Overview of Retirement Issues

This post describes my work on retirement issues that was published in “Defying Magnets: Centrist Policies in a Polarized World”   The book can be found on Amazon and Kindle.

https://www.amazon.com/Defying-Magnets-Centrist-Policies-Polarized/dp/179668015X/ref=sr_1_2?keywords=defying+magnets%3A+Centrist+policies&qid=1550091821&s=amazon-devices&sr=8-2

There are two pillars of retirement income in the United States.   The first pillar involves Social Security a mandatory program covering most workers.   The second system involves voluntary defined contribution pension plans.  This section starts with a basic description of the Social Security system and private defined contribution retirement plans.

The Social Security program has been highly popular with Americans and many retirees are highly dependent on this government-run program.   However, the Social Security system is running shortfalls which will lead to automatic benefit cuts around 2035.   Defined contribution pension plans — 401(k) plans and IRAs — have over the last 40 years become the dominant vehicle for private retirement savings.

Many people forego investing in 401(k) plans and IRAs even though these plans provide generous tax benefits to savers.   The failure of many people to fully invest in tax-deferred retirement plans puzzles many financial advisors.   The analysis presented here indicates that saving for emergencies and reducing debt is and should be a higher financial priority than saving for retirement for many people.

Several changes to rules governing retirement savings accounts which would allow more people to save for retirement while aggressively reducing debt and preparing for emergencies are presented here.

  • Allow tax-free and penalty-free distributions on a portion of total contributions (perhaps 25 percent) for emergencies, student debt reduction, mortgage restructuring and long term care expenses prior to age 59 ½
  • Allow for some tax-free and penalty-free distributions for paying off the mortgage for people over age 50.
  • Eliminate all other distributions from 401(k) plans prior to the age of 59 ½.
  • Prohibit 401(k) loans.
  • Prohibit states from denying Medicaid and food stamp benefits for low-income people with 401(k) assets.

This proposal could be paid for by imposing a haircut on tax exemption for 401(k) contributions.   (Currently, 100 percent of contributions to a 401(k) plan are exempt from income tax.   The new rule would exempt 85 percent of contributions.)

People without access to an employer-sponsored retirement plan have substantially lower retirement savings than people with access to retirement plans at work.  We consider ways to expand retirement savings for people without employer-based retirement savings.   Specific proposals include:

  • Equalization of contribution limits between IRAs and 401(k) plans.
  • A rule change allowing firms without 401(k) plans to contribute to employee IRAs.
  • A rule change allowing firms to compensate contractors and employees of contractors with non-taxed fringe benefits including contributions to IRAs.

According to the trustees of the Social Security system, financial imbalances impacting Social Security stemming from the decrease in the working-age population will result in automatic cuts to Social Security benefits around 2034.   In my view, it will be difficult to implement any compromise that reduces Social Security fiscal imbalances and prevents future automatic cuts to benefits without first increasing private retirement savings and reducing the dependence on Social Security.  The final chapter of this section outlines and reviews some policy proposals related to improving the Social Security system.

Common Ground on Social Security COLAs?

Is there common ground on Social Security COLAs?

Social Security benefits are adjusted for inflation each year.  President Obama is on record for supporting changes to the way the Social Security COLA is calculated.  I suspect that the proposal to modify the current Social Security COLA will receive strong consideration by Congress after the 2014 midterm elections.   

This post has my comments on both the economics and politics of proposals to adjust Social Security benefits:

Comments:

Impact on Beneficiaries:

 Under current law, the Social Security benefit is linked to the traditional CPI.  President Obama is supportive of a change that would link the Social Security COLA to a chained CPI.  On average, the growth in the chained CPI is around 0.2 percentage points lower than the growth in the traditional CPI.  It is of course possible that the percentage point difference between the traditional and chained CPI would be higher in a high-inflation environment.

Estimates in my math blog reveal that the difference in benefits due to the adjustment in the CPI could be around 4 percentage points in a low-inflation environment and around 14 percentage points in a high-inflation environment.

http://dailymathproblem.blogspot.com/2013/04/how-important-is-social-security-cola.html 

 Impact on budget in short term and long term:

Social Security has a major impact on both the current government budget and the future debt to GDP ratio for the nation.

In fiscal year 2013 Social Security accounted for $808 billion in expenditures, around 25% of federal expenditures. It is the single largest federal program.

In calendar year 2010 outlays from Social Security exceeded revenues.   This is the first time outlays have exceeded revenues since the Social Security reform law of 1983.

The size of the this program makes it an important consideration in annual budget talks.  

Impact of future cuts on Social Security:

Under current rules, Social Security benefits are paid solely by Social Security taxes and assets in the trust fund.  However, according to the actuaries at the Social Security Administration the trust fund will be unable to pay full benefits starting at around 2037.

http://www.ssa.gov/policy/docs/ssb/v70n3/v70n3p111.html

“As a result of changes to Social Security enacted in 1983, benefits are now expected to be payable in full on a timely basis until 2037, when the trust fund reserves are projected to become exhausted.1 At the point where the reserves are used up, continuing taxes are expected to be enough to pay 76 percent of scheduled benefits. Thus, the Congress will need to make changes to the scheduled benefits and revenue sources for the program in the future. The Social Security Board of Trustees project that changes equivalent to an immediate reduction (bold added by me) in benefits of about 13 percent, or an immediate increase in the combined payroll tax rate from 12.4 percent to 14.4 percent, or some combination of these changes, would be sufficient to allow full payment of the scheduled benefits for the next 75 years.”

