Loading...
03-04-2002AGENDA PENSION TRUSTEES 03/04/02 ACTION AGENDA - Trustees of the Employees' Pension Fund Meeting Monday, March 4, 2002 - Commission Chambers Item #1 - Call to Order - 9:35 a.m. Item #2 - Approval of Minutes - 2/4/02 regular meeting - Approved. Item #3 - Approve the assumption study report prepared by PriceWaterhouse Coopers and approve changes to the assumptions used in calculating the expected liabilities for the Employees' Pension Plan per the report as follows: increase the expected rate of return on investments from the current 7% to 7.5%, increase the expected salary increase rate from the current 5% to 6%, adjust the expected employee turnover rate to the table recommended in the report which is based upon historical experience, change the mortality table to the 1994 Group Annuity Reserving Table from the current 1983 Group Annuity Table, adjust the expected retirement rates to the table recommended in the report which is based upon historical experience (adjusted to be more conservative), and make no changes to the expected disability rates. - ACTION: Approved. Item #4 - Other Business - None. Item #5 - Adjournment - 9:56 a.m. a02bO2 1 02/21/02 S?? r Employee Pension Plan `Vorksesslon Item #: T u1Sa A) 3 Trustee Final Agenda Item # Agenda Cover Meeting Date: 3/4/02 Memorandum SUBJECT/RECOMMENDATION: Approve the assumption study report prepared by PriceWaterhouseCoopers and approve changes to the assumptions used in calculating the expected liabilities for the Employees' Pension Plan per the report as follows: increase the expected rate of return on investments from the current 7% to 7.5%, increase the expected salary increase rate from the current 5% to 6%, adjust the expected employee turnover rate to the table recommended in the report which is based upon historical experience, change the mortality table to the 1994 Group Annuity Reserving Table from the current 1983 Group Annuity Table, adjust the expected retirement rates to the table recommended in the report which is based upon historical experience (adjusted to be more conservative), and make no changes to the expected disability rates, 0 and that the appropriate officials be authorized to execute same. SUTITMARY: • PriceWaterhouseCoopers is the Employees Pension Plan actuary. This year, they %vere also hired to evaluate the appropriateness of the current actuarial assumptions. Price WaterhouseCoopers has completed this study and has prepared a report detailing their study and making recommendations. A copy of the report is attached. • A meeting was held with City staff on 2/1/02 to review the report with the purpose of reaching a staff consensus of recommendations to take forward to tiie PAC and Trustees. In addition to Finance and Human Resources staff, PAC members, union representatives, and Lee Definer (Plan's attorney) were also present. There are six assumptions that are used in completing an actuary study. The report outlined each assumption, stating what the plan's current assumption is, what the plan's historical experience has been, what other plans use, and made a recommendation as to whether we should change the assumption or if it is currently appropriate. Where the assumptions are set will impact the expected liabilities of the plan and will have an impact on the cost of the plan to the City. The assumptions should be set at what we believe to be the best estimate of future expectations, without regard to how it will impact the cost of the plan. The assumptions and staff recommendations are as follows: • Expected Return on Investments - recommend increasing from 7%n to 7.5%. • Annual Rate of Salary increases - recommend increasing from 5% to 6%. • Employee Turnover - recommend changing to the table recommended in the report, which is based upon historical experience. • Mortality - recommend changing from the current 1983 Group Annuity Table to the 1994 Group Annuity Reserving Table. • Retirement Rates - recommend changing to the table recommended in the report, which is based upon historical experience (adjusted to be more conservative). • Disability Rates - recommend making no changes. • The Pension Advisory Committee reviewed the report and staff recommendations at their monthly meeting on 2/14/02. They voted 5-1 to recommend to the Trustees that the staff recommendations be accepted. The Pension Investment Committee reviewed the proposed investment rate assumption with Bill Badger and John Willoughby (investment consultants) at the quarterly investment meeting on 2/14/02 and all felt that the 7.5% investment return rate assumption was reasonable. Reviewed by: Originating