Agenda Item 4 Attachment 1

Transcription

1 Agenda Item 4 Attachment 1 C A L I F O R N I A P U B L I C E M P L O Y E E S R E T I R E M E N T S Y S T E M E C O N O M I C A S S U M P T I O N S T U D Y FEBRUARY 2012

2 February 20, 2012 David Lamoureux Deputy Chief Actuary California Public Employees Retirement System Lincoln Plaza North 400 Q Street, Room N4340 Sacramento, CA Subject: Dear David: Gabriel, Roeder, Smith & Company ( GRS ) is pleased to present this report with our analysis and recommendation regarding certain economic assumptions used in the actuarial valuation of the California Public Employees Retirement System ( CalPERS or the System ). In our professional opinion, we believe the recommended updates to the price and wage assumptions will provide an improved representation of the System s liabilities and costs. Throughout this report we will document the results of our investigation and findings. We hope the Staff and Board find these items informative and will be happy to discuss this information with you at a time of convenience. Thank you for the opportunity to work on this important assignment. Respectfully submitted, Gabriel, Roeder, Smith & Company Joseph P. Newton, FSA, EA Daniel J. White, FSA, EA Diane Hunt, FSA, EA Senior Consultant Senior Consultant Consultant

3 Table of Contents Table of Contents Page Section I Executive Summary... 2 Section II Price Inflation Analysis... 5 Section III Wage Inflation Analysis... 11

4 SECTION I E X E C U T I V E S U M M A RY

5 Section I E X E C U T I V E S U M M A RY Gabriel, Roeder, Smith & Company ( GRS ) was engaged to perform an analysis of the price and wage inflation assumptions used in the actuarial valuation for the California Public Employees Retirement System ( CalPERS or the System ). A summary of the results of our analysis and recommendations are as follows: Price Inflation Assumption - The current assumption is 3.00% - A reasonable range for the price inflation assumption is 2.50% to 3.00% - GRS recommends decreasing the price inflation assumption to 2.75%. Since the inflation assumption is a component in the development of the other economic assumptions, decreasing the price inflation assumption will also impact the other economic assumptions, including assumed rate of salary increases, wage inflation, the cost-of-living assumption and the valuation interest rate. Wage Inflation Assumption - The current spread between the wage and price inflation assumptions is 0.25%. - GRS recommends increasing the spread by 25 basis points to 0.50% (i.e. an annual wage inflation assumption of 3.25% if the price inflation assumption is 2.75%). We also reviewed the interaction of the recommended wage inflation assumption with the current step-rate salary assumption for consistency and to ensure the combined assumption appropriately models future increases in the member s salary. As a result we also recommend that CalPERS make the following adjustments to the step-rate compensation assumptions (i.e. across-the-board change in the step-rate assumptions): % decrease in the step-rate assumptions for the State Industrial and State Miscellaneous employer groups % increase in the step-rate assumptions for the School and PA Miscellaneous employer groups % increase in the public safety employer groups (i.e. CHP, POFF, Safety, PA CPO, PA Fire, and PA Police). Aggregate Payroll Assumptions - GRS recommends a 3.50% assumption for the year-to-year increase in starting salary for new members (i.e. a wage inflation assumption of 3.25% plus 0.25% for additional increases in productivity). - Decrease the payroll growth assumption from 3.25% to 3.00%. This change recognizes the depressed rate of increase in the employers total payroll expected during the next decade due to the disproportionate number of retirements during this time period (i.e. the baby boomer effect). The relative spread between the price inflation and payroll growth assumption will remain unchanged. 2

