Individual Retirement Account Balances, Contributions, and Rollovers, 2012; With Longitudinal Results : The EBRI IRA Database

Transcription

1 May 2014 No. 399 Individual Retirement Account Balances, Contributions, and Rollovers, 2012; With Longitudinal Results : The EBRI IRA Database By Craig Copeland, Ph.D., Employee Benefit Research Institute A T A G L A N C E As part of the EBRI Center for Research on Retirement Income (EBRI CRI), the EBRI IRA Database is an ongoing project that collects data from IRA plan administrators across the nation. For year-end 2012, it contained information on 25.3 million accounts owned by 19.9 million unique individuals, with total assets of $2.09 trillion. The EBRI IRA Database is unique in its ability to track individual IRA owners with more than one account, thereby providing a more accurate measure of how much they have accumulated in IRAs. The average IRA account balance in 2012 was $81,660, while the average IRA individual balance (all accounts from the same person combined) was $105,001. Overall, the cumulative IRA average balance was 29 percent larger than the unique account balance. Rollovers overwhelmingly outweighed new contributions in dollar terms. While almost 2.4 million accounts received contributions, compared with the 1.3 million accounts that received rollovers in 2012, 10 times the amount of dollars were added to IRAs through rollovers than from contributions. The average individual IRA balance increased with age for owners ages 25 or older, from $11,009 for those ages to $192,961 for those ages 70 or older. Looking at individuals who maintained an IRA account in the database over the three-year period in question, the overall average balance increased each year from $95,431 in 2010 to $95,547 in 2011 and to $106,205 in Males had higher individual average and median balances than females: $139,467 and $36,949 for males, respectively, vs., $81,700 and $25,969 for females. However, the likelihood of contributing to an IRA did not significantly differ by gender within the database, as both Roth and traditional IRAs owned by either males or females (as well as those without a gender identified in the database) had similar probabilities of receiving contributions. IRA owners were more likely to be male. In particular, those with an IRA originally opened by a rollover, or a SEP/SIMPLE IRA were much more likely to be male (57.4 percent of the former, and 58.2 percent of the latter). Younger Roth IRA owners were more likely to contribute to the Roth IRA than were older Roth IRA owners: 43 percent of Roth owners ages contributed to their Roth in 2012, compared with 21 percent of Roth owners ages A monthly research report from the EBRI Education and Research Fund 2014 Employee Benefit Research Institute

2 Craig Copeland is senior research associate at the Employee Benefit Research Institute (EBRI). This Issue Brief was written with assistance from the Institute s research and editorial staffs. Any views expressed in this report are those of the author and should not be ascribed to the officers, trustees, or other sponsors of EBRI, Employee Benefit Research Institute-Education and Research Fund (EBRI-ERF), or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on specific policy proposals. EBRI invites comment on this research. Copyright Information: This report is copyrighted by the Employee Benefit Research Institute (EBRI). It may be used without permission, but citation of the source is required. Recommended Citation: Craig Copeland, Individual Retirement Account Balances, Contributions, and Rollovers, 2012; With Longitudinal Results : The EBRI IRA Database, EBRI Issue Brief, no. 399 (May 2014). Report availability: This report is available on the Internet at Data Security EBRI s retirement databases (the EBRI/ICI Participant-Directed Retirement Plan Database, TM the EBRI IRA Database, TM the EBRI Integrated Defined Contribution/IRA Database TM ) have been the subject of multiple independent security audits and have been certified to be fully compliant with the ISO Information Security Audit standard. Moreover, EBRI has obtained a legal opinion that the methodology used meets the privacy standards of the Gramm-Leach-Bliley Act. At no time has any non-public personal information that is personally identifiable, such as Social Security Number, been transferred to or shared with EBRI. None of the three databases allows identification of any individuals or plan sponsors. Table of Contents Introduction... 4 Data... 4 IRA Types... 4 Average IRA Balances... 5 Contributions... 7 Rollovers vs. Contributions Comparison of Account Balances and Contributions Conclusion About IRAs Endnotes Figures Figure 1, Distribution of IRA Types, by Accounts and Individuals, Figure 2, Distribution of IRA Ownership, by Various Demographic Characteristics and IRA Type, Figure 3, Average IRA Balances for All Accounts and Individuals, by IRA Type, Figure 4, Median IRA Balances for All Accounts and Individuals, by IRA Type, Figure 5, Average IRA Balances for All Accounts and Individuals, by Age, ebri.org Issue Brief May 2014 No

3 Figure 6, Median IRA Balances for All Accounts and Individuals, by Age, Figure 7, Average and Median Individual IRA Balances, by IRA Type and Age, Figure 8, Average and Median IRA Balances for All Accounts and Individuals, by Gender, Figure 9, Average and Median Individual IRA Balances, by Gender and Age, Figure 10, Average and Median Individual IRA Balances, by IRA Type and Gender, Figure 11, Percentage of Those Owning a Traditional or Roth IRA Who Contributed to It and the Percentage of Those Contributing Who Contributed the Maximum Allowable Amount, by All Accounts and Individuals, Figure 12, Percentage of Traditional or Roth IRA Accounts That Received a Contribution, by Age of Account Owner, Figure 13, Percentage of Traditional or Roth IRA Accounts That Received a Contribution that Received the Maximum Allowable Contribution, by Age of Account Owner, Figure 14, Percentage of Traditional or Roth IRA Accounts That Received a Contribution, by Gender of Account Owner, Figure 15, Distribution of Those Contributing to an IRA, by IRA Type, Age, and Gender, Figure 16, Distribution of Those Contributing to a Traditional or Roth IRA, by Age and Gender, Figure 17, Distribution of Those Contributing to a Traditional or Roth IRA and Those Rolling Over to a Traditional IRA, by Age and Gender, Figure 18, Distribution of Those Rolling Over to any Traditional IRA, by Age and Gender, Figure 19, Contributions to Traditional and Roth IRAs and Rollovers to Traditional IRAs, Figure 20, Average and Median Individual IRA Balances, by IRA Type, Age, and Gender, 2010, 2011, and Figure 21, Percentage of Individuals Contributing to Their IRA and of Those Contributing Who Contributed the Maximum, Figure 22, Average Contributions to a Traditional or Roth IRA, by Age and Gender, Figure 23, Distribution and Average and Median Individual IRA Balances of a Consistent Sample of Individuals, by IRA Type, Age, and Gender, 2010, 2011, and Figure 24, Distribution of Individual IRA Balance Changes for a Consistent Sample of Individuals from 2010 to 2012, by IRA Type, Age, and Gender Figure 25, Percentage of Individuals Contributing to Their IRA and of Those Contributing Who Contributed the Maximum for a Consistent Sample of Individuals from Figure 26, Percentage of a Consistent Sample of IRA Owning Individuals from Who Contribute Various Number of Years, by IRA Type, Age, Gender, and Account Balance Figure 27, Percentage of Individuals, by IRA Type, Contributing Who Contributed the Maximum for Various Years, Depending on Number of Years Contributing, for a Consistent Sample of IRA Owners, Figure 28, Average IRA Contribution of Those Who Contribute in Each Year and For Those Who Contribute All Three Years for a Consistent Sample of Individuals from , by Age, Gender, and Account Balance Box Figure A, Sources of Estimated Total U.S. Retirement Plan Assets, ebri.org Issue Brief May 2014 No

4 Individual Retirement Account Balances, Contributions, and Rollovers, 2012; With Longitudinal Results : The EBRI IRA Database By Craig Copeland, Ph.D., Employee Benefit Research Institute Introduction Individual retirement accounts (IRAs) are a vital component of U.S. retirement savings, holding nearly one quarter of all retirement plan assets in the nation. 1 A substantial and growing portion of these IRA assets originated in other taxqualified retirement plans, such as defined benefit (pension) and 401(k) plans, and was moved to IRAs through rollovers from those plans. The Employee Benefit Research Institute (EBRI) has focused on retirement savings since its inception in 1978 and has been particularly detailed in the study of participant behavior and balances in 401(k) plans, and more recently, of IRA account holders as well, notably through the creation of, and updates to, the EBRI IRA Database. EBRI has been able not only to link the account(s) of individuals within and across participating data providers in the IRA database, but also with the account(s) of participants in defined contribution (DC) plans. This is done within a calendar year and longitudinally, permitting the examination of retirement asset holdings both at a point in time and over time as the individuals age and either change jobs or retire. This Issue Brief, the annual, cross-sectional analysis of the EBRI IRA Database, examines the distribution of IRA owners by IRA type, average and median account balances, and contributions and rollovers to IRAs. 2 Added to this crosssectional study is a longitudinal component that looks at account balance changes and contribution behavior of a consistent sample of individual IRA owners from 2010 to Data As part of the EBRI Center for Research on Retirement Income (EBRI CRI), the EBRI IRA Database is an ongoing project that collects data from IRA plan administrators across the nation. For year-end 2012, it contained information on 25.3 million accounts owned by 19.9 million unique individuals, with total assets of $2.09 trillion. 3 For each account within the database, the IRA type, account balance, contributions made and rollovers transferred during the year (if any), the asset allocation, and certain demographic characteristics of the account owner are included (among other items). Based on the richness of the data, the study presents account-level and individual-level results. IRA Types In the EBRI IRA Database, IRAs are classified into five types: (1) traditional originating from contributions (TOFC), (2) Roth, (3) Simplified Employee Pension (SEP)/Savings Incentive Match Plan for Employees (SIMPLE), (4) traditional originating from assets rolled over from other tax-qualified plans (TOFR), such as an employment-based pension or a 401(k) plan (although not all the assets in these accounts are exclusively from employment-based sources, so this should not be used as a proxy for the amount of employment-based dollars in IRAs), 4 and (5) other/unknown. 5 The distribution of the IRA accounts in this database as of year-end 2012 was: 34.8 percent TOFC IRAs percent TOFR IRAs (combined traditional IRAs, 63.4 percent) percent Roth IRAs. 6.9 percent SEPs and SIMPLEs 7.1 percent unknown (Figure 1). 6 ebri.org Issue Brief May 2014 No

