deserve scrutiny, any changes need to be done slowly and carefully.  Moreover, it’s important to recognize that real estate and private equity can play a meaningful if modest role in the pension program.

In defending his decision to terminate managers and raise cash, the Treasurer told the News & Observer that the pension fund’s investments in stocks cost it billions of dollars.  This is not the first time the Treasurer has talked about the losses incurred during the recession.  Admittedly the pension suffered large unrealized losses in the aftermath of the financial crisis.  However, those losses have been completely recouped.   I’m less concerned about the Treasurer’s mischaracterization than what it says about him as an investor.  All of us bring our prior experiences to the market.  Depression era investors of the 1930s shunned stocks.  Inflation era investors of the 1970s shunned bonds and couldn’t get comfortable as the bull market took hold in the 1980s.

I got my start in the wake of rampant inflation and thus was very cautious as stocks began to rise in the early 1980s.  Fortunately, I met Charlie Baehr[13] a veteran broker who got his start in investments in the 1930s.  He taught me how to respect financial history without becoming paralyzed by it.  Without Charlie’s wisdom I don’t think I would have developed the long-term perspective needed to manage portfolios.

In the end the governance of the pension’s investments would benefit from a board of trustees.  Instead of a new treasurer suddenly pulling the pension’s investment program in a new direction and undermining previous programs and work, a board would offer some continuity and balance.  The Treasurer’s voice is important and his views should be given serious consideration.  However, he shouldn’t be the sole decision-maker.  The pension has an Investment Advisory Council, but it has no power.  The professional staff can only make recommendations and serves at the pleasure of the Treasurer.  While a board isn’t a perfect solution, it would be an improvement over North Carolina’s sole fiduciary model.[14]

Treasurer Fowell’s initiative to reduce fees by terminating thirteen equity managers is an excellent case for putting a board of trustees in charge of the investment of North Carolina’s pension assets.





[3] https://www.youtube.com/watch?v=djVVp0NFMYc
[4] Ibid at 18:15.
[5] Memo from CIO SigRist to Treasurer Folwell, Recommendations to Terminate Public Investment Managers as a component of Enhanced Cost-Efficiency Initiative, March 14, 2017, page 1
[6] Full disclosure: I conducted due diligence and recommended hiring two of the three terminated managers, Piedmont and GMO in 2002.  I also served on the board of Piedmont from 2006-2012.  The third manager, Leading Edge, was hired in 2013.
[7] See, memo from SigRist to Supplement Retirement Board, Discussion/Action: Plan Design & Manager Recommendations, June 2, 2017, page 2.

[8] See, NORTH CAROLINA SUPPLEMENTAL RETIREMENT PLANSFOURTH QUARTER PERFORMANCE REVIEW, page 23.  NORTH CAROLINA SUPPLEMENTAL RETIREMENT PLANS FIRST QUARTER PERFORMANCE REVIEW, page 26.  Sands is also A rated by Mercer, page 7.

[9] Memo, Smith to SigRist, “Recommendation to Terminate Active Domestic Large Cap Managers,” May 1, 2017, page 2.  The memo also warns that Hotchkis and Sands might propose increasing their fee in the DC program, since it was based on a broader relationship.
[10] Full disclosure: In 2002 as CIO, I conducted due diligence on Hotchkis for the pension plan and recommended hiring them for a midcap mandate.  Treasurer Moore hired them, but Treasurer Cowell terminated the mandate.  Hotchkis also manages a small cap mandate for the DC plan.
[11] Memo, SigRist to Folwell, “Recommendations to Terminate Public Equity Investment Managers as a Component of Enhanced Cost-Efficiency Program.”
[14] I’ve written extensively about Treasurer Cowell’s aborted effort to examine the board versus sole fiduciary model.  See for example, http://meditationonmoneymanagement.blogspot.com/2015/10/all-or-none-isnt-good-system-of.html

Tuesday, July 18, 2017

Lowering management fees is a laudable goal, but it has to be done prudently

Lowering management fees is a laudable goal, but it has to be done prudently

North Carolina’s State Treasurer ran on a pledge to reduce the fees paid by the state pension plan by $100 million.  By terminating thirteen equity managers and renegotiating reductions with several others, he is well on the way to meeting his pledge.[1]  However, he’s only saving the pension plan money if you ignore a variety of expenses and costs that have been incurred as a result of firing the managers and placing the proceeds of $7.4 billion in cash and bonds. According to data furnished to the News & Observer, Treasurer Folwell has saved $61 million as of mid-June.[2]  In my view his decisions have cost the state’s pension plan far more than he has saved.



Retail and institutional investors are well served by keeping a lid on fees.  However, reducing fees should be the by-product of an investment program and not an independent goal for investors.  Moreover, investors need make sure they carefully account for all the expenses that offset fee savings.[3]  In this post I’ll discuss the transactions costs and opportunity costs associated with the terminated managers as well the steps the Treasurer could have taken to substantially reduce these costs.

Transactions costs are the brokerage commissions, bid-ask spreads, and the market impact of selling securities.  When the Treasurer terminated the thirteen managers he probably hired a specialist to liquidate the investment portfolios as efficiently as possible.  As a result, I’d expect that the commissions were extremely low and great effort was made to minimize the impact of liquidating large stock positions.  According to data furnished to the News & Observer, the first ten portfolios incurred transactions costs of $9 million.[4]  By the time all the portfolios are liquidated, transactions costs are likely to be about $10 million, which is a substantial offset to the $36 million in annualized savings from eliminating those managers.[5]

However, the pension incurred additional transactions costs that further offset the reported savings.  The proceeds were temporarily invested in all sorts of investment fixed income securities and money market instruments.[6],[7] I suspect these investments are very liquid so the transactions costs incurred in purchasing the instruments were probably low; perhaps $5 million.  At some point in the future the fixed income and money market securities will be sold incurring another set of transaction fees (perhaps another $5 million), and then those proceeds will be invested in equities.  As you might guess, the purchase of stocks will trigger yet another round of transactions costs.  Since the Treasurer has shown a sensible preference for index funds, I expect those transactions costs will be low (I’m guessing $5-$6 million) and the management fee on the index funds will be around $1 million (about .01%).  When you add it all up, the pension will have incurred $20-$25 million in transactions costs, while saving $36 million in management fees from firing equity managers.  The net savings to the pension plan will be .01%.

Now we need to consider opportunity costs.  Unfortunately the expenses of terminating the thirteen managers didn’t end with transactions costs.  When the Treasurer terminated the thirteen managers and put the proceeds in fixed income and money market instruments, he incurred opportunity costs.  Opportunity costs represent the difference in return between the newly selected investment and one that was passed up or eliminated.  Any time an investor shifts between different asset classes, he should evaluate the opportunity cost of making the shift.

While opportunity costs aren’t recorded as actual gains or losses on a pension’s books, they represent real economic gains or losses to the funds.  A simple example illustrates the impact of opportunity costs.  Suppose you sold $100 in stock and put the proceeds in cash.  If the stock went up 10% after the sale, the opportunity cost would be $10 (ignoring lost dividends on the stock or income on the cash).  You might still choose to make the sale because you need to raise cash, but it’s important to weigh the opportunity costs.

As I mentioned in yesterday’s post on market timing, the stock market has risen since the Treasurer terminated the equity managers.  Thus one short-term measure of opportunity cost is to look at how the pension would have performed the managers hadn’t been terminated.  Undoubtedly neither the Treasurer nor I can make any more than an intelligent estimate of the short-term opportunity cost, but I as I wrote yesterday, I believe the short-term opportunity cost currently stands at about $250 million.[8]   Since the Treasurer terminated the thirteen managers stock indices have risen, while fixed income has only enjoyed small gains.

Opportunity costs can also be approximated on a long-term basis using the State Treasurer’s own data.  The Treasurer’s strategic asset allocation makes long-term estimates for the various asset classes that comprise the pension plan.  These estimates are important because they determine the probability that the pension will be able to meet its investment requirement of 7.2%.  The strategic allocation calls for stocks to increase by 8% per year, while fixed income increases by about 2.5% and cash by 2%.[9]  Thus, by selling $74 billion in stocks and committing the proceeds to cash and bonds, the pension is incurring a long-term opportunity cost of about 5.5% to 6% (the difference between equity and fixed income returns) or about $350 million.   As the Treasurer has repeatedly warned legislators and the public, achieving the targeted 7.2% return will be difficult in today’s investment environment.  When $7.4 billion is removed from the equity portfolio, the prospect of reaching the target is materially diminished.[10]

We don’t know when the State Treasurer will recommit the capital to the stock market, so we can’t determine the precise opportunity cost associated with his decision to terminate the thirteen managers and investment the proceeds in bonds and money market instruments.  However, I am confident that the opportunity cost will be far greater than the fee savings that the Treasurer has been emphasizing.

Was there are an alternative strategy to reduce fees, minimize costs, and maintain the pension’s equity allocation?  Yes.  If Treasurer Folwell felt a great sense of urgency to fire certain managers, he could have had a transition manager turn the terminated portfolios into the appropriate indices at low cost through a combination of index derivatives and transfers in-kind of securities.    Taking this action would have reduced transactions costs since the pension would not have had to incur the cost repeated purchases and sales of securities, thereby virtually eliminated opportunity costs. 

As a result, he would have left the asset allocation largely undisturbed (avoiding the mistake of market timing) while actually saving the pension substantial fees.

When I was CIO, Treasurer Richard Moore faced a similar problem in early 2002.  As I recall the Treasurer terminated ten managers representing $4.5 billion in equities.  While we had completed our asset allocation work, we had not identified the managers to take over management of the assets.  We employed the exact strategy, I’ve just described.  We transferred as many as securities as possible into a transition account and then used stock index options and futures to create index-like exposure.  As we slowly hired new managers, we had the transition manager undo the transition portfolio and create the new active and index portfolios.[11]  Undoubtedly, we incurred some transaction costs and perhaps some opportunity costs, but they were minimized.

I support the Treasurer’s initiative to lower fees and pare back active management of the equity portfolio.  However, the Treasurer’s approach has cost the pension plan dearly.

Tomorrow I will look at some of the inconsistencies imbedded in the decisions to terminate thirteen equity managers and explain why this is a case study for eliminating the sole fiduciary and putting in a board of trustees.




[1] http://www.newsobserver.com/news/business/article161425553.html
[2] DST Investment Cost Efficiency Initiatives Summary of Progress, June 14, 2017
[3] For example, before an individual investor redeems a higher cost mutual fund for a lower cost one, she needs to carefully consider the tax consequences.  Unless the new mutual fund has extraordinarily better prospects, the tax bill will swamp any fee savings. 
[4] Transaction Costs Associated with Terminating Public Equity Investment Managers in 2017, June 14, 2017
[5] NCRS Public Equity Investment Manager Terminations and Projected Annual Fee Savings, June 14, 2017
[6] Disposition of NCRS Public Equity Manager Terminations, June 14, 2017
[7] Since the fixed income and money market instruments are managed internally, the pension did not incur any management fees.
[10] Treasurer Folwell reduced the target from 7.25% to 7.2% in April.  http://www.pionline.com/article/20170424/ONLINE/170429944/north-carolina-lowers-assumed-rate-of-return-by-5-basis-points.  The Treasurer said:

“When interest rates were high, it was relatively easy to make your assumed rate by just investing in low-risk bonds. Since we have been in an essentially zero interest rate environment for the past 15 years, achieving that same rate in the future is unrealistic,”

[11] As I recall, $1.5-$2.0 billion became index mandates, the remainder were split into various active mandates.