6 Actuarial Audit Report Section II Table Summarizing Recommended Economic Assumptions Assumption Recommendation Comment (1) (2) (3) Price inflation (Inflation) 2.75% Wage inflation Payroll growth rate Increase in average compensation paid to new members Ultimate rate of compensation increase for long-service members Nonpublic Safety Groups Ultimate rate of compensation increase for long-service members Public Safety Groups 3.25% (Inflation %) 3.00% (Inflation %) 3.50% (Inflation %) 3.50% to 4.09% 4.15% to 4.29% A component of the compensation increase assumption for individual members Reflects projected lull in payroll growth due to anticipated shift in workforce demographics Includes inflation plus an increase attributable to long-term productivity Includes wage inflation and longterm productivity increases Includes wage inflation and longterm productivity increases An in-depth discussion of our investigation is in Sections II and III of this report. GRS has also separately provided the Actuarial Office an electronic copy of our analysis of the wage inflation assumption and it s interaction with the step-rate portion of the salary assumption to model future increases in the members salary. 3

7 SECTION II P R I C E I N F L AT I O N A N A LYSIS

8 Section II A N A L Y S I S O F T H E P R I C E I N F L A T I O N A S S U M P T I O N By inflation, we mean price inflation, as measured by annual increases in the Consumer Price Index (CPI-U), which is also the inflation measure referenced in the State Government Code for determining the annual cost-of-living adjustment (COLA) for CalPERS retirees. The inflation assumption also underlies most of the other economic assumptions used in an actuarial valuation, including the investment return, individual salary increases, payroll growth, and COLA assumptions. CalPERS currently assumes a 3.00% price inflation assumption. The last time the inflation assumption was changed was in 2004 when the assumption was decreased from 3.50% to 3.00%. Historical CPI The chart below shows the average annual inflation in each of the ten consecutive five-year periods over the last fifty years: Average Annual Inflation CPI-U, Five Calendar Year Averages 12.00% 10.00% 9.79% 8.00% 6.95% 6.00% 4.00% 4.62% 3.86% 4.43% 2.00% 1.69% 2.87% 2.58% 2.65% 2.15% 0.00% yr Avg. Increase 5

9 Section II The table below shows the average inflation over various periods, ending June 30, 2011: Periods Ending June 2011 U. S. City Average Annual Increase in CPI-U Last five (5) years 2.15% Last ten (10) years 2.40% Last fifteen (15) years 2.46% Last twenty (20) years 2.57% Last twenty-five (25) years 2.94% Last thirty (30) years 3.09% Since 1913 (first available year) 3.23% Source: Bureau of Labor Statistics, CPI-U, all items, not seasonally adjusted As you can see, while inflation has been relatively low over the last twenty years, if we look back over a period of 30 or more years, inflation has averaged slightly above 3.00% per year. However, it is hard to ignore the steady decline in inflation statistics over the last 25 years shown in the charts above. Investment Consulting Firms Most investment consulting firms develop an underlying inflation assumption for their forecasting and derivation of forward-looking capital market assumptions. We examined the 2010 or 2011 capital market assumption sets for six investment consulting firms: Callan, Hewitt Ennis Knupp, New England Pension Consulting (NEPC), Pension Consulting Alliance (PCA), RV Kuhns, and Towers Watson. The average assumption for inflation among these firms was 2.53%, with a range of 2.02% to 3.00%. However, the investment consulting firms typically set their assumptions based on a five to ten year outlook, while actuaries must make much longer projections. Bond Market Another source of information about future inflation is the market for US Treasury bonds. Comparing the yields for conventional Treasury securities and Treasury Inflation-Protected Securities (TIPS) provides a useful measure of the market s expectation of future inflation. Conventional Treasury securities compensate its holders by providing a nominal yield with two components, the real rate of interest plus inflation compensation. Since TIPS already adjust for inflation, the yield only includes the real rate of interest. Therefore the difference roughly reflects the inflation expectation for that maturity horizon. For example, the June 30, 2011 yield for 20-year TIPS was 1.47% plus actual inflation. The yield for 20-year non-indexed US Treasury bonds was 4.09%. Simplistically, this means that on that day the bond market was predicting that inflation over the next twenty years would average 2.62% (4.09% 1.47%) per year. 6