5 Figure 1 Distribution of IRA Types, by Accounts and Individuals, % 72.9% 70% 63.4% 60% Accounts Individuals 50% 40% 34.8% 41.8% 34.2% 30% 28.1% 28.6% 22.7% 20% 10% 6.9% 8.5% 7.1% 6.6% 0% Traditional-Originating from Contributions* Roth Traditional-Originating from Rollovers* SEP/Simple All Traditional Unknown * Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars versus IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. Note: The percentages for individuals add up to more than 100 percent, as an individual may own more than one type of IRA. On a unique individual basis (combining the accounts owned by the same person into a single observation), 41.8 percent of those owning an IRA in the database had a TOFC IRA, 34.2 percent had a TOFR IRA (combined total of 72.9 percent), 28.1 percent had a Roth, 8.5 percent had a SEP or SIMPLE, and 6.6 percent were unknown. Among all IRA owners in the database, nearly one-half (47.2 percent) were ages (Figure 2). However, the age distribution was very different for those owning a TOFC IRA relative to other IRA types. Of those owning a TOFC IRA, 15.7 percent were under age 45, compared with 39.8 percent for those with a Roth, 25.4 percent for those with a TOFR IRA, and 29.1 percent for those with a SEP or SIMPLE. IRA owners were more likely to be male. In particular, those having a TOFR IRA or a SEP/SIMPLE IRA were much more likely to be male (57.4 percent were male for TOFR IRAs and 58.2 percent for SEP/SIMPLEs, recalculated from Figure 2 for those in the database with an identified gender). Approximately half (47.8 percent) of those owning IRAs had less than $25,000 in those accounts at year-end 2012 (Figure 2), while 37.8 percent of TOFC IRA owners were in that category (Figure 2). 7 TOFR IRA owners had the largest percentage of account balances of $100,000 or more at 34.8 percent, while TOFC IRA and SEP/SIMPLE IRA owners had the next-highest percentage (29.7 percent). For all IRAs combined, 24.3 percent of individual owners had balances of $100,000 or more. 8 Average IRA Balances The average IRA account balance in the database in 2012 was $81,660, while the average IRA individual balance (all accounts from the same person combined) was $105,001 (Figure 3). TOFR IRAs had the highest average individual balance at $134,354, ebri.org Issue Brief May 2014 No

6 Figure 2 Distribution of IRA Ownership, by Various Demographic Characteristics and IRA Type, 2012 (All accounts vs. individuals) Age Acct. Ind. Acct. Ind. Acct. Ind. Acct. Ind. Acct. Ind. Acct. Ind. Acct. Ind. Under % 1.7% 0.5% 0.5% 3.1% 3.2% 0.3% 0.3% 0.8% 0.8% 0.4% 0.4% 4.9% 6.3% or older Unknown Gender Female Male Unknown Account Balance Less than $5, $5,000 $9, $10,000 $24, $25,000 $49, $50,000 $99, $100,000 $149, $150,000 $249, $250,000 or more Acct.=Accounts; Ind.=Individuals. All Traditional-Conts.* Roth Traditional-Rlvr* SEP/SIMPLE All Traditional * Traditional-Conts.=Traditional Originating from Contributions; and Traditional-Rlvr=Traditional-Originating from Rollovers. Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars vs. IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. Unknown ebri.org Issue Brief May 2014 No

7 while Roth IRAs had the lowest at $31, The median account IRA balance was $21,191, while the median individual IRA balance was $27,987 (Figure 4). The median TOFR IRA balance was higher than the balances of the other plan types. The overall average and median individual balances were 29 percent and 32 percent higher, respectively, than the overall average and median account balances. The individual balances exceeded the account balances by smaller percentages for each of the plan types, ranging from 3 percent for the SEP/SIMPLE average balances to 9 percent for TOFC IRAs. This suggests that individuals with more than one IRA typically had more than one type of IRA and that not taking into account all IRA holdings would overlook 29 percent of the average individual s cumulative IRA assets. The average individual IRA balance increased with age for owners ages 25 or older (Figure 5). This balance increased from $11,009 for those ages to $192,961 for those ages 70 or older. The median individual balance increased with the owner s age before declining for those ages 70 or older. The median individual balance increased from $3,360 for those under age 25 to $66,852 for those ages (Figure 6). Across each plan type, the average and median individual IRA balances increased with the owner s age at least above age 25 (Figure 7). However, for the TOFC IRAs and unknown category IRAs, the average balance declined for those ages 70 or older, while the median balance declined for TOFC IRAs and all traditional IRAs for those ages 70 or older. For individuals ages 35 and older, the average and median balances for TOFR IRAs were higher than for each of the other plan types, particularly once the account owners reached ages 40 and older. 10 Males had higher individual average and median balances than females: $139,467 and $36,949 for males, respectively, vs. $81,700 and $25,969 for females (Figure 8). Across all ages, males had both higher individual average and median balances than females (Figure 9). The median balance for males reached $102,097 for those ages 70 or older, compared with $56,371 for females of that age. Males had larger average and median balances across each of the plan types as well, with the largest differences being among those with a TOFR IRA, at $170,613 average and $54,212 median for males vs. $94,440 average and $29,807 median for females (Figure 10). For Roth IRAs, average and median individual balances were much closer: $38,647 and $14,353 for males vs. $27,882 and $13,256, respectively, for females. Contributions 11 Focusing only on those owning Roth or traditional IRAs, 10.9 percent of the accounts were contributed to, and 13.1 percent of the individuals owning these IRA types contributed to them in 2012 (Figure 11). Among traditional IRA owners, 6.6 percent contributed, while 25.1 percent of those owning a Roth IRA contributed to it. Of those individuals contributing, 53.5 percent contributed the maximum amount. Almost 6 in 10 (57.9 percent) of those contributing to a 12, 13, 14 traditional IRA contributed the maximum, while 49.0 percent did so with a Roth IRA. Looking at the age of the owners of the IRAs, younger Roth IRA owners were more likely to contribute to the Roth IRA than were older Roth IRA owners (Figure 12): 43 percent of Roth owners ages contributed to their Roth in 2012, compared with 21 percent of Roth owners ages However, the percentage of traditional IRA accounts being contributed to did not significantly vary by the owner s age, as the percentage ranged from 12 percent to 7 percent for account owners up through age 64. In contrast, the percentage of those accounts receiving the maximum contribution to both Roth and traditional IRAs generally increased with the age of the account owner (Figure 13). For Roth accounts, after a drop among owners ages in the percentage of those contributing the maximum amount, the rate increased for owners ages In particular, 44 percent of Roth accounts that received a contribution and were owned by someone ages received the maximum contribution. This number fell to 40 percent among those owned by someone ages 35 39, before increasing to 69 percent of those owned by year-olds. For traditional accounts, the percentage of accounts receiving the maximum contribution increased from 36 percent among those owned by someone ages to ebri.org Issue Brief May 2014 No

