New York City Retirement Systems Part I Experience Study Report – TRS and BERS

November 29, 2023

Table of Contents

    Executive Summary

    This report summarizes the Part I Experience Study performed by Milliman of the five New York City Retirement Systems (“NYCRS”):

    • Section I – New York City Employees’ Retirement System (NYCERS)
    • Section II – Teachers’ Retirement System of the City of New York (TRS)
    • Section III – Board of Education Retirement System of the City of New York (BERS)
    • Section IV – New York City Police Pension Fund (POLICE)
    • Section V – New York City Fire Pension Fund (FIRE)

    This report includes Sections II and III for TRS and BERS.

    The primary purposes of the Part I Experience Study Report are to provide high-level observations of the experience for the indicated systems:

    • This report provides information on key preretirement demographic assumptions – withdrawal, retirement and disability – used in the actuarial valuations performed by the OA.
    • This report provides information on the pre-retirement and postretirement mortality assumptions used in the actuarial valuations performed by the OA.

    Part II Experience Study report will incorporate recommendations for changes to the actuarial assumptions reviewed.

    The experience study includes information for the 10-year period ending June 30, 2021 as provided by the OA.  This includes data from 2012 – 2017 contained in the historical database along with updates for the 4-year period ending June 30, 2021 completed by Milliman.

    TRS

    The following is a summary of our observations regarding the experience of TRS.

    Summary of TRS Observations
    Decrement Observation Potential Impact
    Withdrawal More withdrawals than expected at service less than 15 years or after 25 years. Increasing withdrawal rates generally results in lower liabilities.
    Retirement Change in benefit formula at 20 years of service has a significant impact on rates of retirement.  Fewer than expected if less than 20 years and greater than expected at 20 more years of service.  Differences are most notable for ages 62 and older. While increasing the number of members anticipated to receive the 2% formula benefit generally results in higher liabilities, but impact on accrued liabilities and normal cost may vary.
    Ordinary Disability Eligibility for retirement and change in benefit formula at 20 years of service has a significant impact on rates of disability; nearly none occurred for members with at least 20 years of service and 62 and older, but greater than expected for those with less than 20 years. Modifying rates of disability based on eligibility for retirement and benefit formula generally results in higher liabilities.
    Accidental Disability More accidental disability retirements than expected, especially at ages 55 and older. Increasing accidental disability rates generally results in higher liabilities.
    Ordinary Death Slightly fewer deaths than expected but not enough data for experience to be considered fully credible. Reductions in rates of mortality generally results in higher liabilities.
    Post Retirement Mortality
    Service Retirees Actual number of deaths was generally within 3% of expectations. Further review required to compare experience to the most recent published mortality tables but increases in rates of mortality generally results in lower liabilities.
    Disabled Retirees Actual number of deaths was greater than expected.
    Contingent Beneficiaries Actual number of deaths was greater than expected, especially for male beneficiaries.

    BERS

    The following is a summary of our observations regarding the experience of BERS.

    Summary of BERS Observations
    Decrement Observation Potential Impact
    Withdrawal Actual experience for males and females appeared to be more similar than the current assumption resulting in fewer than expected for males and greater than expected for females. Much higher rates of withdrawal occurred during 2013 and 2017. Increasing withdrawal rates generally results in lower liabilities.
    Retirement Greater number of retirements than expected.

    Change in benefit formula at 20 years of service has a significant impact on rates of retirement as greater percentage of retirements occur after completing 20 years of service.  Much higher rates of retirement in 2013.

    		 Increasing rates of retirement, especially for members anticipated to receive the 2% formula benefit generally results in higher liabilities.
    Ordinary Disability Eligibility for retirement and change in benefit formula at 20 years of service has a significant impact on rates of disability; much fewer than expected for members with at least 20 years of service and 62 and older, but greater than expected for those with less than 20 years. Modifying rates of disability based on eligibility for retirement and benefit formula generally results in higher liabilities.
    Accidental Disability More accidental disability retirements than expected. Increasing accidental disability rates generally results in higher liabilities.
    Ordinary Death More deaths than expected but not enough data for experience to be considered fully credible. Increase in rates of mortality generally results in lower liabilities.
    Post Retirement Mortality
    Service Retirees There were significant variations in the mortality experience over time.  During 2013 – 2015, the number of deaths decreased significantly. This was followed by a large spike in 2019. We don’t know what generated these anomalous results and if the data is reliable. Further review required.
    Disabled Retirees
    Contingent Beneficiaries Actual number of deaths was greater than expected.
    $242 billion
    Aug
    2022