10 Section II Below is a chart with the historical spread between 20-year constant and 20-year inflation protected Treasury bonds. 3.5% Interest Rate Spread 20-Year Constant vs. 20-Year TIPS 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% The historical spread between the constant and inflation protected securities was relatively constant from 2005 up to the beginning of the crisis in the credit market. The decrease in the spread during the collapse of the US investment markets and the subsequent volatility reflect differences in liquidity and the risk premiums that buyers of US Treasury securities require. The Cleveland Fed has developed a model that combines information from a number of sources to address the shortcomings of the "break-even" rate illustrated above. Based on the results of its model, the Federal Reserve Bank of Cleveland reported in November 2011 that it estimates the 10-year expected inflation to be 1.40%, which is less than 2.00% over this next decade. Other Sources of Inflation Forecasts In the Social Security Administration s 2011 Trustees Report, the Office of the Chief Actuary is projecting a long-term average annual inflation rate of 2.8% under the intermediate cost assumption. (The inflation assumptions are 1.8% and 3.8% respectively in the low cost and high cost projection scenarios.) These inflation assumptions have remained unchanged for several years. The Philadelphia Federal Reserve conducts a quarterly survey of the Society of Professional Forecasters. Its most recent forecast (third quarter of 2011) was for inflation over the next ten years to average 2.40%. Most observers expect inflation to continue to be low as the economy continues to try to recover from the recession. (Short-term spikes in energy and food costs are possible, but core inflation remains very low.) However, the Society of Professional Forecasters is implicitly assuming a 2.50% inflation rate from , so it is not just the next 2-3 years that are depressing inflation forecasts. 7

11 Section II Another source of information about this assumption is the Public Funds Survey that is prepared on behalf of the National Association of State Retirement Administrators (NASRA) and the National Council on Teacher Retirement (NCTR). This report surveys about 125 plans, including all of the largest public funds covering state employees or teachers. Price Inflation Assumption Frequency of Plans in Public Funds Survery % 3.00% 3.25% 3.50% 3.75% 4.00% 4.25% 4.50% >4.5% 2 1 Inflation Assumption % The current survey shows that the median inflation rate assumed for large public retirement systems in the U.S. is 3.50%, with the most prevalent assumptions at 3.00% and 3.50%. Approximately 40% of the surveyed systems use an assumption of 3.00% or less. This information is also imperfect because some actuaries imply inflation to mean price inflation while others use their inflation assumption as wage inflation. Also, the information in the Public Funds Survey for many of the systems is more than a year old and it is possible that some systems have subsequently updated their assumptions. We are aware of several statewide public retirement systems that have lowered this assumption in 2010 and Since inflation estimates are at such low levels for the near term, one alternative to setting a single assumption would be to set a select and ultimate inflation assumption. For example, a low inflation rate of 2.25% could be set for the next 5 years, and then ultimately a higher inflation assumption for years after that 5 year period. While this approach may intuitively have some value in reflecting the expected short term inflation projections, it has implications on the rate of return, salary increases, and the development of optional forms of payment that may be unintended. Therefore, we recommend using one inflation rate assumption that reflects a balance between the short-term and long-term projections. 8

12 Section II Recommendation Based on all of this information, we believe a reasonable long-term inflation assumption is between 2.50% and the current 3.00% assumption. We recommend CalPERS decrease their inflation assumption from 3.00% to 2.75%, placing it closer to the levels expected in the financial markets. Note, since the price inflation assumption underlies all the other economic assumptions, including the wage inflation, payroll growth rate, COLA assumption and the valuation s rate of return assumption, the Actuarial Office and Board must consider all the implications before changing. For instance, a 0.25% decrease in the inflation assumption while maintaining a real rate of return of 4.75% would result in a decrease in the valuation s rate of return assumption from 7.75% to 7.50%. 9