8 $140,000 Figure 3 Average IRA Balances for All Accounts and Individuals, by IRA Type, 2012 $134,354 $120,000 Accounts Individuals $126,391 $118,645 $105,001 $106,052 $100,000 $97,286 $89,367 $80,000 $81,660 $60,000 $67,457 $65,417 $60,212 $44,285 $40,000 $30,041 $31,288 $20,000 $0 All Traditional-Originating from Contributions* Roth Traditional-Originating from Rollovers* SEP/Simple All Traditional Unknown IRA Type * Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars vs. IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. $45,000 Figure 4 Median IRA Balances for All Accounts and Individuals, by IRA Type, 2012 $40,000 Accounts Individuals $38,072 $39,172 $35,803 $35,000 $30,000 $27,987 $32,161 $30,346 $33,092 $25,000 $20,000 $21,191 $17,794 $17,570 $15,000 $12,315 $12,796 $10,000 $5,000 $6,443 $5,119 $0 All Traditional-Originating from Contributions* Roth Traditional-Originating from Rollovers* SEP/Simple All Traditional Unknown IRA Type * Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars vs. IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. ebri.org Issue Brief May 2014 No

9 $225,000 $200,000 Figure 5 Average IRA Balances for All Accounts and Individuals, by Age, 2012 $191,208 $192,961 $175,000 $150,000 Accounts Individuals $147,739 $139,597 $152,041 $160,233 $125,000 $108,074 $108,292 $116,795 $100,000 $75,000 $50,000 $25,000 $10,464 $11,165 $10,002 $11,009 $15,258 $17,704 $23,879 $29,202 $33,967 $42,826 $46,513 $59,471 $61,871 $80,525 $80,930 $0 Under or older Unknown Age $80,000 Figure 6 Median IRA Balances for All Accounts and Individuals, by Age, 2012 $70,000 $66,852 $65,419 $60,000 Accounts Individuals $50,000 $49,899 $44,315 $49,847 $43,666 $40,000 $30,000 $20,000 $10,000 $3,194 $3,360 $4,482 $4,721 $6,217 $7,036 $9,390 $11,003 $12,230 $15,770 $16,021 $21,463 $20,320 $28,056 $25,636 $36,363 $33,782 $30,753 $0 Under or older Unknown Age ebri.org Issue Brief May 2014 No

10 64 percent among those owned by year-olds. However, there was a decline for those accounts owned by those ages (the first group with the higher maximum contribution) before a continuation of the upward trend. The likelihood of contributing to an IRA did not significantly differ by gender within the database, as both Roth and traditional IRAs owned by either males or females (as well as those without a gender identified in the database) had similar probabilities of receiving contributions. For example, 10.9 percent of IRAs owned by females were contributed to in 2012, as were 10.8 percent of accounts owned by males (Figure 14). 15 In 2012 almost 2.4 million IRA accounts in the database received a contribution (Figure 15), and the average amount contributed was $3,904. More than one quarter (26.0 percent) of the accounts receiving contributions were owned by individuals in their 50s. Additionally, more accounts owned by males were contributed to than those owned by females, since there are more male account owners (this is different from the likelihood of contributing as described above). The average contribution was highest for accounts owned by those ages In fact, there was a steady increase in the average contribution up through that age, with a noticeable jump when the owners reached age 50, where the maximum allowable contribution is higher under the catch-up contribution rules. Accounts owned by males received slightly higher average contributions ($4,023) than did those owned by females ($3,995). More contributions were made to Roth accounts than to traditional accounts (both types of traditional accounts combined into one category) (Figure 15). However, at $4,120, the average contribution to a traditional account was higher than the $3,756 to a Roth account. Yet, a higher overall aggregate amount was contributed to Roth IRAs ($5.26 billion for Roths compared with $3.97 billion for traditional accounts) due to the fact that there were substantially more Roth owners making any level of contribution. Roth IRAs had a higher percentage of younger individuals contribute to them, as 23.5 percent of the Roth accounts receiving contributions were owned by individuals ages In contrast, only 7.7 percent of the traditional accounts that received contributions were owned by those ages Furthermore, both Roth and traditional accounts receiving contributions were more frequently owned by males, and the average contribution was higher for male account owners to both of the IRA types. Even when accounting for age, the average contributions to accounts owned by males were larger than those owned by females, with the exception of owners 70 or older (Figure 16), although at some ages the average contributions were nearly equal. Yet, the age distributions for the male- and female-owned accounts that received contributions were very similar. For example, 28.4 percent of the female-owned accounts that received a contribution were owned by those in their 50s, compared with 26.0 percent for the male-owned accounts. Rollovers vs. Contributions Rollovers to IRAs overwhelmingly outweighed new IRA contributions in dollar terms. While almost 2.4 million accounts received contributions, compared with the 1.3 million accounts that received rollovers in 2012, 10 times the amount of dollars were added to IRAs through rollovers than from contributions (Figure 17). This is not surprising, considering the annual contribution limit of $5,000 ($6,000 for those ages 50 or older) for IRAs in 2012, relative to the theoretically unlimited amount that could be added via a rollover. The average and median rollover amounts were $71,447 and $15,580, respectively, compared with the average contribution of $3,904. Average and median rollover amounts increased with the owner s age through ages 69 and 64 respectively, after which they began to decrease. Furthermore, the average and median rollover amounts from male-owned accounts were higher than those from female-owned accounts: $83,735 and $15,683, respectively for males, compared with $52,867 and $9,726 for females. Controlling for age, the average and median rollover amounts were still higher among male account owners than among female account owners for those ages 25 and older (Figure 18). The age distribution of those making a rollover was very similar between males and females. ebri.org Issue Brief May 2014 No

11 Average Median Average Median Average Median Average Median Average Median Average Median All $97,286 $32,161 $31,288 $12,796 $134,354 $39,172 $67,457 $17,794 $118,645 $35,803 $60,212 $6,443 Age Figure 7 Average and Median Individual IRA Balances, by IRA Type and Age, 2012 Traditional-Conts.* Roth Traditional-Rlvr* SEP/SIMPLE All Traditional Unknown Under 25 36,067 4,917 8,973 4,964 6,349 1,617 6,216 1,558 26,120 2,909 4,423 1, ,423 3,726 12,427 7,395 7,624 2,726 7,517 2,910 9,282 3,118 3,189 1, ,833 5,226 16,474 10,032 16,589 4,928 14,141 4,951 16,467 5,102 5,338 1, ,372 9,070 19,339 11,044 32,088 11,287 25,021 7,686 29,937 10,255 8,082 2, ,793 13,372 21,883 11,563 48,713 18,722 38,495 11,960 44,836 16,317 12,669 2, ,340 18,570 25,093 12,362 70,232 28,194 51,995 16,872 62,795 23,541 18,684 3, ,832 25,243 29,298 13,459 98,073 38,933 67,470 22,125 85,020 31,546 27,928 4, ,228 33,321 33,928 15, ,328 52,558 86,025 28, ,049 41,070 42,203 6, ,076 44,214 40,322 17, ,109 75, ,133 34, ,481 56,316 68,226 9, ,002 57,609 51,747 20, , , ,040 39, ,563 74, ,530 14, or older 148,701 57,517 79,447 23, , , ,993 42, ,809 73, ,453 19,402 Unknown 116,277 51,506 46,675 12, , ,623 46,375 12, ,163 53, ,262 22,240 * Traditional-Conts.=Traditional-Originating from Contributions; and Traditional-Rlvr=Traditional-Originating from Rollovers. Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars vs. IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. $160,000 Figure 8 Average and Median IRA Balances for All Accounts and Individuals, by Gender, 2012 $140,000 $139,467 Female Male Unknown $120,000 $107,630 $100,000 $80,000 $81,700 $85,230 $60,000 $65,151 $64,700 $40,000 $36,949 $20,000 $20,535 $28,082 $15,003 $25,969 $21,612 $0 Accounts Individuals Accounts Individuals Average Median ebri.org Issue Brief May 2014 No