Dept: Costs Legal N/A Info Srvc N/A M. Simmons 1 4_ v,?i Total N/A Finance Budget N/A Public Works N/A User Dept. Funding Source: Purchasing N/A DCM/ACM Current FY N/A CI Risk Mgmt N/A Other Attachments OP Assumption Study Report Other Submitted by: City Manager, AU 4964 .1 1 a ? None Appropriation Code: Rev. 2/98 Table of Contents Section I - Overview Section II - Economic Assumptions Section III - Demographic Assumptions Section IV - Assumptions Used by Other Florida Government Plans Section V - Recommendations Section I Overview This study was prepared at the request of the City of Clearwater to assist the City in evaluating the appropriateness of the current actuarial assumptions. Generally, in setting actuarial assumptions plan sponsors are guided by a mixture of past experience versus future expectations. Past experience is often believed to be a good predictor of future expectations, however, it must be carefully -,weighed against changes in the underlying environment. For example, past experience in employee turnover or disability rates may be a good predictor of future expectations, however, if there have been changes in the economic climate or other areas that affect these rates the experience may be of limited value. In this review we focused on the following actuarial assumptions: Economic Assumptions Demographic Assumptions Expected Return on Investments Employee Turnover Annual Rate of Salary Increases Mortality Retirement Rates Disability Rates The impact of these assumptions varies from plan to plan and even from individual to individual, making it difficult to make blanket statements. However, we have listed the assumptions below in the general order of their importance to a platys liabilities and costs as well as the correlation between the assumption and the plan's liabilities. • Expected Return on Investments - This generally has the greatest impact on a plan's liabilities. The greater the expected return the lower the present value of future benefits and the lower the plan's funding requirements. • Salary Increase Rate - This assumption is probably second in importance, although turnover could easily be more important depending on the particular circumstances. The greater the expected rate of future salary increases the greater the present value of future benefits and the higher the plan's funding requirements. • Employee Turnover Rates - Probably third in importance, generally, the greater the expected turnover the lower the present value of future benefits and the lower the plan's funding requirements. • Disability Rates - The importance of this assumption depends on the plan's disability benefits. If a plan provides enhanced benefits to disabled employees (such as immediate payment.of unreduced benefits or more than the regular accrued benefit) the impact can be greatly increased. In Clearwater's case, we expect that the greater the expected rate of disability the higher the present value of future benefits and the higher the plan's funding requirements. Section 1 Overview (Continued) • Retirement Rates - The importance of this assumption also depends on the features of the plan. If early retirement benefits are heavily subsidized or the plan has a limit on service this assumption is more important. If a plan subsidizes early retirement benefits, the greater the expected early retirement rates the higher the present value of future benefits and the higher the plan's funding requirements. Generally, the greater the rate of delayed retirement the lower the present value of future benefits and the lower the plans funding requirements. In Clearwater's case we expect that delaying retirement would reduce costs. • Mortality Rates - This is usually the assumption with the least impact unless the plan in question has been using a mortality table that is seriously out of date (which is not the case for the City of Clearwater). Unless the plan has an unusual death benefit, the greater the expected mortality the lower the present value of future benefits and the lower the plan's funding requirements. Please note that the ranking of the assumptions above is a rough estimate based on general trends for a variety of pension plans. The actual rank for a particular plan will vary depending on the plan provisions, the employee demographics and current assumptions. Also, for some of the assumptions, the impact can vary - for example, in certain circumstances increasing employee turnover can increase finding requirements. 