13 SECTION III WA G E I N F L AT I O N A N A LY S I S

14 Section III A N A L Y S I S O F T H E W A G E I N F L A T I O N A S S U M P T I O N The salary assumption can be thought of as consisting of wage inflation (that part of the pay increase which is given to all employees) and an additional component to reflect step increases and other increases correlated with service. CalPERS updated the step-rate increase assumptions when the Actuarial Office performed a demographic assumption review in The results of this analysis and the updated assumptions are documented in its Experience Study Report dated April Our analysis will focus on developing a recommended wage inflation assumption that, when combined with the updated step-rate increase assumptions previously developed by CalPERS, results in an appropriate assumption for projecting the total increase in the member s salary to their retirement age. After identifying a recommended wage inflation assumption and verifying the appropriateness of the total compensation assumption, we will analyze the assumptions regarding the rate of increase in the average starting pay for new members and the payroll growth assumption. These are not used in calculation of the System s liability as of the valuation date, but are necessary for performing open group projections and for calculating the amortization costs for the established amortization bases. Wage Inflation s Impact on Individual Compensation Increases Many of the factors that result in pay increases are largely inapplicable or have diminished importance for longer-service employees. Generally, the step or service-related increases have stopped or are minimal. Promotions occur with less frequency. Additional training or acquisition of advanced degrees usually occurs early in the career. In theory then, salary increases for longer-service employees are almost entirely driven by wage inflation. Wage inflation is the increase in the average wage of all members of the workforce. The current annual wage inflation assumption for all groups in CalPERS is 3.25%, which is a 25 basis point spread over the price inflation assumption. Historically, wage inflation almost always exceeds price inflation. This is because wage inflation is in theory the result of (a) price inflation, and (b) productivity gains being passed through to wages. The current 3.25% assumption can be thought of as comprised of (a) a 3.00% inflation rate, plus (b) an additional 0.25% for productivity gains. For the ten years ending in 2010, for the economy as a whole, wage inflation has outpaced price inflation by about 0.40% as measured by the difference between increases in the National Average Wage (a statistic used by the Social Security Administration) and increases in the Consumer Price Index. However, this wage inflation measure has exceeded price inflation by about 0.90% for the last twenty years and about 1.00% since the year The Social Security intermediate assumption set assumes wage inflation of 4.00% (2.80% price inflation plus an additional 1.20%). Differences in the price inflation at a national level and a regional geographic location exist due to things such as the cost of housing and can influence short-term wage inflation. Therefore, we also examined inflation information released by the Department of Labor for the Western Region of the United States as well as certain metropolitan areas located in California. For the ten-year 11

15 Section III period ending June 2010, the CPI-U for the Western United States was four basis points higher than CPI-U at a national level (i.e. 2.41% versus 2.37%). The ten-year average of the CPI-U measures for the San Francisco-Oakland-San Jose and Los Angeles-Riverside-Orange County geographical regions were also less than 50 basis points different than the price inflation measure at a national level. Based on these measures, there were not significant regional pressures in price inflation to influence wage increases during this observation period. We also analyzed CalPERS s experience for the ten-year period ending June 30, Below is a table with the average annual increase in salary for continuing active long-service members, i.e. members with 25 or more years of service during the observation period. Non-Public Safety Groups Group Average Salary Increase Price Inflation Salary Increase in Excess of Inflation (1) (2) (3) (4) School 4.26% 2.37% 1.89% Industrial 3.38% 2.37% 1.01% Miscellaneous 3.14% 2.37% 0.77% PA Miscellaneous 4.61% 2.37% 2.24% Public Safety Groups Group Average Salary Increase Price Inflation Salary Increase in Excess of Inflation (1) (2) (3) (4) CHP 6.23% 2.37% 3.86% PA POFF 4.58% 2.37% 2.21% PA Safety 4.52% 2.37% 2.15% PA CPO 6.52% 2.37% 4.15% PA Fire 5.62% 2.37% 3.25% PA Police 6.08% 2.37% 3.71% The above analysis shows that average salary increase for long-service members was significantly higher for the public safety groups than the non-public safety groups. Some of this difference is explained by the disparity in occupations and responsibilities of public safety employees versus general employees. Based on our experience working with other large public retirement systems that also cover public safety employees, public safety employees have generally benefited from above average across-the-board increases in compensation during the last decade. However, we must also recognize that economists are forecasting that the economy will continue to face significant challenges over the next several years or even decade. This will almost certainly mean that the financial pressures local governments currently face will persist, which 12