12 Female Male Unknow n Average Median Average Median Average Median All $81,700 $25,969 $139,467 $36,949 $85,230 $21,612 Age Under 25 12,700 4,447 13,841 4,593 8,569 2, ,964 4,900 12,828 5,384 8,772 3, ,971 7,376 20,416 8,423 14,967 5, ,931 11,174 34,103 12,829 25,270 9, ,010 15,457 51,570 18,921 36,171 12, ,289 20,619 73,092 26,862 47,967 16, ,824 26, ,275 36,778 62,516 20, ,084 33, ,467 49,048 82,828 26, ,433 43, ,151 69, ,374 36, ,313 55, ,932 97, ,898 50, or older 142,013 56, , , ,022 47,153 Unknow n 84,068 15, ,191 17, ,585 44,405 Figure 9 Average and Median Individual IRA Balances, by Gender and Age, 2012 The distribution of contributions was concentrated near the maximum amount, with 55.2 percent of those contributing $5,000 $6,000 (Figure 19). 16 On the other hand, 37.5 percent of the rollovers were less than $5,000, and more than half (57.1 percent) were less than $25,000, such that the majority of rollovers were on the lower end of the rolloverdistribution amounts. However, 19.2 percent of the rollovers were $100,000 or more, meaning that while a large number of rollovers were relatively small, almost one-fifth of those rolling over qualified plan distributions to an IRA in 2012 did so with a six-figure balance or more. 17 Comparison of Account Balances and Contributions While each year s database is a unique snapshot (cross section) of that year s IRA balances and contributions, it is informative to compare the results between years to consider changes in account balances and contribution trends. The first comparison conducted is by examining each year s snapshot. The second comparison focuses only on those individuals who have an account in the database in each year of the analysis ( ). This not only allows the analysis to focus on the activity within these accounts over an extended period of time, but also controls for changes in the aggregate and averages resulting from the additions and subtractions from the database because of new data providers into the database, as well as accounts being opened or closed. Furthermore, the distribution of the growth in the balances across each account holder in the study can be deduced, as well as the persistence of contributions by these individuals. 18 Snapshot Comparison The average balance for each year s full sample decreased from $91,864 in 2010 to $87,668 in 2011, before jumping to $105,001 in 2012 an increase of 14.3 percent from 2010 to 2012, and 19.8 percent from 2011 to 2012 (Figure 20). The median followed the same pattern, dropping from $25,296 to $23,785 and then rebounding to $27,987, representing increases of 10.6 percent between 2010 and 2012 and 17.7 percent between 2011 and This same down, then-up pattern in average balances occurred for each gender and among traditional IRAs. However, the average balance continued up in both 2011 and 2012 for those accounts owned by 30- to 49-yearolds. Above that age, the pattern of a decrease in average balance in 2011 and an increase in average balance in 2012 is seen. The average balance for Roths and SEP/SIMPLEs increased in both 2011 and The percentage of individuals who contributed to their IRA in each year was consistent across years at 12.1 percent in 2010, 13.2 percent in 2011, and 13.1 percent in 2012 (Figure 21). The percentage of individuals owning traditional IRAs that contributed to them rose from 5.2 percent in 2010 to 6.6 percent in In contrast, Roth owners had the highest percentage contributing in percent, compared with 24.0 percent in 2010 and 25.1 percent in ebri.org Issue Brief May 2014 No

13 Figure 10 Average and Median Individual IRA Balances, by IRA Type and Gender, 2012 $200,000 $180,000 $160,000 $154,107 $170,613 Roth SEP/SIMPLE $140,000 $120,000 $100,000 $80,000 $60,000 $40,000 $20,000 $27,882 $48,061 $89,560 $77,588 $94,440 $37,645 $38,647 $84,143 $116,389 $88,067 $23,758 $54,247 $100,909 $96,437 $107,285 $60,130 $13,256 $13,258 $29,245 All Traditional Traditional-Conts.* Traditional-Rlvr* Unknown $27,291 $29,807 $8,187 $14,353 $23,617 $45,155 $34,303 $54,212 $10,846 $10,363 $14,025 $34,126 $35,148 $26,210 $6,383 $0 Female Male Unknown Female Male Unknown Average Median * Traditional-Conts.=Traditional-Originating from Contributions; and Traditional-Rlvr=Traditional--Originating from Rollovers. Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars vs. IRA-only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. 70% Figure 11 Percentage of Those Owning a Traditional* or Roth IRA Who Contributed to It and the Percentage of Those Contributing Who Contributed the Maximum Allowable Amount, by All Accounts and Individuals, % 50% All Traditional* Roth 51.9% 57.2% 57.9% 53.5% 48.3% 49.0% 40% 30% 24.4% 25.1% 20% 10% 10.9% 13.1% 6.0% 6.6% 0% Accounts Individuals Accounts Individuals Contributing Of those Contributing--Contributing the Maximum * Traditional IRAs in this figure include all traditional IRAs. ebri.org Issue Brief May 2014 No

14 50% Figure 12 Percentage of Traditional* or Roth IRA Accounts That Received a Contribution, by Age of Account Owner, % 46% 43% All Roth Traditional* 40% 35% 37% 35% 30% 28% 27% 25% 20% 20% 23% 22% 23% 23% 21% 23% 15% 10% 5% 0% 15% 15% 12% 13% 12% 11% 11% 9% 10% 9% 8% 8% 8% 7% 7% 7% 7% 7% 7% 5% 4% 2% 1% Under or older Unknown * Traditional IRAs in this figure include all Traditional IRAs. 80% Figure 13 Percentage of Traditional* or Roth IRA Accounts That Received a Contribution that Received the Maximum Allowable Contribution, by Age of Account Owner, % 60% 50% 40% 36% 36% 38% All Roth Traditional* 43% 44% 36% 42% 41% 44% 44% 40% 53% 48% 42% 56% 52% 47% 59% 52% 49% 55% 60% 59% 61% 64% 66% 63% 66% 69% 64% 66% 66% 66% 30% 20% 10% 0% Under or older * Traditional IRAs in this figure include all Traditional IRAs. ebri.org Issue Brief May 2014 No

15 Figure 14 Percentage of Traditional* or Roth IRA Accounts That Received a Contribution, by Gender of Account Owner, % 26.4% 25% Female Male Unknown 23.2% 24.3% 20% 15% 10.9% 10.8% 10.9% 10% 6.3% 6.0% 5.7% 5% 0% All Traditional* Roth * Traditional IRAs in this figure include all Traditional IRAs. While the percentage of individuals contributing remained consistent across the three years, the percentage of contributors who contributed the maximum rose from 43.5 percent in 2010 to 53.5 percent in 2012 (Figure 21). Increases occurred for each IRA type, with owners of traditional IRAs having higher likelihoods of contributing the maximum in each year. Given that the percentage of individuals who contributed the maximum increased, it follows that the average contribution also increased. In 2010, the average contribution was $3,335, increasing to $3,723 in 2011 and to $3,904 in 2012 (Figure 22). This increase in the average contribution occurred for each known age and gender group of the contributing owners of IRAs. Furthermore, the average contribution increased with the age of the IRA owners through ages for each year, with the exception of 2011, when the increase stopped at ages 60 64, and in 2010 for those ages Consistent Sample Comparison In order to compare the experience of the same account owners longitudinally, only the individuals owning an IRA with a positive account balance in the database in each year ( ) were included to form a consistent sample of individuals. 19 Each individual s accounts were studied to determine the change in his or her IRA balances and contribution behavior during This provides a more accurate picture of account growth, rather than relying on aggregate database totals, which might include new individuals or might exclude individuals who no longer have an account. This allows for a better understanding of how accounts grow for those maintaining their IRAs, as well as the contribution activity. For this consistent sample, the overall average balance increased each year from $95,431 in 2010, to $95,547 in 2011, and to $106,205 in 2012 (Figure 23). This increase occurred across each owner age group and IRA type, except for owners ages 70 or older and for TOFR owners, who experienced a decline of their balances in 2011 preceding ebri.org Issue Brief May 2014 No

16 Figure 15 Distribution of Those Contributing to an IRA, by IRA Type, Age, and Gender, 2012 All Contributions Traditional Contributions* Roth Contributions Number Percent Average Total Number Percent Average Total Number Percent Average Total (in thousands) (in millions) (in thousands) (in millions) (in thousands) (in millions) All 2, % $3,904 $9, % $4,120 $3,971 1, % $3,756 $5,264 Age Under , , , , , , , , , , , , , , , , , , ,295 1, , , ,532 1, , , ,713 1, , , , , , or older , , , Unknown , , , Gender Female ,995 3, ,191 1, ,854 1,797 Male ,023 3, ,220 1, ,889 2,244 Unknown ,584 2, , ,414 1,223 * Traditional IRAs in this figure include all Traditional IRAs. Figure 16 Distribution of Those Contributing to a Traditional* or Roth IRA, by Age and Gender, 2012 Female Male Unkown Number Percent Average Total Number Percent Average Total Number Percent Average Total (in thousands) (in millions) (in thousands) (in millions) (in thousands) (in millions) All % $3,995 $3, % $4,023 $3, % $3,584 $2,116 Age Under , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , or older , , , Unknown , , , * Traditional IRAs in this figure include all Traditional IRAs. ebri.org Issue Brief May 2014 No

