Each year, CalPERS continues to innovate by providing local agencies with new information and additional resource tools to support their actuarial valuations. Staying informed and up to date with these valuations is a crucial aspect of our profession. On September 18, 2024, we hosted a webinar to review some of the key features and results of the new valuations. You can replay the session by clicking HERE.
The 2023 valuations extensively utilize hyperlinks to facilitate navigation within the document and to direct readers to tools and resources on the CalPERS website. Although the Table of Contents may not appear to be hyperlinked at first glance, clicking on any of the topics will take you directly to the corresponding pages within the report. Additionally, external website links are prominently displayed on page two of the valuation:
CalPERS Website Links
Required Employer Contribution Search Tool
Public Agency PEPRA Member Contribution Rates
Pension Outlook Overview
Interactive Summary of Public Agency Valuation Results
Public Agency Actuarial Valuation Reports
COMPARATOR AGENCIES RESOURCE DATASET AND VISUALIZATION TOOL
The “Interactive Summary of Public Agency Valuation Reports” is an excellent resource for comparing local agency valuation results across both Pooled (Risk Sharing) and Non-Pooled plans. The main page features a Power BI report with numerous filters, including Classification, Agency Type, and County. For more detailed analysis, the entire dataset can be downloaded, filtered, and organized into a pivot table. Please note that, as of this newsletter, the links still reflect 2022 valuation results, but 2023 data should be available soon.
RISK SHARING POOLED PLANS VS AGENT NON-POOLED PLANS
During our September webinar, we dedicated considerable time to discussing practical risk-sharing strategies and how the larger risk pool’s gains and losses are allocated to individual plans. We highlighted that most agencies' risk-pool plans represent only a small fraction of the total risk pool, meaning that an adverse experience loss in any single plan has only a minimal impact on its own unfunded accrued liability. See webinar, slides 30-40 for further detail.
Of the 4,138 total CalPERS local agency valuations, a disproportionate number of them are Risk-Sharing pooled plans. These are plans typically had less than one hundred employees, when they were initially created only contain single benefit tier. Agent, non-risk sharing plans had greater than one hundred employees when initially created and may contain several benefit tiers within a given valuation.
Some agencies may receive as many as nine or ten separate valuations. Without consolidating these valuations, it can be challenging for agencies to gain a comprehensive understanding of their overall pension obligations. Additionally, single rate plans have not benefited from actuarial projections of workforce transitions from Classic employees to PEPRA employees, as non-pooled plans have. Fortunately, for the 6/30/24 valuations, CalPERS is considering a significant consolidation of risk-pool plans, grouping them by type (e.g., Safety and Miscellaneous). As of this newsletter, the details of this consolidation are still being developed and reviewed.
BLENDED CONTRIBUTION RATES VS. CONTRIBUTION RATES BY BENEFIT GROUP
During our September webinar, we also discussed some of the pros and cons of using blended rates vs individual benefit group rate information.
Key Take Aways:
For annual budgeting purposes, we recommend using employer rates specific to individual benefit groups rather than blended rates. Blended rates may no longer accurately reflect your agency’s current salary distribution across benefit tiers, as they were developed based on employee salaries and composition from two years ago. The transition from Classic to PEPRA employees may have occurred more rapidly in some departments, leading to potential overbudgeting of normal costs and larger budget variances that can be challenging to explain to departments or auditors.
For long-term forecasts, however, the 10-year blended rate projections can be useful in capturing the expected downward trend in normal cost rates as Classic employees are gradually replaced by Tier Two and PEPRA employees over time.
HISTORICAL TREND INFORMATION
Two years of data does not reveal a trend! In the last few years, CalPERS has added ten years of historical data including assets, liabilities, UAL, normal cost contribution rates, required UAL contributions and even Additional Discretionary Payments (ADPs). Previously, gathering this information could be a chore as each valuation only contained the current and prior year’s data. Kudos to CalPERS for making this information more accessible!
INVESTMENT RETURN STRESS TEST SCENARIOS
(Pg. 20 Risk-Sharing Plans & Page 38 Non-Pooled Plans) CalPERS valuations offer at least two measures of Investment Risk that agencies should take note of.
Long-term Scenario Forecasting
The first stress test scenario is intended for a long-term range-forecasts where CalPERS offers a pessimistic scenario where their portfolio only earns 3% on average for the next 20 years and an optimistic scenario, CalPERS earns 10.8% on average for the next 20 years.
The 5th percentile indicates that only 5% of the possible outcomes are expected to be worse than this scenario, while the 95th percentile indicates that only 5% of the outcomes are expected to be better. This means that 90% of the potential outcomes are expected to fall between these two values.
Over a long-term period, such as 20 years, investment returns are usually expected to average out and show less variability than in shorter periods. By looking at the 5th and 95th percentiles, the analysis is considering the more extreme ends of the potential distribution of returns.
Using the 5th and 95th percentiles allow the organization to assess the impact of relatively rare but possible outcomes, helping to understand the potential financial strain in worst-case scenarios (5th percentile) and the potential benefits in best-case scenarios (95th percentile).
Since the projections imply that only 10% of outcomes fall outside of this range (5% on each end), it suggests these scenarios are unlikely but still within the realm of possibility, making them useful for stress-testing and risk management.
In the example local agency below, a consistent 3% return increases the UAL payment by 66% within four years while a consistent 10.8% return only decreases their UAL payment 13.3% during that same four-year period but the UAL is over a fairly short time horizon thereafter.
Short-term Volatility Stress Test
Standard Deviation Analysis: The report considers a second risk, potential extreme single-year returns, illustrating how an investment loss of one or two standard deviations below the expected return would increase required contributions in future fiscal years.
CalPERS’s stated average expected investment return is 6.8% with a standard deviation loss of ± 12.0%. Here is how to interpret schedule above:
One Standard Deviation CalPERS’ investment return is expected to be within one standard deviation of the average expected return in about 68% of the years. This means the investment return would fall within a range of 6.8% ± 12.0% (between -5.2% and 18.8%) most of the time, with 12% representing the potential funding loss or gain to the plan.
Two Standard Deviations: The investment return is expected to be within two standard deviations of the average in about 95% of the years. This means the investment return would fall within a range of 6.8% ± 24.0% (between -17.2% and 30.8%) in nearly all cases, with 24% representing the potential funding loss or gain to the plan.
Since one standard deviation captures 68% of all outcomes (returns between -12% and +12%), there is a 32% chance of experiencing a return outside this range. Focusing on downside risk, this translates to a 16% probability of a market loss greater than -5.2% (a 12% experience loss) in any given year, and less than a 2.5% chance of a market loss exceeding -17.2% (a 24% experience loss). These thresholds provide a reasonable basis to illustrate the likelihood of a significant one-year investment loss and its potential impact on an agency’s UAL repayment schedule as follows:
The illustration above shows that at a 68% confidence level (less than a 16% chance) the UAL repayment schedule could increase by more than $9.5 million due to a single one-standard deviation loss. At a 95% confidence level (less than a 2.5% chance) that the UAL repayment schedule could increase by more than $19 million by FY 2031-32, when it is projected to peak.
Since many member agencies struggle to maintain enough surplus to improve their funded status by even 1%, experiencing a one-standard deviation (12%) funding loss or more, plus interest, can be particularly challenging. Therefore, it is crucial for agencies to quantify their potential exposure to investment losses and ensure sufficient liquidity to meet the mandatory repayment schedule.
In an upcoming newsletter, we will explore this topic in greater detail, focusing on practical methods to quantify potential loss ranges specific to an agency’s plan, evaluate risk-based funding strategies to mitigate market volatility, and provide guidance on setting reserve targets, determining when to use reserves, and other related policy considerations.
For questions regarding any of these details, please contact the GovInvest Customer Sucess Management Team via email: CSMTeam@govinvest.com.