16 Section III will likely depress the rate of future salary increases that employees will receive. Combining these observations we expect that future salary increases for long-service public safety employees will continue to be greater than the compensation increases that non-public safety employees will experience, but will still be much lower than the increases granted the last ten years. Recommendation Fundamentally, the wage inflation assumption should be consistent for all the membership groups of the Retirement System and the step-rate merit assumption should continue for those employer groups that provide additional increases to members with extended years of service. Based on all the information discussed above, we recommend increasing the difference between the wage and price inflation assumption from 25 to 50 basis points (i.e. an annual 3.25% wage inflation with an annual 2.75% price inflation assumption). We have also analyzed the effect of combining the recommended wage inflation and the current step-rate increases to confirm that the combined assumptions will appropriately model future increases in the member s salary. As a result of this investigation, we recommend CalPERS make the following adjustments, i.e. across-the-board increases or decreases, to the current steprate assumptions: % decrease in the step-rate assumptions for the State Industrial and State Miscellaneous groups % increase in the step-rate assumptions for the School and PA Miscellaneous employer groups % increase in all the public safety employer groups (i.e. CHP, POFF, Safety, PA CPO, PA Fire, and PA Police) Combining these recommended changes will result in total assumed increases in salaries for long-service employees to average about 1.00% above inflation for the non-public safety employees (ranging from 0.75% to 1.34%), and about 1.50% above inflation for the public safety employees (ranging from 1.40% to 1.54%). These recommendations would also result in assumed salary increases for long-service employees that are in proximity of the assumptions used by the Chief Actuary of the Social Security Administration (1.20%) and the historical average difference between the National Average Wage and CPI. Attached is an exhibit with the development of the recommended wage inflation assumption and adjustments to the current step-rate assumption. 13

17 Section III Development of the Recommended Wage Inflation Assumption and Adjustments to the Step-Rate Merit Assumption Group Recommended Difference in the Wage and Price Inflation Assumption Current Step-Rate Merit Assumption 1 Recommended Adjustment in Step-Rate Merit Assumption 2 Total Assumed Increase in Salary in Excess of Inflation (2)+(3)+(4) Actual Average Salary Increase in Excess of Inflation 3 Difference between the Proposed Assumption and Actual Experience (5) (6) (1) (2) (3) (4) (5) (6) (7) Non-Public Safety Groups School 0.50% 0.38% 0.25% 1.13% 1.89% (0.76%) Industrial 0.50% 0.60% (0.25%) 0.85% 1.01% (0.16%) Miscellaneous 0.50% 0.50% (0.25%) 0.75% 0.77% (0.01%) PA Miscellaneous 0.50% 0.59% 0.25% 1.34% 2.24% (0.90%) Public Safety Groups CHP 0.50% 0.40% 0.50% 1.40% 3.86% (2.46%) POFF 0.50% 0.40% 0.50% 1.40% 2.21% (0.81%) Safety 0.50% 0.54% 0.50% 1.54% 2.15% (0.61%) PA CPO 0.50% 0.49% 0.50% 1.49% 4.15% (2.66%) PA Fire 0.50% 0.50% 0.50% 1.50% 3.25% (1.75%) PA Police 0.50% 0.49% 0.50% 1.49% 3.71% (2.22%) Footnotes: 1 Weighted average increase in the step-rate merit increase. 2 Recommended across-the-board change (increase/decrease) in the step-rate component of the compensation assumption. 3 The actual average salary increase in excess of inflation for long-service members (i.e. members with more than 25 years of service)

18 Section III Year-to-Year Increase in Compensation for New Entrants The actuarial valuation performed each year for CalPERS is based on the members as of the valuation date and does not include future members. However, in performing projections, it is necessary to include new cohorts of employees to replace the current active members as they terminate employment or retire. Therefore, it necessary to have an assumption regarding the rate of increase in the average starting pays for each year s cohort of new active members. Generally, it is expected that the year-to-year increase in compensation for new members will increase at a rate higher than price inflation. These additional year-to-year increases are primarily from gains in productivity, which are seen in the wage inflation assumption. Employers must also remain relatively competitive to the compensation packages provided by other employers that are competing for the same workforce, including those employers in the private sector. In the year 2000, the average compensation for members with one year of service was $28,836 for non-public safety members and $36,936 for public safety members. By the year 2009, the average compensation for members with one year of service had increased to $38,965 for nonpublic safety members and $57,214 for public safety members, implying an average annual increase of 3.40% for non-public safety members and 4.98% for public safety members. Since price inflation during this time period was 2.37%, the average increase in new-hire compensation above inflation was 1.03% and 2.61%, respectively. Recommendation We recommend a 3.50% assumption for the year-to-year increase in compensation for new entrants in the non-public safety employers. This is comprised of 2.75% for price inflation, plus 0.50% for wage inflation, and 0.25% to reflect the additional productivity increase (that is part of the long-term step-rate component of the compensation assumption). The new entrant compensation increase assumption is 0.75% higher than the recommended price inflation and approximately 0.25% lower than the actual experience for the last ten years. The difference between the recommended assumption and the actual prior experience reflects the fact that the stresses from the current economy and declines in government revenues will put some downward pressure on future increases in salary. We also recommend using the same 3.50% assumption for public safety employees. While the actual year-to-year increases in average compensation for new members was significantly higher during the study period, this is the same period that public safety employees also received historically high compensation increases. We believe it would be inappropriate to assume those increases would continue indefinitely into the future. For instance, public safety divisions face the same budgetary pressures that non-public safety divisions are dealing with, which will depress the rate of increase in future compensation for new employees in this employer group too. While it wouldn t be unreasonable to assume a slightly higher assumption for public safety employees, we recommend the assumption for both types of employees should be the same. Otherwise, in theory, relative wage difference between the two groups would diverge to a level that ultimately would result in the entire workforce shifting away from the low compensation occupations to the occupations with the higher compensation. 15

19 Section III Payroll Growth Rate Since the contribution rates are calculated to be a level percentage of payroll, as payroll increases over time, the dollar amount of the amortization charges/(credits) also change to remain constant as a percentage of payroll. Thus, the annual amortization cost is dependent on the rate at which total payroll is anticipated to increase and the actuarial valuation must use an assumption regarding the projected increase in total covered payroll. Theoretically, the long-term total payroll for a population of constant size should grow at the rate that starting pays increase. However, because of the baby boomer retirements expected over the next years, actual payroll growth will be lower. Payroll projections assuming a constant membership and using the recommended assumptions described in this report indicate the average expected rate of increase in total payroll to be as follows: Projection Period Average Expected Increase in Projected Total Payroll Non-Public Safety Groups Public Safety Groups All Groups Combined 5 years 2.51% 3.21% 2.68% 10 years 2.65% 3.18% 2.77% 20 years 2.90% 3.23% 2.98% 30 years 3.07% 3.29% 3.12% 50 years 3.25% 3.37% 3.27% Based on the results of these projections, we recommend decreasing the payroll growth assumption from 3.25% to 3.00%. The relative spread between the inflation assumption and the payroll growth rate will remain constant. Also, the recommended assumption will continue to remain close to the projected average increase over the number of years that is approximately equal to the blended amortization period for the non-public safety groups. This assumption will include some margin for the public safety groups. The recommended payroll growth rate is 0.25% less than the recommended wage inflation and 0.50% less than the year-to-year increase in compensation for new members. A payroll growth assumption that is 0.25% or 0.50% higher than the recommended assumption may also be considered reasonable. It is supported theoretically, but it does not reflect the anticipated lull for the next several years due to employer budget constraints and the anticipated change in membership demographics. 16