17 Figure 17 Distribution of Those Contributing to a Traditional* or Roth IRA and Those Rolling Over to a Traditional IRA, by Age and Gender, 2012 Contributions Rollovers Number Percent Average Total Number Percent Average Median Total (in thousands) (in millions) (in thousands) (in millions) All 2, % $3,904 $9,234 1, % $71,447 $15,580 $92,930 Age Under , ,059 1, , ,877 2, , ,491 4,770 1, , ,071 10,000 3, , ,565 14,010 5, , ,596 16,936 7, ,295 1, ,422 20,000 10, ,532 1, ,924 25,000 15, ,713 1, ,228 37,943 22, , ,694 35,538 15, or older , ,507 25,851 9,922 Unknow n , ,566 27, Gender Female ,995 3, ,867 9,726 21,553 Male ,023 3, ,735 15,683 47,343 Unknow n ,584 2, ,359 23,959 24,035 * Traditional IRAs in this figure include all Traditional IRAs. an increase in Female IRA owners experienced a continual increase in average balances, whereas both male and unknown-gender IRA owners saw declines in 2011, followed by increases in The median values followed a continual upward trend across all IRA owners, except for those ages 65 and older and for TOFR IRAs. The comparison of the averages and medians, while instructive, does not show the distribution of the changes in the individuals IRA balances, an important consideration, as different individuals could experience significantly different changes between years, particularly in view of the varying levels of contributions to and withdrawals from the accounts as well as the asset allocation within the accounts. Using the experiences of these same individuals in the consistent sample, the 25 th percentile, median, and 75 th percentile of the resulting percentage changes of these individuals balances are presented in Figure 24. The median percentage change in the account balances for all individuals was an increase of 10.2 percent from 2010 to This means that half of the individuals had an increase greater than that amount and the other half had a smaller increase, no change, or a decline. Furthermore, with no percentage change at the 25 th percentile, less than 25 percent of the individuals who consistently owned an IRA had a smaller balance in 2012 than they did in 2010, while the highest (fourth) quartile of growth surpassed 19.5 percent. The growth rates for Roth IRA balances were higher, both overall and for each age and gender. The median Roth IRA increase was 16.6 percent from 2010 to 2012, compared with 7.9 percent for all traditional IRAs. A major factor in these different rates of increase was that new contributions make up a larger proportion of the Roth IRA balances than they do for traditional IRAs, which magnified the impact of contributions. The significant differences in the distribution of percentage change in the balances at ages 70 or older is due to the required minimum distribution (RMD) rules that affect individuals of these ages, which require them to make withdrawals from traditional IRAs starting April 1 of the year following the calendar year in which they reach age 70-½. These rules do not apply to Roth IRAs, which likely explains the continued increases found at this age for the Roth owners. Even with the required withdrawals, half of these traditional owners had balances in 2012 that were at least as large as they were in 2010, meaning that the returns they received during those years were equal to, or larger than, the amount they may have withdrawn. 20 ebri.org Issue Brief May 2014 No

18 Figure 18 Distribution of Those Rolling Over to any Traditional* IRA, by Age and Gender, 2012 Female Male Unkown Number Percent Average Median Total Number Percent Average Median Total Number Percent Average Median Total (in thousands) (in millions) (in thousands) (in millions) (in thousands) (in millions) All % $52,867 $9,726 $21, % $83,735 $15,683 $47, % $73,359 $23,959 $24,035 Age Under ,510 1, ,108 1, ,901 1, ,424 2, ,396 2, ,968 3, ,297 3, ,613 4, ,717 7, ,385 6, ,955 9,288 1, ,597 15,345 1, ,107 8,850 1, ,059 14,241 2, ,798 20,632 1, ,087 10,641 1, ,253 17,401 3, ,409 25,384 1, ,487 12,637 2, ,967 20,920 5, ,981 29,622 2, ,773 15,141 3, ,403 26,362 7, ,422 37,982 3, ,700 24,342 4, ,009 40,178 11, ,205 54,503 5, ,731 24,183 3, ,107 34,242 8, ,604 53,246 3, or older ,170 19,304 2, ,229 22,030 5, ,653 40,613 2,395 Unknown ,042 27, ,310 5, ,601 27, * Traditional IRAs in this figure include all Traditional IRAs. Figure 19 Contributions to Traditional* and Roth IRAs and Rollovers to Traditional* IRAs, 2012 Number Percent (in thousands) Contributions All 2, % Less than $1, $1,000 $1, $2,000 $3, $4,000 $4, $5,000 $6,000 1, Rollovers All 1, Less than $2, $2,000 $4, $5,000 $9, $10,000 $24, $25,000 $49, $50,000 $74, $75,000 $99, $100,000 $149, $150,000 $249, $250,000 or more * Traditional IRAs in this figure include all Traditional IRAs. ebri.org Issue Brief May 2014 No

19 Figure 20 Average and Median Individual IRA Balances, by IRA Type, Age, and Gender, 2010, 2011, and 2012 Average Median All $91,864 $87,668 $105,001 $25,296 $23,785 $27,987 Type Traditional-Conts.^ 88,403 78,051 97,286 29,756 24,721 32,161 Roth 24,798 25,741 31,288 11,471 11,344 12,796 Traditional-Rlvrs^ 123, , ,354 38,138 31,944 39,172 SEP/SIMPLE 55,733 56,479 67,457 15,471 15,711 17,794 All Traditional 103,346 98, ,645 32,647 28,457 35,803 Unknown 96,441 83,062 60,212 18,815 21,982 6,443 Age Under 25 21,986 11,434 11,165 5,782 3,238 3, ,290 12,278 11,009 4,769 4,488 4, ,236 18,106 17,704 7,229 6,612 7, ,683 27,664 29,202 10,819 10,072 11, ,968 38,354 42,826 14,745 13,751 15, ,998 51,006 59,471 19,329 18,312 21, ,046 66,771 80,525 24,505 23,216 28, ,196 86, ,074 31,762 29,080 36, , , ,739 42,998 38,838 49, , , ,208 58,965 50,122 66, or older 162, , ,961 56,198 49,994 65,419 Unknown 108, , ,233 35, ,475 43,666 Gender Female 71,112 66,529 81,700 23,246 21,642 27,826 Male 120, , ,467 32,752 30,704 40,103 Unknown 85,037 76,604 85,230 22,820 19,916 26,589 ^ Traditional-Conts.=Traditional--Originating from Contributions, Traditional-Rlvr=Traditional-Originating from Rollovers. Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars versus IRA-only dollars. The traditionaldollars.the traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. 70% Figure 21 Percentage of Individuals Contributing to Their IRA and of Those Contributing Who Contributed the Maximum, % 57.9% 53.5% 50% 47.2% 50.7% 48.7% 49.0% 43.5% 43.6% 40% 39.3% 30% 26.0% 24.0% 25.1% 20% 13.2% 13.1% 12.1% 10% 5.2% 6.1% 6.6% 0% All Traditional Roth All Traditional Roth Contributing Of Those Contributing Who Contributed the Maximum Source: EBRI IRA Database. ebri.org Issue Brief May 2014 No

20 The likelihood of contributing by the owners of IRAs in the consistent sample decreased from 13.7 percent in 2010 to 13.2 percent in 2011, and slightly increased to 13.3 percent in 2012 (Figure 25). For traditional IRA owners the likelihood of contributing decreased in 2011 and 2012, from 6.1 percent in 2010 to 6.0 percent and 5.9 percent, respectively. Among Roth owners, there was also a decrease in 2011 from 25.7 percent to 24.4 percent before a drop to 22.8 percent in Of those contributing in each year, the likelihood of contributing the maximum increased each year among both IRA types, reaching almost 55 percent for traditional IRA owners and 46 percent for Roth owners. Using the longitudinal nature of the consistent sample, the persistence of contributing to IRAs or the number of years each of those in the sample contributed was investigated. First, the percentage of these consistent account owners who did not contribute to their IRA in any of the years was 80.2 percent (Figure 26). The remainder broke down as follows: 7.4 percent only contributed one year, 4.6 percent contributed two years, and 7.8 percent contributed all three years. Looking at the different IRA types, considerable differences resulted in the likelihood of contributing to the IRA and in the number of years contributions were made. Among traditional IRA owners, 90.4 percent did not contribute to the IRA in any year, while 3.0 percent contributed all three years. In contrast, 65.3 percent of Roth IRA owners did not contribute in any year and 15.0 percent contributed all three years. 21 Roth IRA owners ages were the most likely to contribute in any year and all three years, at 56.1 percent and 24.3 percent, respectively. These percentages continued downward as the age of the Roth IRA owners increased, reaching 13.9 percent contributed in any year, and 4.1 percent in all three years among those ages 70 or older. There were no major differences by age below age 70 of traditional IRA owners, as 2.2 percent to 4.2 percent contributed all three years and 8.4 percent to 13.9 percent contributed in any year. Furthermore, no significant gender differences were found in the years contributed. However, by account balance, those with balances in the $25,000 $99,999 range exhibited the highest likelihood of contributing all three years, while the account balance groups just above and below these amounts had the next highest levels of contributing all three years. Those IRA owners with the lowest and highest balances were the least likely to contribute. Roth IRA owners were more likely to contribute any amount, but traditional-ira owners who do contribute are more likely to contribute the maximum allowed amount (Figure 27). Of the traditional IRA owners who contributed all three years, 44.5 percent contributed the maximum all three years. For comparison, 32.6 percent of the Roth IRA owners who contributed all three years contributed the maximum amount all three years. This same result followed for those that contributed only two years, but just barely for those only contributing one year. The higher average IRA contribution for traditional IRAs relative to Roths is likely due to more IRA contributors maxing out their contribution. The average traditional IRA contribution in 2012 for all those making a contribution was $4,127, compared with $3,766 for all Roth IRA contributors (Figure 28). The average traditional IRA contribution was also higher than the Roth average in 2010 and 2011, and the average contribution for each IRA type increased from Traditional IRA owners who contributed all three years had an average contribution larger than that of all traditional IRA owners who contributed in each year. However, for Roth IRA owners, the average contribution for those contributing all years was only higher in 2011 than for less-consistent Roth IRA contributors. However, these differences across each IRA type were small, with none being larger than $106. Among traditional IRA owners at each age group, the average contribution for those contributing all three years was higher than any who contributed in each year except for those ages in There was no consistent pattern by age for Roth owners. IRA owners identified as males had slightly larger average contributions than those identified as females for both IRA types. Furthermore, for accounts associated with a known gender, the average contribution was higher for those who contributed all years than for anyone contributing in a particular year. The larger the account balance, the higher the average contribution, and those with the larger account balances among those contributing all three years had higher ebri.org Issue Brief May 2014 No

21 average contributions. In contrast, among those with the lowest balances, the all-three-years contributors had lower average contributions than anyone who contributed. Conclusion This study provides results for the fourth year (2012) of data available from the EBRI IRA Database, with the last three years being contiguous. The results show the importance of being able to measure not only account balances, but an individual s combined account balances to determine the potential total retirement savings he or she may have in multiple accounts. Indeed, the overall, cumulative IRA average balance was 29 percent larger than the unique account balance. Therefore, databases that are not able to link accounts owned by the same individual within and across data providers will likely understate the total IRA assets, and thus the retirement accumulations held by individuals. With three years of contiguous data now available in the database, a more detailed examination of the longitudinal changes in these individuals IRAs was possible. While the overall average balance increased 11.3 percent for those individuals with an account in each year from , 25 percent (regardless of age) had no increase or a decline in their balances since On the other hand, changes among the highest 25 percent of balances exceeded 19.5 percent. Roth IRA owners experienced a much higher distribution of increases, with the lowest 25 percent of the balance changes for IRAs topping out at 7.6 percent, and the highest 25 percent exceeding 36.3 percent. An annual-snapshot percentage of those contributing to IRAs doesn t show whether the same individuals were contributing over time, or if different people contributed in different years. However, a consistent longitudinal sample allows for this examination. For example, among traditional IRA owners, approximately 6 percent contributed to the IRA annually, but over a three-year period approximately 10 percent of traditional IRA owners contributed. The percentage contributing also increased for Roth IRA owners, from approximately 25 percent in any one year to almost 35 percent over the three-year period. As the EBRI IRA Database continues to expand and mature, further examinations of the longitudinal changes will be conducted. The first such examinations included in this study were investigations of the changes in IRA balances for the same set of individuals and of the persistence of contributions among these same individuals. Now that the linking of the IRA Database with DC account data is underway, results on the combination of individuals assets in these accounts can be determined, along with the growth and movement of dollars both within and between these accounts. Ultimately, under the auspices of the EBRI Center for Research on Retirement Income, the tracking of dollars from DC plan accumulations to IRA rollovers, and eventually through decumulation, will be measured to assess whether retirees seem positioned for a remaining lifetime of financial well-being. ebri.org Issue Brief May 2014 No

22 About IRAs Individual retirement accounts (IRAs) were created by the Employee Retirement Income Security Act of 1974 (ERISA) as a way to provide workers who did not have employment-based pensions an opportunity to save for retirement on a tax-deferred basis. The Economic Recovery Tax Act of 1981 (ERTA) extended the availability of IRAs to all workers with earned income, including those with pension coverage. The Tax Reform Act of 1986 (TRA 86) restricted the tax deductibility of IRA contributions to those with incomes below certain levels and created nondeductible IRAs (where contributions are not taxdeductible but earnings still accrue tax-deferred), and partially (or wholly) deductible IRAs, depending on income. The Taxpayer Relief Act of 1997 (TRA 97) created a new type of nondeductible IRA the Roth IRA and allowed nonworking spouses to contribute to an IRA, subject to certain income restrictions. As an account type, IRAs currently hold the largest single share of U.S. retirement plan assets, largely from rollovers from other types of plans (see Box Figure A). Nonemployment-based IRAs. There are two basic types: Traditional IRAs: Anyone with earned income, as well as a nonearning spouse of an earner under certain conditions, can contribute. Contributions are tax deductible (or not) depending upon the contributor s income and participation in an employment-based retirement plan. Earnings in these IRAs accrue tax-deferred, and withdrawals after age 59-½ are taxed as ordinary income. Minimum withdrawals from a traditional IRA must commence by April 1 of the calendar year after the year the individual turns age 70-½. Roth IRAs: This type of IRA offers tax-free investing for retirement: No taxes are paid on investment returns or on withdrawals made after age 59-½, as long as the Roth IRA has been held for at least five years. Contributions to Roth IRAs are not tax-deductible, but there are no mandatory withdrawals after age 70-½ (as there are with traditional IRAs). Certain income limits restrict eligibility for contributing to a Roth IRA. The current maximum annual contribution to a traditional or Roth IRA is $5,500 for those under age 50 at the end of This limit can be split between a traditional and a Roth IRA, but the combined limit is $5,500. Those age 50 or older during 2014 can make an additional $1,000 catch-up contribution, for a combined annual limit of $6,500. The maximum contribution to a Roth IRA and the maximum deductible contribution to a traditional IRA may be reduced depending upon an individual s modified adjusted gross income. Employment-based IRAs Simplified Employee Pension (SEP) plans allow employers to make contributions on a tax-deferred basis for their employees and allow self-employed individuals to make contributions for their own retirement. Savings Incentive Match Plans for Employees (SIMPLE) plans also allow for tax-deferred, employer contributions plus allow salary-reduction contributions by the employees. The employers must make matching contributions or nonelective contributions to the plans. Traditional originating from rollovers (TOFR) IRAs or traditional originating from contributions (TOFC) IRAs: In the EBRI IRA Database, traditional IRAs are separated into two categories to highlight the amount of IRA assets that have moved from other tax-qualified plans (including defined benefit (DB) and defined contribution (DC) plans, and prior IRAs) and were subsequently rolled over to new IRAs those originating from rollovers and those originating from contributions. However, this cannot be construed as an estimate of the dollars originating in the employment-based system and transferred to the IRA system, as both types of accounts could have received rollovers or contributions subsequent to their establishment. Additionally, a rollover could have been an IRA-to-IRA rollover without any money originating in the employment-based system. This distinction is important for those interested in seeing the relative contribution of the employment-based retirement system vs. that funded solely by IRA contributions. As the longitudinal aspect of this database is developed, a more refined measure of these dollars will be established. The Internal Revenue Service reports these accounts as a single category called traditional IRAs. The tax treatment is the same for these IRAs once the dollars are in the IRA. ebri.org Issue Brief May 2014 No

23 Figure 22 Average Contributions to a Traditional* or Roth IRA, by Age and Gender, All $3,335 $3,723 $3,904 Age Under 25 2,496 2,814 2, ,754 3,095 3, ,752 3,135 3, ,794 3,209 3, ,923 3,315 3, ,078 3,466 3, ,667 4,122 4, ,970 4,347 4, ,205 4,500 4, ,319 4,471 4, or older 4,192 4,360 4,625 Unknown 3,282 4,110 3,549 Gender Female 3,453 3,755 3,995 Male 3,630 3,831 4,023 Unknown 3,096 3,431 3,584 * Traditional IRAs in this figure include all Traditional IRAs. Figure 23 Distribution and Average and Median Individual IRA Balances of a Consistent Sample* of Individuals, by IRA Type, Age, and Gender, 2010, 2011, and 2012 Average Median All 100.0% $95,431 $95,547 $106,205 $27,212 $27,454 $30,996 Type Traditional-Conts.^ ,095 83,115 92,821 26,878 27,272 30,918 Roth ,641 26,692 32,274 11,277 12,136 14,803 Traditional-Rlvrs^ , , ,591 35,916 35,306 39,013 SEP/SIMPLE ,201 60,949 69,951 16,386 18,038 21,164 All Traditional , , ,633 32,159 31,932 35,499 Age # Under ,894 11,794 14,272 4,529 5,266 6, ,451 10,720 13,607 4,934 5,537 6, ,323 15,628 19,358 6,879 7,681 9, ,052 25,292 30,476 10,525 10,979 12, ,845 36,982 43,621 14,832 15,170 17, ,650 51,536 59,846 20,232 20,451 23, ,148 70,119 80,641 26,370 26,625 30, ,387 94, ,689 33,474 33,917 38, , , ,769 45,944 46,166 51, , , ,146 62,976 62,502 68, or older , , ,325 71,928 68,573 71,116 Unknown , , ,997 70,738 69,041 69,250 Gender Female ,600 66,585 75,140 23,952 24,383 27,826 Male , , ,718 35,231 35,309 40,103 Unknown ,426 92, ,014 23,837 24,005 26,589 * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ^ Traditional-Conts.=Traditional Originating from Contributions, Traditional-Rlvr=Traditional-Originating from Rollovers. Both of these accounts could have received contributions or rollovers after their origination, so these are NOT proxies for employment-based dollars versus IRA only dollars. The traditional-originating from rollovers do provide an estimate of the dollars that have been moved into a new IRA. # The individual's age is from ebri.org Issue Brief May 2014 No

24 Figure 24 Distribution of Individual IRA Balance Changes for a Consistent Sample* of Individuals from 2010 to 2012, by IRA Type, Age, and Gender Total Traditional^ Roth Percentile Percentile Percentile 25th Median 75th 25th Median 75th 25th Median 75th All 0.0% 10.2% 19.5% 0.0% 7.9% 16.2% 7.6% 16.6% 36.3% Age # Under or older Unknown Gender Female Male Unknown * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ^ Includes all Traditional IRAs. # The individual's age is from % Figure 25 Percentage of Individuals Contributing to Their IRA and of Those Contributing Who Contributed the Maximum for a Consistent Sample* of Individuals from % % 52.7% 50.0% 54.4% 52.8% 50.7% 46.0% 45.0% 43.7% 40% 30% 25.7% 24.4% 22.8% 20% 13.7% 13.2% 13.3% 10% 6.1% 6.0% 5.9% 0% All Traditional Roth All Traditional Roth Contributing Of Those Contributing Who Contributed the Maximum * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ebri.org Issue Brief May 2014 No

25 Figure 26 Percentage of a Consistent Sample* of IRA Owning Individuals from Who Contribute Various Number of Years, by IRA Type, Age, Gender, and Account Balance Total Traditional^ Roth One Two All Three One Two All Three One Two All Three None Year Years Years None Year Years Years None Year Years Years All 80.2% 7.4% 4.6% 7.8% 90.4% 4.3% 2.4% 3.0% 65.3% 11.5% 8.2% 15.0% Age # Under or older Unknown Gender Female Male Unknown Account Balance Less than $5, $5,000 $9, $10,000 $24, $25,000 $49, $50,000 $99, $100,000 $149, $150,000 $249, $250,000 or more * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ^ Traditional includes both traditional types. # The individual's age is from % Figure 27 Percentage of Individuals, by IRA Type, Contributing Who Contributed the Maximum for Various Years, Depending on Number of Years Contributing, for a Consistent Sample* of IRA Owners, Years 2 Years 1 Year None 53.2% 53.6% 50% 44.5% 47.5% 46.1% 46.8% 46.4% 40% 37.2% 36.9% 39.7% 41.5% 32.6% 30% 20% 16.9% 18.8% 10% 11.3% 11.2% 7.0% 8.7% 0% Traditional Roth Traditional Roth Traditional Roth Of Those Contributing Three Years Who Contributed the Maximum Of Those Contributing Only Two Years Who Contributed the Maximum Of Those Contributing Only One Year Who Contributed the Maximum * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ebri.org Issue Brief May 2014 No

26 Figure 28 Average IRA Contribution of Those Who Contribute in Each Year and For Those Who Contribute All Three Years for a Consistent Sample* of Individuals from , by Age, Gender, and Account Balance Traditional^ Roth All All All All All All All Years All Years All Years All Years All Years All Years All $4,008 $4,057 $4,061 $4,167 $4,127 $4,173 $3,689 $3,653 $3,746 $3,766 $3,766 $3,725 Age # 2012 Under 25 2,618 2,622 2,649 2,821 2,725 2,837 2,608 2,750 2,831 2,970 3,054 3, ,634 2,626 2,605 2,822 2,751 2,838 3,188 3,231 3,258 3,375 3,306 3, ,950 2,890 2,984 3,065 3,082 3,084 3,218 3,224 3,241 3,329 3,235 3, ,372 3,353 3,411 3,479 3,479 3,465 3,194 3,178 3,215 3,265 3,196 3, ,563 3,589 3,603 3,675 3,656 3,669 3,227 3,207 3,262 3,291 3,263 3, ,667 3,715 3,710 3,791 3,754 3,779 3,384 3,369 3,428 3,446 3,434 3, ,007 4,086 4,229 4,339 4,404 4,453 3,837 3,811 4,045 4,056 4,170 4, ,422 4,515 4,472 4,594 4,524 4,597 4,410 4,387 4,485 4,490 4,511 4, ,568 4,664 4,602 4,732 4,650 4,715 4,730 4,706 4,783 4,785 4,777 4, ,620 4,772 4,614 4,835 4,653 4,802 4,873 4,901 4,878 4,936 4,870 4, or older 4,438 4,624 4,400 4,659 4,398 4,574 4,808 4,932 4,752 4,911 4,732 4,817 Unknown 4,758 4,750 4,707 4,912 4,572 4,805 4,463 4,543 4,612 4,686 4,441 4,496 Gender Female 3,967 4,075 3,983 4,195 4,060 4,209 3,788 3,801 3,815 3,910 3,841 3,876 Male 4,001 4,160 4,028 4,276 4,085 4,276 3,846 3,900 3,873 4,009 3,892 3,963 Unknown 4,058 3,926 4,168 4,022 4,238 4,024 3,369 3,166 3,501 3,290 3,511 3,248 Account Balance Less than $5,000 1,679 1,243 1,312 1,288 1,403 1,219 1,415 1,063 1,119 1,060 1, $5,000 $9,999 2,298 1,259 1,834 1,325 2,123 1,364 1,940 1,290 1,575 1,293 1,770 1,335 $10,000 $24,999 3,074 2,549 3,129 2,726 3,214 2,733 2,912 2,534 3,019 2,745 3,054 2,736 $25,000 $49,999 3,901 3,809 3,983 3,944 4,051 3,956 3,887 3,830 3,971 3,964 3,975 3,903 $50,000 $99,999 4,422 4,470 4,510 4,586 4,561 4,595 4,481 4,509 4,548 4,610 4,562 4,570 $100,000 $149,999 4,700 4,787 4,744 4,889 4,802 4,899 4,643 4,692 4,702 4,784 4,724 4,746 $150,000 $249,999 4,844 4,965 4,880 5,037 4,943 5,057 4,806 4,871 4,863 4,955 4,883 4,929 $250,000 or more 5,139 5,288 5,170 5,356 5,247 5,358 5,170 5,251 5,197 5,311 5,199 5,255 * The consistent sample has only the individuals with at least one account in each year ( ) of the database. ^ Traditional includes both tradtional types. # The t individual's age is from ebri.org Issue Brief May 2014 No

27 Box Figure A Sources of Estimated Total U.S. Retirement Plan Assets, 2012 (Total $23.7 trillion) Federal Government 14% Private Defined Benefit 13% State & Local Government 20% Private Defined Contribution (including 401(k)s) 17% Private Insured 12% Individual Retirement Accounts 24% Source: Board of Governors of the Federal Reserve System, "Financial Accounts of the Untied States: Flow of Funds, Balance Sheets, and Integrated Macroeconomic Accounts." Fourth Quarter Endnotes 1 See Box Figure A at the end of this study. 2 See Craig Copeland, Individual Retirement Account Balances, Contributions, and Rollovers, 2011: The EBRI IRA Database, EBRI Issue Brief, no. 386 (Employee Benefit Research Institute, May 2013) for the most recent prior year of the database. 3 Below is a comparison of the EBRI IRA Database with numbers from the Internal Revenue Service and the Federal Reserve s Financial Accounts report as referenced in the Box Figure A. EBRI Database EBRI Database Internal Revenue Flow of Funds Service 2008 Data 2012 Data Total Assets $1.46 trillion $2.09 trillion $3.7 trillion $5.6 trillion Percentage Traditional Assets 85.3% 85.6% 88.5% Average Rollover Amount $72,398 $71,447 $48,508 Average Traditional/Roth Contributions $3,723 $3,904 $3,161 The above percentage of traditional assets is adjusted for known assets. With the unknown assets included, the traditional IRA asset percentage is 82.4 percent. Based on this asset comparison, the database includes about 37 percent of the assets. The number of individuals owning IRAs in the database represents about one-third of all IRA owners, accounting for growth from the 54.5 million individuals the Internal Revenue Service reported owning an IRA in For 2007, the IRS tabs showed ebri.org Issue Brief May 2014 No

