The Effect of Pension Accounting on Corporate Pension Asset Allocation: A Comparative Study of UK and US

Transcription

1 The Effect of Penson Accountng on Corporate Penson Asset Allocaton: A Comparatve Study of UK and US By El Amr* London Busness School Regent s Park London NW1 4SA eamr@london.edu Yanlng Guan School of Busness The Unversy of Hong Kong ylguan@busness.hku.hk Denns Oswald Ross School of Busness Unversy of Mchgan Ann Arbor, MI dennso@umch.edu December 2007 * Correspondng author. We thank Katherne Schpper (Edor), Anonymous Revewer, Davd Blake, Elroy Dmson, John Pckles, Laurens Swnkels, Rchard Taffler and semnar partcpants at Tel Avv Unversy and Unversy of Warwck for many useful comments. We also thank London Busness School's research fundng for ths project. Address correspondence to El Amr, London Busness School, eamr@london.edu

2 The Effect of Penson Accountng on Corporate Penson Asset Allocaton: A Comparatve Study of UK and US Abstract We nvestgate whether new penson dsclosures and subsequent full penson recognon under FRS 17 and IAS 19 had any mpact on penson asset allocaton of UK companes. We also compare penson asset allocaton of UK companes to that of US companes pror to and durng the adopton of SFAS 158. Both FRS 17 and IAS 19 requre penson assets and lables to be valued by reference to market condons and the total surplus/defc n the penson scheme to be recognzed on the balance sheet. Addonally, perodcal actuaral gans/losses are requred to be recognzed mmedately n comprehensve ncome. Therefore, these standards ntroduce a large element of volatly nto company balance sheets and comprehensve ncome. The requrements n FRS 17 and IAS 19 are smlar to those of SFAS 158, whch replaced SFAS 87 n December We dentfy a Dsclosure perod as the perod n whch UK companes had to dsclose all the requred data under FRS 17 n the notes to the fnancal statements whout formally recognzng the full penson surplus/defc on the balance sheet. We also dentfy a Full Recognon perod startng one year pror to adopton untl one year subsequent to the formal adopton of eher FRS 17 or IAS 19. We hypothesze that there exsts a shft of penson assets from equy to debt secures by UK sponsorng companes durng the Dsclosure perod of FRS 17 due to the hgher vsbly of pensons n the UK and the antcpaton of full recognon. We also predct a declne n penson funds allocated to equy secures durng the Full Recognon perod, around the adopton of FRS 17 and IAS 19. Smlarly, we predct a declne n penson assets allocated to equy secures durng the adopton of SFAS 158. We fnd that UK companes modfed ther penson asset allocaton polces by shftng assets from equy to debt secures durng both the Dsclosure and the Full Recognon perods. We also fnd that pror to the adopton of SFAS 158, US companes mantaned a stable allocaton to eques and bonds. However, there s a shft from equy to debt secures durng the SFAS 158 Full Recognon perod. Fnally, we fnd that UK and US frms wh relatvely larger penson schemes and larger magnudes of actuaral gans/losses shft more penson assets from equy to debt secures. Keywords: Defned Benef Penson Plans, Penson Asset Allocaton, Penson Surplus/ Defc, Actuaral Gans/Losses, FRS 17, IAS 19, SFAS

3 1. Introducton The purpose of ths study s to nvestgate the effect of dsclosure and subsequent full recognon of penson surplus/defc on the allocaton of penson funds to equy and debt secures. Corporate sponsored penson funds play a sgnfcant role n capal markets due to the large sze of ther nvestment portfolos. Research by Internatonal Fnancal Servces London shows that corporate penson fund assets n the UK amounted to 896 bllon as of December 2005, representng more than one quarter of the 3,450 bllon assets managed by all types of UK funds (Seb, 2006). Thus, allocaton of assets by corporate-sponsored penson funds has receved consderable attenton n both the academc lerature and the popular busness press due to potentally large economc transactons that are nvolved. Ths study focuses on defned benef penson plans. Durng the last two decades the number of defned benef plans has been decreasng whle at the same tme the number of defned contrbuton plans has been ncreasng. Stll, 47% of penson funds n the Uned States (about $1.7 trllon) are lnked to defned benef plans. Also, penson asset allocaton by defned benef penson plans has a sgnfcant mpact on the fnancal statements of sponsorng corporatons because of the key assumptons (e.g., dscount rate, expected return on penson assets, etc.) that must be made by corporate managers n the fnancal reportng process (Amr and Benartz, 1998; Blake, 2001). Pror research relates penson asset allocaton of US corporatons to penson fundng levels through tax and other nstutonal and regulatory settngs, such as mnmum fundng requrements, the tax-deductbly of penson contrbutons and the exstence of the US Penson Benef Guaranty Corporaton (PBGC). 1 Penson asset allocaton s also related to employee demographcs and to frm rsk (Fredman, 1983; Bode et al., 1984). In addon to 1 See Black (1980), Tepper (1981), Harrson and Sharpe (1983) and Bader (1991). In August 2006, The Penson Reform Act was sgned nto law n the Uned States. Ths Act requres defned benef penson plans to attan full fundng status whn seven years by makng addonal contrbutons to the plans. 2

4 fundng levels, demographcs and frm rsk, Amr and Benartz (1999) argue that US corporatons that sponsor defned benef penson funds modfy ther penson asset allocaton to avod recognon of mnmum penson lably under Statement of Fnancal Accountng Standards (SFAS) No. 87 (FASB, 1985). Specfcally, they argue that sponsorng companes that are closer to mnmum lably recognon requrements would nvest more n bonds to allow better matchng between penson assets and lables, thus reducng the lkelhood of lably recognon. They provde emprcal evdence n support of ths argument. In November 2000, the Accountng Standards Board (ASB) n the UK ssued Fnancal Reportng Standard (FRS) No. 17, Retrement Benefs (ASB, 2000), whch nally had to be adopted for perods endng after 23 June In 2002, ASB extended the transonal perod of FRS 17 to fscal years startng on or after January 1 st, 2005, whch concded wh the adopton of Internatonal Fnancal Reportng Standards (IFRS) n Europe. Effectvely, UK companes had to adopt the revsed Internatonal Accountng Standard (IAS) No. 19 (IASB, 2004), whch was vrtually dentcal to FRS Durng fscal years UK companes were requred to provde detaled dsclosure on the status of ther penson plans. Unlke s predecessor, Statement of Standard Accountng Practce (SSAP) No. 24 (ASC, 1988), FRS 17 requres that penson assets are measured at market value and penson lables are dscounted usng the prevalng market yeld on AA-rated corporate bonds. In contrast, under SSAP 24 the valuaton of both penson assets and lables rely prmarly on the expected rate of return on assets. Furthermore, FRS 17 requres any actuaral gans or losses arsng durng the year to be recognzed mmedately n the statement of total recognzed gans and losses (comprehensve ncome), whle under SSAP 24, actuaral gans/losses are deferred and amortzed to ncome over the average 2 IAS 19 allows companes to use the 10% Corrdor method nstead of full penson recognon. Only a handful of UK Companes elected ths opton and we decded to delete them from our sample. 3

5 remanng servce lves of employees. Fnally, FRS 17 requres the total penson surplus/defc (the dfference between the market value of penson assets and the present value of the projected penson lables) to be recognzed on the balance sheet as an asset or a lably (net of deferred tax). Under SSAP 24 such surplus/defc s kept off balance sheet and dsclosure n the notes s requred. FRS 17 s also dfferent from SFAS 87 (US GAAP untl December 2006) n three major respects. Frst, under SFAS 87, although penson assets are measured at market value, the dscount rate n estmatng penson lables s based on an actuaral bass. Second, under SFAS 87 actuaral gans/losses (partcularly dfferences between expected and actual returns on penson assets) are deferred and amortzed to ncome over the average remanng servce lves of employees subject to the 10% corrdor method. Fnally, SFAS 87 requres recognon of the penson surplus/defc on the balance sheet only f a threshold s exceeded (e.g., addonal mnmum lably). In September 2006, the Fnancal Accountng Standard Board (FASB) ssued SFAS 158 (FASB, 2006) on defned benef and other postretrement plans. The new standard, whch s effectve for annual fnancal statements ssued n December 2006, requres mmedate recognon of all actuaral gans/losses and pror servce cost through other comprehensve ncome and the full recognon of the penson surplus/defc, smlar to FRS 17 and IAS By requrng recognon of net penson surplus/defc on the balance sheet and the mmedate recognon of actuaral gans/losses n comprehensve ncome, FRS 17 ntroduces materal volatly to UK companes balance sheets, especally f penson assets are mostly nvested n equy secures. Frstly, reportng actual, rather than smoothed, penson returns njects volatly nto shareholders equy. Furthermore, the recognzed net penson asset/lably could be a sgnfcant porton of a company s book value and market 3 FRS 17 stll dffers from SFAS 158 n the estmaton of the dscount rate. 4

6 capalzaton. For example, n 2002 Rolls Royce dsclosed a FRS 17 defc of 580 mllon n s penson fund, more than 25% of the assets of the company. The volatly that can therefore be ntroduced nto corporate balance sheets s evdenced by the fact that at the end of 2001 the combned surplus for the FTSE 100 was 5 bllon, but by md July 2002 wh the collapse n world stock markets ths fell to a defc of 25 bllon (Reynolds, 2002). Matchng penson assets wh penson lables s the most effectve method of reducng the volatly of the penson defc/surplus and the effect of FRS 17 on the volatly of comprehensve ncome and shareholders equy. FRS 17 requres penson lables to be dscounted usng the prevalng market yeld on AA-rated corporate bonds. Consequently, frms can reduce mark-to-market volatly by selectng a portfolo of penson assets whose far value s posvely correlated wh the far value of the penson lably, namely shft penson assets from equy to debt secures. 4 Alternatvely, companes could react to the new accountng rules by termnatng penson plans (Klumpes and Whtngton, 2003) or by convertng defned benef penson plans nto defned contrbuton and cash balance plans (D Souza et al., 2004). 5 However, these actons may be costly due to tax and labour negotaton costs. Ths study nvestgates the mpact of new penson dsclosures and subsequent full penson recognon under FRS 17 and IAS 19 on penson asset allocaton of sponsorng companes n the UK. Specfcally, we examne changes n UK companes penson asset 4 That s, an ncrease (decrease) n the yeld on AA-rated corporate bonds wll decrease (ncrease) the value of the penson lably. If the penson assets are prmarly nvested n bonds, the value of the penson assets wll smlarly decrease (ncrease) wh an ncrease (decrease) n the yeld on AA-rated corporate bonds. In 2001, Boots plc. lqudated all of s equy holdngs n s 2.3 bllon penson fund and moved the proceeds nto long-dated bonds. The sterlng bond market receved a boost n the week that ths news was announced. The Company lsts plan matury, plan sze, FRS 17, rsk reducton and plan management costs as possble reasons for swchng to a bond portfolo. Although the company's head of corporate fnance rejects the vew that FRS 17 motvated the swch, he stated that "[T]he Boots FRS 17 surplus at 30 September 2001 was 300 pretax. Had we remaned at the 30 March 2000 asset allocaton, ths would have been a 50 mllon defc." For addonal detals see McLesh (2001) and Ralfe (2002). 5 Swnkels (2006) observes that certan Dutch companes swched from defned benef penson plans to defned contrbuton plans followng the adopton of IFRS standards n

7 allocaton over the perod , durng whch UK companes had to dsclose penson nformaton n the notes to the fnancal statements and later recognze the full penson surplus/defc on ther balance sheets under eher FRS 17 (pror to January 2005) or IAS 19 (subsequent to January 2005). We dentfy two dstnct accountng regmes for UK companes: A Dsclosure perod rangng from fscal 2000 untl one year pror to the adopton of FRS 17 / IAS 19. We also dentfy a Full Recognon regme startng from one year pror to adopton untl one year after the adopton of FRS 17 / IAS 19. We predct that UK companes that sponsor defned benef penson plans wll shft penson assets from equy to debt secures durng the Dsclosure perod. Ths s because of the ncreased vsbly of penson plans due to marketbased dsclosures and the antcpaton of the effect of full penson recognon on the volatly of shareholders equy and comprehensve ncome. We also predct a declne n penson assets allocated to equy secures durng the Full Recognon perod due to the recognon of penson surplus/defc as a lably on the balance sheet and the hgher antcpated volatly of shareholders equy and comprehensve ncome. Thrd, we expect the shft to bonds to be more sgnfcant n companes wh larger relatve penson schemes and larger actuaral gans and losses. A shft from equy to debt secures could occur ndependently of penson accountng requrements. If penson asset allocaton s affected by global capal market changes rather than by new accountng requrements, s reasonable to expect a smlar shft n the asset allocaton of US corporate penson funds. Although, there are sgnfcant dfferences between the UK and US equy and debt markets, observng a smlar shft n asset allocaton for US companes n the absence of accountng changes weakens the argument that changes n accountng requrements are the man cause for the shft. Thus, we dentfy two penson accountng regmes n the US: a Partal Recognon regme startng from fscal 2000 untl 6

8 November 2006, and a Full Recognon regme startng from one year pror to adopton of SFAS 158 untl the year of adopton (fscal 2006). We compare the asset allocaton of UK companes to that of US companes and predct that the shft from equy to debt secures wll be more pronounced n UK companes than n US companes. To test our hypotheses, we use penson asset allocaton data for large UK companes (FTSE 350) over the perod We also collect smlar data for large US frms covered by Pensons and Investments over the perod The fnal sample conssts of 1,351 frm-year observatons for the UK sample and 1,906 frm-year observatons for the US sample. We fnd that durng the FRS 17 Dsclosure perod, UK companes changed ther penson asset allocaton by shftng funds from equy to debt secures. Durng the same perod, US companes mantaned a relatvely stable allocaton to eques and bonds. Ths sgnfcant dfference between UK and US companes supports the argument that new penson dsclosures nfluenced penson asset allocaton. We also fnd that UK companes shfted penson assets from eques to bonds durng the Full Recognon perod. Ths shft reduces the effect of the new standards on the volatly of shareholders equy and comprehensve ncome. Cross sectonal tests show that UK companes for whch the negatve mpact of full recognon was expected to be stronger shfted relatvely more funds from equy to debt secures. Results based on US companes also ndcate a shft away from eques to bonds durng the SFAS 158 Full Recognon perod, albe a smaller shft than documented for UK companes; smlar to the UK ths shft for US companes s related to the estmated mpact of the new penson accountng standard. Collectvely, the results corroborate the clam that penson accountng standards have a sgnfcant mpact on corporate penson asset allocaton. Ths study contrbutes to the lerature n several ways. It examnes the determnants of penson asset allocaton whn an nstutonal settng outsde the US. In addon, s 7

9 conducted durng a perod of a penson accountng change n the UK and n the US, whch s a powerful settng for testng the man hypothess n the paper. Furthermore, the emprcal evdence n ths study s mportant n understandng the possble effects of full penson recognon accountng on capal markets. The remander of ths study s organzed as follows. Secton 2 provdes nstutonal background. Secton 3 develops the hypotheses to be tested. Secton 4 descrbes the research desgn. Sample selecton and descrptve statstcs are ncluded n Secton 5. Secton 6 provdes the emprcal results whle Secton 7 provdes concludng remarks. 2. Instutonal Background SSAP 24, whch was used by UK companes for defned benef pensons snce 1988, came under pressure n the late 1990s for s poor dsclosure requrements, lack of transparency n the fgures produces, and s nconsstency wh US and nternatonal standards on penson accountng. ASB responded by ssung FRS 17, Retrement Benefs, n November Orgnally, FRS 17 was meant to be adopted by companes wh accountng perods endng on or after 23 June In November 2002, the ASB ssued an amendment to FRS 17 n response to the Internatonal Accountng Standard Board s (IASB) plan to reconsder the provsons of IAS 19. The amendment states that the requrements of FRS 17 would become mandatory for accountng perods begnnng on or after January 1 st, The purpose of that extenson was to allow an orderly transon to revsed standards algned wh those of the IASB whout mandatng two changes n accountng for retrement benefs whn a perod of two years. Consequently, whle some UK companes adopted FRS 17 pror to 2005 (early adopters), most adopted IAS 19, whch s very smlar to FRS 17, except for one mportant aspect: IAS 19 allows companes to use the 10% corrdor method. 6 See Appendx 1 for a lst of events pertnent to the ntroducton of FRS 17. 8

10 However, only a handful of companes elected ths opton. Durng the entre FRS 17 transonal perod, UK companes had to dsclose the present value of the penson oblgaton, far value of plan assets, penson asset allocaton, estmaton parameters and other nformaton n the notes to the fnancal statements. FRS 17 dffers from SSAP 24 n three major aspects. Frst, FRS 17 requres both penson assets and penson lables to be valued by reference to current market condons. Specfcally, penson assets should be measured at market value at the balance sheet date and penson lables are measured usng a dscount rate based on the return avalable on AA corporate bonds at the balance sheet date. Under SSAP 24 the valuaton of both penson assets and lables rely prmarly on actuaral assumptons. Second, FRS 17 requres any actuaral gans or losses arsng durng the year to be recognzed mmedately n the statement of total recognzed gans and losses. Ths mples that under FRS 17 actual rather than expected return on penson assets s reflected n shareholders equy and comprehensve ncome. Fnally, FRS 17 requres the total surplus/defc n the penson scheme to be recognzed as an asset or a lably (net of deferred tax) on the balance sheet. Under SSAP 24 any penson surplus/defc was kept off balance sheet. Untl December 2006, accountng for defned benef plans n the Uned States was based on SFAS 87 (and SFAS 106 for other postretrement benefs). SFAS 87 requres that penson assets are measured at far market value and penson lables are estmated on an actuaral bass. Also, actuaral gans/losses are deferred and sometmes amortzed to the ncome statement over the average remanng servce lves of employees, subject to the 10% corrdor method. In addon, SFAS 87 requres recognon of the penson surplus/defc only f a mnmum lably threshold s exceeded. In December 2006, FASB ssued SFAS 158, whch requres full recognon of penson surplus/defc on the balance sheet and 9

11 recognon of actuaral gans and losses n comprehensve ncome. 7 Thus, accountng for defned benefs plans n the US as of December 2006 s smlar to that of UK companes, namely full penson recognon. 8 Overall, FRS 17 and IAS 19 that followed make the effect of defned benef penson schemes on companes fnancal statements more transparent. The standards am at mprovng the reportng qualy of penson accountng by recognzng the off-balance-sheet funded status of penson plans and by movng from an actuaral bass to a market-based approach n valuaton. However, the use of market values at the balance sheet date, the full recognon of a penson surplus/defc, combned wh the mmedate recognon of actuaral gans/losses, ntroduces materal volatly nto the measurement of penson assets and lables and therefore volatly n the sponsorng company s shareholders equy and comprehensve ncome. To examne the effect of FRS 17 and IAS 19 on penson asset allocaton of UK companes, we dvde the sample perod nto two sub-perods: A Dsclosure perod rangng from fscal 2000 untl one year pror to the adopton of FRS 17 / IAS 19. For most frms n our sample, ths year s fscal The second s a Full Recognon perod rangng from one year pror to adopton of FRS 17 / IAS 19 untl one year subsequent to adopton. For most frms n our sample ths perod s between fscals 2004 and For comparson, we also dentfy two sub-perods for US companes. However, here the dstncton s between partal and full recognon, because SFAS 87 contaned mechansms that lmed the magnude of unrecognzed penson defc. Specfcally, we defne the 7 The purpose of the 10% corrdor method s to prevent accumulaton of actuaral gans/losses above 10% of the larger between market value of penson assets and projected benef oblgaton. Amortzaton of actuaral gans/losses s lmed only to the porton that exceeds the corrdor. The presumpton n SFAS 87 s that over the long run actuaral gans/losses wll revert back to a mean of zero. Mnmum lably threshold s the dfference between the Accumulated Benef Oblgaton (ABO) and the market value of penson assets. An addonal lably under ths requrement s recognzed aganst an ntangble penson asset or other comprehensve ncome. 8 See Appendx 2 for a comparson of SSAP 24, FRS 17, SFAS 87, SFAS 158 and IAS

12 perod as the Partal Recognon perod and the perod as Full Recognon perod Hypotheses Development Followng the ssuance of FRS 17 n 2000, UK companes were requred to dsclose the present value of the penson oblgaton and the market value of penson plan assets, as well as actuaral assumptons and detals on asset allocaton. These new dsclosures meant greater transparency of the penson scheme as well as ncreasng nvestor scrutny of the mpact of penson schemes on fnancal statements. Whle FRS 17 dsclosures provde comfort to nvestors n over-funded plans, under-funded companes are perceved as rsker because penson defcs are a form of debt. 10 Also, the exstence of Mnmum Fundng Requrements ncreases the lkelhood of addonal penson contrbutons, especally n under-funded plans when the economy s not dong well. Hgher future penson contrbutons and the uncertanty assocated wh ther tmng and magnude also ncreases the perceved rsk of sponsorng companes. Furthermore, under FRS 17, employees have access to more accurate dsclosures on the penson scheme. Thus, dsclosng penson defcs could trgger pressure by employees to reduce plan rsk, as the employees bear the defc, whle the company enjoys the surplus. Thus, reportng penson defcs n the notes to the fnancal statements s costly and companes would prefer to hedge themselves aganst potental penson defcs by reducng the volatly of the penson surplus/defc and mprovng the matchng of penson assets and lables. Better matchng s acheved by allocatng more penson assets to debt (.e., bonds and other nterest bearng secures) nstead of equy secures. As the market value of 9 We plan on extendng the sample perod to nclude December 2007 data as the post-adopton year. 10 See Harrs et al. (2001) for a dscusson of the effects of FRS 17 on fnancal statements and nvestors valuaton of sponsorng companes. 11

13 bonds s generally less volatle than that of stocks, and snce penson lables under FRS 17 are measured usng the prevalng market yeld on AA-rated corporate bonds, such a polcy would result n lower volatly of the penson defc/surplus. In addon, a shft from equy to debt secures s lkely to be exacerbated by the antcpated adopton of FRS 17 and the recognon of the full penson lably on the balance sheet. In contrast, shftng penson assets to bonds s lkely to result n hgher future penson contrbutons as the rate of return on bonds s expected to be lower than that on equy secures. Thus, the shft from equy to debt secures wll contnue as long as the cost of reportng a penson defc s hgher than the present value of addonal future contrbutons. US companes, on the other hand, have been dsclosng market-based penson nformaton n the notes to the fnancal statements under SFAS 87 snce Thus, defned benef plans have been under nvestors and employees scrutny pror to fscal We, therefore, expect ltle or no shft from equy to debt secures for US companes durng Ths analyss leads us to our frst research hypothess: Hypothess 1: UK companes that sponsor defned benef penson plans wll shft penson assets from equy to debt secures durng the FRS 17 Dsclosure perod. Durng the same perod, any shft by UK companes wll exceed any shft by US companes that sponsor defned benef penson plans. Adopton of FRS 17, or s equvalent standard, IAS 19, meant recognon of the entre penson defc/surplus on the balance sheet, recognon of all pror servce costs n net ncome and recognon of all actuaral gans/losses n comprehensve ncome. Therefore, n addon to ncreasng, on average, the amount of debt on the balance sheet, adopton of the new standards was expected to ncrease the volatly of shareholders equy and 12

14 comprehensve ncome. There are several contractual consequences to the adopton of full penson recognon. Frst, as many contracts are based on verfable balance sheet fgures, hgher recognzed debt on the balance sheet ncreases the lkelhood of volatng exstng debt covenants, hence the cost of debt renegotaton. Second, hgher volatly of shareholders equy ncreases the probably of volatng equy-based debt covenant because the balance sheet s now exposed to market volatly through nterest rate changes and changes n market values of eques. Thrd, the recognon of any penson defc would be set off aganst the company s dstrbuton reserves. Therefore, the adopton of FRS 17 may decrease dstrbutable reserves and may have a negatve effect on the company s ably to pay dvdends or to mantan a stable stream of dvdends. Adopton could also have a negatve effect on stakeholders ably to evaluate management performance. For example, Return on Equy (ROE), a wdely used management performance measure, may be dstorted as shareholders' equy becomes more volatle. If stakeholders consder comprehensve ncome n the numerator of ROE, then the volatly of ROE s even hgher. Also, hgher volatly of shareholders equy and comprehensve ncome could create a percepton of weaker management control. To mgate the effect of adopton on exstng contracts and to faclate better performance evaluaton, we expect managers of UK companes that sponsor defned benef penson schemes to shft penson assets from equy to debt secures durng the adopton of full penson recognon. Followng the adopton of SFAS 158, US companes are also expected to shft penson assets from equy to debt secures to reduce the effect of the new standard on total debt and the volatly of shareholders' equy and comprehensve ncome. However, the effect of adoptng SFAS 158 n the US s not expected to be as dramatc as that of adoptng FRS 17 / IAS 19 n the UK. Ths s because pror to adopton, US companes had to recognze part of 13

15 the penson lably due to the SFAS 87 mnmum penson lably requrement. Our second hypothess s thus: Hypothess 2: UK companes that sponsor defned benef penson plans wll shft penson assets from equy to debt secures durng the Full Recognon perod (adopton of FRS 17 or IAS 19). The shft of UK companes durng the Full Recognon perod wll exceed any shft by US companes durng the SFAS 158 Full Recognon perod. Next, we consder the cross-sectonal varaton n the mpact of new penson dsclosures and full penson recognon on UK companes. For a company wh a large penson scheme, the effect of a change n the market value of penson assets could be sgnfcant. For example, n s annual report for fscal 2003, Charter Plc, a UK-based engneerng company, reported penson assets wh a market value of mllon, whle s shareholders equy amounted to 24.9 mllon. Ths means that a declne of 5.4% n the market value of penson assets could elmnate the company's entre dstrbutable reserves on the balance sheet restrctng s dvdend payout ably. Smlarly, a declne n the yeld on the AA-rated corporate bonds could elmnate shareholders equy, as the penson oblgaton ncreases when the dscount rate declnes. Thus, we expect UK companes wh larger penson schemes relatve to shareholders equy to shft more assets from equy to debt secures durng the Dsclosure and Full Recognon perods. Companes wh larger penson schemes also experence larger actuaral gans/losses, especally f a majory of penson assets are nvested n equy secures. These companes are more sensve to the volatly effects of FRS 17 and would therefore have stronger motvaton to shft penson assets from equy to debt secures. In contrast, companes wh smaller actuaral gans/losses would have a weaker motvaton to swch penson funds to 14

16 debt secures, snce shftng to less rsky assets also mples a lower expected return on penson assets on the ncome statement (Fernandez, 2002). Therefore, we expect the shft from equy to debt secures durng both the Dsclosure and Full Recognon perods to be more sgnfcant for UK companes wh a larger rato of actuaral gans/losses to shareholders equy. Consequently, we hypothesze that the shft from equy to debt secures wll be more sgnfcant for UK companes wh a larger rato of penson assets/lables to shareholders equy and larger actuaral gans/losses relatve to shareholders equy. As for US companes, we have no pror reason to expect these relatons durng the Partal Recognon perod. However, we expect these relatons to exst n US companes durng the Full Recognon perod. However, we also expect these relatons to be stronger for UK companes than for US companes. Hypothess 3a: The shft of penson assets from equy to debt secures by UK companes durng the Dsclosure perod s posvely correlated wh the relatve sze of the penson scheme and the magnude of actuaral gans/losses. Ths relaton s not expected for US companes durng the Partal Recognon perod. Hypothess 3b: The shft of penson assets from equy to debt secures n UK companes durng the Full Recognon perod s posvely correlated wh the relatve sze of the penson scheme and the magnude of actuaral gans/losses. Ths relaton s expected to be stronger for UK companes than for US companes durng the Full Recognon perod. 4. Emprcal Desgn Although penson assets could be nvested n a varety of asset categores, dsclosures n 15

17 fnancal statements usually classfy penson assets nto three man categores: stocks, bonds, and others. Stocks and bonds together account for about 90% of total penson funds n our UK and US samples. Others often nclude such assets as mortgage-backed secures, venture capal, prvate placement, propertes, etc. To test hypotheses 1 and 2 we analyze penson asset composon over the perod for a sample of UK companes and compare ths composon to a sample of US companes. Specfcally, we are nterested n whether UK sponsorng companes have shfted penson funds from equy to debt secures durng the Dsclosure perod and the Full Recognon perod, and whether ths swch has occurred n US sponsorng companes durng the Partal Recognon and the Full Recognon perods. To test hypotheses 3a and 3b, we construct a model that explans the cross-sectonal varaton n the percentage of penson funds allocated to equy secures (requity ). The nal model s: requity + β DIVP 6 = β + β IMPACT 0 + β TAXR β SDCF 8 + β FUND 2 + β SIZE 9 + β FUND β FAGE 10 + β HOR 4 + β CLOSE 11 + β LEV 5 + ε (1) The dependent varable n Equaton (1), requity, s the market value of penson assets allocated to equy secures dvded by the market value of total penson assets for frm n year t. The man test varable n Equaton (1) s IMPACT, whch measures the potental mpact of the new accountng standards on company n year t. As the mpact relates prmarly to the relatve sze of the penson plan, we use the followng four measures: () EXPOS1 : The far value of penson assets deflated by book value of shareholders equy for frm n year t. A change n the market value of penson assets affects shareholders equy drectly, ceters parbus. Therefore, EXPOS1 captures the company s exposure to the volatly n the market value of penson assets and net 16

18 penson surplus/defc. () EXPOS2 : The Projected Benef Oblgaton (PBO) deflated by book value of shareholders equy for frm n year t. A change n the dscount rate wll drectly affect the projected penson oblgaton, and dependng on the penson asset mx may affect shareholders equy. Therefore, EXPOS2 captures the company s exposure to the volatly n dscount rates. () ACTGL1 : The absolute value of realzed actuaral gans/losses deflated by the shareholders equy for frm n year t. A hgher magnude of actuaral gans/losses s prmarly the result of a larger dfference between expected and realsed penson asset returns; ceters parbus, ths should lead to more volatle shareholders equy. Therefore, ACTGL1 captures the frms exposure to changes n penson asset values. (v) ACTGL2 : ACTGL1 dvded by the percentage of penson funds allocated to equy for frm n year t. A lower percentage of penson funds nvested n eques would result n a lower magnude of actuaral gans/losses. Hence, there mght be a mechancal posve relaton between ACTGL1 and requity. The scalng of ACTGL1 by the percentage of penson funds allocated to equy should mgate ths mechancal relaton. Therefore, ACTGL2 proxes for the expected volatly reflected n actuaral gans/losses assumng all penson funds are nvested n eques. In calculatng EXPOS1, EXPOS2, and ACTGL1 we adjust shareholders equy by undong the mmedate recognon of actuaral gans/losses and the recognon of net penson surplus/defc for early adopters. We nclude FUND and FUND 2 as a measure of the fundng status of frm s penson fund n year t. These varables are ncluded to capture the tax and regulatory nfluences on penson fund asset allocaton. In general, the tax-deductbly of penson contrbutons n the UK and the US should nduce companes to pre-fund ther penson plans; companes that are 17

19 subject to hgher tax rates should have even greater ncentves to pre-fund ther penson plans. However, pror lerature s mxed as to the nfluence of fundng status on penson asset allocaton. Black (1980) and Tepper (1981) argue that snce returns on penson assets are not taxed, these assets should be nvested n the most heavly taxed secures, presumably bonds. Ther argument suggests no assocaton between fundng levels and asset allocaton as all companes nvest n bonds regardless of fundng levels. Harrson and Sharpe (1983) argue that the exstence of the Penson Benef Guaranty Corporaton (PBGC) n the US provdes a (US) company wh a put opton on s extremely under-funded penson oblgaton. Together wh lmed tax deductbly n the case of extremely over-funded plans, they argue that fundng and asset-allocaton decsons are jont and extreme. To maxmze tax benefs on one hand and the value of the PBGC opton on the other hand, companes should eher over-fund the penson plan and allocate all the assets to bonds, or under-fund and allocate all the assets to eques. Although n practce fundng/asset-allocaton decsons are rarely extreme, ths argument supports a negatve relaton between fundng levels and allocatons to eques. A serous caveat s that the value of the PBGC put opton n the US has declned over tme and n partcular snce the 1986 Tax Reform Act. Also, an nsurance company such as the PBGC dd not exst n the UK untl Therefore, the ncentve to allocate penson assets to equy secures n cases of extreme under-fundng may be of a second order nature. Bader (1991) argues that companes strve to mnmze the volatly of future penson contrbutons. These contrbutons are farly predctable for moderate fundng levels, but less predctable for more extreme levels. To reduce the volatly of penson contrbutons, Bader (1991) argues that extremely over-funded and under-funded plans should nvest n bonds, whle only moderately funded plans should ncrease allocaton to eques. Hs argument 18

20 suggests an nverted U-shape relaton between fundng levels and the allocaton to eques. Based on these arguments, we nclude both FUND and FUND 2 to accommodate the possbly of a non-lnear relaton between fundng levels and asset allocaton. 12 The fundng status s measured as the far value of penson assets over the ABO for frm n year t. 13 We nclude HOR to control for the horzon of the penson oblgaton. Specfcally, penson fund managers that am at reducng the volatly of penson defcs should take nto consderaton the horzon of the penson oblgaton. Whle penson oblgatons to retrees are relatvely short-term and are prmarly affected by nterest rates, penson oblgatons to actve employees are relatvely long-term and are prmarly affected by salary ncreases. As bonds are more correlated wh nterest rate changes and stocks are more hghly correlated wh salary ncreases, companes wh a relatvely young workforce should nvest more n stocks and less n bonds. Consequently, we expect a posve correlaton between nvestment horzon and allocaton to equy secures. HOR, s measured as the natural logarhm of the rato of PBO to current servce cost for frm n year t. Our measure dffers from Amr and Benartz (1999), who use ABO n ther calculaton, snce ABO s not avalable for UK frms. However, we beleve that our measure s a reasonable proxy for nvestment horzon snce a workforce whch s closer to retrement should generate a larger servce cost relatve to a young workforce. We also recognze the fact that an older workforce would lkely lead to a larger PBO, ceters parbus; however, we beleve ths mpact s not as nfluental as the servce cost mpact. Overall, an older workforce should lead to a smaller rato of PBO to servce cost, ndcatng a shorter nvestment horzon. To summarze, we expect a posve 11 A government regulatory body such as the PBGC dd not exst n the UK durng the sample perod, but exsts today (see Fnancal Economcs Roundtable, March 2005). 12 Amr and Bernatz (1999) documented an nverted-u relaton between fundng status and the percentage nvested n eques. 13 We defne fundng status n ths way to be consstent wh pror studes (Amr and Bernatz, 1999). Snce ABO s not avalable for UK companes, we calculate ABO based on the formula proposed n Amr and Bernatz (1999): ABO = PBO / (1+G) N, where G s the assumed projected salary ncrease and N s penson fund s nvestment horzon. 19

21 relaton between HOR and the amount nvested n eques. We acknowledge that our horzon varable (HOR ) may be subject to measurement error. To mgate ths problem, we nclude a varable that measures frm age (FAGE ). More mature frms are lkely to have more mature workforce and hence a shorter nvestment horzon. Thus, we expect a negatve relaton between frm age and percentage allocaton to eques. Frm age s measured as the natural logarhm of number of years snce the year of ncorporaton. If the date of ncorporaton s not avalable, then the year n whch the company appeared on the database (Datastream for UK frms and CRSP for US frms) was used as a proxy for year of ncorporaton. We nclude two control varables that capture the nfluence of debt contracts and dvdend payout polcy. As defned benef penson plans are reported on sponsorng companes' balance sheets, the penson lably and correspondng nvestment portfolo may be affected by certan contractual arrangements. In partcular, companes that are closer to the volaton of debt covenants have stronger motves to mprove asset/lably matchng n order to reduce penson defcs on the balance sheet. Better asset/lably matchng would also reduce the volatly of shareholders' equy and future penson contrbutons, whch n turn would reduce the volatly of dvdends. We expect companes wh tghter debt covenants and hgher dvdends payout ratos to be more concerned about penson asset/lably matchng. Therefore, to capture the effect on leverage and dvdend polcy on penson asset management, we nclude fnancal leverage (LEV ) and dvdends payout rato (DIVP ); we expect the coeffcents on these varables to be negatvely assocated wh the amount nvested n equy. Fnancal leverage (LEV ) s, measured as long term debt dvded by the sum of long term debt and market value of equy for frm n year t. Dvdend payout rato (DIVP ) s measured as dvdends per share dvded by retaned earnngs per share for frm n year t. If current retaned earnngs are negatve, then the varable s measured as the 20

22 average dvdends per share over the current and past two years dvded by average retaned earnngs per share over the current and past two years. Snce returns on penson funds are not taxed, penson assets should be nvested n the most heavly taxed secures, presumably bonds. Furthermore, companes subject to hgher tax rates should have greater ncentves to allocate more penson assets to bonds. Therefore, we nclude the company's effectve tax rate (TAXR ), measured as total tax expense dvded by pre-tax ncome for frm n year t. If current pre-tax ncome s negatve, we use the average tax expense over the current and past two years dvded by the average pre-tax ncome over the current and past two years. We also nclude the volatly of operatng cash flows and frm sze to capture frm rsk. It has been documented by Fredman (1983) and Bode et al. (1984) that companes tend to offset hgh corporate rsk by nvestng more of the penson assets n bonds. Ths polcy of offsettng rsk through the penson fund may reflect management preference to avod makng contrbutons to the penson fund when operatng cash flows are low. 14 Thus, we expect a negatve correlaton between the varably of operatng cash flows and equy allocaton. In addon, companes wh more dversfed operatons would prefer to assume more rsk n ther penson fund. To the extent that larger frms are more dversfed, we would expect a posve assocaton between frm sze and allocaton to eques. The volatly of operatng cash flows (SDCF ) s measured as the standard devaton of operatng cash flows over the current and past four years, deflated by book value of common equy for n year t. Frm sze (SIZE ) s measured as the natural logarhm of market value of equy for frm n year t. Fnally, as many companes move away from defned benef to defned contrbuton 14 Specfcally, f operatng cash flows are volatle and the penson assets are nvested n eques, the plan s lkely to become under-funded when operatng cash flows are low. As a result, the company would have to make large contrbutons to the penson fund n tmes of low operatng cash flows. 21

23 penson plans, many defned benef penson plans are closed to new entrants. Ths feature could have a sgnfcant effect on penson asset allocaton, although we are unable to predct the drecton of ths effect. We nclude n our model an ndcator varable (CLOSE ) that s equal to one f the prncpal defned benef plan s closed to new entrants, and zero otherwse. To drectly test our hypotheses, we defne two dependent varables: The frst one computes the change n the percentage of assets allocated to eques durng the Dsclosure perod. For UK companes, ths perod begns n year 2000 and ends one year pror to adopton of FRS 17 or IAS 19, as follows: rdisclose = requity (year 2000) - requity (Pre-Adopton year). 15 The second dependent varable computes the change n percentage of assets allocated to eques durng the Full Recognon perod. For UK companes, ths perod begns one year pror to adopton of FRS 17 / IAS 19 and ends one year subsequent to adopton, as follows: radopt = requity (Pre-Adopton year) - requity (Post-Adopton year). Smlarly, we compute rdisclose and radopt for US companes durng the Partal Recognon and Full Recognon perods, respectvely. These varables are defned as follows: rdisclose = requity (year 2000) - requity (Pre-Adopton year), radopt = requity (Pre-Adopton year) - requity (Adopton year). We estmate the followng emprcal models: rdisclose = β + β IMPACT + β FUND + β FUND + β DIVP + β TAXR + β SDCF + β SIZE β FAGE β HOR + β LEV + β CLOSE ε (2) 22

24 radopt = δ + δ IMPACT δ DIVP + δ TAXR δ SDCF 8 + δ FUND 2 + δ SIZE 9 + δ FUND 3 + δ 10 FAGE 2 + δ HOR 4 + δ CLOSE 11 + δ LEV 5 + η + (3) where the ndependent varables are measured at the end of the relevant perod. In addon, we estmate equatons (2a) and (3a) usng frst dfferences of the ndependent varables. Each ndependent varable s measured as the dfference between the level of the varable at the begnnng of the perod (e.g., Dsclosure perod) and s level at the end of the perod: rdisclose + β ΔLEV 5 + β ΔFAGE 10 + β ΔDIVP + β ΔTAXR = β + β ΔIMPACT β ΔCLOSE ε + β ΔFUND 2 + β ΔFUND + β ΔSDCF + β ΔSIZE β ΔHOR 4 (2a) radopt + δ ΔLEV δ ΔFAGE 10 5 = δ + δ ΔIMPACT + δ ΔDIVP + δ ΔTAXR δ ΔCLOSE η + δ ΔFUND 2 + δ ΔSDCF 8 + δ ΔFUND δ ΔSIZE 9 + δ ΔHOR 4 + (3a) We estmate equatons (2), (2a), (3) and (3a) for UK and US companes. Table 1 provdes defnons of all of our varables. (Table 1 about here) We use ordnary least square (OLS) to estmate the above equatons. Petersen (2006) argues that resduals may be correlated across frms or across tme n panel data and therefore OLS standard errors may be based. In ths study, s possble that the unspecfed determnants of the dependent varables are correlated both over tme and across frms. For example, f the equy market as a whole s volatle, then all frms may shft away from 15 Although the Dsclosure perod begns n year 2000, n some of our tests we use year 2001 as the startng year because many observatons for year 2000 are mssng. 23

25 volatle eques to less volatle bonds. On the other hand, the strategy of penson asset allocaton could also be drven by some unspecfed frm-specfc factors, whch wll gve rse to a frm effect n the error terms. The presence of posve correlaton among error terms results n underestmated standard errors and thus nflated t-statstcs. To address such econometrc concern, we employ two-dmenson clusterng suggested by Petersen (2006) to accommodate the possbly of both tme effect and frm effect n such panel data. 5. Sample Selecton and Descrptve Statstcs To obtan our nal UK sample we selected those frms from the FTSE 350 that sponsor defned benef penson plans over fscals Approxmately 250 of the FTSE 350 sponsor a defned benef penson plan. Informaton on market value of penson assets, actuaral present value of penson lables, penson actuaral assumptons, actuaral gans/losses and detals of penson asset allocaton are collected from annual fnancal statements. 17 All other fnancal data for UK companes are from Datastream. Data for US companes' penson asset allocaton untl 2004 are collected from Pensons and Investments, a perodcal survey that covers the largest 1,000 penson funds n the US. Of ths 1,000, approxmately 300 penson funds relate to defned benef plans for publcly traded frms (the remander are sponsored by prvate frms, unons, or government entes, or are foregn companes lsted n the US). Asset allocaton data for are collected from notes to the annual fnancal statements. Fnancal data for US companes are from Compustat. After removng observatons wh mssng data, the sample conssts of 3,257 frm-year observatons, ncludng 1,351 observatons for the UK sub-sample and 1,906 observatons for the US sub-sample. 16 Fscal 2006 accordng to Compustat defnon ncludes data for the frst half of Data on penson asset allocaton for 2000 for UK frms s from Penson Funds and ther Advsors. 24

26 Table 2 provdes descrptve statstcs for the UK and US sub-samples. Over the entre sample perod, UK companes allocate, on average, 63% of ther penson assets to eques, whereas US companes allocate, on average, 62% to eques. The magnude of actuaral gans/losses (ACTGL1, ACTGL2) s slghtly hgher n US than n UK companes. However, the sze of the penson plan relatve to shareholders' equy s larger n UK than n US frms, as reflected by hgher means and medans of EXPOS1 and EXPOS2 (statstcal tests not reported). As for other varables, the UK and US sub-samples are smlar to each other n terms of fundng status (FUND), penson horzon (HOR), frm age (FAGE) and effectve tax rates (TAXR). US companes are less rsky as reflected by lower volatly of cash flows (SDCF) and larger frm sze (SIZE). US companes also have lower dvdend payout ratos and are more hghly leveraged than UK companes. 18 (Table 2 about here) By end of fscal 2006 a total of 136 UK companes n our sample (58%) have closed ther prmary defned benef penson plans to new entrants. In contrast, only 76 US companes n our sample (21%) have closed ther prmary defned benef penson plans to new entrants by 2006 (not tabulated). The dfference between US and UK companes (sgnfcant at the 0.01 level) s consstent wh our clam that UK companes take actons to reduce the effect of penson accountng on fnancal statements. 19 Fgure 1 presents the mean/medan percentage of penson assets allocated to equy and debt secures for UK and US companes. Focusng on the means, the top two lnes show that UK companes reduced ther allocaton to eques durng the sample perod whle US companes mantaned a relatvely stable allocaton to eques. The bottom two lnes show 18 All contnuous explanatory varables are wnsorzed at 1% and 99% to mgate the effect of extreme observatons, except TAXR and DIVP whch are wnsorzed at 5% and 95% to remove negatve values. 25

27 that UK frms ncreased the allocaton to bonds whle US frms dsplay a relatvely stable allocaton to bonds durng the same perod. Plots based on medans exhb smlar patterns. These results are consstent wh the argument that changes n asset allocaton of UK companes durng the Dsclosure and Full Recognon perods of FRS 17 / IAS 19 were drven by the accountng change rather than by general economc condons. (Fgure 1 about here) Fgure 2 depcts the trend of penson asset allocaton around the adopton of FRS 17 / IAS 19 for UK companes. For each company, we dentfy the adopton year and denote as AY(0). For the majory of our sample, the adopton year s fscal 2005, however, 54 companes n our sample elected to adopt FRS 17 pror to the mandatory year. Then, we plot the mean (medan) percentage of penson assets allocated to eques and bonds from year AY(-3) to AY(3). Both Fgure 2a and Fgure 2b exhb a declne n penson funds nvested n eques, and an upward trend n the percentage nvested n bonds. Overall, Fgure 2 exhbs a consstent shft away from eques to bonds throughout the sample perod. (Fgure 2 about here) When comparng asset allocatons across US and UK, we rely on the assumpton that the correlaton between equy and debt markets n these countres s suffcently hgh. Also, s possble that asset allocaton s nfluenced not only by corporate actons but also by changes n market values of stocks and bonds. 20 Table 3, Panel A presents annual stock and bond returns for the US and UK markets, based on Dmson et al. (2002, 2004, and 2007). As 19 Greenwch Assocates report that 22% of large defned benef penson plans are closed to new entrants n the US (McSherry, 2006). 20 Informal dscussons wh penson experts n the UK suggest that penson asset management has changed over tme. Untl the md 1990s, when balanced management was the man nvestment strategy n the UK, the penson fund trustees set a target strategc asset allocaton (SAA) for fund managers (e.g. 70% equy, 20% bonds, 10% cash), but they also set some bounds around that (e.g. %equy = 70% +/- 5%). Good equy performance could lead to the 75% upper bound beng breached. The manager would then have to argue for the bound to be ncreased (temporarly) or be asked to sell eques and redstrbute the funds to brng the SAA back n lne. Snce the late 1990s, the suaton changed as the mplcaton of penson defcs became more sgnfcant. Lably-drven nvestng replaced the SAA that was set ndependent of lables. Currently, there s less flexbly over the SAA than prevously, and rebalancng occurs more frequently. 26