2 Section II Economic Assumptions The economic assumptions that we reviewed both have a component based on general inflation and a separate "real" component. For salary increases this is generally known as the merit increase component and for investment returns it is considered the real rate of return. Since both assumptions are impacted by inflation they tend to be fairly closely correlated. For most pension plans with an investment mix that is 40% to 60% equities the difference between the expected return on assets and the salary increase rate is in the range of 2% to 3%. Of course this will vary greatly depending on the individual employer's industry, geography or other factors. However, the spread between these two assumptions is frequently considered a reasonability measure. An expected asset return of 9% with a salary increase assumption of 3% (a 6% spread) might be deemed inconsistent - the investment return implies higher underlying inflation than the salary scale. Another issue that we feel should be weighed here is the tendency of many people to underestimate the real rate of increase in salaries. The actuarial assumption is intended to predict the annual increase in compensation as defined by the plan over the long term. It includes the effects of many factors, including (depending on the plan) cost-of-living, merit, promotions, bonuses, overtime, and market adjustments. We have literally had situations where clients have stated that "no one received more than a 5% salary increase" where there were dozens of individuals whose compensation - as defined by the pension plan - increased by 10% or more. Expected Return on Assets Following is a comparison of the historical rate of return earned by the City's pension plan and the assumed rate: Year Assumed Rate Actual Rate Deviation 2001 7.0% (6.2)%* (13.2)% 2000 7.0% (3.4)% (10.4%) 1999 7.0% 18.6% 11.6% 1998 7.0% 16.7% 9.7% 1997 7.0% 17.5% 10.5% 1996 7.0% 14.8% 7.8% 1995 .7.0% 23.4% 16.4% 1994 7.0% 0.9% (6.1%) 1993 7.0% 9.3% 2.3% 1992 7.0% 6.5% (0.5%) 1991 7.0% 28.5% 21.5% 1990 7.0% 6.2% (0.8%) 1989 7.0% 20.8% 13.8% 1988 7.0% 9.1% 2.1% 1987 7.0% 10.8% 3.8% 1986 7.0% 13.2% 6.2% Average 7.0% 11.7% 4.7% *Estimated - final number not available Section II Economic Assumptions (Continued) The table above summarizes the Citv's experience over the past 16 years. While this is generally considered useful in attempting to set the assumptions it is important to remember: "past performance does not guarantee future results". Annual Salary Increases Following is a comparison of the historical salary increases earned by employees covered by the City's pension plan and the assumed rate: Year Assumed Rate Actual Rate Deviation 2000 5.0% 5.8% 0.8% 1999 5.0% 4.2% (0.8%) 1998 5.0% 7.4% 2.4% 1997 5.0% 5.6% 0.6% 1996 5.0% 6.7% 1.7% 1995 5.0% 6.4% 1.4% 1994 5.0% 4.4% (0.6%) 1993 5.0% 1.2% (3.8%) 1992 5.0% 6.8% 1.8% 1991 5.0% 6.1% 1.1% 1990 5.0% 5.3% 0.3% 1989 5.0% 8.7% 3.7% 1988 5.0% 9.1% 4.1% 1987 5.0% 5.9% 4.9% 1986 5.0% 7.4% 2.4% Average 5.0% 6.1% 1.1% The table above summarizes the City's experience over the past 15 years. As noted above regarding the expected asset return assumption, past results may not be the best predictor of future events. Also, please bear in mind that this is a long-term assumption predicting salary increases as far away as 30 years in the future for some employees. 4 Section III Demographic Assumptions Employee Turnover This is generally the demographic assumption with the greatest impact on the plan's liabilities and funding. We reviewed the turnover within the plan for the past 9 years and three trends were clear - total turnover was low compared to most other groups, total turnover was lower than the assumptions would predict and turnover has been increasing gradually over this period. Total turnover for the past 9 years has increased gradually from 23% to 63% over the period. We reviewed the turnover statistics for this period based on age, years of service. sex and hazardous/non-hazardous duty status. The current turnover assumptions are shown below: Age Males Females 20 14.9% 37.4% 25 9.9% 22.4% 30 6.9% 14.9% 35 4.9% 10.4% 40 2.8% 7.4% 45 1.7% 4.3% 50 0.4% 2.7% 55+ 0.0% 0.9% The percentages shown above represent the probability that an employee in the group indicated will terminate employment for a reason other than death, disability or retirement within the next year. For example, a 30 year-old female is expected to have a 14.9% chance of terminating employment within the next year. Traditionally, the actuarial assumptions are customized for any group within a plan when there is credible evidence or expectation that the experience will be different. The most common differentiator is age. It is fairly common to also adjust the turnover assumption based on service (for example, a newly hired 40 year-old might be twice as likely to terminate employment in the next year as a 40 year-old with five years of service). The last differentiator that is typically used is employee groups - for example hazardous duty versus non-hazardous duty. Our analysis showed that years of service had very little impact on turnover. The rates varied very little within an age group regardless of the employees' years of service. We did, however find meaningful differences based on sex and hazardous duty status. The tables below show the turnover rates based on these variables. Section III Demographic Assumptions (Continued) Non-Hazardous Duty Males Females Age Range Exposure Turnover Rate Exposure Turnover Rate <25 48 20% 34 9% 26-30 122 7% 34 24% 31-35 168 8% 93 12% 36-40 239 4% 108 6% 41-45 353 5% 1,16 , 8% 46-50 281 5% 133 ( 4% 50+ 302 4% 209 10% Total 1,513 6% 757 9% The number of female hazardous duty employees was too low to provide any meaningful infannation by age. For the two-year period we ultimately used to prepare these tables there were only 77 units of exposure, therefore we did not break out this group by sex. Hazardous Duty Age Range Exposure Turnover Rate <25 30 4% 26-30 127 1 % 31-35 165 3% 36-40 172 2% 41-45 105 2% 46-50 125 2% 50+ 86 1% Total 810 2% As in other areas, past experience can be helpful in setting future expectations, however, it is not the only factor to be considered. The City's management should be consulted regarding the past experience and the degree to which future trends may be different. Retirement Rates The City's pension plan has several different eligibility criteria for normal retirement (i.e., unreduced benefits). For hazardous duty personnel this is the earlier of the date they have 20 years of service or the date they attain age 55 with 10 years of service. For other employees this is the earliest of age 55 with 20 years of service, age 65 with 10 years of service or 30 years of service regardless of age. The current actuarial assumption is that all employees will retire as soon as they are eligible for unreduced benefits. This is a common actuarial practice in part because there is often very little Section III Demographic Assumptions (Continued) cost impact of delayed retirement and it is balanced in part by early retirement, which is also disregarded. Tile following tables summarize the actual experience for City employees over the period reviewed. Hazardous Duty Age Range Exposure Retirement Rate <45 11 900 45-49 62 11% 50-54 58 9% 55+ 27 26% N on-Hazardous Duty Age Range Exposure Retirement Rate <55 6 17% 55-59 18 50% 60-64 26 31% 1 65+ 17 47% Mortality Rates This is the assumption that usually varies the least from plan to plan or group to group. This is because, with rare exceptions, there is little reason to predict differences in mortality between any two groups of employees. Also, the data from one employer is usually not reliable due to the limited sample size. Finally, it is not generally an assumption that has a significant impact on plan costs and liabilities. We did review the City's mortality experience over the past 9 years and it was generally lower than would be expected. The sample size was very small to draw any meaningful conclusions, however, we did review it to reach a comfort level that nothing extraordinary appeared to be happening. The City is currently using the 1983 Group Annuity Table for Males, with a 6-year setback for females. This table is widely used today and in fact is required for private plans for sonic purposes by the Internal Revenue Service. Given general improvements in mortality in the U.S., however, it does seem appropriate to adopt newer tables as they become available. The Society of Actuaries recently published the 1994 Group Annuity Reserving Table and this will be required by the IRS in many cases and will likely be the most common table in the very near future. The State of Florida Division of Retirement may also recommend or require it in the near future. 7 Section III Demographic Assumptions (Continued) Disability Rates Over the past 6 years there have been a total of 12 disability pensions approved, an average of 2 per year. Of these disabilities, 25% have been in the line-of-duty. The actuarial assumptions would predict approximately disabilities per year. It is also assumed that 100% of future disabilities will be in the line-of-dutv. 8 Section IN' Assumptions Used by Other Florida Government Plans As requested by the City. eye have enclosed information regarding the actuarial assumptions used by a number of other Florida government pension plans. Please: note that the information included is based on the most recent actuarial valuation report available, which is as of different dates for some plans, ranging from 1999 to 2001. We do wish to point out that while this information may be interesting., it is very difficult to draw anv usefill conclusions from it. For example, knowlllg W11- expected rate of return another City is using is of very limited value without also knowing their investment policy and portfolio mix. Investment Return Plan Expected Rate of Return City of Clearwater 7.00% City of Gainesville - General ( 9.25% City of St. Petersburg - General 8.00% City of St. Petersburg - Police 8.50% City of St. Petersburg - Fire 8.00% City of largo - P&F 8.33% City of Lakeland - I`ire 7.50% City of Lakeland - Police 8.00% City of Orlando - General 8.00% City of Orlando - Police 8.00% City of Orlando - Fire 8.00% City of Tampa - P&F Unavailable City of Tampa - General 8.00% Salary Increases Plan Annual Salary Increase Assumption City of Clearwater 5.0% City of Gainesville - General 4%-7.0% City of St. Petersburg - General 4.5%-8.0% City of St. Petersburg - Police 5.0%-9.5% City of St. Petersburg - Fire 5.0%-9.5% City of Largo - P&F 6.0% City of Lakeland - Fire 6.0% City of Lakeland - Police 6.5% City of Orlando - General 5.5%-8.7% City of Orlando - Police 6.0%-8.0% City of Orlando - Fire 5.5%-7.5% City of Tampa- P&F 5.1%-9.7% City of-1'arnpa - General 6.0% 9 Section IV Assumptions Used by Other Florida Government Plans (Continued) Turnover Plan Rate at Age 30 Rate at I Age 35 Rate at Age 40 Rate at Age 45 City of Clearwater 6.9% - 1\-1 14.9%-F 4.9% - N1 110.4%-F 2.8% - N1 7.4%-F 1.7% - N1 4.3%-F City of Gainesville - General * 4.0% 3.0% 15% 2.0% City of St. Petersburg - General * 3.2% - M 5.0%-F 3.0% - iV1 (4.0%-F 2 - M 12.5%-F 2.5% - M 2.0%-F City of St. Petersburg - Police 3.0% 1.7% 0.1% 0.2% City of St. Petersburg - Fire 1.4% 2.2% 1.5% 0.4% City of Largo - P&F 5.0% 3.8% 2.6% 1.6% City of Lakeland - Fire 3.0% N/A 1.0% N/A City of Lakeland - Police 5.0% N/A 2.0% N/A City of Orlando - General 6.1% 5.3% 4.4% 3.5% City of Orlando - Police * N/A 2.7% N/A 0.0% City of Orlando - Fire * 1.0% ( 0.5% 0.3% 0.3% City of Tampa - P&F * 0.0% - F 2.0%-P 0.0% - F 0.6%-P 0.0% - F 0.5%-P 0.0% - F 0.5%-P City of Tampa - General 9.7% 3.7% 1.4% 0.00/0 * Rates are increased in the first five years of employment Mortality Plan Mortality Table City of Clearwater 1983 GAM City of Gainesville - General 1983 GAM City of St. Petersburg - General UP 1994 - with improvements City of St. Petersburg - Police UP 1994 - with improvements City of St. Petersburg - Fire 1983 GAM City of Largo - P&F 1983 GAM City of Lakeland - Fire 1983 GAM City of Lakeland - Police 1983 GAM City of Orlando - General 1983 .GAM City of Orlando - Police 1983 GAM City of Orlando - Fire 1983 GAM City of Tampa - P&F 1983 GAM City of Tampa - General 1983 GAM 10 Section V Recommendations The intent of this study was to assist the City in determining the most appropriate assumptions to be used for '200? and beyond. The overriding principle in this process is that each individual assumption should reflect the actuary and plan sponsors best estimate of future expectations. It is, however, quite common for plan sponsors to be somewhat consen•ative in this area. There are two primary reasons for this approach: • If it turns out that the "best estimate" of future expectations was too optimistic the plan sponsor will be faced with cost increases. If it turns out that they were too conservative future costs will decrease or benefits can be increased, which is usually a much more palatable situation. • Even if the assumptions of future expectations are very accurate, there could be timing differences that create problems. For example, assume that the long-term rate of return assumption is set at 8.0% and the City does realize this return over the next 30 years. If, however, the returns are well below this level for the first 5 years the City may see significant finding increases until the assets recover. The proper balance of the best estimate of future expectations versus cautiousness can only be determined by the City's management. After all, it is the City that must balance its budget, provide the best possible retirement benefits to its employees and serve its taxpayers. It is also the City that is taking the risk inherent in sponsoring the pension plan. Expected Return on Investments There are a number of different approaches that can be used to establish the expected rate of return on plan investments. These include historical reviews of prior performance and prospective analyses based on the portfolio mix and long-terns economic projections. The historical returns of the City's plan are an obvious benchmark that can be used to establish the assumption for the future. There are a number of factors that must be carefully considered in this context, including: • The investment portfolio may have changed during the period under review. For example, the portfolio may have been 70% equities in the past but 50% currently. Also, the exposure to international, small cap or high tech stocks may have changed significantly. As a result, the past experience may be less relevant as a barometer of future expectations. • The general economic climate may have changed considerably during the period. For example, inflation is currently considerably lower than several years ago and dramatically lower than during the 1980s. • The historical return for the City's plan depends in large part on what period you include in the review. The average annual return for the 16-year period for which we have data is 11.7%. However, the average for the last 10 years is 9.8% and the average for the last 5 years in only 8.6%. For the past 2 years the average return is a negative 4.8%. Section V Recommendations (Continued) Another approach which can be used to establish the expected rate of return involves building on current expectations for future inflation and economic growth. Historically. long-term bond yields have been in the 5.5% to 6.5% range on average. These yields can be looked at as a function of the expected real growth in the economy. the expected long-term inflation rate and a risk prerniunl. Nlost estimates of real growth in the U.S. economy are in the 1.5% to 2.0% range while inflation is generally expected to renlain in the 2.0% to 2.5% range. (This is consistent with the U.S. government economic forecast of nonlinal chain--veighted GDP growth of 3.=4% during 2000-2010). The historical risk premium for these investments has been between 1.5% and 3.0% for most periods. These factors combined predict long-term bond yields of 5.0% to 7.5%. Other long-term fixed income securities would be expected to have a similar expected return range. Return on equity investments can be looked at as the expected long-term bond yield plus an added risk premium. The historical risk premium has been in the range of 3.0% to 5.0% on average. Using the expected long-term bond yield from above. this equates to expected long- term equity returns of 8.0% to 12.E%. Combining the results above, a portfolio with an equity/fixed income mix of 60%/40% would have an expected long-term rate of return of 6.8% to 10.5%. Based on the City's current investment policy and portfolio nlix, we would be comfortable with an expected rate of return anywhere in the range of 7.0% to 8.5%. This change would decrease the Plan's estimated liabilities. Annual Rate of Salary Increases As in the area of the expected return on investments this assumption is often established based on a mixture of past experience and future expectations. One other point that is often considered is the difference, or spread, between the expected return and the salary scale. Based on the City's past experience and current economic forecasts we recorrunend increasing the salary increase assumption to at least 6.0%. We also recommend that the spread between this assumption and the expected return be limited to no more than 2.0%. This change would increase the Plan's estimated liabilities. 12 Section V Recommendations (Continued) Employee Turnover Based on the City's experience «•e recommend the following rates: Recommended Current Age Range Non- Hazardous Duty Males Non- Hazardous Duty Females All Hazardous Duty Male Female <25 15% 15% 5% 14.9% 37.4% 26-30 10% 15% 3% 9.9% 22.4% 31-35 5% 10% 3% 6.9% 14.9% 36-40 5% 10% 2% 4.9% 10.4oio 41-45 5% 5% 2% 2.8% 7.4% 46-50 5% 5% 0% 1.7% 4.3% 50-54 2% 5% 0% 0.4% 2.7% 55+ 0% 0% 0% 0.0% 0.9% This change is expected to increase the plan's estimated liabilities. 13 Nlortality We recommend that the City adopt the 1994 Group Annuity Reserving Table. This change would increase the Plan's expected liabilities. Retirement Rates Based on the City's experience and our experience with other plans we recommend the following rates: Age Range Non-Hazardous Duty Hazardous Duty <45 0% 20% 45 0% 20% 46 0% 20% 47 0% 20% 48 0% 20% 49 0% 20% 50 20% 50% 51 20% 50% 52 25% 75% 53 25% 75% 54 25% 75% 55 50% 100% 56 50% 100% 57 50% 100% 58 50% 100% 59 50% 100% 60+ 100% 100% This change is expected to decrease the Plan's estimated liabilities. Disability Rates We recommend no change to the disability rates at this time. 14