28 an average rollover amount of $72,896. See Victoria L. Bryant, Accumulation and Distribution of Individual Account Arrangements, Statistics of Income Bulletin, Spring 2012, pp for complete IRS tabs of IRAs. Also see the discussion in the About IRAs box about the differences in IRA types. 4 Traditional IRAs are broken down into categories based on how the accounts originated with the data providers either through contributions made by the account owners or through rollovers from other tax-qualified vehicles, such as traditional pension or 401(k) plans. Since both of these account categories could have received contributions or rollovers at some point after their origination, these are NOT proxies for employment-based dollars vs. IRA-only dollars. The TOFR IRAs do provide an estimate of the dollars that have been moved into new IRAs, regardless of the source. 5 Some of the IRAs could not be classified into the four stated categories by the data providers. Consequently, these accounts are classified as unknown in this database. 6 For those with a known account, 37.5 percent were TOFC IRAs, 30.7 percent TOFR IRAs (combined 68.2 percent), 24.4 percent Roth, and 7.4 percent SEP/SIMPLE. 7 This isn t surprising considering that Roth IRAs didn t exist until Furthermore, Roth IRAs have a higher percentage of younger contributors who, on average, contribute less annually. Additionally, Roth IRAs have generally not received large rollovers from other tax-qualified plans, such as 401(k) plans, that traditional rollover IRAs have received. However, this could change over time with the growing availability of a Roth option in 401(k) plans and changes in tax laws that allow for more ease in converting traditional IRAs into Roth IRAs. 8 The individual account-balance distribution could shift toward a larger percentage of individuals having balances greater than $100,000 in the future as the database expands to include other data providers. However, this will depend on the number of new individuals added relative to the number of new accounts added for individuals already in the database. 9 In 2010, the ability of individuals to convert their traditional IRAs to Roth IRAs was significantly eased. As a result, after 2010, the differences in the average balances between TOFR, TOFC, and Roth IRAs could become smaller, depending on the number of individuals taking advantage of these new Roth-conversion rules. 10 TOFR IRAs and those of unknown type, where owner age was unknown, had significantly higher average balances than those of all other ages for those IRA types. This reflects a specific segment of the market for which some of the data administrators do not have age records. The likelihood of having more than one IRA was also much higher among this group with an unknown age. 11 Contributions to SEP and SIMPLE IRAs are not considered in this section due to the differing limits in them relative to the nonemployment-based IRAs and the potential incentives to contribute to SIMPLEs through matching contributions. 12 In 2012, the maximum contribution to an IRA was $5,000 for those younger than age 50 and $6,000 for those ages 50 or older, due to the additional $1,000 catch-up contribution allowed individuals of that age. 13 In 2001, 69.9 percent of those making a contribution to a deductible, traditional IRA made the maximum contribution ($2,000), but in 2005, only 26.8 percent were found to have made the maximum contribution ($4,000 for those under age 50 and $4,500 for those ages 50 or older) to this IRA type. See Craig Copeland, Ownership of Individual Retirement Accounts and 401(k)-Type Plans, EBRI Notes, no. 5 (Employee Benefit Research Institute, May 2008): The percentage of individuals making the maximum contribution could be higher if individuals are contributing to more than one account and the other account(s) is (are) not in the database. However, within the database, of those contributing to a Roth IRA in 2012, only 3.2 percent also contributed to a traditional IRA. Of those contributing to a traditional IRA, only 4.6 percent contributed to a Roth IRA. Furthermore, 98.1 percent of those contributing to either a traditional or Roth IRA in 2012 contributed to only that IRA. Thus, virtually all of those contributing to an IRA contribute to only one account. 15 Those accounts owned by individuals with an unspecified gender in the database had a similar overall likelihood of receiving contributions than that of those with a known gender. However, Roth accounts of those with an unknown gender were slightly more likely and traditional accounts of those with an unknown gender slightly less likely to receive a contribution than were accounts owned by those with a known gender. Yet, overall the results are very similar across gender possibilities in the database, so there is no conclusive evidence that the male female split is not representative overall despite the percentage without a known gender. ebri.org Issue Brief May 2014 No

29 16 This number was larger than the percentage contributing the maximum due to some individuals ages 50 and older contributing more than $5,000 but not reaching the maximum of $6,000 for their age. 17 While only 19.2 percent of the rollovers were valued at $100,000 or more, the assets coming from these rollovers amounted to 77.7 percent of the assets rolled over in All of the values in the longitudinal section are nominal dollars. 19 This sample includes 11.0 million individuals with $1.16 trillion (2012 value). Compare Figures 2 and 23 to see the distribution of the individual owners by age and gender. The distributions are very similar. 20 The distribution of the percentage balance changes became more negative for the oldest IRA owners. For example, those owners ages had a 25 th percentile of change at -6.6 percent and a median of 2.0 percent, owners ages had a 25 th percentile of -8.4 percent and a median of -4.2 percent, and those ages 85 or older a 25 th percentile of percent and -4.7 percent for a median. 21 In an earlier EBRI publication, the persistence of contributions was investigated. A different formulation of the persistence statistic based on the current-year contributors to see what percentage of those contributed in both of the prior years was used. It was found that 21.8 percent of those making deductible contributions to an IRA in 1998 also made them in 1996 and Furthermore, three fourths of those who contributed all three years made the maximum contribution in 1998, compared with 70.4 percent of those who made a deductible contribution in See Craig Copeland, IRA Assets and Characteristics of IRA Owners, EBRI Notes, no. 12 (Employee Benefit Research Institute, December 2002): 1 9. ebri.org Issue Brief May 2014 No

30 Where the world turns for the facts on U.S. employee benefits. Retirement and health benefits are at the heart of workers, employers, and our nation s economic security. Founded in 1978, EBRI is the most authoritative and objective source of information on these critical, complex issues. EBRI focuses solely on employee benefits research no lobbying or advocacy. EBRI stands alone in employee benefits research as an independent, nonprofit, and nonpartisan organization. It analyzes and reports research data without spin or underlying agenda. All findings, whether on financial data, options, or trends, are revealing and reliable the reason EBRI information is the gold standard for private analysts and decision makers, government policymakers, the media, and the public. EBRI explores the breadth of employee benefits and related issues. EBRI studies the world of health and retirement benefits issues such as 401(k)s, IRAs, retirement income adequacy, consumer-driven benefits, Social Security, tax treatment of both retirement and health benefits, cost management, worker and employer attitudes, policy reform proposals, and pension assets and funding. There is widespread recognition that if employee benefits data exist, EBRI knows it. EBRI delivers a steady stream of invaluable research and analysis. EBRI publications include in-depth coverage of key issues and trends; summaries of research findings and policy developments; timely factsheets on hot topics; regular updates on legislative and regulatory developments; comprehensive reference resources on benefit programs and workforce issues; and major surveys of public attitudes. EBRI meetings present and explore issues with thought leaders from all sectors. EBRI regularly provides congressional testimony, and briefs policymakers, member organizations, and the media on employer benefits. EBRI issues press releases on newsworthy developments, and is among the most widely quoted sources on employee benefits by all media. EBRI directs members and other constituencies to the information they need and undertakes new research on an ongoing basis. EBRI maintains and analyzes the most comprehensive database of 401(k)-type programs in the world. Its computer simulation analyses on Social Security reform and retirement income adequacy are unique. EBRI makes information freely available to all. EBRI assumes a public service responsibility to make its findings completely accessible at so that all decisions that relate to employee benefits, whether made in Congress or board rooms or families homes, are based on the highest quality, most dependable information. EBRI s Web site posts all research findings, publications, and news alerts. EBRI also extends its education and public service role to improving Americans financial knowledge through its award-winning public service campaign ChoosetoSave and the companion site EBRI is supported by organizations from all industries and sectors that appreciate the value of unbiased, reliable information on employee benefits. Visit for more th Street NW Suite 878 Washington, DC (202)

31 CHECK OUT EBRI S WEB SITE! EBRI s website is easy to use and packed with useful information! Look for these special features: EBRI s entire library of research publications starts at the main Web page. Click on EBRI Issue Briefs and EBRI Notes for our in-depth and nonpartisan periodicals. Visit EBRI s blog, or subscribe to the EBRIef e-letter. EBRI s reliable health and retirement surveys are just a click away through the topic boxes at the top of the page. Need a number? Check out the EBRI Databook on Employee Benefits. Instantly get notifications of the latest EBRI data, surveys, publications, and meetings and seminars by clicking on the Notify Me or RSS buttons at the top of our home page. There s lots more! Visit EBRI on-line today: