Investing money has become a major public sector enterprise in Minnesota and other states. The return on pension fund investments is a vital source of funding for public employee pensions. Over the last decade, the assets of state and local government employee pension funds have grown nationwide from under $200 billion to over $700 billion. In addition, the investment of state funds has become a significant source of revenue for state governments. In fiscal year 1990, investments produced over $200 million in income for the funding of state obligations in Minnesota. Managing all these investments is thus an important and complex job.
In Minnesota, the State Board of Investment (SBI) is responsible for investing the assets of the statewide public pension funds and various state government funds and accounts. SBI's investments are undertaken within a legal framework established by the Legislature. At the end of fiscal year 1990, the market value of SBI's portfolio was $16.3 billion, including $12.9 billion in pension assets, $3.0 billion in state government funds, and $0.4 billion in the Permanent School Fund.
This report examines the rates of return earned on funds under SBI's control over the last decade. We consider the impact of SBI's management, as well as legal and other constraints, on investment results. The report focuses on the following questions:
In general, we found that the Board and its staff, along with the Investment Advisory Council, have done a good job in fulfilling their statutory and fiduciary responsibilities. SBI has restructured both the Basic Retirement Funds and the Supplemental Investment Fund and consequently improved their investment performance. In addition, SBI was among the first institutional investors to implement a performance-based fee system for compensating its external stock managers. Finally, SBI combined numerous cash accounts under its control into two large cash pools, thus improving the efficiency with which it managed the accounts.
However, statutory constraints restrict SBI's ability to maximize the earning power of the state's retirement funds and the Permanent School Fund by discouraging SBI from holding common stocks. Over the long run, these constraints are costing the Post Retirement Fund at least $35 million to $50 million annually and costing the Permanent School Fund between $3 million to $9 million per year. Statutory changes to the formula used to calculate post-retirement benefit increases and to the accounting restrictions governing the Permanent School Fund are needed so that SBI can increase its stock holdings and the state can benefit from sound long-term investment strategies.
To put SBI's investment performance in perspective, it is first necessary to review the general performance of financial markets over the last decade. In absolute terms, the 1980s was a good decade for investors.
United States stock markets had an average annualized return of 15.9 percent over the ten years ending June 30, 1990. Despite the stock market crash in October 1987, annual stock returns still averaged 15.5 percent over the last five years -- only slightly lower than stock returns in the early 1980s. Bond returns were lower than stock returns but also reached the double-digit level by averaging 11.8 percent per year. Real estate investments did well during the early 1980s but cooled off considerably in the last several years. Over the full decade, real estate investments had an average annual return of 10.0 percent. Even 91-day Treasury bills -- one of the safest and most liquid investments -- provided a return far in excess of the rate of inflation. T-bills returned 9.2 percent per year while inflation increased only 5.0 percent annually.
Returns on financial assets during the 1980s were high compared to historical averages but fairly consistent with historical spreads between stock and bond returns. Over the last 64 years, stocks have increased at an annual rate of 10.3 percent while bonds increased 5.2 percent annually. Inflation during the 1980s was lower than during the 1970s but still high compared to long-term historical averages. Over the last 64 years, the inflation rate averaged 3.0 percent.
To discuss SBI's performance, it is also necessary to understand Minnesota's atypical retirement fund structure. Unlike nearly all other states:
In Minnesota, the Basic Retirement Funds serve as the accumulation pools for the pension contributions of employees and employers during the employees' working years. SBI has a 30 to 40 year time horizon over which to invest the assets of the Basic Funds and thus can focus on generating higher total returns through long-term appreciation in the market value of the assets it holds. The state has set contribution rates so that contributions, plus investment earnings averaging 8.5 percent annually on the Basic Funds, are expected to eliminate the funds' unfunded liability by the year 2020.
At the time of an employee's retirement, funds are transferred from the Basic Funds to the Post Retirement Fund in order to finance the retiree's expected retirement benefits. The Post Fund must earn an average of 5 percent annually in order to support the promised benefits. The Post Fund requires a different investment approach than the Basic Funds. Unlike the Basic Funds, the Post Fund has a shorter time horizon over which to invest and must pay benefits on an ongoing basis. Particularly restrictive on the fund's investment approach is the statutory formula for calculating permanent benefit increases. Retirees receive permanent benefit increases to keep up with inflation only to the extent that the fund's realized earnings exceed 5 percent in a given year. This statutory formula provides a disincentive for SBI to hold stocks in the Post Fund. Holding more stocks in the Post Fund would provide greater benefit increases over the long run, but benefit increases would be more volatile from year to year and would not likely be granted every year under the current formula.
Consequently, SBI has taken a very conservative approach with the Post Fund's investments. In order to guarantee a 3 percent annual benefit increase, SBI substantially lowered the proportion of the Post Fund invested in stocks as interest rates fell over the 1980s. Stock holdings, which were 43 percent of the fund's market value ten years ago, are now less than 10 percent of the Post Fund.
Given the two funds' respective statutory purposes, SBI's performance in investing the Basic and Post Funds has been satisfactory. The Basic Funds did not perform well in the early 1980s but improved their performance over the last five years. For the five years ending June 30, 1985, the Basic Funds increased 13.0 percent annually compared to 14.8 percent for a composite of financial market indices. In addition, excluding alternative assets such as real estate and venture capital, the Basic Funds were outperformed by a majority of the over 800 public and private pension funds reporting to Wilshire Associates' Trust Universe Comparison Service (TUCS).
However, the restructuring of the Basic Funds in 1983 and 1984 began to pay off in the second half of the decade.
For the five years ending June 30, 1990, the Basic Funds had an average gain of 13.0 percent annually, matching a composite of financial market indices. Excluding alternative assets, the Basic Funds gained 13.9 percent annually compared to 13.3 percent for the median pension fund performer. These comparisons may somewhat overstate how well the Basic Funds have performed over the last five years. If alternative assets had been more appropriately weighted in the composite index, the Basic Funds would have slightly underperformed the adjusted composite index. In addition, it could be argued that the Basic Funds should have outperformed the median pension fund performer by more than they did during this period since the Basic Funds held about 20 percent more stocks than the typical pension fund. Nevertheless, the performance of the Basic Funds was considerably improved over the early 1980s.
Given statutory constraints, the Post Fund performed well during the 1980s.
Such strong performance relative to the inflation rate was made possible by the high interest rates of the late 1970s and early 1980s and the significant gains realized on the fund's once substantial stock holdings. However:
SBI staff estimate that benefit increases in the near future will likely range between 2.5 and 4.0 percent per year. The reduction in the fund's stock holdings, decreases in interest rates on bonds, and changes in the fund's cash flow will prevent the Post Fund from generating the large benefit increases retirees grew accustomed to in the 1980s.
Although the Basic and Post Funds have performed satisfactorily given their respective statutory purposes, it is clear that current law does not encourage SBI to maximize the long-term earning power of the Post Fund. To illustrate this point, it is instructive to compare the combined performance of Minnesota's two funds to that of other pension funds which are not generally subject to the same legal restrictions. Such a comparison shows that:
Excluding alternative assets, the combined Basic and Post Funds had an estimated 13.2 percent annual rate of return over the last ten years. The median performer among public and private pension funds reporting to TUCS gained an average of 14.2 percent annually. For the last five years, the difference in performance is smaller but the estimated rate of return on Minnesota's two funds still trailed the median performer among other pension funds. Minnesota's two funds gained 12.8 percent per year compared to 13.3 percent per year for the median performer. Among public funds reporting to TUCS, the median performer for the last five years gained 13.8 percent annually.
There are two reasons for Minnesota's lower rate of return. First:
Excluding alternative assets, Minnesota's pension funds had a combined asset mix of 42 percent stocks and 58 percent bonds and cash equivalents at the end of fiscal year 1990. The median asset mix for funds reported to TUCS was more aggressive: 51 percent stocks and 49 percent bonds and cash equivalents. Since stocks have outperformed bonds, Minnesota's more conservatively invested pension funds have underperformed the median pension fund nationwide.
The performance of stocks held by Minnesota's two funds lagged behind both the general stock market and the median performer among pension funds. Over the last ten years, SBI-held stocks increased 14.4 percent per year, compared to 15.9 percent for the Wilshire 5000 Equity Index, 16.9 percent for the Standard & Poor's 500 Stock Composite Index, and 17.0 percent for the median equity performer among pension funds.
The lower than average stock performance was partially offset by stronger than average bond performance. SBI-held bonds gained 12.0 percent per year compared to 11.8 percent for the Salomon Broad Investment Grade Bond Index and 11.4 percent for the median performer among pension fund bond portfolios.
Issues and Recommendations
In light of investment results over the last decade, the following questions are pertinent:
The current two-fund system, along with the existing benefit increase formula, has produced very good benefit increases for retirees during the 1980s. However, the system is unlikely to perform that well in the future and has two fundamental problems in the long run. Given current law and investment strategy for the Post Fund:
Most other pension systems have one fund, instead of two funds, and base benefit increases on inflation or provide automatic increases independent of investment performance. As a result, other systems are able to increase their stock holdings and their long-term rates of return.
Since this study did not focus on pension policy, we do not offer specific recommendations for changing Minnesota's two-fund system and its atypical benefit increase formula. However, we do recommend that:
The current benefit increase formula should be replaced with one that 1) permits SBI to modestly increase the allocation to stocks in the retirement funds, and 2) provides retirees with an annual increase which is more sensitive to the inflation rate. We estimate that a modest increase in stock holdings -- up to the median level held by other public and private pension funds -- would provide the retirement funds with $35 million to $50 million in additional funding annually over the long run. With care, a new system and formula could be devised so that:
These other objectives may include higher initial pensions for future retirees, lower contribution rates for current employees or employers, or a reduction in the period of time before the various pension systems achieve full actuarial funding.
SBI's lower than average stock performance has resulted primarily because the Basic Funds' stock portfolio has been more reliant on small, growth-oriented stocks than stock market indices and other pension funds. Although small capitalization stocks have substantially outperformed larger capitalization stocks since 1925, the last seven years have seen small capitalization stocks underperform the general stock market. Consequently, SBI's stock performance has trailed that of market indices and the majority of other pension funds, which are generally less reliant on small capitalization stocks.
After study by the Investment Advisory Council and its staff, the State Board of Investment decided in June 1990 to alter the stock portfolio of the Basic Retirement Funds. The portfolio had consisted of a Wilshire 5000 index fund, comprising 60 percent of the portfolio, and active manager accounts, comprising the remainder. In October 1990, SBI's passive stock manager began the two-year process of changing the index fund into a <169>tilted<170> index fund. The tilted fund will be constructed so that, when combined with the stocks held by SBI's active managers, the overall exposure of the fund's stock portfolio will generally approximate the Wilshire 5000. The only significant difference between the performance of the stock portfolio and the Wilshire 5000 will be in how the active managers perform relative to the segments of the market in which they invest.
This new plan can be viewed as a compromise between SBI's current stock portfolio and that of most other pension funds. The new portfolio will eliminate the previous small stock bias relative to the Wilshire 5000. However, it will hold more small, growth-oriented stocks than most other pension funds, which have stock portfolios more closely tied to the S&P 500 and, consequently, to larger capitalization stocks. So, SBI's stock performance will be more reliant on small, growth-oriented stocks than most pension funds, but not any more so than the Wilshire 5000. In addition, the stock portfolio will continue to rely more on passive management than most public funds but will not go to the extreme of indexing the entire stock portfolio.
Since the success of this new strategy depends significantly on the ability of the active managers to add value to the stock portfolio, it is particularly important that the Board and others receive a clear indication of how the active stock manager group is doing relative to the alternative of passive management. In the past, SBI staff have not provided the Board with a clear comparison of the aggregate performance of the active stock manager group to the group's aggregate benchmark. Consequently, we recommend that:
Short-term results, whether favorable or unfavorable, should not be used to justify significant changes in the Basic Funds' stock strategy. However, over the long run, the Board and its staff need to track the overall success of active management in order to determine whether their strategy is working.
Supplemental Investment Fund
The $477 million Supplemental Investment Fund is structured much like a family of mutual funds, providing investment options for a number of public employee groups. The Fund consists of six different accounts with rather different investment objectives and asset mixes. Overall, we found that:
The performance of the four newer accounts is in line with market expectations. In addition, the Income Share Account's performance improved since its stock portfolio was indexed to the Wilshire 5000 in 1988. The Growth Share Account has continued to underperform stock market indices because of its small stock bias, but will likely outperform market indices when small capitalization stocks once again lead the market.
In November 1984, Minnesota voters passed a constitutional amendment that removed investment restrictions on the stock and bond holdings of the Permanent School Fund from the Minnesota Constitution. Following passage of the amendment, SBI no longer had to restrict stock holdings to 20 percent of the fund or corporate bond holdings to 40 percent of the fund. In addition, the type of stocks and bonds that could be purchased was broadened.
However, instead of increasing stock holdings, the State Board of Investment (on the advice of SBI staff and with agreement from the Investment Advisory Council) eliminated stocks from the Permanent School Fund within one year of the amendment's passage. It is not entirely clear why stocks were eliminated from the fund. Staff advised the Board that the accounting restrictions still in place in statutes and in the Constitution were still too restrictive and that holding any stocks would produce unacceptable volatility in the earnings of the fund. In addition, it has been suggested that policymakers in the administration and the Legislature needed the predictable income that bonds would produce in order to balance the budget.
Whatever the reasons for eliminating stocks, the consequences for the Permanent School Fund have been unfortunate. While the stock market increased 15.5 percent annually over the last five fiscal years, the Permanent School Fund has returned only between 8.5 and 9.5 percent per year. Since 1985, the fund has been invested almost entirely in bonds. Since the bonds are generally held to maturity, the fund's principal source of earnings is the interest on the bonds. Those earnings have been used each year to reduce the amount of General Fund appropriations that are necessary to finance K-12 education.
It is clear that:
We estimate that, if the fund had held 50 percent stocks and the state had sacrificed approximately $9 million to $11 million in income over each of the last five years, the Permanent School Fund would have been approximately $115 million larger at the end of fiscal year 1990. The fund would thus be more capable of generating income during a severe budget crisis such as the one we are now experiencing.
We estimate that, by not holding stocks, the state is losing $3 to $9 million annually over the long run. In order to achieve these higher returns, it will be necessary to forego some income during the time stock holdings increase in market value. Given the state's budget crisis, it would be very difficult to alter the fund's portfolio over the next biennium. But now is the time to plan for future changes. As with the Post Retirement Fund, it appears necessary to change the statutory restrictions governing the Permanent School Fund in order to bring about a change in investment strategy. Consequently, we recommend that:
SBI has the responsibility for investing more than 400 state agency accounts with the objective of providing competitive money market returns while preserving capital. Most of these accounts are combined into the Treasurer's Cash Pool, which had an average daily balance of $2.2 billion in fiscal year 1990. <FN - Appendix B of this report provides information on the funds invested by state agencies other than Board Investment.> In addition, SBI is responsible for investing the Trust Fund Pool, which consists of the cash balances of the retirement-related funds and the Permanent School Fund. The Trust Fund Pool had an average daily balance of $0.2 billion last year.
Performance of the cash pools appears to be satisfactory compared to the staff's performance target. Over the last three years, the Treasurer's Cash Pool returned 8.6 percent per year and the Trust Fund Pool averaged 8.3 percent, compared to the target of 7.5 percent for 91-day Treasury bills.
The target, 91-day Treasury bills, is not representative of SBI's cash portfolios and is too easy a target to beat. As of June 30, 1990, the Treasurer's Cash Pool held 40 percent commercial paper, 8 percent corporate notes, and only 19 percent Treasury securities. The Trust Fund Pool held 65 percent commercial paper and only 5 percent Treasuries. In addition, the two pools had average maturities of 213 and 111 days respectively. Consequently, the pools were invested in securities which are both riskier and longer in their maturities than 91-day Treasury bills. Generally, riskier and longer cash portfolios achieve higher yields than 91-day T-bills.
SBI is not alone in its need for an adequate performance measure for cash equivalents. Across the nation, development of such measures has been much slower than for stocks and bonds. However, a number of government units have developed and are using several methods worth considering. We recommend that:
SBI staff provide the Board and the Investment Advisory Council with quarterly and annual reports on investment performance. These reports are very useful and provide extensive data. In fact, the reports provide more information than many other pension plans provide in their annual reports.
However, these reports can be improved in a number of respects. In particular, we recommend that: