What happens to households with spending “spikes” that lack the income and cash reserves to support spending volatility?  This Issue Brief provides a unique analysis of 401(k) plan participants’ finances by linking 401(k) plan data with consumer banking data to better understand how 401(k) participants behave when faced with irregular expenses. Changes in credit card utilization, 401(k) plan contributions, and/or 401(k) plan loan use were examined after these participants experienced a significant spending spike.

Figure 7 Figure 12 Appendix Figure 3 Distribution of Annual Spending Amounts Above Income and Figure 10 Tur DAT m A inc S Figur C por r cc ik om e nin te A g dit o s c e ag unt P s a g e a R e t s o n b C )IV nd , unf ar tB e ACY he t e unde d Utili alanc q nur uit spe : e e JP nding d s la Mor spe (re Fi g zs e g at g nding ur rra a a en ion and e tnd Plan lia o C 16 tiv ha in y spik ). ese tFu e o ea ha sr inc r t (h t s a spe 4 w e om Loans r0 o m num , nding e 1 or a . ( nd W ek) b , hi e spik 71 su rle P .1 p of r35 lan g ee secu oc s e prno e ccLoans ur re rctit nt e ry c e nt ov nc of p rof e e ot tr he e (t oc = he dse ols in p 1 if th b hous yp a inc rte ic e om la hold a ipa c e spik e nt a w ob nd s w hic e serv oc c it h a ah csh ur a no rt e rions r e e d rd ser e e si v in y ha olv v ge ne d s e ing ), a d spik r 35 t2 c o r e e (e p s s of d ns e eitr e ur c ce A 25 a e nt p r d p a p of e ll d e nd e rc bix e t nt Figure 2 Data Sources How F Likeliho inanc odial of Fa Increasing ctorCredit s Out Card sid Debt e of , Decrea a 40 sing 1(k) 401(k) Plan Con Can tributio I ns, mp andact Summary Statistics Cash Reserves and Above Income Alone 8 Incidence of an Unfunded Spending Spike, by Credit Card Utilization in Year One The $150,000 income threshold is the limit set for participants to partake in the tax-preferred, in-plan emergency savings c Fi ha or hous us g d la ur t o a re e g m hold s b ee a r2 r, l aa d 28 nc ob nd at e a p ser 3 e of is rfo c v k $1 e are nt t ions p su 00 tha m ,000 confide d m w spi a e rro y ek r nt e st found m s of a ia or tl a is et50 ,ic nd tc o s, om p secu ha c eom r pv ca e er r p nt e e ha le d . or tR d w e e a it la v ath o a sona rle rgia e a nly b rst b le a lo 41 nd e d ne p e .8 p hy 22 fin msi it ont e pc ions re a crl, h w e cnt e ,e F nt le a ihe of n nd cha a trr te n hose onic p d c r a ob spik in unfun a la it w C nd e rit s of e h h su a pc r d rlt o r a e 75 e s c c d ) d e t . it d spe e p ur r e cira s a nding crt l sa e d ic nt ut s fe or ili o spik gza fu la ta t h rion re g e d es oc r S of a ( a cr Fi ur m e 80 g rus p ur e –ld 10 e e e.d 5) I 0 n tha . t Appendix Figure 2 C The redle it v cTaking e al o rd f th utili eza a a New tcion is a count 401(k) b not alahe ncre im Plan Loan is p aor lso tac nt or fa rec la tin or te d Year in wditeh t Tw eloa rmn inc ini o, ng by ide w Unf he nctehe un ar nd a ded inc 401( r e Spend k a) ses plain cr n p ing aerd tic Spike it ipa car nt d t b O aakla ccurre es nc a e s. pla A nce ns t loa hen . As (Households With Incomes of $150,000 or More) Re tirement Readiness Figure 1, Demographic Characteristics of the Sample ............................................................................................... 6 Data Definitions EBRI/ 60ICI 4 % 01(k) Plan Database — This is a participant-level database constructed from the administrative records vehicle under SECURE 2.0, and $2,500 is the maximum they can have saved in the account. On a p a r d ee r d c a it d e ion, e d nt si olla .g ne 8.2 p r d b atsi o es rc c ,om ew nt hic p of lyh is w hous ithow h e fe hold d te he S ra A ob l st m Sa ser a O xim nda u vtaum p trions u dts t a fo llow o ha r p Pd ra r ot b o thr le e bc i e tta e o m a n nd or ount T m lim ak or for iit n e g a m a c a c n in ont e N ss e hs - w p to la of P p n e le a the rn m sona L se eo rg a spi l infor e nnc ke ys sa m (Fi a vting gion. urs a e Ther 4) cc. oun Ov e a e t r rin SEC e a ll, th seve e U r a RE l sh acc ow ount n in bFi ala gur nce e 2 inc , 35 rea .7 p ses, ertche ent lik of elih the ood hous of e inc hold rea s ha sing d cno redrit e vcolv ard ing dec brte d de itc c re aa rd ses ba (la Fingcur ese a 17 t t)he . Sp end ecifi of caylly ea , rof one those , whil we it h Figure 1 Variable Median Mean More Than $7,500 $1–$1,000 >$1,000–$2,500 >$2,50 F0i– g $4,000 ure 8 >$4,000–$7,500 For the unique addition of the financial factors outside of the plan, the higher the credit card utilization, the more likely Depeof nde 40 nt1( Va kr) ia p blla e:ns Ne w at P tlhe an Le ond an iof n Y e ea arc T h ye wo ar, representing a large cross section of 401(k) plans. The database represents a By Craig Copeland, Michael Conrath, and Sharon Ca 401rs (k)on Acc ount Balance a k 2.0 e ve y ris c aont gset, e rnum ols a 52 b p nd eerr c of pe olic nt the ie of se s in p the spik hous la ecse w e w a hold hic s th a w ob o.r ser e dv ea sitg ions ned ha to de spik nsur ee s not cust o co mve err e dd a tb ay is inc sa om fe,e secu and re ca , sh and rea ser nony ves mla ous: (1 rger th )a n 9 40 Fi 22 g.5 p 1( urke) e 2, a rc ce c Fi oun nt na ha nc t b d ia a l C cla re nc ha dit er s of a cc atredr le iut st ssic ili ts of za hatn ion t$2 he of ,000 Sa 80 m , –p10 54 le 0 p ................................. 9 p ee rr cc ee nt nt . Thi incrs c earsed edit t he c ................................ ar ird cut reili dza it d tion ebt p w la he ys a n a ................................ n im spik pe or oc tant cur rrole ed. in t Thihe s ...... 6 100% Age2 43.0 43.4 This D w em ould og b re a p eh xp ic ec C teh d a sr inc ac et e the ris thre ticss h oofld th fo er S the am sp pik lees was 25 percent, so meeting that requirement for higher-income SpendingM — e d Tot ian a l sp Dee m nding ogra is p h the ic a an nnu d F al sum inanc of ialt he Fa c m to ont rshly , b y spe Unding nfund ce ad pt S urp ee dn thr dioug ng S h cr pik ed eit O ac nd cud re re bit n c ce ards, September 70 14 %, 2023 • No. 591 No Spike in Year Two Spike in Year Two Numa be ne r ofw O b psla en loa rvationn s U iss t ea dk e n a 58s ,6w 59 e ll as the higher the change in revolving credit card debt held from year one to year broad range of participants — including those who are young, old Le , sne s tw ha n to $2 the ,00ir 0 jobs, or have been with their cur 4r.e 8nt % Figure 15 B $2 e,500 fore J.P agg . rMo egr ag te ad n As ove set Ma r the y na ea gre m and ent75 (JP pMA ercM) r ent ha ece div spe es t nding he da no ta, t a cll se overle ec dt e bd y d inc atom a ise highly alone a ag bg ov re eg tahis ted t hr and esh aold ll unique d pe ar ctlin icipa ed nt to t48 ak.4 ing pe ar p ce la nt n lo for a nt, hose as lew ss it h b avaaila lab nc ilie ty s of of $2 cre0,00 dit c0 a– r$4 d b 9,9 orr99 ow a ing nd ctould o 38 .1 for p ce er c the ent p for art ic those ipantw s it to h b taa kla e nc a e ps of lan Tenv2 Figure 4.4 3 7.7 households would result in much higher spike amounts than for those with lower incomes. elect12 ronic payment transactions, Chase checks, and cash across 10 specific spending categories: apparel & services, 3 Nam Fi e g our f D ei s3, tribP ue tiro cn e nt a g e of NorH m ous al ehold s With Monthly Spending 25 Percent or More Than the Median Spending of the Prior two. Starting mortgage payments in year two is also positiv 8ely associated with an increased likelihood of taking a plan $2,000–$4,999 7.5% employ 90e %rs for manyMedian years. Contribution Rates, by Credit Card Utilization in Year One a ide mnt ong ifia tb hose le infw or itm h inc ation, ominc es lu of dle ing ss na tha mn $150 es, acc,00 ount 0 (num Figur be e r6) s, a . d For dre ty ss -one es, d pa etrecs of ent of birtth h a osnd e wSoc ith sp ial Se ending curit y not num cov be errs, is ed by loa $100 n to ,00 50% p 0 o ay Per re m xpe cent orns e. e age s notof cov Hou eredseho ba yg inc rsilds nom c2 eW or ith ca sh Mont rese hly $r5 v.e 7 s. I 2 Spend 2 n facting , the 25 $ 7 m .1 ePer 2 d 3ian cr cent ed it or More card ratio Th of an t61.4% hose the taking a Age Age 42 43 10 Log e Ld ikuc elia hto ion, od e nt e r t-a 17 inm 487e .8 nt 53, 2f7ood & beverage, health care, housing, transportation, travel, charitable contributions, and 12 Months N (Pa ot rtic Cov ipa ernt ed s b W yith Inc Inc om ome e ................................ s of $75,000 –$99,999 a................................ nd Tenure of More Than ................................ 15 Years) ........... 7 These participants were not likely to both increase credit card debt and take a new plan loan in the analysis 50.3% year, as only loa Int n. rFu oduc rthert m ion ore, the occurrence of a spending spike in year tw $5 o ,0is 0 0p –osit $9,9iv 99 ely associated with an increased like 1lih 1.2 ood % inc rem om ove e d a. lone (2) JP haMA d tM ota ha ls s p larut ge rp r tiv ha an $7,50 cy protoc 0. ols For in p hous laceehold for sit w s r ite h inc searom cheers of s. Re m sea ore rc the harn $150 s are o,000 bliga, te 8d 1 p to erus ceent the coul da dt a not 60% tbal2 $30,663 $84,407 plan loan was 63.5 p Median ercent com Spend pareding with 16.7 p of the ePr rce ior 12 Mont nt for those w hs ho Not did n Cov ot taered ke a pby lan loa Inco n. me Ho Lessw thanF 30inancial Factors Ou 1t 1.s 1%ide of a 401(k) Plan Can Impact 10% Gross Wage Income $55,305 $64,822 other. Chase 80% Data — JPMorgan Chase Bank, N.A. (Chase) serves nearly half of America’s households with a broad range of 7.9 percent of those with spikes and 2.7 percent of those without spikes did both. of taking a plan loan. The spending ratio is positively associated$ w 10it ,0 h ta 00–k $ing 14,9a 9 9loan but statistically insignificant. 8.7% The Work p 25 errob % s’ a fin bili atnc y e of s tca ak n fa ing ca e p m laan ny loa cha n inc llenge rease s ov d sh er atr he ply ir tco ar12 ee.8 p rs inc erluding cent for ir rp ea grula ticipa r ent xps w ensit eh s, 40 whic 1(kh a ) ar ce c ou som nt e bta im lae nc s q es of uite fun solely for d spending appr ov spik ede rs eov sea errc $ h a 2,50 nd 0 aw rA eit Eh th oE bC liga Te 1irt einc d not om et o and re- ic $ da 2 esh .nt 83 ify 8 re ser any ve indiv s, and idua $5 .94 39 l r 2 peeprrce esnt ent ce ould d in t not he ddo atso w a. (3) ith th JPMA eirM does Fig 30–39 ure 4, Percentage of Households With Spending 29 Sp .8% ikes of 25 Percent or More Than the Median Spending of the Prior S A pending n a Rly ats iis o o Ye f M ar aO xn im eum Like9.1% lihood Parameter Estimates 1.18 1.03 R etirement Readiness financial services including checking, savings, investments, cred$ it1 c 5a ,0r0 d0s–,$ a 1nd 9,9 9 loa 9 ns. Chase’s scale and wide reach a 6.9l% lows Consequently, the financ9ial factors outside of what is known within the plan are important considerations in 401(k) plan 11 E $2 lavre ,000 gn a e. H m –ow $4 ong ,999 w tor he fr k e om hous rs d 2.6 e ea hold l w peit rs ch cov e w nt it h a for er ing spe those tnding hew se it 22.5% h b e spik xpe ae la ns , nc highe es a es o nr d f le chow rss ed it t ha t he ca n $2,00 r yd a ut ffe ilicza 0. The t ot tion herw p aa e spe s rcceor cnt ts of ra eg laet tet he d a kir w ing it fin h loa aa nc high nia s l g inc eoa rr e lik ls ae se a lih rd eood to not incom allow e 9% alone 12 the Mont p (Fi ublic gur hsa e tN ion o 7) ot. C ov f ae ny re d inf bor y A E m Inc R aC tom ion a T1e a bnd outC a an ind sh Re iv $ser 1 idua .0v 7e 5l o s ................................ r entity. $ A2 ny .1 8d 6ata point ................................ included in any public .............. ation 7 40–49 The contribution data in the 401(k) plan database 2is 6 .i6 n %dollar amounts, not the percentage of income, and income is not 59.0% Spending R 70 at% io Year T Standard 95% Confidence Chi- wo 1.26 90% experience this spending 0.99 Income — Since all the spending data are at the household level, the income used in this study is also at the 40% 50% 41.9% for a comprehensive view 47.5% of household finances. In this analysis, $20 the ,00 0 C–ha $4se 9,9d 99 ata sample is restricted to the hous 23e .0 hold % s participant decisions to take a plan loan. of 21 ripe .2 ta a kpr ing e eracse a nt of p la a for n lo na tly hose asi n. s, p Of th w ait rth b ic eula hous ala rly nc ew e hold s of ith r s w $2 espe 0,00 ith a ct 0 tspik o –$r49 ee tir ,999 ha em ving e , nt be no p for rer e pe atvr a olv a pte ions ing ring . cr I off for e n ma dit cny aarc d c c a ount dses, ebt b in y w aor laenc kaere r s’ on s a only eb, ov 10 e sour .4 p thac e te ra cof m ent ount . By Craig Copeland, Michael Co ccranra t1 th, and Shar 2on 0.7%Carson 35.9% b av a 50–59 a sed ilab le on fo cr us mo tom st e orf d the ata 401( may k )only particip reflect ants a . g Th grus eg , a a2 tny e 4 .2 inf p % aor rtic mip aa tion. nts w (4) ho The had d un at sa ta is ble st inc ore od me on s, a s uc secu h as re fro ser m vre erd a uc nd ed c an be Parameter DF Estimate Error Limits Square Pr > ChiSq Credit Card Utilization Year One 44.60% 10.60% household level, but the use of single-customer households matches the income to the 401(k) p44.9% articipant. Figure 5 Figure 5, Spending Relative to the Median of the Prior 12 Months $5 of 0,0Sp 00– e$ nding 99,99 9— Percentage Above Specific Thr1 e5sh .3old % s in 2016 8% –2020 who use Chase as their primary banking institution, and 38.5% their A v tot erag al hous e of 3 m ehold onth s spe with nding this sp thr enoug dingh all 19.7% 19.6% 60% accdelta2 $0.000 $0.242 80.0% t si ook a gnific a pnt lan loa savings i n in y s a ean e r tw mop loy com mp ea nt re -d b aw sed ith 24.5 p retiremeernt ce nt sa v of ings p those lan, hat vy ing pica ally spik a 40 e w 1( itk h cr ) pla ed n. it Thu cards, ut som ilizaet ion 401( ofk )80 –100 a ov c60–65 e ce rtim ss 20e % ed oonly r sma unde Intercept 1 -1.7622 0.0359 -1.8326 -1.6918 2407.93 <.0001 ller b ro st nu ric se ts se if ctho urit sy e p wre oc ree p da ur rt eo s. Re f covsea 8e.re 3% rdc he comp rs a erns e a not tion p , ew rm ould itte ha d tvo e e ha xpo d a rt d te he cre d aa stea in out the si ddeo ll of ar J.P amo . Mo untr g oa f n Credit Card Utilization Year Two 7.4% 42.10% 5.90% Spending Relative to the Median of the Prior 12 Months of Spending — of Median Monthly Spending .................................................................................................................... 8 E ven with the additional factors, the variables within the plan ge $ne 100 ra ,0lly 00 follow or mor ethe results from EBRI/ICI 401(k) 2p 2la .7n % payment mechanisms (selected credit and debit card transactions, electronic payment transactions, check and cash 35.4% NMRTG2 4.6% p pe ar rt cic ent ipa (nt Fis gur could e tbal20 1 -0.9559 0.0653 -1.0840 -0.8279 214.14 <.0001 14 t)a . kIe nt a e rloa estn ingly, it from t he ap p pe la an rs t orha ad t jtus het hous their echold ontrs ibut are ions like tly o t to hetse ake p la on nsa, dw dhil itiona e otl he crr esd it could card a d ce ce bss t b c erfor ede it contrib40 ution % s but would not necessarily have had a decrease in the percentage of income contributed. The dollar amounts are C hase’s (JP Net MC I) n c sy ost m ee m — s. The This sy is st the em o c bom serp vlie eds w deit ph a ositll JP ed a MC mount Infor fr m om ation Te the Cha chnolog se datya Ris fork sa Ma lana rieg s a em nd ent w a re gq euir s te ha mte is nt ne s t Credit Card Revolving Balance Year One $1,689 $599 Percentage Above Specific Thresholds of Median Monthly Spending 13 Figure 17 7% Craig Copeland is Director of Wealth Benefits Research at the Employee Benefit Research Institute (EBRI). Michael cross-50 sec % tional studies. For account balance, the omitted category is $5,000–$9,999, so those with balances less than payments) and sources of income including wage income, Social Security, annuity, pensions, etc. can be linked to the 30% spndrat2 1.07 1.23 Gross Income tbal50 1 -0.2072 0.0369 -0.2795 -0.1349 31.56 <.0001 Fi t out aa ls g ko ur ing si ne d ee t t6, he of of D a a p is ny la p trn loa la ibut re nc o ha 34.1% ion o n, r ra us ace s te f A a som riz p nnu p ar tie oxim o a ns c l Sp om fro ae b te nding ina m ly the t55 ion o – A p56 rior mf a ount pe y ll t e rc a hr s A e r nt e db e ue ov of t o to e those c Iov dnc ise c om r rimi w unu e it na h cr asu nd tion a elly d Cit a te sh high c sa ting r Re d e . se ut xpe As ili rv za a e ns s a re tion ess nd ult, . B of A athe b la >ov nc 0 – d eing 79 e I cnc re p ta he om esre c se s e e nt in A dlone e c inc c ois ntrib rions easution e is d a s for the m of onit any or ing taxe as a nd nd secu derd ituc y of tions da ttaa. k(e5) n o JP ut MA be M for pr eov the ide p s v aya clua hec bkle is ins de ight posit s teo d. polic ymakers, businesses, and financial Credit Card Revolving Balance Year Two $1,952 $457 Percentage Who Increased Credit Card Debt or Took a New 401(k) Plan Loan Conrath is the Chief Retirement Strategist and Head of the C Re redtitir C ea m rd e/nt Lim Iins t Right atios St — Y ra ea te r g Oy n e Team for J.P. Morgan $5,000 are less likely (negative coefficient) to ta Ak bo e va e I loa ncom n ethan th Abovose e Inc om we it an h d Ca bala snc h Res es of erves$5,000–$9,999, while those with EBRI/ICI Database. For more infor 29.6% mation about Chase, visit the following website: tbal200 1 0.1166 0.0285 0.0607 0.1724 16.71 <.0001 shck2 35.1% $20,000–$29,999 6.1% t k ahe re ey ir no c om ctr e ap d ll one it (the H couseho ant rre d s of d ult ep b lds a o t, f rta w ic W n hil ipa it ah I e cnt ti only o nc s’ n o fin bm y aa e the sm nc s L ia a e pl w ll inc a ss rticip e Tha llbe rea a n $150 nt ing. si.ng pe ,000 rcent ) ................................ age took a new plan loa ................................ n with that level cre....................... dit card 8 Sourc pe r:ofe Esss ti6% m iona atesls f, rob m ut t hte he Ese BRins I/ICight I 401 s c (k)a P nno lan tD caom tabe a sa et atn he d se expe lect ns Chea s of e d ca onsu ta. Fo mre m r o prr eiv in afc oy rm . a W tie on t,a sk ee e e th ve e r Dya tp ar e Sc oa uut rce ion to e s box inns thur e te e xt. When a Spending Spike Occurred, by Account Balance 15 40 % % 0% 35.7% Ass 30e %t Management. Sharon Carson is a Retirement Strategist on the J.P. Morgan Asset Management Retirement highe r balances than that are more likely (positive coefficient) to take a loa 13.7% n. However, for balances of $100,000 or https://www.cha tbal500 1 0.1574 0.0273 0.1040 0.2108 33.34 <.0001 se.com/digital/resources/about-chase. $30,000Gros –$49,9 s9 I 9ncome — This is an estimate ba2 sed 8.2 % on net income with the addition of estimated federal income and 90% ut ilization. Source: Estimates from the EBRI/ICI 401(k) Plan the confide 25% or nce and security of our account holders’ private information. 12 25.0% 14.9% 5.0% In a different specification using a simple ordinary least squares ( >OL 0% S– )1 re 9% gression of the same variables, a spike occ 1urre 3.8% nce Figur Ie ns 7, ight Dis s St tribut rate ion o gy te f A am nnu . Thi al Sp s Iss ending ue Br ie Af mw ount as w s A ritb te ov n w e Iitnc h a om sse is a ta nd nc e C a fr sh om Re the ser I vns es a titut nd e’A s r bov esea e Irnc ch a omnd e A elone ditor ial more, the likelih tbal1000 1 0.1533 0.0305 0.0936 0.2130 25.30 <.0001 ood of taking a loan is not different (insignificant coefficient) from those with balances of $5,000– More Thi $5 s st 0,0udy a 00–$74 im ,9s t 99o provide a unique analysis of 4031( 0.k 5) % plan participants’ finances by linking 401(k) plan data with 5%Federal Insurance Contribut Dions atab aA sc et a (nFI d C s35% e Al) e ctta C xes ha sfor e d a tthe a. F hous or me orhold e info , rp mlus atio a nny , 401(k) plan contributions for the 30% Figure 14 Increase Credit Card Debt New Plan Loan 20% and an increase in outstanding credit card debt are also strongly as2 s0 o% cia –te 39d % with increases in the likelihood of taking a 1 0 loa .5% n. Data priv (a Hc ouseho y is fulds lly prot With I ec nc ted om.e s of No p$1 ersona 50,000 lly id ore Mor ntifie a) b................................ le information is conta................................ ined within the data a....................... nd all spending 9 staffs. Any tbal5000 1 0.0275 0.0327 -0.0366 0.0916 0.71 0.4001 views expressed in this report are those of the author and should not be ascribed to the officers, $9,999. Thus, a hump-shaped line would result based on the probability of taking a loan as the account balance see the Data Sources box in the text. c onsu $75 60 ,0 m % 0e 0– r $b 9a 9nk ,99ing 9 data to better understand how1 40 6.01( %k) participants behave when faced with irregular expenses. In This resea y Per era ch p rcent . apeage r was of p rTh oduc ose ed tW hrith oug Unf h a cun olla ded bora Spend tion betw ing een Spikes the EmpW loy ho ee To Bene ok fit a Re New searc 401(k) h Institut Plan e and J.P. 20% 9.9% 7.0% 17.2% 4 40%–59% 9.1% 10% agrsinc2 1 -0.0138 0.0019 -0.0175 -0.0101 53.36 <.0001 and tsa rus ving tees, or attribut othe es a r sponsor nalyzeds o in t f EB his RI re, sea Em rp ch a loye re e k Be ep ne t c fit om Re ple sea terly c h I anony nstitm utou e-s E.du cation and Research Fund (EBRI- 13 20 4% % 54.9% increases. $100,000 or more 54.2% 19.2% 71% pa For rticula exa rmp , cha le,nge sees in Ho ld cr ee n, d it S a craarh, d ut Ste iliv za en tion Ba, ss 40 , a 1( nd k )Cra pla ig n cont Coper la ibut nd,ions “401( , aknd ) P/or lan 40 As1( sek t )All po la cn loa ation, n use Acco u ant re Ba exa lam ncine es,d a nd aft er Morgan As Lo 50% or an set Ma or Increased nagement. J.P Credit . MorgaCard n Asset Ma Debt na in geYear ment Tw is the o, bby rand Credit for theCard Utilization asset management in bus Year iness O of ne Figure 8, Median Demographic and Financial Factors, by Unfunded Spending Spike Occurrence ................................ 10 8.3% 52.6% 12.8% 60%–79% 8.4% age20 1 -0.1168 0.0292 -0.1741 -0.0595 15.97 <.0001 51.9% 9.7% More ERF), or their staffs. Neither EBRI nor E11.7% BRI-ERF lobbies or takes positions on specific policy proposals. EBRI invites Unfunded Spending Spike — A10.5% n unfun28% ded spending spike is determined to occur when a household’s monthly t Lhe oase n Ac pa tiv rtity icipa in nt 2020, s expe ” EBR rienc I Is es a ue si Bri gnif efic , a no nt . 576, spending and ICI “spik Ree se .”a rc h Perspective, vol. 28, no. 11 (November 2022). 50.3% JPMorgan Chase & Co. and its affiliates worldwide. age40 1 0.1000 0.0199 0.0610 0.1391 25.18 <.0001 New Plan Loan 8 In 0c % re– a1 se 0 Cred 0% it Card Debt 6.4% 22.5% 10 3% % A similar hump shape results for age, as those in their 20s are less likely to take48.4% a loan than those in their 30s, while 10 10% % Figur ce om 9, m Ie nc nt ide on ncte his of re asea n Ur nfu ch.nde d Spending Spike, by Spending Ratio .................................................................. 10 Tenu 50 re% With Employer 4.4% spending is at least 25 percent or more than the previous 12 months’ median monthly spending, and this spending 60% 6.5% Endnotes 14 age50 1 0.0997 0.0213 0.0579 0.1415 21.90 <.0001 Board of Governors of the Federal Reserve System, Consumer Credit G.19, 4.2% 8.7% those in their 40s and 50s are more likely to take a loan before falling back to the same level as those in their 30s for 54% A lt Lhoug ess5% th h an 40 21( ye ka)r splan loans are a source of leakag3 e0 .f6 r% om retirement savings if they are not paid back in full when a cannot 75 be % or covered by the household’s income and cash reserves (checking and savings accounts), i.e., it is not funded EBRI is not affiliated with JPMorgan Chase & Co. or any of its affiliates or subsidiaries. 1.7% 43.4% 2% 0% Figure 10, Incide age60 1 -0.0742 0.0318 -0.1365 -0.0118 5.43 0.0197 nce of an Unfunde d Spending Spike, by Credit Card Utilization in Year One ..................................... 11 Longitudinal Sample Construction Spending-to-Income Ratio https://www.federalreserve.gov/releases/g19/current/default.htm. More 1 those in their 60s. For tenure, the lik 22% elihood increases with tenure through 10–19 year 56.3% s and then levels off for those 56.1% pa 2r– t4 ic ipa yeant rs leaves their em Ap bo loy ve Ie nc r, om the e an yd Re do sp errvov eside2 flexib 055.9% .4% ility that can lead to higher Ap bo ar vt e I icnc ipa omte ion and contributions. As a Suggested Citation: Copeland, Craig, Michael Conrath, and Sharon Carson, “How Financial Factors Outside of a by immediate liquid assets. Year two is the test year of whether a spending spike occurs. 1 54.7% 0% ten2 1 -0.1041 0.0227 -0.1485 -0.0597 21.10 <.0001 For example, see Holden, Sarah, and Jack VanDerhei, “Contribution Behavior of 401(k) Plan Participants.” EBRI Issue Brief Less than 0.80 13.3% 38.1% 15 50% wit 5– h 20 o 9 40 y % earr s more years of tenure. As income increa 1se 7.5 s% , the likelihood of taking a loan decreases. Contributions are not Figure 11, Incide Incrnc ease e Cr of eda it Ca n U rd nfu Deb nde t d Spending Decreas e Sp Con ike tri, bub tiy ons Gross Income N e................................ w Plan Loan ................................ Any Action .. 11 re su Emp lt, loy pla en loa e Bene nsfit arR ee a se c aom rch m In on stitop ute tion for and Gre 40 e1( nw ka )ld p la Re ns se . aRe rch, su lt 2023 s from Re tire them E eB nt RICon /ICfid I 40 enc 1(ek )S urve Plan D y, EB ata Rb I aCha se sh rtbow oo kt he 40 1% 1(k) Plan Can Impact Retirement Readiness,” EBRI Issue Brief, no. 591 (Employee Benefit Research Institute, Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. 0% ten5 1 -0.1502 0.0251 -0.1995 -0.1010 35.78 <.0001 In this study, spending and 401(k) plan data from 2016–2020 at the household level are examined. In order to create no. 238 (October 2001). Available at https://www.ebri.org/publications/research-publications/issue- 0.80–0.94 19.6% Figure 9 41% 100% or C sir g 1ed nif 0–1 it ic 9a C y nt e aa ra rd st tU he ti li 99 0% za p tieo rn ce — nt Thi leve s l. >0 is –m 19% easured by t2he 20 0.9 – r 39% % atio of the rev40 olv –59% ing credit card60 b–a 79% lances in the la 80st –100% month of the Appendix inc (Emp ide loy nce ee of Beloa nefit ns R aend seatrc he h loa Insn a titute mount , April s a 27, cross 2023) spe . cific participant demographic factors but not in regard to the September 14, 2023). ten10 1 0.0715 0.0235 0.0255 0.1175 9.27 0.0023 0% this hous So euhold rce: Es v tim ie aw tes, fro the m th follo e EBRI/I w CI 4 ing 01st (k) e Plp as n Dat we abrae s et a ank de sn elet co t Cha mse er dg ae ta . t Fhe or m spe ore inn fod rm ing ationd , sa et ea th fr e Da om ta S C ouha rcese s boa x ind n the40 tex1( t. k) plan data from briefs/co Mo nte rent/full/contribution-behavior-of-401(k)-plan-participants-154; Munnell, Alicia H., Annika Sundén, and Catherine Incidence of an Unfunded Spendin0 g .9 Sp 5–1 ike, .04 by Spending Ratio 14.8% Figure 12, Likelihood of Increasing Credit Card Debt, Decreasing 401(k) Contributions, and Taking a New 401(k) Plan 17% 45.7% Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. 2 20 or 0% more years0 1 2 10.5% 3 4 5 6 or More year to the credit limit on those cards. A ratio of 0 percent means that the household has no revolving credit card debt, participants’ ov ten20 1 0.0742 0.0300 0.0154 0.1330 6.11 0.0134 erall finances or pot ential reasons for taking the loans. 16 E Tay BRI lor /I , 40 C “W I % 40 ha1( t De k) te Pla rmine n Da st a 401 ba(se k) to Pa rticip creataetion thea ful ndl Con dattrib a sa uti mo pns le? : ” 1 CR .0R 5– W 1.4 o9 rking Paper, no. 2000-12. Chestnut Hill, MA: 35 .4% “Field of LDre oan in amsY ? eMe ar aTw suring o, 0%b y the Unfunde Impact do Sp f Fiena nding ncial Sp We iklle b e Oc ing c ur Initia renc tie v e ................................ s on 401(k) Plan Utiliza ................................ tion,” EBRI Issue Brief..... , no .12 20–39% 80–100% 30% 70% while C o 10 p0 p yrig erh ct In ent m form eans a ti tha on t :t he This hous repe or hold t is c ha op s us yright ed e th d e bful y tl a hellow Ema pb loy le ecer eBdeitne on fit tRe hesea ir crre cd h I it ns catrit du (s). te (EBRI). You may AEECT1 1 0.0015 0.0007 0.0002 0.0029 4.88 0.0272 Source: Estimates from the EBRI/ICI 401(k) Plan Database and Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. Center for Retirement Research at Boston College, December 2000.1 .Av 50a –il 1a .9 b9 le at https://crr.bc.edu/wp- 10.2% 554 (Employee Benefit Research Institute, March 10, 2022). 18% 200% o So r urce: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. Figure 13 Conclusion The analysis presented here, which links 401(k)-plan data and banking data, builds on those results to 64.4% evaluate the selec ctop Ch ya , sp er int dat, aor . F o drow monloa re ind fo rtm his at io re np , o se rt e solely for the Data S o pu errcsona es l and noncommercial use, provided that all hard copies retain Figure 13, Med AERCT1 1 -0.0058 0.0022 -0.0101 -0.0014 6.60 0.0102 ian Amount of New 401(k) Plan Loans and Increases in Credit Card Debt, by Unfunded Spending Spike 36.2% 1) U nfunde Using thed uni Sqpe ue nd parting icipant Spik /custe os m ’e I r m idepac ntifiet r (o not n Fi penanc rsonally ial ideB nte ifihavior able infor mation) in each dataset, the More 2.00 or more content/uploads/2000/12 8% /wp_2000-12.pdf; and US General Accounting Office, “401(k) Pension Plans: Loan Provisions 6.8% Figure 11 Appendix Figure 1 1 in 3 experien Median ced at lA ea m st ou on nte of sp New ending 401(k) spike Plan Loan above the s iand r inc Increases ome and cash 21.2% in Credit reserv Card Debt es, an 14d most , of this New box inP th lan e te L xto . an — Plan loan data are only available at year end, so 20.7% any participant who has no outstanding loan im Rep va olv cting of fin cons anc ccrat1 1 0.6943 0.0201 0.6549 0.7338 1189.68 <.0001 um ial fa er c crte or ds out it has g side row ofn the at 9 p la pn e, rcsu ent ch a pes ove r yearra d ll sp uring ending the st le udy p vels a end riod d e frbom t a c20 cum 16 20 ula – .22020. %tion, on In btehe ha v 2023 ior any 30 % aOc nd ca ur ll c reop ncy er ight ................................ and other applicab ................................ le notices contained t................................ herein, and you may c ................................ ite or quote small por........... tions of 12 individuals in both sets of data are established. These 19.1% individuals with both the spending and the saving data are then The Enhahous nce Peahol rticip ds w ation ith sp but ikMa es yha Aff d elow ct Inc er o inc me om Se ecs a urity nd fo high r So eme r spe .” L nding etter R ra etpios, c ort, GAO/ reditH cE aH rd S 59.3% -ut 98ili -5 za(tOc ion to , b aend r 1997 revolv ). W ing ashi crng edto it n, Incidence of an Unfunded Spending Spike, by Gross Income Longitudinal Data Sample and Segmentation group 20 rai %sed cash using credit cby ards Unf or un th ded eir 4Spend 01(k) ing S oSpike urce: EsO tim ccurre ates from nce the EBRI/ICI 401(k) Plan Database and Figure 16 balanc 60 e% at the end of year one but has a balance at the end of year two is considered to have a new loan in year two. accdelta2 1 0.0109 0.0019 0.0072 0.0145 33.21 <.0001 24.5% ins ide the plan. 16.9% Figure 4 Retirte he m e rnt ep or C 0% onfide t provnc ide e 10 d S % tur ha vte y you (20 RC % do S)so ve , alm 30 ost rb %a ttim wo a -tnd hir 40% d ws of ith pw ro 50 or p% k ee r rc s sa itatiion. d 60 % the Air ny d us eb 70 e t% is be a y ond prob 5 80 tle % he m sc aop nde 90 ha of %lf of the w for 10 or0% e kg eoing rs sa id grouped into households using Chase’s method for determining members of a household. The unit of observation in 60% DC: c ard US baGe lanc ne era s l in b Accot oun h ye ting ar s Offic one e .a Av nd ail tw abole t ha at n house www.gahold o.gos w v/asitsh eout ts/h e spi hsk-e 98 s, -5. but pd fa.g e was not significantly different (Figure select Chase data. For more information, see the Data Sources Figure 14, Percentag 1e in of 3 e Tho xpese rien W ce itd h U at nfu leasnde t ond e s Sp pee nnding ding s hSp ocik k e as Wh nd mo o stTook of the a m N to eo w k 40 act1( ion k ) t oP rla ain Loa se casn o h r Incr 1e 4.6 ased % Credit Distribution of Account Balances, by Credit Card Utilization in Year One NMRTG2 1 0.1688 0.0332 0.1037 0.2339 25.83 <.0001 15 In addition, Per any cent parage ticipaof nt Hou who ha seho s a highe lds W r out ithst Spend anding ing loan b Spikes alance of at t25 he e Per nd of cent yea r or More two than Th the an y ha th de at the end debt is nega So tiv urce ely : Es im timp ate asc fro ting m thet he EBRI/I ir a CI 4 bili 01t(k y) Pl to ansa Datv ae ba for se a nr de st eir lee ct m Cha ent se d .ata . Thi For s st moreudy informfur ationt, he seer thv ea Dat lid aa So te us t rceshis box isent n the te im xt.ent by looking requires EBRI’s prior e12.8% xpress permission. For permissions, please contact EBRI at permissions@ebri.org. 19.5% this study is the household. The number of people in these hous boe x hold in ths m e tex aty . not truly reflect the exact household size, 8) . Specifically, the me (Pa dia rtic n g ipa ross nts inc With omInc e of ome those s of $ w 7it 5h a ,000 spik –$99e ,9 w 99 a s $55 and Te ,305 nure v of s. $6 More 4,8 Tha 22 n for 15 tY hose ears)without a spike. The Card Debt in Year Two, by Credit Card Utilization in Year One .................................................................. 13 2 20% spndrat2 1 0.0220 0.0090 0.0044 0.0396 6.02 0.0142 This study is part of a joint effort between the Employee Benefit Resea 17.3% rch Institute (EBRI) and J.P. Morgan Asset of For yea the r one Median mo sis t re als co e Spend nt consi results de ing r,e s de of e to H ha th old v e e en Pr t,a S k ior 12 Mont a era n a h, n Se te w v eloa n Ba n hs s in y s,Not aend ar Cov Cr twao ig .ered Copela by nd ,Inco “401( me k) Pand lan As Cash set AlloReserv cation, Ac es count $4,000 at the relationship between spending spikes, credit card utilization, and 401(k) loan usage. The tight link between these as the household size can only be approximated based on the number of unique individuals who have Chase accounts. No Spike in Y 15.0% ear Two Spike in Year Two median spending ratios in years one and two were nearFigure 1.0 for 6 those without a spike, compared with roughly 1.2 for 10% shck2 1 0.4640 0.0156 0.4335 0.4945 888.75 <.0001 50% Management to deliver data-driven research to better understand how the financial factors faced by 401(k) plan Balanc Rep es,o arnd t Availabili Loan Activty ity: in Thi 2020, s re Le p ” s or EB s T t ha R isI n $ a Is v 20,0 saue ila 00 b Bri lee on f, no the . $20,000 5 76, inte a rne – nd $99,999 t ICI at w Re wsw ea .e rc bh r$1 i.or P00 ers g ,00 p 0 e or Mo ctiver,e vol. 28, no. 11 (November 70% v Fiagrur iab ele 15 s su , Me ggd eist an C s tha ont t rribut etire ion meRa nt tp ela s, b nni yng Cre isd p ita C rta rof d Uti a holis lizattion in ic finaY nc ea iarl p One lanni (Png artjic our ipane nty s Wi . In fa th I cnc t, om pare tic s of ipat $7 ing 5,00 in 0 a – 12.0% As an example, if only one spouse has a Chase account, this will be considered a one-person household. This 100% Distribution of Annual Spending Amounts Above those with a spike 10.4% . The median credit card utilizations were particularly higher among those with a spike at over 40 There are th 5r0 e% e groups of dummy variables to account for the impact of the nonlinear correlation between them and taking a new plan loan: account Spending Rati64.9% o — 47.8% This is the ratio of total annual spending to annual net income. $3,508 44.5% p 2022) artic.ipa nts outside of their 401(k) plan impact their retirement prep 16arations. Thus, the aim is to provide unique fact- $99,999 and Tenure of More Than 15 Years) ........................................................................................... 14 2.6% budget webinar is associated with higher 401(k) plan contributions. Programs to help with workers’ overall finances — $3,500 10% 35% with spikes balanS hous ce pe , ae gnding e hold , and uni tent u S rob ep . ik sere va st ion n Inco ecess me itatand es the Cash defining Reserv of spe es cifiand c dat a A v bao ria ve ble Inco s. me Alone p ercent vs. 10 percent or less. Similarly, the likelihood of experiencing a spike increased with the spending ratio and b Tabl ased e ins o ight f C s o to nt he elp nt bs uil d a stronger retirement system by policymakers, plan sponsors, and plan providers. 90% 3 for example, financial wellbeing benefit (Hsouse — cholds ould b W eith Inc indisp ome enss a b Le le s. sThe Than de $c 1is 50 ion to a ,000) take a plan loan is not just For more information on the EBRI/ICI 401(k) Plan Database and the findings from the database see Holden, Sarah, Steven 60% Acco T uen nt B u 0% a rle a n — ce Thi : tba s l2is 0 — the le s nu s tm hab ne $r2 of ,00y 0,e ta br as t l50ha — t $t2 he ,00 40 0–$1( 4,k 99 ) 9p , la tbn al 2p 0a 0r — tic$ ipa 10,nt 00 0 of –$t 1he 9,9 9 hous 9, tba el5 hold 00 — ha $2s b 0,0e 00 e– n w $49it ,9h th 99, te ba irl 1c 0ur 00r — ent Spending surges can play havoc on a household’s finances and possibly lead to the need to access more funds. Thus, a Fi crg eur dite 40 c16 a %rd , D ut isili trza ibut tion ( ion o Fig f A urcecs ount 9 and Ba10 lanc 42.2% ), ew s, b hile y tC he re d lik it eC lih arood d Uti dliz ecarte ion in ased as Ye a grr oss One inc (Pom arte ic ipa incnt res ase Wd it h I (Finc gur om e e 11 s of ). Average number of spikes: 2 Introduction Less ................................ Than $2,000 $2,000–$4,999 ................................ $5,000–$9,999 $10,000–$................................ 14,999 $15,000–$19,999 $20................................ ,000–$49,999 $50,000–$99,999.......................... $100,000 or More 3 $50,0 2) d0e 0 p I –n o e $9 nd 9r ,e 9 d9 nt e 9r, on a to nde w tns bha alur 5t0 e ha 0 0t ha p — pte $ 1 ns t0 he 0 ,in t 0d 0a 0he toa r m s pa la om re n b p (o le m ut ionly t ton ed c tinc a he telude g toot rya — s l fina hous $5,0nc 0e 0ia hold –$l 9p ,9s w r9 ofile 9).he of re tthe he p Ca ha rtic seipa da nt ta . has all or the majority of Bass, and Craig Copeland, “401(k) Plan Asset Allocation, Account Balances, and Loan Activity in 2020,” EBRI Issue Brief no. $3,000 $1–$1,000 >$1,000–$2,500 >$2,500–$4,000 >$4,000–$7,500 More Than $7,500 employer. $7 5,000–$99,999 and Tenure of More Than 15 Years) ............................................................................. 14 study o 80% f irregular spending spikes can provide insight into 401(k) plan participants’ financial decisions, including taking Specifically, 401(k) plan participants who experience spending spikes are compa $2,804 red with those who do not experience 41.8% 0% 40% 100% Age:t he age ir2 0 spe — nding ages le , sfil s t th ea rn s 3 a0r,e a a ge p4p 0lie — d4 t 0o –4t9 he , a ghous e50 — ehold 50–5s t 9, o anm d a ege et6 0 the — full 60 or( m old aejror (oit m yi) tt e spe d ca ndi tegng ory c — ritae greia s . 30 These –39). filters include 576, and ICI Research Perspective, vol. 28, no. 11 (November 2022). These spending spikes have a clear impact on the likelihood of 401(k) plan participants taking a plan loan and Data Sources ......................................................................................................................................................... 4 0% a plan loan and/or increasing cred >0 it –c 19% ard debt. A spe 20nding –39% spike could 40– b 59% e a result of a 60 n –une 79%xpected expe 80–100% nse, e.g., a car t hem 50 on %their credit card utilization, 401(k) plan contributions, and 401(k) plan loan use. First, the households with Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. Given the impact of participants’ overall finances on the need for a plan loan, it appears clear that prohibiting plan loans Figure 17, Percentage Who Increased Credit Card Debt or Took 58.1% a New 401(k) Plan Loan When a Spending Spike 70% Tenure: ten2 — less than 2 years, ten2 — 2 to less than 5 years, ten10 — 10 to less than 20 years, ten20 — 20 or more years (omitted category — 5 to but are not limited to: all 12 months of spending data, households with spending more than 50 percent of their increa 90 si% ng their credit card debt in the year of the spik 7 e, but the spikes do not have a clear impact on the likelihood of 4 $2 30 ,50 %0 33.6% repair, or an expected expense su23.2% ch as a vacation. p a Srete ic ipa Lucnt ass, L woho ri, J ha acv ke V a a nD spe erhe nding i, Ke spik lly H e aa hn re, ide Je nt Oh, ifi e ad nd . Onc Livia e S ide alon nte ifine,d “Th , the e 3% imp Diff acte of renc the e: spe Wha nding t Lea d spik s toe H s ig on hec r redit card Longitudina So l Sa urcem : Esp tile ma te Csonst from th ru e c EBRI/I tionCI 4 ................................ 01(k) Plan Database a26.8% nd selec................................ t Chase data. For more informatio................................ n, see the Data Sources box in the........................... text. 4 less than 10 years). would notOc ne cc ur ess rea dr, ily by im Ac p 71.1% c roun ove tp B aa rtla icnc ipa ent ................................ s’ retirement security................................ . Without the option o................................ f taking a plan loan, p................ articipants 15 estimated gross income, and households with credit card spending outside of Chase of less than 30 percent of their decreasing 401(k) contributions. Of those with a spending spike in the analysis year (year two), 17.2 percent took a Household Demographics Retire60 m% ent Spending?” Employee Benefit Research Institute & J.P. Morgan Asset Management Research Collaboration debt, 80 40 %1(k) plan loan usage, and 401(k) plan contribution levels are then assessed. 29.2% 40% 41.7% agrin w could 2: gro see ss in kc o loa mens in6 out yearsi td we o it nhe $1 0 p,la 00n to fill spe 0s. nding gaps, and those loans may have terms more expensive than those of Data Definitions ..................................................................................................................................................... 5 30% overall spending. As noted, a monthly unfunded spending spike is defined as a spike at least 25 percent above the prev 27.8% ious 12 months’ new plan loan and 47.5 percent increased their credit card debt, compared with 6.5 percent and 34.1, respectively, of A ppendix Figure 1, Longitudinal Data Sample and Segmentation ............................................................................. 17 As shown in Figure 1, the single-customer household participants are widely distributed across ages, incomes, and availa $b 2,0 le0 0 at https://am.jpmorgan.com/us/en/asset-management/adv/insights/retirement-insights/the-3-difference-what-leads- AEEC P Tr 1o : ebi mpt lo yR eee cs oults ntributi oo nsn F in ye a act r ono e r ins $ 1 A ,0ff 00e s.cting the Probability of Taking a Plan Loan a plan loan. 50 70% % 10 20% 15.4% m H those ow edia evn sp w eit r, hout w ending he a n spe c rte ha d nding it t c ca annot r d spik ut ili b e e za in t fun t1 ion 7 ha .7 d% t e r e d sa a b c m y he e t he s 80 p ye a hous r (e Fire g chold e ur nt e, 12 ’s inc the ). lik om H eow lih e a ood end ve ra of , vta inc he ilarrb e ele a w si c ang a s no shc r re e sid ser gitnif v ce a ics in t ra dnt d e dha b iff t te d r m e ec ont nc ree ah. ses in the to 46 Household Demographics ....................................................................................................................................... 5 t to e- nur hige he s rw -re ith th tireme eirnt c-ur sarv eing nt s e-m rap te loy s/ earnd . For Va nD exa em rhe ple i, ,J a 11 ck.1 p , and er Ke cent lly a Hraehn le,ss “In tha Dat n a a g Th e e30 re, Is 26 Tru .6 p th: er c Und ente rs arta e nd aging es 40 Ho– w49 , and AERCT1: employer contributions in year one in $1,000s. 3) Ap p Onc endix e the Fig se urhous e 2, e SA hold S Out s ap re ut ide for nt P ifi rob edit , ton he Ta sam king ple ais N re ew st rP icla te n Loa d to n si ngle ................................ -customer household ................................ s — a household tha ..... t 18 In addition to a cross-sectional analysis, a probit regression can be used to see the impact of various factors on the 30% 11 $1,507 percent, while the increasing trend of taking a new plan loan goes up by 5 percentage points (19.5 percent to 24.5 Se sha praer aw te ho ly, dsepce rnding eased tcha ont t r cib annot utions b ein t fun he d eadna bly y si the s y hous ear. e hold (For’ s inc a com om pe le taelone bre a is k d dow iscus n o sed f th . e sample data related to 60% 8.3 p House eho rce ldnt s Ac artua e allgye S s 60 upp– o65 rt . Sp For end inc ing om in eR se , tire 6.1 me pe nt, rc” ent EB ha RI v Is es inc ue om Brie efs of , no.$2 531 0,00 (Emp 0–$2 loy 9,9 ee99 Be ne and fit 19 Re.2 p searc erh ce Ins nt tit ha ute ve, June ccrat1: revolving credit card debt to card credit limit ratio in year one. This $1 r4e 0 ,50 sea %0 rch found that the lack of income and cash reserves to support spending spikes is likely to result in higher Spending Spikes .................................................................................................................................................... 6 p only rob a ha bili s one ty of p ae p rson artica ipa ttant che tadk ing to the a p b laan lo nk a an. ccount This a (s)llow . Thi s s p fore a rson w ll of the ould fa catls oro s e be xa am 40 ine 1( dk ) to pb ae rt ic cont ipant rolle wd he for re a to loa see n i s 11.1% 13.3% 43.1% p conne ercent ct)ions from be ctrw ee de itn ha card v ing utilia za spe tion nding of 60 spik –79e p a end rcetnt ak ting o craecd tion to fund it card utiliza the tion o spik f80 e, – see 10 0 p Ape pr ecndix ent. Fi Ng o ur otehe 1. r) inc . reases in 20% accd24, A ep ltp a 2021) e : ndix chang fo Fi e g r inur mo ree vo re 3, lv in info Sum g crrma ed m it a tion cr ay rd St a db e ab o ttis ut frto ithe c ms ................................ e n EB d o Rf I/ ye JP arMo onrg e ta on e n Ads s oe f t y................................ e Mana ar twog e inme $1,nt 000 re s s(e pa os rc ith ive................................ c v oallla ue b o isra ation n inc.r ease in out.................. standing debt). 19 incomes of $100,000 or mor $1,172 e. Approximately one-third of the participants have tenure of less than two years and 10.5 credit card debt. This higher debt can have a long-lasting impact on retirement security, since higher credit card 50% available in their plan and are aged 65 or younger. The demographic and financial characteristics of this person are which ones have a statistically significant impact. Probit only allows for two options for the dependent variable, in this Unfunde By any 10 m %de Sp asu ernding e, househo Spikes’ ld Im mont pachl t y on spe Fina nding ncia is l B g ee ha ne vrior ally ................................ highly variable. In fa................................ ct, 90 p 20.4% ercent of the hous ......................... ehold 9 30% 18.3% the likelihood of taking a new plan loan when credit card utilization increased went up by more than 3 percentage NMRFu TGr2 the : d20 u rm m %m or ye v, artihe able re = 1 w ie f r m eo rno tgag si eg p nif ayic ma ent nts c fiha rst nge occu s in loa rred in yn use ear two a (nd no m co re rtd ga itg e c a pr ad y m dee nb ts t iin t n yehe ar oyne ea ).r after the spike occurred, as percent have 20 or more years. ut 5 ilization is correlated with lower 401(k) plan contributions and account balances, even when controlling for tenure and $1,000 Data privacy of customers and contractual relationships with re 33.2% cordkeepers have been carefully protected, and no data was t chose ase ta us king ed in t a loa hen o ana r not lysi s. taT khe ing si angle loan. -c us The tom pe ro r bhous abiliteyhol of dt a ek ns ing ure as t ne ha wt ptlhe an d loa em nog in y rap ehic ar t w fao c tis or s ass from um etd he t o ba bnk e a ing data observations were found to have had at least one month where their spending was 25 percent or larger than their 40% spndrat2: total spending to total net income ratio in year two. p the oint ps. art icipants appeared to need money immediately to cover the spikes in spending. Credit Card Utilization and 401(k) Plan Loans ......................................................................................................... 13 22.5% incom 2e 0. % Thus, the availability of emergency savings to cover spending spikes is a critical factor in preventing or stalling transferred to JPMorgan Asset Management. EBRI has no access to personally identifiable information. match the 401(k) participant. In contrast, a multi-customer household might include additional 401(k) assets that are function of specific parameters within the 401(k) plan such as the account balance (a series of dummy variables for m edian spending in the p21.9% rior 12 months and was not covered by that month’s income (Figure 3). Twenty-eight percent shck2: dummy variable =1 if a spending spike occurred during year two. As far as financial factors, 4.8 percent of the participant households have 401(k) accou 12.8% nt balances of less than $2,000 8.6% 30% 10% a cycle 10 of % increasing debt that can signific antly impact retirement readiness. Account Balances and Plan Loans .......................................................................................................................... 15 $500 not various cap 0% tb ur ae ladnc w eit thin hre tsh his old ds) at, aset the. age of the participant (a series of dummy variables for various age thresholds), gross of 6 the households had four or more months where their spending was 25 percent or larger than the median spending Thos In loo ek w ing ith a at tspe he nding potent spik iallye long not -only lasting had im ap g ar ce t aof ter c rlik edeitlih cood ard d of eb tt a, kh ing igha er ne crw e dpitla c n ar lo da u n tili and zation i incres a asiss ng occia re te dd it w caitrh d ld ow ebe t, r 10% a nd See7.5 p the Ap erc pe ent nd ix ha in veV b aa nD lae nc rhe es of i, Ja$2 ck,,000 and– Ke $4ll,y 99 H9 a, hn w,hil “In e 22 Da.7 p ta Th erecre ent Is ha Tru ve th: b aUnd lanc ee rs s of tand$1 ing 00 H,000 ow H or ous m eor hoeld (sFi Ac gur tua e ll2) y . 4.3% Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. Less Than 0.80 0.80–0.94 0.95–1.04 1.05–1.49 1.50–1.99 2.00 or More 15.1% 20% 2.2% A variable with a Pr > t of greater than 0.0100 is not considered to have a significant effect at the 99% level. income, the tenure of the participant with their current employer (a series of dummy variables for various tenure b c P of ont ut rob t he trit he ibut Re p r m ior ion r su ed lt 12 ia s on an tm easont m , Fa e ount c v hs te or n w , s A w of ithe h a ff the n cont ecn a tloa ing vn ert rollin he aan gd e P g t rof he ob for t am hr b te ili e enur d e ty ia m of n eont inc a Ta nd hs rkeing .ia nc G se om iv a 8.7% in e P n th elc a . rn Loa For ea dtit ne ecxa n aa r................................ r m dly p d a le ell of b , tfor wte he p re a ro a tic b lso ser ipa 1.0% la v nt rag s tions e ................................ w r it tha h te ha n th d nur this ee ys w ir of e re rm e g ul or for ae r ly .. 15 0.7% Support Spending in Retireme 7.0% nt,” EBRI Issue Brief, no. 531 (Employee Benefit Research In15.6% stitute, June 24, 2021) for an Over one-third (35.7 percent) of these households have no revolving credit card debt in year one, while 22.5 percent Source: Estima So te us rc f ero : Es mti m thae te EB s froR mI /th IC eI EBRI/I 401(k) CI 4 Pl 01 a(k n) D Pla an ta Dat baas be as a en ad nd sse elle ec ct t Cha Chsa es de a ta d. aFo tar . m Fo ore r m info orm rea iti nofn o, rm see a th tio e n Dat , sae So e u th rce e s D ba otxa i nSo the u te rce xt.s box in the text. 4) Since the status of many of the variables must be known before and after the study year (year two), these 0% 0% 25.6% EBRI Issue Brief is registered in the U.S. Patent and Trademark Office. ISSN: 0887 –137X/90 0887 –137X/90 $ .50+.50 threshold $0s), and employee and employer contributions. Also, this analysis allows for the unique variables outside of the high spending, significant differences in participant behaviors based on this spending not covered by income alone were t those han 15 ye without ars a a nd spe inc nding ome spik s of e$7 . T 5,00 hose 0–w $it 99 h sp ,999 ik, es ha the m d e ad m iae n de ia m np out loye stea nding contribut new ion r baa latnc e w e,a or s 9.1 p addit eiona rcent l b for ala nc those e for w ith examp 10 le % schematic of how the overlap of the Chase data with data from an EBRI database is determined. are using 80–100 percent of their credit card limit. Just over half (52.4 percent) of the households have spending ratios 1 0% (Median Balance= $184,357) 20%-39% (Median Balance=$122,051) 80%-100% (Median Balance=$79,803) Conclusion ................................ 0 1................................ 2 ................................ 3 4................................ 5 ........................... 6 or More 16 Increase in Credit Card Debt New Loan Amount Likely households include those who spend less than their income; spend consistently more than their income; experience <25% of the last 12 households must have three contiguous years in the sample to be included. Thus, each instance of a household having 0% © 2023, Employee Benefit Research Institute —Education and Research Fund. All rights reserved. 1 9.5% 401(k) plan to be tested as well, including the credit card utilization at the end of year one, the change in the revolving t not chose red found. it ha car vd ing H ut ow ili a za loa ev te ion n inc r, tof his re 0 p a dse eefin rfr ce it om nt ion , tw he ishic a p n im h d riore py cor e re a ta a r,ses nt of m t$o e 3,50 a5.0 p sur 8 ev ets. $2,80 ro ce cnt onsi for d4e trfor hose for t hose pw olic ith cr y w m ita e ho k de it ut r s a c a a rnd spik d ut pe ili la za (n Fitsponsor g ion ur eof 13 80 s a ).– The 1 s t 00 he y of 1.05 or more, while one-third have ratios of 0.95 or less. mon 7ths' median. $100,000 or More 0% $20,000–$29,999 $30,000–$49,999 $50,000–$74,999 $75,000–$99,000 In future research, the cause of the spike will be examined to see if the effects are different depending on the cause of the three contiguous years of complete data during 2016–2020 is an observation for this analysis. This results in 58,915 EndnotesSo ................................ urce: Estimates from the EBRI/ICI 4 ................................ 01(k) Plan Database and selec................................ t Chase data. For more informatio ................................ n, see the Data Sources box in th............................. e text. 19 Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. credit card balance from end of year one to end of year two, mortgage payments started in year two (=1 if a new 2 grapple with the appropriate size of emergency reserves for the working population, because what is held outside of p meerd ce iant n inc So (Fi ug rc re ur ea : Es se e ti15 min ate ). sc r fro C eor m d it th re eA c sp EBRI/I bo arond vd e I CI 4 d nc e ingly om b 01t(k e for an ) , Plt a d Re he nt hose Dat sm e ab re v aes sd w eia ait nn a h a d sec l espik c ct oun Chae st e w b daa a tas $1,5 la . Fo nc r m eo d 07 ree inc fo c rrm e om a ases tip ona , r sfr e e ed om th w e Dat it A $1 h bo a So 84 v $1 e I u,357 ,172 rc nc eom s bo e for xfor in t htt ehose hose text. w wit ith cr houe t d ait spik care d. Likely households include those who increase spending by at least 25% at least one month in any given year. spike. observations from 35,184 unique households. Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. t ut he ili za 40t1( ion k)of pla 0 p n is not k ercent to now $7n9,80 by 3 for plan sp those onsor wsit. h credit card utilization of 80–100 percent for those with the same Source: Estimates from the EBRI/ICI 401(k) Plan Database and select Chase data. For more information, see the Data Sources box in the text. e e e e e e e e e e e e e e e e e e eb b b b b b b b b b b b b b b b b b br r r r r r r r r r r r r r r r r r ri. i. i. i. i. i. i. i. i. i. i. i. i. i. i. i. i. i. i.o o o o o o o o o o o o o o o o o o or r r r r r r r r r r r r r r r r r rg g g g g g g g g g g g g g g g g g g IIIIIIIIIIIIIIIIIIIs s s s s s s s s s s s s s s s s s ss s s s s s s s s s s s s s s s s s su u u u u u u u u u u u u u u u u u ue e e e e e e e e e e e e e e e e e e B B B B B B B B B B B B B B B B B B Br r r r r r r r r r r r r r r r r r rief ief ief ief ief ief ief ief ief ief ief ief ief ief ief ief ief ief ief A re • • • • • • • • • • • • • • • • • • • s e S S S S S S S S S S S S S S S S S S Sa e e e e e e e e e e e e e e e e e e erp p p p p p p p p p p p p p p p p p p cttttttttttttttttttth e e e e e e e e e e e e e e e e e e e mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe mbe repor r r r r r r r r r r r r r r r r r rr t14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 14 fro ,,,,,,,,,,,,,,,,,,, m 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 t2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 h3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 e E • • • • • • • • • • • • • • • • • • • B N N N N N N N N N N N N N N N N N N NR o o o o o o o o o o o o o o o o o o o...................I 591 591 591 591 591 591 591 591 591 591 591 591 591 591 591 591 591 591 591 Edu cation and R esearch Fund © 2023 Employee Benefit Research Institute 13 20 15 16 11 10 17 12 14 19 18 2 6 5 4 9 8 7 3

Volume 591

Pages 13

EBRI Issue Brief

Sept 7, 2023

Craig Copeland

Michael Conrath

Sharon Carson

Retirement