Would an adjustment to the Social Security Administration substantially delay the future benefit cuts?

I have not seen any work on the number of years an adjustment in the Social Security COLA would delay the future benefit cuts.  I believe the short answer is that COLA adjustments by themselves would result in a relatively small delay in future forced reductions in Social Security benefits.   

The reason the delay in benefit cuts is likely to be small is that much of the impact of the COLA adjustment occurs after 2037. Note that the Social Security projections on extending the life of the Trust Fund to 75 years is based on a scenario that assumes an immediate reduction in benefits.   The COLA adjustment does not result in an immediate benefit reduction.

One year after the adjustment to the COLA the estimated impact on all Social Security beneficiaries is only 0.2 percentage points.  As noted above, 20 years above the impact in a low-inflation scenario for those who have lived 20 years in retirement will be around 4 percentage points.  It will take around 38 years for the maximum annual impact of the COLA adjustment to be realized.

The long run solution to the Social Security problem will involve benefit cuts, revenue increases dedicated to Social Security, and shifts of revenue from the general fund to Social Security.  Population aging will inevitably lead to an increase debt to GDP ratio.  

The proposed COLA is a substantial decrease in benefits which does not preclude other cuts in the future. In fact future benefit reductions would still be mandatory once trust fund assets expired.  Moreover, the proposed benefit reduction is not linked to a commitment for needed revenue increases of any kind.

The Retirement Crisis:

It is increasingly obvious that the current approach to saving for retirement is not working for a large number of workers.

This interview with Teresa Ghilarducci provides some evidence on this point.

http://www.pbs.org/wgbh/pages/frontline/business-economy-financial-crisis/retirement-gamble/teresa-ghilarducci-why-the-401k-is-a-failed-experiment/

 

Research by the Employee Benefit Research Institute found that between 4% and 14% additional baby boom workers had their retirement become at risk due to the 2008/2009 crash.

Future crashes would also lead to an increase in workers with retirement at risk.

The COLA adjustment would also increase the percent of workers, both current and future, who would end up with inadequate retirement savings.

Senator Warren from Massachusetts appears to be one of the few politician who has grasped both the severity of the current retirement crisis and the impact of the proposed COLA adjustment on the adequacy of retirement savings.  Below is a link to a recent speech where Elizabeth Warren proposes to expand Social Security.

An expansion of Social Security is not likely to occur.  However, liberals can and should insist on substantial improvements in the nation’s retirement system, which must coincide with and offset future cuts to Social Security.

Politics:  

President Obama broached the issue of changes to the Social Security COLA during the fiscal cliff negotiations.  My first post at this blog was about President Obama’s COLA adjustment proposal offered during the fiscal cliff debate.  It was my first post at this blog.

A new crisis over the debt limit is likely to occur after the November election.  Many in the House and the Senate will only have a few more months to serve, either because of  a planned or unplanned retirement.  Given the possibility of a debt default and the lame duck status of many in the House and Senate it is likely there will be considerable support for a COLA adjustment after the 2014 elections. 

Conclusion:

President Obama has placed liberals at disadvantage by endorsing the COLA cut without firm concessions on future revenues and restrictions on future future cuts. Most in the Republican party oppose the use of any additional revenue to offset Social Security imbalances.  In my view, the long run solution to the Social Security problem will involve benefit cuts, revenue increases dedicated to Social Security, and shifts of revenue from the general fund to Social Security.  Moreover, even if a comprehensive Social Security reform plan is implemented sooner rather than later it is likely that population aging will still lead to a substantial increase in the debt to GDP ratio over the next few decades.

The passage of a COLA adjustment does not prevent future cuts to Social Security, which will be automatically triggered when the trust fund is depleted.  A case can be made for adjusting the Social Security COLA but only if this change is made in conjunction with other changes that guarantee the survival of Social Security and improve the current retirement system.

Impact of altering indexation of Social Security for females vs males

Question:   What is the expected value of lifetime Social Security benefits for females and for males when benefits are linked to the traditional CPI and when benefits are linked to the chained CPI.

Discuss the reasons why women might prefer a switch to the chained CPI over proposals to partially privatize Social Security.

Short Answer:

Answer is contingent on several assumptions laid out below.  I find that changing from the traditional CPI to a chained CPI would reduce the expected value of lifetime Social Security benefits by around $16,000 for males and $21,000 for females.

The actual impact is invariably different from the expected impact.  Regardless of gender, people with the longest life span get the most from Social Security.

However, Social Security is really essential for females because private annuities are more expensive.  See my previous post on this topic.

http://dailymathproblem.blogspot.com/2014/01/gender-differences-in-life-expectancy.html

Analysis:

Key assumptions:

The key assumptions in this analysis are

  1. Person retires at age 62 and receives an initial Social Security retirement benefit of $15,000 per year
  2. Traditional CPI grows at 2.42% per year
  3. Chained CPI grows at 2.09% per year.
  4. In year of death person receives ½ year Social Security Benefit
  5. Probability of surviving from age 62 to age y> 62 is determined by the CDC life tables for females and males.

Readers interested in the discussion of assumptions on difference between traditional and chained CPI might want to look at this post.

http://dailymathproblem.blogspot.com/2014/02/comparing-traditional-and-chained-cpi.html

The expected lifetime Social Security benefit is E(SSB)=Sum(Pyr x CByr)  where Pyr is the probability of surviving to a particular year and CByr is the cumulative benefit from the retirement age at 62 to the year of death.

The logic behind the calculation of the probability a retiree survives to a specific date is similar to the logic behind the geometric distribution.   The probability of surviving to age y > 62 is the product of the probability of surviving to age y-1 and the probability of dying at age y.

Calculations:

The chart below has data on likelihood of surviving to age y+0.5 for males and females and the cumulative Social Security Benefit to age y+0.5 under both the existing COLA and a chained CPI COLA.

Survivor Probabilities and Cumulative Benefits
Age y Probability of surviving to exactly age y+0.5 for males Probability of surviving to exactly age y+0.5 for females Cumulative Benefit With Existing COLA Cumulative Benefit With COLA linked to chained CPI
62 0.01321 0.00831 $7,500 $7,500
63 0.01405 0.00896 $22,682 $22,657
64 0.01496 0.00965 $38,230 $38,130
65 0.01599 0.01044 $54,156 $53,927
66 0.01713 0.01133 $70,466 $70,054
67 0.01830 0.01227 $87,171 $86,518
68 0.01946 0.01322 $104,281 $103,327
69 0.02062 0.01422 $121,805 $120,486
70 0.02178 0.01526 $139,752 $138,004
71 0.02306 0.01647 $158,134 $155,889
72 0.02458 0.01783 $176,961 $174,147
73 0.02620 0.01929 $196,244 $192,786
74 0.02780 0.02077 $215,993 $211,816
75 0.02935 0.02224 $236,220 $231,243
76 0.03079 0.02380 $256,936 $251,076
77 0.03230 0.02547 $278,154 $271,323
78 0.03392 0.02732 $299,885 $291,994
79 0.03557 0.02926 $322,143 $313,096
80 0.03691 0.03112 $344,938 $334,640
81 0.03791 0.03288 $368,286 $356,634
82 0.03876 0.03473 $392,198 $379,088
83 0.03954 0.03677 $416,690 $402,011
84 0.03996 0.03858 $441,774 $425,413
85 0.04032 0.04017 $467,464 $449,304
86 0.04010 0.04170 $493,777 $473,694
87 0.03929 0.04275 $520,727 $498,594
88 0.03787 0.04322 $548,328 $524,015
89 0.03584 0.04302 $576,598 $549,967
90 0.03325 0.04209 $605,551 $576,461
91 0.03021 0.04041 $635,206 $603,509
92 0.02681 0.03798 $665,578 $631,123
93 0.02321 0.03490 $696,685 $659,313
94 0.01957 0.03128 $728,544 $688,093
95 0.01604 0.02730 $761,175 $717,474
96 0.01276 0.02314 $794,596 $747,469
97 0.00984 0.01903 $828,825 $778,091
98 0.00734 0.01513 $863,882 $809,353
99 0.00529 0.01163 $899,788 $841,269
100 0.01012 0.02606 $936,563 $873,851
1.00000 1.00000

The expected value of lifetime benefits for males/females under traditional/chained CPI is simply the dot product (the sum product function in EXCEL or NUMBERS) for the relevant probabilities and cumulative benefits.

Impact of Change in COLA by Gender
Males Females Difference Females- Males
Traditional CPI $392,077 $463,804 $71,727
Chained CPI $376,005 $442,772 $66,767
Difference Traditional-Chained CPI $16,072 $21,032

The change in the COLA formula from the traditional CPI to the chained CPI leads to a reduction in expected lifetime benefits of $16,000 for males and $21,000 for females.

Social Security still provides longevity protection under a chained CPI.

http://dailymathproblem.blogspot.com/2014/01/gender-differences-in-life-expectancy.html

This is especially important for females because of their longer life expectancy.

Concluding Thoughts:

The issue of the Social Security COLA is important and complex.   I am of the view that a change in the COLA could be part of a package of Social Security and retirement reforms.  Social Security reform must also encompass additional revenues and rule changes that eliminate future automatic cuts in Social Security benefits.   Pension reform must encompass improvements t0 401(k) plans and additional sources of low-cost annuity income.

Some readers might be interested in my views on the politics of the COLA debate.

http://policymemos.blogspot.com/2014/01/common-ground-on-social-security-colas.html

Impacts of shift from traditional to chained CPI on Social Security Benefits

Question: What is the potential annual impact and cumulative dollar impact of a policy change that links Social Security benefits to the chained CPI rather than the traditional CPI?

Assumptions:

Analysis presented here pertains to a single retiree who retires at age 62 with a $1,250 per month Social Security retirement benefit.

The traditional CPI grows at 2.42% per year.  The chained CPI grows at 2.09% per year.     These statistics were based on BLS data over the 1999 to 2013 time period.  Economists at the Bureau of Labor Statistics at the Department of Labor informed me that they did not have data on the chained CPI for years prior to 1999.

See the link below for statistics on the traditional and chained CPI.

http://dailymathproblem.blogspot.com/2014/02/comparing-traditional-and-chained-cpi.html

Analysis:  Information on the growth of the annual Social Security Benefits adjusted for the traditional CPI and adjusted for the chained CPI is presented for a retirement potentially spanning from age 62 to age 100 is presented in the table below.

In this table, the first column is age, the second column is the Social Security benefit adjusted by the traditional CPI, the third column is the Social Security benefit adjusted for the chained CPI, and the fourth column is the cumulative change in the Social Security benefit due to the adjustment process.

Path of Social Security Benefit With adjustment based on the traditional CPI Path of Social Security Benefits with Adjustment based on the chained CPI Reduction in Benefits for Age Due to Switch from Traditional to Chained CPI Cumulative Reduction in Benefits
$15,000 $15,000 $0 $0
$15,363 $15,313.50 $50 $50
$15,735 $15,633.55 $101 $151
$16,116 $15,960.29 $155 $306
$16,506 $16,293.86 $212 $518
$16,905 $16,634.41 $271 $788
$17,314 $16,982.06 $332 $1,120
$17,733 $17,336.99 $396 $1,516
$18,162 $17,699.33 $463 $1,979
$18,602 $18,069.25 $533 $2,512
$19,052 $18,446.90 $605 $3,117
$19,513 $18,832.44 $681 $3,797
$19,985 $19,226.03 $759 $4,557
$20,469 $19,627.86 $841 $5,398
$20,964 $20,038.08 $926 $6,324
$21,472 $20,456.88 $1,015 $7,338
$21,991 $20,884.42 $1,107 $8,445
$22,523 $21,320.91 $1,202 $9,647
$23,068 $21,766.52 $1,302 $10,949
$23,627 $22,221.44 $1,405 $12,355
$24,198 $22,685.86 $1,513 $13,867
$24,784 $23,160.00 $1,624 $15,491
$25,384 $23,644.04 $1,740 $17,231
$25,998 $24,138.20 $1,860 $19,091
$26,627 $24,642.69 $1,985 $21,075
$27,272 $25,157.72 $2,114 $23,189
$27,932 $25,683.52 $2,248 $25,437
$28,608 $26,220.31 $2,387 $27,824
$29,300 $26,768.31 $2,532 $30,356
$30,009 $27,327.77 $2,681 $33,037
$30,735 $27,898.92 $2,836 $35,873
$31,479 $28,482.01 $2,997 $38,870
$32,241 $29,077.28 $3,163 $42,033
$33,021 $29,685.00 $3,336 $45,369
$33,820 $30,305.41 $3,515 $48,884
$34,638 $30,938.79 $3,700 $52,583
$35,477 $31,585.42 $3,891 $56,475
$36,335 $32,245.55 $4,090 $60,564
$37,215 $32,919.48 $4,295 $64,859

Some observations:

  • The annual impact of the change from the traditional to chained CPI grows over time.
  •  The annual impact is $463, at age 70 $1,302 at age 80 at age 80, $2,532 at age 90, and $4,294 at age 100.
  • The cumulative impact of the change in the COLA formula is $1,979 at age 70,  $10,949 at age 80, $30,356 age 90, and $64,859 at age 100.

Some Implications:

  • The change from a traditional to chained CPI would have a very large impact both on household and national finances.
  • The fiscal impact of the change in the COLA formula would grow for 38 years until it reaches a constant rate.   (After 38 years the new COLA fully impacts all retirees based on their age.)
  • The change phases in slowly which gives people time to respond and change spending patterns.
  • The annual and cumulative impacts are largest for people near the end of their life when expenses both from increased medical needs and a need to change living arrangements are largest.

You may be interested in my policy blog on the Social Security COLA.

http://policymemos.blogspot.com/2014/01/common-ground-on-social-security-colas.html

Reconsidering Use of 401(k) Funds for Emergencies

Question:

Current tax rules governing 401(k) plans allow workers to distribute funds for hardship expenses prior to age 59 ½. The distributed funds are fully taxed at the ordinary income tax rate and subject to a 10 percent penalty.

Should rules governing 401(k) plans be changed to allow for the allocation of limited 401(k) funds to an emergency fund and new restrictions on the bulk of 401(k) contributions prior to age 59 ½?

In many states, people with assets in 401(k) plans and IRAs are not eligible for government benefits including food stamps.

Should these rules be altered so that funds not available for immediate disbursement do not impact eligibility for government benefits?

Specifics of the proposed rule changes are as follows:

  • Workers may allocate 20 percent of their 401(k) contribution or IRA contribution to an emergency fund.
  • Funds placed in the emergency fund could be distributed for emergency expenses without being subjected to income tax or a tax penalty.
  • Disbursements from all 401(k) or IRA contributions not in the emergency fund prior to the age of 59 ½ would be prohibited.
  • Future 401(k) loans would be eliminated.
  • Funds in 401(k) plans and IRAS that were not available for immediate disbursement would not be counted toward Medicaid or food stamp eligibility.

Background:  Many workers, especially those entering the workforce, must choose between establishing an emergency fund or saving for retirement through a 401(k) plan.   The financial advantages associated with 401(k) plans, both in the form of an employer match and tax savings can be substantial.   However, many workers have insufficient funds to pay rent, health bills and outstanding loans.  Some workers with limited liquidity choose to create a retirement fund and delay saving for retirement.

The IRS allows workers to withdraw money from their 401(k) plan prior to retirement but the practice is discouraged through the imposition of a 10 percent penalty.

Around 87 percent of 401(k) plans allow workers to borrow funds form their 401(k) plan.  Around 18 percent of workers in plans allowing 401(k) loans had loans outstanding.   The average outstanding loan balance for 401(k) loans at the end of 2015 was a bit lower than $8,000.

Go to the ICI web site for some information on 401(k) plans.

https://www.ici.org/policy/retirement/plan/401k/faqs_401k

Workers with 401(k) plans who leave their current position must either repay their loan or be subject to additional tax and a financial penalty.

In many states, people with 401(k) plans are not eligible for food stamps or Medicaid.

https://www.brookings.edu/wp-content/uploads/2016/07/03_increasing_saving.pdf

 

https://www.agingcare.com/articles/how-401k-accounts-and-iras-affect-medicaid-eligibility-208382.htm

These rules discourage low-income workers with variable income from making 401(k) contributions.

Discussion

Contributions to 401(k) plans, unlike contributions to traditional pensions or Social Security, are voluntary.  Some people with limited liquid assets and high debt levels are reluctant to tie up funds in a retirement account.   The early disbursement option (albeit with taxes and penalty) and the loan option can encourage some people to contribute to their 401(k) plan.

Individuals who disburse funds from 401(k) plans prior to age 59 ½ may have insufficient resources in retirement.  Moreover, individuals who take disbursements prior to age 59 ½ and pay taxes and a penalty may become worse off than individuals who never contributed to their 401(k) plan.

 The emergency fund feature of the new 401(k) would provide substantial incentives for people with limited liquidity and large debts to make 401(k) contributions.   However, the prohibition against early disbursements from the bulk of 401(k) contributions could reduce the number of people with insufficient funds in retirement.

 Under the new rules, 401(k) or IRA funds not available for immediate disbursement would not affect eligibility for any government benefits.   This change would encourage low-income people to make additional contributions to 401(K) plans and IRAs.

 

 

 

 

 

A House Equity and Mortgage Payoff Spreadsheet

A House Equity and Mortgage Payoff Spreadsheet:

Question:   A person buys a house and plans to either sell and move or pay off the mortgage in twelve years.

The person is considering taking out a 15-year or a 30-year fixed rate mortgage.

The assumptions on the home purchase, house equity growth, the cost of selling and moving, and the cost of funds for the payoff of the mortgage are presented in the table below.

Table One: Assumptions for 30-year vs 15-year FRM Comparison:

Label 30-year FRM 15-year FRM
Purchase Price of House $500,000 $500,000
Down payment percentage 0.9 0.9
Initial Loan Balance $450,000 $450,000
Mortgage Term 30 15
House appreciation rate 3.0% 3.0%
Mortgage Interest Rate 4.0% 3.3%
Years person owns house 12.00 12.00
Cost of selling and moving to a new home as % of house value 9.0% 9.0%
Tax Rate on Disbursements from 401(K) Plan 30.0% 30.0%

 

Create a spreadsheet that provides estimates of house equity after the sale and move or mortgage payoff amounts after twelve years when the house buyer uses a 30-year FRM and when the house buyer uses a 15-year FRM

Base your mortgage payoff calculation on the assumption that the source of funds for the mortgage payoff are fully taxed funds from a 401(k) plan.

Spreadsheet:

http://wp.me/a2WYXD-4i

 

 

Results:

The results for the comparison of the 15-year and 30-year FRM for the assumptions presented in table one are presented in Table 2.

Table Two: Results for the 30-year vs 15-year FRM Comparison:

 

30-year FRM 15-year FRM
House Equity after Selling and Moving Costs $318,303 $540,109
Forecasted Mortgage Payoff Amount -$472,025 -$155,160

 

Observations on the 30-year vs 15-year FRM comparison:

The person taking out the 15-year FRM mortgage has around $222,000 more in house equity at the end of the 12-year holding period.

The mortgage payoff calculation when funds are disbursed from a 401(k) plan includes tax on the disbursements.   Inclusive of the tax bill, the mortgage payoff amount is $317,000 higher for the buyer who uses the 30-year FRM than for the buyer who uses the 15-year FRM.

Other Applications for the House Equity or Mortgage Payoff Spreadsheet:

 Modify the mortgage payoff calculation to allow for a situation where funds for the mortgage payoff are obtained from three sources – (1) a savings account, (2) sales of common stock, and (3) disbursements from a 401(k) plan.   Treat tax rates as an endogenous variable in the new model.

Compare results for both mortgage types under the 90% LTV assumption to results under an 80% LTV assumption.

Run the model on 15-year and 30-year FRMs for holding periods ranging from 1 to 15 years.   How does the advantage of the 15-year FRM vary with holding period?

Authors Note:   This problem was discussed further in the post below.

Essay Nine: Retire Mortgage Debt or Accumulate in Your 401(k) Plan:

https://financememos.com/2015/10/09/essay-nine-retire-mortgage-debt-or-accumulate-in-your-401k-plan/

Essay nine points out that many financial advisors stress accumulation of wealth in 401(k) plans rather than mortgage balance reductions even when their clients are nearing retirement.  The major banks employing the same financial advisors issue mortgages and sponsor 401(k) plans.   As a result, the interests of the financial advisors and the interests of their clients are not automatically aligned.

This approach can backfire when stock markets underperform nearing retirement.

During working years. the tax code favors people with large mortgages and people who are contributing to their 401(k) plan.  However, after retirement the person who must disburse funds from a 401(k) plan often has a hefty tax bill.

 

 

Retire Mortgage Debt or Accumulate in Your 401(k) Plan?

Increasingly, many Americans nearing the end of their work life find they have a large mortgage and must choose between paying off the mortgage or contributing more funds to their 401(k) plan. A large number of financial advisors advise their clients to increase 401(k) contributions rather than pay off their mortgage.

http://www.forbes.com/sites/nancyanderson/2014/01/03/7-reasons-not-to-pay-off-your-mortgage-before-you-retire/

http://www.kiplinger.com/article/real-estate/T040-C000-S002-should-you-pay-off-your-mortgage-before-you-retire.html

http://www.washingtonpost.com/news/get-there/wp/2015/03/26/the-case-for-not-paying-off-your-mortgage-by-retirement/

http://www.foxbusiness.com/personal-finance/2014/06/18/maybe-shouldnt-pay-off-your-mortgage-before-retirement/

http://business.time.com/2013/05/28/the-new-retirement-why-you-dont-have-to-pay-off-your-mortgage/

http://online.wsj.com/ad/article/privatewealthretirement-mortgage

http://www.schwab.com/public/schwab/nn/articles/Should-You-Pay-Off-Your-Mortgage-Early-Before-You-Retire

http://www.principal.com/planningcenter/retirementplanning/nearingretirement/retirementnews/mortgagequestion.htm

My view is that it is essential for people nearing retirement to eliminate their mortgage debt even if this goal requires some reduction in 401(k) contributions.   I have two reasons for this view.   First, as noted and explained in the previous section 401(k) plans are not capable of mitigating the impact of market down turns at the end of a career or during retirement.   Intuitively, a person with no debt is much better able to withstand market downturns than a person with a mortgage.   The Wall Street analysts always say don’t sell on a panic the market will come back.   Well retirees with a large mortgage often have no choice but to sell.

Second, the financial risk considerations interact with another factor, the tax treatment of 401(k) plans. During working years mortgage interest and 401(k) contributions reduce income tax burdens.   During retirement a person with a large mortgage payment and most financial assets inside a 401(k) plan will pay more in tax than a person without a mortgage.

All disbursements from a 401(k) plan are fully taxed at the ordinary income tax rate.   A person with no mortgage disburses enough to cover discretionary expenses and taxes   A person with a mortgage must disburse enough to cover discretionary expenses, the mortgage and taxes.

The disbursement to cover the mortgage leads to additional taxes because all disbursement from the 401(k) plan is taxed. MOREOVER, THE DISBURSEMENT ON FUNDS USED TO COVER THE TAX ON THE 401(K) DISBURSEMENT IS ALSO TAXED.

Part of Social Security is taxed for people with income over a certain threshold. A quick way to find out if part of Social Security benefit is taxable is to compare your income to the threshold for your filing status — $25,000 for filing status single and $32,000 for filing status married.

http://www.irs.gov/uac/Newsroom/Are-Your-Social-Security-Benefits-Taxable

Higher disbursements from the 401(k) plan can increase your adjusted gross income beyond the threshold and increase the amount of the Social Security benefit subject to tax.   Of course any 401(k) disbursement used to pay the income tax is also taxed.

So let’s take a household with all financial assets in their 401(k) plan with a $30,000 annual mortgage.   This monthly mortgage is $2,500, not huge.   Let’s assume that the person has to pay around $1,000 more in tax on Social Security benefits because of the additional disbursement to pay down the mortgage. A first order approximation of the amount of additional tax needed because of the additional $31,000 disbursement is $31,000 x the marginal tax rate for the taxpayer.   For most filers the marginal tax rate would be around 25 percent in 2014.

So the taxpayer with the mortgage and the additional tax burden because of the additional 401(k) disbursement will probably disburse $39,000 more per year from their 401(k) plan.

This analysis puts a whole new wrinkle on the question how much money does one have to save in their 401(k) to have a secure retirement.   The answer is much more if you have not paid off your mortgage.

Note that the disbursement to cover the unpaid mortgage must occur whether the market falls or rises.

Many people who choose to add to their 401(k) plan rather than pay off their mortgage prior to retirement are going to have sell their home and downsize. Downsizing may make sense but most people don’t want to downsize until they are fairly old.

Some people who end up selling their home may choose to rent rather than buy a new home.   The main risk of choosing to rent throughout retirement is that home prices and rents may rise.   This exacerbates longevity risk.

Downsizing should be a choice not an outcome from a failed financial plan or worse the result of financial advisors putting their interests over your interests

Concluding thoughts on mortgage debt in retirement: An increasing number of households are retiring prior to their mortgage being entirely paid off. Surprisingly, the existence of a mortgage in retirement is often consistent with a financial plan developed by a financial planner.   Many financial analysts and planners advise their clients to increase savings in their 401(k) plan rather than retire their mortgage.

These financial planners are not being upfront with their clients.   Retirees with mortgage debt and all or most financial assets in a 401(k) plan are at the whim of the market and have a substantial tax obligation.   The advice that put people in this position is in my view a form of malpractice.

Appendix to Essay on Mortgage Debt and 401(k) Assets in Retirement

The issue of whether to pay off a mortgage or contribute to a 401(k) plan for an older worker is related to the issue of mortgage choice, especially for older homebuyers.   The following question addresses the interaction between mortgage choice and 401(k) investment strategy for an older worker.

Question:   A 50 year-old person is buying a house and must choose between a 15-year mortgage and a 30-year mortgage.   The mortgage choice will impact how much money the person can contribute to his or her 401(k) plan.

The person makes $80,000 per year. The initial mortgage balance is $400,000.   The person’s 401(k) balance at age 50 is $200,000. The 30-year FRM rate is 3.9 percent and the 15-year FRM rate is 3.1 percent.

Discuss the advantages and disadvantages of two strategies (1) taking the 30-year FRM and investing 15% of salary in the 401(k) plan and (2) taking the 15-year FRM and investing 5% of salary in the 401(k) plan.

Analysis:

Let’s start with a reiteration of mortgage choice issues a subject previously broached in essay four.

Observations and Thoughts on Mortgage Choice Issues:

  • The monthly payment on the 30-year FRM is nearly $900 less than the monthly payment on the 15-year FRM. The higher mortgage payment on the 15-year FRM will all else equal require the person who chooses the 15-year FRM to make a smaller 401(k) contribution than the person who chooses the 30-year FRM.
  • After 15 years the 15-year FRM is completely paid off.   The remaining loan balance on the 15-year FRM is around $257,000.
  • Note that gains from the quicker pay down on the 15-year mortgage are not dependent on market fluctuations.   The gain from debt reduction occurs regardless of whether the market is up or down and regardless of when bear or bull makers occur.
Analysis of 15-Year Versus 30-Year FRM
  30-Year 15-Year
Mortgage Interest Rate 0.039 0.031
Mortgage Term 360 180
Initial Loan Balance 400000 400000
Payment -$1,886 -$2,781
Loan Balance after 15 years $256,799 $0.00

 

Observations and Thoughts on Two 401(k) Contribution Strategies:

As noted in essay eight, the final 401(k) balance after 15 years depends on both the rate of return of the market and the sequence of the returns in the market.   Outcomes are presented for two market scenarios.   The first involves 7% returns for the entire 15-year period.   The second involves 7% returns for 10 years followed by -4% returns for 5 years.

  • The difference in the final 401(k) balances (high contribution minus low contribution) under the 15-year bull market scenario is around $211,000.
  • The difference in the final 401(k) balance (high contribution minus low contribution) under the 10-year bull and 5-year bear scenario is around $131,000
Analysis of Different 401(K) Contribution Strategies
5% Contribution Rate 15% Contribution Rate Difference
7.0% Return for 15 Years $675,442 $886,752 $211,309
7.0% Return for 10 years followed by -4.0% return for 5 years $394,339 $525,090 $130,751

The initial balance in the 401(k) plan for both scenarios is $200,000.

 

Additional insights on the tradeoff between 401(k) contributions and mortgage retirement:

 

  • The 30-year mortgage/high 401(k) contribution strategy results in major tax savings during working years compare to the 15-year mortgage/low 401(k) strategy.   All mortgage interest is tax deductible and the 401(k) contribution is not taxed.
  • The 15-year mortgage/low 401(k) contribution strategy results in major tax savings during retirement compared to the 30-year mortgage/high 401(k) strategy.   The previous example in this section demonstrated exposures for the person with mortgage debt in retirement when the person is dependent on 401(k) disbursements.   Remember all disbursements from a conventional 401(k) plan including disbursement used to pay the mortgage and disbursements used to pay taxes are fully taxed at ordinary income rates.   By contrast, the money gained from paying off the mortgage and most capital gains on owner-occupied real estate is not taxed.
  • Financial risks associated with a bull market persist through retirement as long as the saver allocates 401(k) assets into equity.

 

 

Understanding the four percent rule

Question:   Is the 4.0% rule an appropriate guideline for determining the amount of savings a retiree should spend each year?

Background on the 4.0% rule:  Under the four percent rule (as I understand it ) the retiree’s expenditure in her first year of retirement is four percent of wealth in certain accounts.  It is more difficult to apply the 4.0% rule to total household wealth because house equity, a major component of wealth is not liquid.  (The application of the 4.0% rule to total wealth including house equity would at some point require the sale of the home.)

Whether strict adherence to the 4.0% rule leads to a smooth, stable, and sustainable consumption pattern for the household depends on asset returns, inflation, and the timing of inflation and asset returns.  A sharp decrease in returns at  the beginning of retirement could lead to a relatively quick depletion of assets.  A sharp increase in returns at the beginning of retirement could allow retirees to spend more than allowed or provide a bequest to heirs.

The 4.0% rule may result in retirees too quickly depleting their resources in the current financial environment where the risk free return is lower than inflation.

http://financememos.com/2013/02/15/chasing-yield-by-investing-in-long-term-bonds/

Illustrating the 4.0% rule:  We consider four scenarios — (1) 2.00% returns and 3.0% inflation all years, (2) 4.0% returns and 3.0% inflation rate for all years, (3) a -20% return the first year followed by 2.0% returns and 3.0% inflation, and (4) -20% return the first year followed by 4.0% return and 3.0% inflation.

We calculate the number of years it would take for the retiree to deplete all assets under the four scenarios.  Results presented in Table Four indicate that years until depletion range from 19 years for scenario three to 30 years for scenario two.

Adequacy of resource for 4% rule under four scenarios
Shock Return Inflation Rate Year Balance goes to $0
None 2.00% 3.00% 23
None 4.00% 3.00% 30
-20% first year 2.00% 3.00% 19
-20% first year 4.00% 3.00% 23

I suspect that these calculations understate the financial risk associated with adherence to the 4.0% rule.   Some analysts suggest that investors who use the 4.0% rule should maintain a larger portion of their portfolio in equites.  I disagree.  The worse case scenario for the 4.0% rule involving poor stock returns in a period of inflation actually occurred during the 1970s.

Initial Balance $100,000
Disb. Rate 4.00%
Rate of return Inflation rate Real Value of initial balance Beginning of year balance Disbursement End of year balance
1 2.00% 3.00% $100,000 $100,000 $4,000 $98,000
2 2.00% 3.00% $103,000 $98,000 $4,120 $95,840
3 2.00% 3.00% $106,090 $95,840 $4,244 $93,513
4 2.00% 3.00% $109,273 $93,513 $4,371 $91,013
5 2.00% 3.00% $112,551 $91,013 $4,502 $88,331
6 2.00% 3.00% $115,927 $88,331 $4,637 $85,460
7 2.00% 3.00% $119,405 $85,460 $4,776 $82,393
8 2.00% 3.00% $122,987 $82,393 $4,919 $79,122
9 2.00% 3.00% $126,677 $79,122 $5,067 $75,637
10 2.00% 3.00% $130,477 $75,637 $5,219 $71,931
11 2.00% 3.00% $134,392 $71,931 $5,376 $67,994
12 2.00% 3.00% $138,423 $67,994 $5,537 $63,817
13 2.00% 3.00% $142,576 $63,817 $5,703 $59,390
14 2.00% 3.00% $146,853 $59,390 $5,874 $54,703
15 2.00% 3.00% $151,259 $54,703 $6,050 $49,747
16 2.00% 3.00% $155,797 $49,747 $6,232 $44,510
17 2.00% 3.00% $160,471 $44,510 $6,419 $38,982
18 2.00% 3.00% $165,285 $38,982 $6,611 $33,150
19 2.00% 3.00% $170,243 $33,150 $6,810 $27,003
20 2.00% 3.00% $175,351 $27,003 $7,014 $20,529
21 2.00% 3.00% $180,611 $20,529 $7,224 $13,715
22 2.00% 3.00% $186,029 $13,715 $7,441 $6,548
23 2.00% 3.00% $191,610 $6,548 $7,664 -$985
24 2.00% 3.00% $197,359 -$985 $7,894 -$8,899
25 2.00% 3.00% $203,279 -$8,899 $8,131 -$17,208
26 2.00% 3.00% $209,378 -$17,208 $8,375 -$25,927
27 2.00% 3.00% $215,659 -$25,927 $8,626 -$35,072
28 2.00% 3.00% $222,129 -$35,072 $8,885 -$44,659
29 2.00% 3.00% $228,793 -$44,659 $9,152 -$54,704
30 2.00% 3.00% $235,657 -$54,704 $9,426 -$65,224