Social Security (SS) today faces a massive actuarial under funding that will, left uncorrected, result in a nearly 25% cut in benefits around the year 2034. This article discusses how we got to such a problem. When the source of the problem becomes apparent, the solution will also.

The problem stems from two sources: Legacy Costs and Solutions to the 1982 SS Crisis. We will discuss what these problems are, their magnitude, their impact on the SS trust fund and who has borne their costs

Legacy Costs

To quote the SS administration (SSA), “A number of studies have used estimates of historical and projected lifetime net transfers (benefits less taxes accumulated or discounted using market-based interest rates) by birth cohort under the Old-Age and Survivors Insurance (OASI) program to calculate and compare the aggregate present-value sum of those transfers for selected birth-cohort groups. Using historical and projected OASI Trust Fund interest rates, such calculations confirm that, from a program accounting perspective, the earliest generations of program participants received positive lifetime net transfers, while later generations are projected to experience negative lifetime net transfers.” The cumulative net transfer to the earliest cohorts is referred to as a “legacy debt”. The costs incurred to service this debt have been borne by later birth cohorts. Legacy costs are basically the costs of getting the OASI program up and running so that the outflow to each birth cohort would be generally reflective of their payment.

How Big is the Legacy Debt

Without going into great detail. the previously mentioned SSA article concludes that “Although the program may have created a real legacy debt borne by later program participants in the form of a lower capital stock or below-market lifetime rates of return, it is also possible that no real legacy debt was created, or that it is substantially smaller than is often suggested.” Other studies ( have concluded that the trust fund was already out of balance by 1957 by nearly a quarter trillion dollars.

So long as the OASI program was entirely a pay as you go system it is hard to make the case that there are any legacy costs, However, from the beginning, and expanding greatly in 1982, the OASI program has been partly a prepaid program. If it were entirely a prepaid program there would be a large legacy cost.

The original costs to fund the retirement benefits (SSRB) should have been borne not just by working Americans – those with earnings that were paying the Federal Income contributions Act (FICA) taxes – but by all Americans; whether their income was earned or unearned. America as a nation was better served by eliminating the overwhelming poverty of many elderly after WWII and as such America as a nation should have contributed to the cost of establishing SSRBs for the earliest birth cohorts – not just the working class, i.e. those that contributed FICA to the SS trust fund.

To put it succinctly, the cost of establishing the OASI program should have been borne by the general fund, not the SS trust fund. Had that been the case the trust fund in 1957 might have had in the neighborhood of an additional $250 Billion as a result of transfer payments from the general fund to the SS trust fund

Impact of legacy costs on the SS Trust fund

To put the value of the legacy cost into current dollar we can use just the values of CPI-U. CPI-U in 1957 was 27.6. In 1982 it was 94.3 and at the beginning of 2018 CPI-U is 246.5. Using only the value of CPI-U to adjust the value of $250B to 2018 (1982) dollars, we find that the SS trust fund would be better off to the tune of $2.2Trillion ($850Billion).

To put those numbers in perspective, the 1982 SS crisis that resulted in the Greenspan Commission would not have happened – at least not then. Rather than having a paltry 24 Billion dollar trust fund balance in 1982, the balance would have been nearly a trillion dollars. Today’s trust fund (current balance nearly $3 Trillion) would be over $5 Trillion.

Without doing the math, it is clear that a onetime infusion of $2.2Trillion to the SS trust fund would have a significant effect on its ability to pay scheduled benefits in the future. But…, Legacy costs were not paid by the general fund and were in fact entirely borne by working class Americans that paid into the SS trust funds through FICA taxes. Unearned income (and income in excess of income caps) did not contribute anything to the establishment of the OASI program.

The 1982 Social Security Crisis

As mentioned above, the startup costs of the OASI program were not funded by general tax receipts but only by the low to moderate income working class through FICA withholding. The result was that by 1981, the trust fund had diminished to a mere 24 billion dollars. Estimates were that the trust fund would be depleted by 1983. Ronald Reagan tapped Alan Greenspan to head The National Commission on Social Security Reform (informally known as the Greenspan Commission after its Chairman) to look at the problem, and come up with solutions.

The problems were two-fold. First, there was a cash flow shortage in both the near and far term and the second was the demographics that indicated the number of working vs. retirees would continue to drop to as low as two. The end result was raising the payroll tax and shaving benefits to address the cash flow problems and to very gradually raise the retirement age on future retirees to address the worker/retiree ratio problem.

The commission claimed that these changes would ensure the solvency of the SS trust fund for 75 years and they might have but for some major changes. A. The staggering growth of income inequality meant that FICA was not collected on 90% of the national wage base as used in the baseline by the commission. B. Lifetime longevity has increased faster than the commission expected. And C. congress has expanded benefits.

The Social Security Trust fund would, over the next 25 years, generate nearly $2.5 Trillion in surplus revenue. Revenue that was “borrowed” to run the government and mask the true deficits generated by the massive tax cuts to the rich under Ronald Reagan. Basically, the rich were given tax cuts (marginal rates for wealthy individuals dropped from 70% to 28%) while the shortfall in revenue was made up, in part, by the increase in FICA withholding on those with modest incomes.

Once again, unearned income was not called upon to fund any of the OASI program and was generally given a nice tax cut while the trust fund made up the shortfall. In addition, the commission gave short shrift to the worker/retiree ratio problem.

Problems with the Greenspan Solution

There were several problems with the solutions adopted as a result of the Greenspan commission. The most glaring (IMO) was the inadequacy of the changes to Full Retirement Age (FRA) to address the worker/retiree ratio problem. The second was the failure to index the income wage cap on FICA taxes to ensure the collection of 90% of the national wage base.

However, I think the biggest problem arose because of the increase in FICA taxes that generated the huge surpluses. Generating excess revenue through the increased FICA tax rates changed the OASI program from a pay as you go program to, at least partially, a prepaid retirement savings account. The average American, for the first time felt, probably not incorrectly, that he was actually saving for retirement through his FICA contributions. People began to count on SSRB for their retirement and treated their expected SSRB as a major part of their retirement plan.

Where We Are Today

Fast forward to today. The trust fund now has nearly $3Trillion. Due to the growth of income inequality and extended lifetimes the trust fund is expected to be depleted by 2034 (a “mere” 52 years after the Greenspan commission). Today, many in congress want to severely cut back on benefits. Congress continues to cut general revenue taxes, Increase spending and now wants to blame “entitlements” like OASI program for the huge deficits.

Note and note well (N.B.) To date the OASI program has not contributed one penny to the national debt. The only reason to try to cut back on SSRBs is so the general budget no longer need pay back the borrowed monies from the SS trust fund – to date nearly 3 trillion dollars.

So, have we seen the solution?

Clearly three huge problems that undercut the actuarial balance of the SS trust fund are: A. the worker/retiree ratio, B. legacy costs, and C. improperly indexed salary caps. Clearly correcting these three problems can have a huge impact on the actuarial lifetime balance of the SS trust funds.

First, raise the FRA. The FRA is in the middle of slowly increasing from 65 to 67. There is nothing wrong with allowing the FRA to continue to rise until the worker/retiree ratio reaches some appropriate and sustainable ratio. The FRA can then be indexed to maintain an appropriate worker/retiree ratio.

Interestingly, as the ratio increases, the actual tax rate contributed to FICA can decrease – that is basically reversing some of the income transfer (from young to old) that funds the OASI program.

Increasing the FRA is a serious cut in benefits; one that might be long overdue. Current retirees at age 66 can expect to have nearly two decades of retirement. That is far more than was originally planned and given the demographics of society a lot more than society can handle. Raising the age to 69 and indexing it for longevity could account for more than a third of the shortfall.

Second, correct the improper indexing of the salary cap. Removing the cap entirely would correct nearly 76% of the imbalance. Raising the cap to capture FICA on 90% of earned income would eliminate nearly a third of the problem.

Third, begin to repay some of the legacy costs that the SS trust fund has been solely responsible for. As earned income below salary caps has already been taxed for this cost, it should be borne entirely by higher earned incomes and unearned income.

There is no reason why there should be such a disparity between earned and unearned income tax rates. There is nothing wrong with letting the hard working American keep more of the profits from his labors, but when Messieurs Reagan, Bush and Trump professed this belief they clearly had no intent on allowing that to happen. Rather they cut taxes for unearned incomes and high incomes that were not subject to FICA while either raising or neglecting FICA taxes – those that are significant to the hard working American.


Social security is basically a social contract that guarantees income to the elderly at the expense of younger people. Beginning in 1982, the social contract changed. The government collected taxes far in excess of the needs of the trust fund. This allowed the government to cut taxes elsewhere –specifically Reagan funded a large tax cut for the wealthy – incurring a self-funded debt to the trust fund. It is time for the rates on the wealthy and specifically on unearned income to rise to repay the trust fund as it is just now beginning to reach its maximum and will begin to drain resources from (rather than contributing to) the general budget.

Rates on income not subject to FICA should increase again, not just to pay back the trust fund for borrowed money but also to repay legacy costs.

FRA and Salary caps should be properly indexed to maintain appropriate ratios.

Maintaining proper ratios for Salary caps and FRA can solve nearly 70% of the OASI solvency problem. Transfers from general revenue, raised by taxing income not subject to FICA, to OASI to cover the legacy costs – costs that should have been paid long ago – can cover the remaining shortfall without severely impacting tax rates or wealth accumulation.

The retirement problem in the US is not so much one that people have not planned or saved (though in fact they have probably not done enough of either) but that they have been led to believe that they were in fact saving for their retirement through FICA. It is the government’s unwillingness, not inability, to fund these promises that are the major cause of retirement insecurities in the country.

SS        Social Security

SSA     SS Administration

OASI   Old Age and Survivors Insurance

SSRB   SS retirement Benefits

FRA     Full Retirement Age


Across the Grand Canyon in Two Years, Three Months, 4 Days and 6 Hours

Shawn P. Heneghan

In June of 1996, my wife Kathy and I started what turned out to be a two year adventure to cross the Grand Canyon. The following is a description of the hike, with details explaining how it can take more than two full years.

Our western canyon hiking adventure (a.k.a. Pro Fun Tour ’96) began in earnest at the top of the south rim of Grand Canyon National Park on the morning of June 6, 1996. The south rim, at an altitude of about 7200 ft., is about 5000 ft. above the Colorado River. There are two maintained trails down to the river from the south rim, the Bright Angel Trail and the South Kaibab Trail. While the South Kaibab is shorter (6 miles compared to 9 miles), the Bright Angel is by far the easier hike due to the presence of water, and a rest stop halfway down. Our plan is to hike Bright Angel down about 3000 feet to Indian Gardens campground, and finish the downward trip to Phantom Ranch the following morning.

We start at 8 AM, because the forecast is for hot and sunny weather. In the Grand Canyon, when they say hot, they mean HOT!. It is not unusual for the high temperature at the bottom to exceed 120F, despite the pleasant 60F at the beginning. There are four water stops along Bright Angel Trail from the rim to Indian Gardens. The source of all this water is actually a spring/water fall on the north rim. In fact all of the water used by the facilities on both rims comes from this one spring on the north rim.

Fortunately, the piping for the water to the south rim parallels this portion of the Bright Angel Trail, and the Park Service has supplied these water stops. In the heat, sun and low humidity these rest areas are lifesavers. We find out later how literal this is, as four scouts and a scoutmaster actually meet their demise in the area of the South Kaibab Trail at the very same time that we are on the Bright Angel.

By the time we get to Indian Gardens Campground at about noon, the temperature is already over 100F and still rising. We sit in the shade of a cottonwood, continually soak ourselves with water, rest, and await the setting of the sun. At sunset we walk out to a viewpoint to take in the sunset over the Grand Canyon from about 1500 feet above the waters of the Colorado. We spend the early evening watching deer come to the springs for food and water.

Arising early the following morning, we continue deep into the inner gorge, arriving at Phantom Ranch at about 10:30 AM. The expected high temperature is 125F. The heat here is not intolerable however due to Glen Canyon Dam some 110 miles upstream. What has that to do with the seemingly unbearable heat here at Phantom ranch? The water running out of Glen Canyon Dam is a pretty constant 46F. Here at Phantom Ranch, the Colorado River runs a cold 48 degrees. Sitting along the river with your legs in the water it is easy to stay cool.

For the next 5 days, we will not stray far from the river as we meet the guides from Diamond River Rafting and head downstream through the lower half of the Grand Canyon along the Colorado River. Rafting in the Grand Canyon is a unique adventure. The trip we have chosen is a 5 day motorized trip through the lower portion. This portion of the canyon includes some of the more famous rapids including Hermit Falls and Lava Falls. Nights are spent on riverside beaches, and the river guides prepare three meals a day. Each day is spent both rafting the rapids and hiking up the myriad of canyons that open into the Colorado. The highlight of these hikes is certainly the beautiful turquoise waters of Havasu Creek.

The first part of Grand Canyon adventure ends at the Havasupai Indian Reservation where we are bused out to meet a charter flight back the North Rim Grand Canyon Airport.

Jumping ahead to the end of Pro Fun tour 98, we join my brother and his wife at Diamond River Rafting in Page, AZ. We start down the Colorado in the upper half of the Grand Canyon; a four day trip to Phantom Ranch. This upper portion contains most of the significant Anasazi ruins. One place in particular contains the more than 40 original home sites where shards of ancient pottery are literally everywhere. Vasey’s Paradise and Redwall Cavern are also on this section of the river.

Arriving at the beach at Phantom Ranch on the morning of 9 September 98 we have finished rafting all of the Grand Canyon. It now remains to only finish the Trans-Canyon hike that we started two plus years ago.

The North Rim sits on the Kaibab Plateau at 8200 feet, just about 6000 feet above the Colorado River and 14 miles away via the North Kaibab Trail. Warning signs are plentiful: warning us that we should not attempt to hike out of the canyon from here unless we start before 6AM. Of course we have no plans to hike the entire trail in one day, but rather plan to spend the night at Cottonwood Camp about 6 miles upstream.

We walk for just about four hours to reach Cottonwood Camp. After rousting a sleeping rattler from our designated campsite, we spend the afternoon, relaxing in the shade of the cottonwoods and bathing in the clear waters of Bright Angel Creek. The waters of the Colorado have been running red due to thunderstorms and flash flooding in the side canyons throughout our rafting trip, so the clear water is well received.

Arising at 6AM for our final climb, and one last good drenching from a thunderstorm, we proceed up the North Rim. The North Kaibab Trail passes two impressive waterfalls, one of which (roaring springs) is the water source mentioned earlier. The pump station nearby is one of only two water stops along the last six miles and nearly 5000 feet of hiking. Today, the clouds have arrived early and the temperature is not more than 90F.

This last hike is the most grueling portion of the trans-canyon hike. For more than six hours we climb on the cliff faces, continually tramping upwards. We eat through cartons of cereal, loaves of bread, and hoards of candy bars, but still we are hungry and not yet at the top. The effort required to hike out of the canyon is staggering, but in just over eight hours we arrive at the North Rim.

And so,. the adventure ends. We have finally completed our trans-canyon hike; a mere two years, three months, 4 days and 6 hours after it began. I think this may be a record for the longest time required to cross the canyon. This summer we hope to try again, finishing in about 6 days.

The Canyon is a special place. The views are phenomenal, but the hiking is not for the faint of heart or weak of leg. Rafting, while significantly easier, is by no means easy. For those who cannot or will not hike into or out of the canyon there are two alternatives. You can ride a mule (a not altogether pleasing prospect), or you can raft the entire 200 plus miles of the Colorado (an eight day ordeal). The last possibility of course is to skip it entirely, but then you will miss one of the natural wonders of the world.




Figure 1: Map of Main Trails at Grand Canyon National Park

Table 1. Itinerary for Trans Canyon Hike

Date Place Altitude Activity
6 June 96 South Rim 7200 Begin Trans-Canyon Hike
Indian Gardens 4400 Camp
7 June 96 Phantom Ranch 2200 Begin Rafting Trip
9 Sept. 98 Phantom Ranch 2200 End Rafting Trip, Start hike up
Cottonwood Camp 4000 Camp
10 Sept. 98 North rim 8200 End Trans-Canyon Hike



Table 2. Grand Canyon Facts and Figures

Distance Travel mode Time required
North Rim to South Rim 10 miles dragonfly 2 hours
200 miles Auto 5 hours
23 miles Hike 2-4 days
Colorado River 277 Miles Raft 6-10 days
Average Temperature Rim Summer 50-80
Phantom Ranch Summer 80-110
Park Size 1900 square miles
Canyon Depth Maximum 6000 feet
Canyon Depth at Grand Canyon Village 5000 feet
Visitors 5 million/year


Retirement Planning


Many authors (Michael Kitces and Wade Pfau among others) consider retirement planning to be among the most difficult problems to solve. Their reasons are many but mainly it has to do with the myriad of unknown (and some essentially unknowable) quantities. Among these are lifestyle and the associated costs, health and associated health care costs, life expectancy, inflation, safe withdrawal rates, and future rates of return. These variables are all just related to the costs of retirement and have not addressed the “plan” that was followed in order to get to retirement — that is, how and when was the funding of that retirement gathered: after tax vs. pre-tax, risk analysis, and return on investments.

What I intend to do here is to assemble some of my observations that will, hopefully, help the reader to understand some of the problems, and show some assumptions that will ameliorate some of the difficulties. The ultimate goal is to allow you, the reader, to set your own goals and enjoy a nice retirement without having to depend on a community of financial planners (FPs) for whom I have more than a bit of disdain.

We will start by establishing the primary retirement goals and establish some planning guidelines/assumptions (lifetimes, returns on investments, inflation, the 4% rule). Then, look at how much income will be needed in retirement. The amount of income needed in retirement is intimately related to how much you make today and how much you save today. Once you know how much you plan to spend in retirement, then we can establish how that income will come to be from various sources: social security, pensions, savings, work etc. And, of course that income in retirement must exist after taxes – and so we will discuss taxes while working and in retirement.

One of the major problems associated with retirement – especially early retirement is how to fund health care. This almost intractable problem has almost innumerable variables: costs, personal considerations, government programs, and employer options. I have really given short shrift to health care. The reason is I don’t understand it – haven’t had to deal with it, and it is changing so rapidly that I cannot (nee will not) attempt to keep up.

It is my belief that retirement planning, and actually retiring is significantly easier than FPs would have us believe. I base this belief on the fact that 10,000 baby boomers a day are currently retiring. According to essentially all of the literature, these people are unprepared for retirement. The fact that they do it doesn’t seem to bother the FP community at large. They continue to produce articles, books, videos, and interviews that espouse their view that these people are unprepared – and yet boomers continue to retire.

My feeling, one that I hope to present here, is that the outrageous savings demands that are put forth by the FP community are arbitrary and well beyond the means or the needs of most average people.

Financial planners

I mentioned that I have disdain for the whole community of FPs. Why? Mostly because they have sold the American public a lie: specifically, that they need to save huge amounts of money to fund their retirement. The basis for these claims lies in the assumption that you will need 80-100% of your pre-retirement income in order to retire comfortably.

The 80-100% rule is absurd.

If you think carefully about what this means you discover the source of this rule and its fallacy.

First lets assume that we will work with a dual income couple, filing “Married Filing Jointly (MFJ), paying FICA (7.65%), State (5%) and local (2.5%) taxes and using 2017 tax rates.

A couple making 75K each might face a total tax bill as high as 47K (MFJ – non itemizing): 11.5K FICA, 4K local, 7.5K state, and 24.5K to the feds. This leaves them with 70% of their salary after taxes. The assumption from here is that you need the same income, after taxes, in retirement that you had before retiring. Therefore you need 80-100% of your pre-retirement gross income (GI) in retirement. It is simple, easy to understand — and wrong.

Here is the basic problem – If while you are working you spend everything you earn after taxes, you sure don’t need a financial manager. Any kind of retirement plan, or even a simple savings account is beyond your means. If you spend it you don’t save it.

If you do save it – say 15% for this couple (this is just an initial guess) in a 401k, then they might be spending only 55% of their 150K income. In addition, a couple making that kind of income may well be saving in after tax accounts an additional 10-15% of their income.

There are other factors that can decrease this percentage even further: mortgages and college loans are a couple of the biggest corrections — assuming that the prudent couple has planned so that these things disappear at or near the time of retirement. We’ll discuss some of these matters further into this document but it is easy to see that retirement planning for this couple may revolve around replacing as little as 40% of their income; a far cry from 80-100%.

Taxes will have an impact on your retirement income also, but… Retirement income is generally treated significantly differently – read much better — than working income. We’ll discuss the differences below.

The bottom line is that retirement may not cost near what the FP community has led us to believe. Why? I believe Darrow Kirkpatrick put it best. “Financial advisers can have serious conflicts of interest when it comes to computing your retirement “number.” Unfortunately, the profession has strong incentives for being overly conservative, and lots of cover for doing so. For starters, the most prevalent compensation model for advisers is still a percent of assets under management. And guess what happens when you retire? Yes, you start spending those assets that your adviser is collecting fees from, and their income goes down. Secondly, liability for the adviser is more clearly connected to you running out of money than you having too much money in retirement. There is little risk to an adviser in you working five or 10 years longer and dying with a few extra million on hand. (And your kids will love them for it.) However, if you run low during retirement due to professional advice you received, there will be hell, and possibly lawyers, to pay.”

Primary Retirement Goal

The primary goal of any retirement plan should contain the following two ideas.

  1. (PRGA) It should have a probability of success greater than 95%; preferably ~99%.
  2. (PRGB) It should have sufficient income in retirement to maintain the lifestyle that you enjoyed prior to retirement.

These two seemingly simple statements (PRGA and PRGB) have many implications that we will discuss further below. As to PRGB, there is nothing that prevents a retiree from wanting to improve his lifestyle while in retirement vs. working; having more free time can allow for a lot of new or expanded (expensive) activities. It is just meant to set a floor to your retirement – if you do not have sufficient income to maintain the lifestyle you are used to, you are likely to be very unhappy in retirement.

Planning Horizon

One thing that everyone should understand is that there is a distinct possibility that you will live to 90 plus years old. That’s right, male or female there is a 5-15% probability that you will be alive at 95. While those probabilities may seem high, the lifetime of a couple (when the second death will occur) is astonishing. There is a 25% chance that one of a couple of 65 year-olds will live to be 95.

It is statistics like this with which you must deal in order to satisfy PRGA. If you only consider an average lifespan you are basically planning to fail half the time. In order to guarantee PRGA (>95% chance of success) you must plan for a lifetime that can occur with a frequency of as low as 5%. To wit, plan on a 30-35-year retirement. Anything less and there is a significant chance that one of you (for a couple) will be living in poverty for the last few years of life.

If you have a 5% chance of financial ruin in year 30 of retirement, and a 5% chance of living 30 years in retirement, then you will only fail when both remote possibilities come true. If you have planned for each of these your plan will be successful 99+ % of the time.

To get an accurate estimate of your own personal planning horizon I suggest . This tool will personalize the lifetime probabilities for your situation. It will (likely) clearly show that to guarantee PRGA you should be planning on a long lifetime.

Inflation and Rates of Return

Two of the variables that many planners throw around unnecessarily are inflation and rates of return. At first glance they each seem to be important, unrelated, and highly variable. However, I have done the spreadsheet analysis of various scenarios using fixed and variable inflation rates and discovered an interesting fact. The inflation rate is not an important variable if you use real rates of return. A real rate of return (RR) is a rate of return above inflation. Basically ignoring inflation and working in present day constant dollars. It is simple to understand and produces accurate savings growth predictions. Predicted future dollars have essentially the same purchasing power as today.

If you are invested in stocks or fixed return bonds the RR can vary as inflation increases or decreases or as the market goes up or down. However, Treasury Inflation Protected Securities (TIPS) yield a RR. Currently that return is about 1% for 30-year bonds.

For planning purposes I use a 2% RR for savings before retirement. This should be an easily achievable real ROI while working. There is no serious need to be overly conservative while saving for retirement as you not only have a long time horizon for savings, but you still possess a wonderful resource: YOU. If you are saving for retirement, then you are working. Earning a salary etc. You have your job and skill set as a resource to help get over market downturns.

However, PRGA dictates retirement income should be guaranteed for a long time and you have given up yourself as a major resource to fund your lifestyle. To this end, I always use RR of guaranteed instruments and currently that rate for 30 year TIPS is 1%. At a minimum, this 1% RR can be used as a goalpost – something that you may beat, but do not want to miss. A gauge if you will that can be used to measure your returns against.

Clearly investing in equity markets have (in the past), can (in the present), and probably will (in the future) produce better returns. There is nothing wrong with making a better RR in your retirement account than 1%, I just don’t feel that a prudent plan should include these outsized gains. I personally knew several people that retired in 2000-2001 after the market surges of the late 90s on the assumption that they would continue to beat the then current TIPS returns of 2-3%. They didn’t and their retirement dreams went up in smoke.

Sequence of Return Risk

One of the biggest risks, IMO, in retirement is Sequence of Returns (SOR) risk. SOR problems come from average up markets that might experience a few down years. The earlier in your retirement these down years occur – the more devastating the effect on your retirement plan.

As mentioned above, many people retired in 2000 – and watched the market tank for a few years. It has since come back. We will use these people as an example for our study of SOR risk.

The SP500 in Jan 2000 stood at 1450. In January 2017 it stood at 2700. That is an average gain (not including dividends) of 3.7%/yr. CPI-U over that period has risen from 169 to 243 – an average increase of 2.1%/yr. Clearly, the average return of the markets has beaten inflation by more than 1%. And yet, people that were in the markets and retired in 2000 were most disappointed.

What happened? The markets dropped from 1450 to 850 in January of 2003. Assuming you started with 100K and withdrew 5K/yr. (inflation adjusted) to fund your retirement, you watched your savings balance drop from 100K to 47K. Your 5K withdrawals represent a much larger percentage of capital withdrawal as the markets dropped.

The markets began to recover – reaching 1420 in January of 2007 (nearly back to the original 1450) but your balance recovered to only 52K. The next sell off nearly annihilated your savings. In January 2009 its balance was so low. 22K, that it would not recover. By 2017 your plan would fail. These trends are listed in Table 2.

Table 2. Savings balance invested in the SP500 using a $5000/yr. (inflation adjusted) withdrawal (dividends not considered)


SP500 Date cpi withdrawal Balance
1450 Jan-00 168.8 100000
1160 Jan-01 175.1 4500 75500
1150 Jan-02 177.1 4668 70181
850 Jan-03 181.7 4721 47152
1120 Jan-04 185.2 4844 57286
1180 Jan-05 190.7 4937 55417
1300 Jan-06 198.3 5084 55969
1420 Jan-07 202.4 5286 55849
1322 Jan-08 211.1 5396 46599
800 Jan-09 211.1 5628 22571
1170 Jan-10 216.7 5629 27382
1325 Jan-11 220.2 5777 25233
1410 Jan-12 226.7 5871 20980
1570 Jan-13 230.3 6042 17319
1872 Jan-14 233.9 6139 14511
2070 Jan-15 233.7 6236 9810
2060 Jan-16 236.9 6230 3533
2700 Jan-17 242.8 6316 -1686

As a point of comparison, TIPS could be purchased in 2000 that guaranteed a RR of 1.6% (this was easily doable in 2000). After 17 years withdrawing identical amounts and earning inflation plus 1.6%, an initial 100K investment would have a balance of 66K.

This, in a nutshell, is SOR risk. Sure, the markets have beaten 1%RR, but the volatility, and specifically the early years of low RR, reduced your savings significantly. You could never recover sufficiently to fund the remainder of your retirement. Avoiding SOR risk – or at least investing so that it will not ruin your plan is critical.


Top Down Planning – How much do you Spend

When it comes to the question of “How much do I need in retirement?” it is tempting to fall back on easy answers or “expert” opinions.” See above for my feelings about FPs – the “experts” and their quality estimates. And yet, this is probably the most critical factor to determine in retirement planning. How much you need in retirement impacts how much you need to save while working, what kind of return you need to achieve and how you will invest in retirement.

It makes perfect sense to me that the day after I quit working essentially the same bills will show up in the mailbox, I will still want to go on my summer vacation, take the wife out to eat on occasion, go to the theatre, and pay the plumber when necessary. Life won’t change all that fast so what I was spending last year while working is what I will want to spend next year while retired. So, to meet PRGB you should want to ensure that your after tax retirement income (ATRI) meets or exceeds your current SPENDING (Capitalized because this is a critical number in retirement calculations. It will occur frequently in this discussion).

I am a fan of simplicity. I think the easiest way to answer the question “What is SPENDING?” is to take a top down approach. That is, start with your income (a figure with which most people are familiar), and then subtract known amounts that do not contribute to your daily living activities. The idea is to determine how much you spend for the lifestyle you are living today. You start by subtracting large known amounts (taxes and savings) and fine-tune this estimate by subtracting other costs that will not continue into retirement (e.g. a home mortgage). I would not recommend trying to put too fine a point on the calculation.

The advantages to this top down approach are: A. The numbers are known and B. the estimate is actually a maximum. This means that the calculation can be done accurately (and simply), and the answer will overestimate how much you need (underestimating SPENDING can be disastrous).

Taxes and Retirement Savings

Taxes (FICA, Federal, local and state) and savings (401k, IRA, after tax) are probably the largest expenditures that do not contribute to your lifestyle. Subtracting these generally well known amounts from GI gives a zeroth order approximation to answer the question “How much do I spend now?”. The reality of the situation is you only spend what is left from your paycheck. You don’t spend the money you send to the governor or the IRS. You don’t spend the money you put into your retirement accounts or even into your after tax investment accounts. Therefore, what is left, after subtracting these reasonably easily knowable amounts, is what you spend while working – aka SPENDING.

Generally speaking (for anyone actually saving for retirement) this estimate will be 60% of GI or less. I have run this analysis for several people and have found SPENDING to be in the range of 45 -55% of GI.

Mortgages and other costs

Something I refer to as imputed income (or imputed savings) is not really income, but reflective of something that you no longer need to purchase. The idea is that if you did purchase it you would need ATRI to do so. The most significant item that qualifies – and is surely in nearly everyone’s retirement plan is housing. Yes, in retirement you need housing. Many folks have paid off their mortgage and own their primary residence outright and some even own a vacation home.

Clearly you can get by on less income if you don’t have to pay for housing. With a national average home price of nearly 200K and interest rates at 5% this is 10 – 12 thousand dollars that was needed in yearly income while working and buying a home that is not needed in retirement. Yes there are any variables here: tax rates, interest rate, balance, location, housing choice, opportunity cost – what you would get in an investment account rather than investing in housing etc. The bottom line is that buying a home while working is either an investment that pays off (as imputed income) in retirement or it is a cost of living while working that does not carry over into retirement (imputed savings).

It is, I believe, easier to consider the mortgage you pay while working (if it is planned to disappear in retirement) as a deduction to SPENDING. That is, it is a cost while working that will not continue into retirement. Then, in keeping with PRGB, this amount is not needed in ATRI. Accounting for mortgage payments can easily reduce the fraction of GI that is actual SPENDING by an additional 10-20%.

There may be other costs that do not continue into retirement that can also reduce SPENDING: Union Dues, Commuting Costs, Student Loans, and College Tuition. I do not suggest that the calculation of SPENDING be finessed too much. However, some of these other costs can be significant. Many people are considering retirement even while they are paying tuition for their children; some are even still paying off student loans that are significant drains on cash flow. Commuting costs can be real, but if you are not willing to decrease the number of automobiles that you own, insure, maintain and operate there is probably little to be gained there.

The bottom line on SPENDING is that a simple top down calculation can easily show that you may be happy in retirement with as little as a third of your pre-retirement GI. Peter Hodes (a friend) did this calculation and, in fact, determined that he only needed 35% of his GI. Doubting this, he did a bottom up calculation (a much more difficult calculation to perform accurately). Having kept reasonably meticulous records for several years he was able to determine that the top down calculation was very close to his bottom up calculation.


As mentioned earlier, healthcare is clearly the most complicated and indecipherable part of an early retirement plan. Once you have reached age 66, the costs can be determined (Medicare parts B and D or Part C and possibly a medigap policy). But before age 65 – who knows. Who paid how much before retirement and who will pay how much after retirement?

I have basically punted on healthcare in early retirement. Courtesy of the US Governments involvement in vietNam in the late 60s and early 70s, and a conscription that was the law of the land, I am eligible for VA care. This provided health care for my early retirement.

You will have to determine how much insurance you want, what the cost will be, how much you are paying now, and then fit the difference into your plan. The bottom line is that healthcare can increase SPENDING by 10-12K/year until you are eligible for medicare.

After eligibility, costs do not disappear. Part B is currently 134/month. Part C supplement is 45/month. For a couple, that works out to about 4000 a year. There may also be copays and deductibles. Of course there might have been some costs before retirement including insurance payments, copays and deductibles. There may be little change in cost or costs could go up several thousand.

As a first order cut add to SPENDING $4,000 less what you spend for insurance and copays today.

Well, hopefully, now you understand why I gave up.


Now comes the really interesting observation. The more you save (whether in retirement accounts, imputed savings, or after tax accounts) the less you spend. The less you spend, the less you need to fund in retirement. So as you step up your savings rate you hone in on your retirement goal in two ways. 1. You are increasing your savings faster and 2. You are decreasing the amount needed in order to fund your retirement because you are decreasing SPENDING.

Therefore if you are setting up a plan, be sure to have a feed back mechanism that allows your goal to change based on how much you save. This is one of the major mistakes many people make. They assume that they spend 70% of their income today and need to replace that with ATRI. To save a sufficient amount to replace that 70% after taxes, FPs assume they need to have income from retirement accounts that replaces about 85% of GI.

Without going into details the savings plan to achieve that 85% goal may require that the couple save 15-20% of their salary. But, if they are saving 20% of their salary, they need replace only 50% of their salary. This would result in a lower savings goal and obviously a lower savings rate. Feedback is important.

Conclusion: Establishing Your Retirement Goal

The first step in retirement planning is to determine what you are spending to maintain the lifestyle you are living today. Using a top down approach can calculate this number reasonably accurately, giving an upper estimate. Knowing what I have called SPENDING is crucial to establishing a savings plan and making retirement decisions (see PRGB). Feedback is important as SPENDING decreases as savings increases.

It is highly likely that you spend significantly less than your GI. Fractional spending might be on the order of a third to a half of your GI. This makes saving for retirement significantly easier than the FP community declares. Anecdotally, I and my wife Kathleen are living very happily with a better lifestyle, and have been for 18 years, on 40% of what we earned while working.

Finally, once you have a good understanding of SPENDING then you have established your goal. Specifically, you will want to ensure that ATRI is equal to or greater than SPENDING. That is what guarantees PRGB.

Funding Retirement

Now that you have established a goal (or at least the mechanism for determining the goal) for retirement income it is possible to begin understanding how to achieve the necessary savings that will fund that goal. There are several major sources of “income” that can be used to fund a retirement: Social Security (SS), retirement savings, after tax savings, pensions (not considered here) and lifestyle changes (imputed income).

One of the most important for many people will be Social Security. SS income is tax advantaged, guaranteed, and inflation indexed. It is hard to beat that combination. So, it is important to understand how much to expect from SS.

The second source of income comes from savings – whether from retirement accounts or after tax, Roth or traditional IRA, 401k etc. Many people consider that their savings must make up for the total shortfall between SPENDING and SS retirement benefits (SSRB).

I personally like to consider a third source: Imputed income from lifestyle savings.

Social Security

Social Security is in reality an annuity program that provides inflation-indexed payments to the beneficiary – and often it is a two-life policy, supplying benefits to the surviving spouse as well as possible benefits to current spouses, ex spouses and children. At a minimum SS payments are at least half what your spouse gets while he (or she) is alive and increases to the full amount for the surviving spouse.

Estimating your SS benefit

Simply, your SS benefit depends on two factors: Your Average Indexed Monthly Earnings (AIME) and your age when you claim your SS benefit relative to your full retirement age FRA).

For those that are interested, I give reference to and explanation of how to calculate AIME and your primary insurance amount (PIA). For many it may be best to use the calculator at . However, I have been told that this is not always straightforward when estimating SSRB in the future for an early retirement today. Regardless, obtaining an accurate estimate is critical as SSRB is likely to be half or more of your retirement income.

I think it is important to understand the mechanics of the SSRB calculation as working even just one additional quarter may have an out-sized effect on your PIA, See, Knowing whether one more year of work will replace a “zero” in your highest 35 years, or whether the increase in PIA will be 15% of the increased AIME or 32% of the increased AIME may play a part in your decision to work/retire.

Your AIME is a complicated formula that indexes your yearly incomes using the national average wage index (NAWI) to determine your indexed yearly income. The NAWI is the ratio of the national average wage to the average wage for the year the income was earned. SS can easily supply a list of the NAWI used for a given retirement year at .

Once the NAWI has been determined, then your yearly income is multiplied by the NAWI and the 35 highest years are used to calculate your AIME (sum of the highest 35 years divided by 420). Then from AIME, SS determines your primary insurance amount (PIA).

For an individual who first becomes eligible for old-age insurance benefits or disability insurance benefits in 2018, his PIA will be the sum of:

(a) 90 percent of the first $895 (firsts bend point) of his average indexed monthly earnings, plus

(b) 32 percent of his average indexed monthly earnings over $895 and through $5,397 (second bend point), plus

(c) 15 percent of his/her average indexed monthly earnings over $5,397 (up to the maximum AIME subject to SS benefit withholding.

The bend points change yearly.

Then, finally, The PIA is adjusted lower for retiring before FRA and increased for delaying beyond FRA. If you claim SSRB at age 62, you receive something less than 75% of your Full Retirement Age (FRA) benefit. Waiting until age 70 supplies you with nearly 32% more than FRA benefits. The exact numbers are currently in flux as the FRA is changing from age 66 to age 67.

The current values for bend points and FRA and total description is available at .


When to claim SSRB

First, note that there is a distinct possibility that SSRB will be curtailed around year 2033. This can have a significant impact on the calculations, especially for people younger than 62. I am not going to consider the myriad possibilities of varying FRA and possible declines in SS payments in 2033 or 2034 etc. However, for those that have an FRA of 66, Wade Pfau has a nice description of the “return” expected by delaying SS claiming decision in Forbes –

In the referenced article Mr. Pfau clearly shows that the return on waiting starts off negative (if you die between age 62 and 70 it is negative 100%) but by age 80 begins to show a positive real return that increases with longevity. In keeping with PRGA, you should consider that you will live to age 90-95 and get a real return from SS for delaying your claiming decision until age 70 of nearly 6%.

It should be easy enough to replicate Mr. Pfau’s work with your FRA, and different SS curtailment scenarios as the situation becomes more clear.

Delaying receipt of SSRB can easily be shown to be a good “investment”. It is a very cheap form of longevity insurance. Where would you be able to get a guaranteed RR of 5-6% today? Note well, this discussion has so far only discussed the dollar value of delaying claiming. There are other considerations than just dollars. And, You do need to have funds available to spend to delay 5-8 years.

I delayed claiming for a couple of reasons: A. I was eligible to claim a spousal benefit at age 66. B. Delaying my SS retirement benefit until age 70 will garner my spouse a significant increase in her SS benefit if, as is likely, she outlives me. I wish to make it as easy for her as possible.

It is possible that neither of my reasons is applicable to your situation. Also, you must decide if longevity insurance is something you need. If you will have sufficient annuity and SSRB available at age 90 given you claim at 62 then consider claiming early.

That leaves only the question of whether you want to claim early and have a lot of flexibility (vacationing/investing) or get a nice guaranteed return and claim later.

There is no question that delaying benefits is a monetary winner given the recipient lives until age 85 or so (and of course maintaining PRGA requires you at least plan to live beyond 85). But, there are other considerations than just maximizing the bucks.

Among those other considerations are: Your health and the likelihood of a long life. Will your spouse outlive you and need extra income. Are you interested in maintaining a significant legacy? Or maybe you just want to buy a boat.

How much income does SS replace

The percentage of GI that SS replaces varies from more than 100% for very low single income married couples to about 30% for a single individual that earns near the SS salary cap ($120,000 per year). The replacement percentage continues to drop for higher income workers as income above the salary cap is neither taxed by SS nor used in the AIME calculation.

For average married two income families with a total household income of $55,000/yr. that are kind of the rule in America these days, the replacement percentage is around 55% of GI. Table 1 shows the expected SS income at FRA for various households’ incomes and the income replacement percentage versus GI or GI after taxes.

Table 1. Income, SSRB, and SS replacement percentages for various incomes

income social security SS replacement of GI
his hers total after tax total his hers total total after tax
10,000 10,000 9,235 9,000 4,500 13,500 135% 146%
20,000 20,000 18,470 12,560 6,280 18,839 94% 102%
20,000 10,000 30,000 26,735 12,560 9,000 21,560 72% 81%
20,000 20,000 40,000 34,878 12,560 12,560 25,119 63% 72%
40,000 40,000 34,878 18,960 9,480 28,439 71% 82%
40,000 20,000 60,000 50,348 18,960 12,560 31,519 53% 63%
60,000 60,000 50,348 25,360 12,680 38,039 63% 76%
60,000 30,000 90,000 73,553 25,360 15,760 41,119 46% 56%
60,000 60,000 120,000 94,038 25,360 25,360 50,719 42% 54%
90,000 45,000 135,000 104,140 30,545 20,560 51,105 38% 49%
90,000 45,000 135,000 104,140 30,545 20,560 51,105 38% 49%

After tax total income is calculated assuming 7.65% FICA and standard federal tax deductions MFJ return. No state or local income taxes are considered.

Summary of Social Security Benefits

SSRB are generally a really good deal for most Americans. With an average household income in the US making just under $50,000, you can easily see from Table 1 that a married couple can expect that SS will replace 2/3rds or more of your after tax working income (and, obviously even more of ATRI). It is probably for this reason that so many people are actually retiring successfully today. Even for upper middle class folks (those making upwards of 100K/yr.) SS will likely replace half of your after tax income – and surely more than half of SPENDING.

Since SSRB are likely to be a major fraction of funding towards achieving your ATRI goal, it is critical that you: understand how SS works, when is best to claim your benefit, how much you (and your survivors) will get, and how important it is to you and the entire retirement community.

Saving for Retirement Roth vs Traditional IRA (or 401k)

So, now we get around to saving for retirement. Once you know how much ATRI is needed, and you know how much will be replaced by SSRB, you will need to replace the shortfall from retirement accounts or after tax savings.

There seems to always be questions raised concerning whether it is better to pay taxes now and invest in a Roth IRA or invest in a traditional IRA (or 401k type) and pay the taxes later. It can be easily shown that if tax rates didn’t change – it wouldn’t matter. Michael Kitces clearly points this out in on slide 7 page 4. To quote what he refers to as the tax equivalency principle “A certain amount of pre-tax income results in the same amount of after-tax wealth in the end, regardless of which account type it goes to, whenever tax rates remain the same”.

Taxes are not likely to remain the same in retirement as when working. There are several reasons for this:

  1. Taxable income will likely be considerably less than total income resulting in lower taxes.
  2. Unearned income is generally taxed more favorably than earned income.
  3. Elderly get larger personal deductions.
  4. Savings are generally taken at the highest marginal tax rates.
  5. SS income is tax advantaged.
  6. Tax rates continue to fall despite the overwhelming agreement among economists that rates need to rise.

In addition to the clear possible savings by funding retirement plans pre-tax, there is the flexibility of withdrawals that allows the taxes to be paid at a time of your convenience. For instance if you decide to retire early and fund part of your retirement with after tax savings and part from your traditional IRA, you may avoid taxes on the withdrawal completely. Another example is rolling over much of your savings in the year that you have a disaster claim – or huge medical bills. It happens.

Bottom line, pay the taxes later – always better. Michael Kitces, Dirk Cotton, Wade Pfau and other financial writers agree – as do most FPs.

Safe Withdrawal Rates and Annuities

In keeping with the safe and guaranteed return that ensures PRGA I suggest following the 4% rule. This basically says that you can draw down your savings by 4% every year in constant dollars and expect your retirement accounts to last throughout your lifetime. To guarantee a monthly income of $100 requires 300 ($100*12/0.04) times that amount in savings; about $30,000. Funding a retirement for thirty years or so can be expensive.

It is a fairly easy exercise to set up a spreadsheet that starts with 100 dollars that increases by inflation plus 1% every year and decreases by 4 dollars (increased by inflation every year). Starting with inflation set at 0% it shows that this simple “portfolio” would last for 30 years. Interestingly, (as stated above in Inflation and Rates of Returns) maintaining the real rate of return at 1% and increasing inflation to 4% shows that the portfolio has a 32-year lifetime. That is: A. It essentially satisfies PRGA and B. As contended above the inflation rate has little effect on the outcome so long as one works with real returns.

Many people find it simpler and easier to fund an annuity. Annuities can currently be purchased for 65-year-old males that “return” about 6% per annum. First note that this “return” is not a return on investment (ROI) but the yearly income you get as a fraction of the annuity purchase price. That is, if you give the annuity company $100,000 they will give you $6,000/yr, essentially guaranteed for your lifetime.

That $6,000 is not inflation indexed. It is more realistic for planning purposes to assume that the “real” value of that $6,000 is actually $4,000. That coincides with the 4% rule above. The remaining $2,000 should be reinvested to help ameliorate the effects of inflation in later years.

Variable rate Annuities can also help create inflation indexed annuity income. However, I think in keeping with PRGA that basic levels of income should be covered by fixed returns. If a third of the total annuity is invested in variable units that can earn ROI that increases its payout by 3 times inflation then the total annuity will produce inflation-protected income – at minimal risk. With inflation currently running in the neighborhood of 1-2% it is not too difficult to select a variable annuity investment that will beat inflation by a factor of three.

This introduction to annuities is far too brief to adequately address the cost, benefits, and options available. Interestingly just as I was writing this (7 Nov 2017) Dirk Cotton gave a nice description of some of the annuity world – he does it better than I do so … .

Taxes in Retirement

As mentioned previously, taxes in retirement are significantly different from taxes while working. There is no FICA to pay. Many localities do not tax unearned income; Social Security (SS) benefits are tax advantaged not only at the federal level but for many states as well. Finally, there is no need to maintain your residence in a high tax area – follow the lead of George W. Bush and Dick Cheney and declare residence in Texas (or Wyoming or Nevada) and avoid state and local income taxes altogether.

I will present here a couple of simple examples to demonstrate the dramatic tax differences between working couples and retired couples. In each example, I assume that 10% of salary is saved in tax advantaged 401k type retirement plans, and there are state (3%) and local taxes (2.5%) much like I experienced while working in Ohio. These are just examples based on 2017 tax rates and standard deductions.

Example 1. A couple making about 75K (50K for him and 25K for her)

While working this couple would pay: FICA of $5700; Local and state taxes of about $4000; and Federal taxes of $6000 – assuming that they used the standard deductions. The total tax bill for this working couple is nearly $16,000 — nearly 20% goes to taxes.

In retirement after saving diligently and producing 14K of taxable retirement income to go with their 36K of SSRB they would discover that they have essentially no tax liability. That’s it. No federal tax on 14K of taxable income. No local tax on unearned income, and no FICA because they have no salary income. Without having looked carefully, the state of Ohio would certainly not get very much if anything as Ohio exempts SSRB from taxation.

Example 2 A couple making 150K (75K each)

They might have a tax liability of nearly 40K (27%) while working – and yet after retiring and earning 43K of taxable retirement benefits (to go along with their 57K of SSRB) would have a federal tax liability of less than $7000 (7%). Their ATRI and SPENDING would be nearly identical at 94K. Again their state and local tax liabilities should be negligible.

The Tax Torpedo

One of the favorite topics for FPs to talk about is the SS tax torpedo. Why? Because it can in some cases put ordinary folks into a marginal 46% tax bracket, and can be at least somewhat avoided with careful planning.

There are people in the 25% income tax bracket that will have an additional 85 cents added to their taxable income (as more of their SSRB becomes taxable). The combination works out to a 46% marginal tax. There is a baby torpedo at lower incomes that raises the nominal 15% tax bracket to 22.5%. Many people will find themselves in this bracket.

Understanding the tax torpedo

The tax torpedo exists because the government has given tax breaks on SS income to lower income folks and removes this break as income increase. At low incomes, Social Security Income is completely tax-free. But as income increases from other sources (Dividends, Interest, and taxable retirement account withdrawals), the tax-free status is phased out. Until the tax break is fully “recovered”, the marginal rate can be excessive.

To determine the amount of SSRB subject to income tax you need to calculate a provisional or Modified Adjusted Gross Income (MAGI) that comprises your taxable income, your tax-free income from municipal bonds and half of your SS benefit. Then there are two thresholds for determining how much of your SS benefit becomes taxable as you income increases.

For couples with MAGI below 32K (25K for singles), no SS benefit is taxable. For each dollar above MAGI of 32K and below 44K a half a dollar of SS benefit is taxable up to half of your total SS benefit.

Once MAGI exceeds 44K (34K for singles) each additional dollar of MAGI adds $0.85 of SSRB to taxable income – up to 85% of total SS benefit.

Avoiding the tax torpedo

If you read through the above you probably have figured out that the tax torpedo is complicated – the reason why FPs love it. There are reasonable ways to avoid/ameliorate the tax torpedo effects. Professional FPs may have an advantage in this type of planning. The key to avoiding the torpedo is understanding that Roth IRA distributions are exempt from MAGI calculations and only 50% of the SSRB contribute to the calculation. So, clearly moving savings into Roth IRAs and increasing SSRB relative to taxable retirement incomes helps to minimize the effect.

Delaying claiming SSRB: If you delay SSRB you increase the amount you get from SS. But, of course, along the way you must live on something and that generally means drawing down retirement savings. This allows future earnings to be more heavily weighted by SSRB – which, as mentioned above, are advantaged. Those years before claiming SS allow time for Roth conversions.

Lumping taxable income in alternate years: In one year you hit the max for SSRB to be taxed and then roll over additional funds from a traditional to a Roth IRA. The additional funds are taxed at the marginal rate and do not trigger additional torpedo taxes as there is no more SS benefit to tax. Then in the following year take withdrawals just sufficient to hit the lower threshold.

Lifestyle changes: Imputed income in retirement.

Imputed income is savings that you achieve in retirement that were not available while working. These lifestyle changes decrease the amount of income needed in retirement without necessarily adversely impacting your retirement. For those that have not planned well, lifestyle changes that do impact your retirement will be forced upon them. In retirement, there can be many significant lifestyle changes.

Many retirees find that in retirement things that they did while working suddenly cost less or are non-existent. Why? Well partly because places like to give us old folks a price break. But also because we get to chose our times more carefully. For example:

  1. Without the constraints of working I no longer have to take a vacation during spring break. Skiing is just as much fun (and maybe better) the next week and the prices have dropped considerably.
  2. Camping in national Parks, National Forests, and BLM rec lands is half priced and you can go on weekdays when it is less crowded.
  3. You no longer need to attend theatre on weekends or evenings. Matinees are often cheaper.
  4. Golfing is cheaper on weekdays
  5. Senior dining specials
  6. Taking advantaged of last minute airfare/travel specials that working would have precluded.
  7. Traveling mid week when hotel rates are lower.
  8. Eliminating one or more automobiles
  9. Downsizing your home
  10. Finally paying off the mortgage or student loans

In retirement hobbies may have a significant impact on your quality of life – even while not actually producing income. Many of these hobbies/activities are just a bit too time consuming to really pursue while working, but in retirement time is aplenty. Many of these save significant sums of money and actually improve your life, for example:

  1. I personally found that cooking as a hobby more than adequately replaces dining out.
  2. Gardening is one of retirees favorite past times and gives better, fresher and cheaper food than the supermarket.
  3. Home/auto repair saves lots of money.
  4. Walking instead of working out at the gym.
  5. Reading instead of going out to shows.

Finally setting a savings goal

Finally, remember, as you adjust your savings rate to achieve a certain income goal that the income goal will decrease/increase as savings increase/decrease. That is the feedback discussed above. Adjust SPENDING to account for any significant lifestyle changes – difficult to assess but there may be significant knowable adjustments. Subtract pensions and social security and what remains is the ATRI you need to generate from your retirement savings.

You will probably need more than 25 times this annual amount in savings to guarantee PRGA and PRGB. That is using the 4% rule and you will need to generate more to account for taxes – low though they may be they are unlikely to be zero. Again, feedback in your planning will be necessary, as taxes in retirement will change with the savings plan. But with careful analysis, a final necessary savings figure can be determined.

For many folks with SPENDING at 70% of GI (saving very little for retirement) and SS replacing 55% of GI, ATRI will need to be about 15% of GI. Using the 4% rule would indicate that savings should be in the neighborhood of 4 times your annual salary — a not particularly onerous savings goal and a far cry from the 20 times annual income recommended by FPs. And of course, when savings are increased the ATRI needed will decrease and the subsequent result for savings necessary will decrease.

Retirement Investing

It is clear that getting a real 1% return will secure your retirement for about 30 years if you follow the 4% rule. However, following that rule can make for an expensive retirement and it is not easy to fund guaranteed investments that will yield a real 1%. Diversifying into equities or corporate bonds can easily beat the 1% real return but come with risks.

Losses in retirement can be devastating to a retirement plan. While working, you have a huge asset that does not show on the balance sheet – your ability to work. Retire early and if things head south you can always go back to work – may not be optimal but it is always doable. The largest risk (my opinion) that retirees (especially early retirees) face SOR risk.

It is precisely for these reasons that I think retirees should fund their basic needs with guaranteed returns: annuities, government bond ladders, certificates of deposit (CD) etc. Once an adequate income is guaranteed then, and only then, should a retiree begin to acquire equity exposure.

Certificates of Deposit

CDs can be an important part of building a bond ladder — particularly for years 1-5. A quick google search (March 2018) can easily show that returns from 1% to nearly 3% can be achieved using 1-5 year CDs.These yields are comparable to or slightly better than similar duration treasuries.

CDs are easy to purchase – whether from the bank directly or in a brokerage account. They are generally guaranteed by the Federal deposit Insurance Corporation. CDs don’t fluctuate in value – you can always cash them in for face value plus interest earned (less penalty).

CDs are also quite flexible. Yes, you will have to take a small hit to exercise the flexibility, but cashing a CD early generally costs 3-6 months of interest only. So, buying a  5-year, 2.8% CD (one of the highest available as of March 2018) and redeeming it after three years (with a 6 month penalty) would be essentially the same as purchasing a 3-year, 2.35% CD, which is only slightly worse than the best available 3-year CD  (2.55%). So, the fact is there is really not much of a penalty.

CD advantages over treasuries: A. Ease of purchase, B. protection of capital, C. possibly higher yields.


Bonds can be purchased individually. Government bonds can be bought at for no fee. The nice part of owning individual bonds is known and guaranteed return. However, without significant effort (buying and selling and paying the appropriate fees) Bond Ladders are probably the most expensive way to fund a retirement. However, they also offer the least risk and the greatest flexibility.

If you want to fund retirement savings or a retirement with guaranteed bonds (good for guaranteeing PRGA and PRGB) then the question of course is which ones and how.

There are basically four types of US Government Issue bonds: Savings bonds in I and EE varieties, Treasury Bonds, and TIPS. They each have specific disadvantages and/or advantages. There are also municipal bonds and of course bond funds. I own three different types TIPS, EEs and I-Bonds.


EE savings bonds

You can purchase up to $10,000 of these bonds each year for you and your wife. The basic interest rate that they pay is minimal; currently 0.1%. However, if you hold these bonds for 20 years they are guaranteed to double in value, which is equivalent to 3.5% per year. There is a minimal holding period (1 year) and 3-month interest penalty if cashed in prior to 5 years.

EE advantages over treasuries: A. they pay better if held for 20 years. B. There are no tax implications until these are redeemed. And C. They never lose value – if interest rates rise you can redeem them at purchase price plus accrued interest. There is no decrease due to rising interest rates that might be experienced with Treasuries.


These bonds pay a fixed rate (usually small and currently 0.1%) plus a variable rate that is based on the latest inflation figures. The interest rate adjusts every six months. Again there is a minimal holding period (1 year) and 3-month interest penalty if cashed in prior to 5 years. There are no tax implications until these are redeemed.

I-bond advantages/disadvantages vs. TIPS: A. They never lose value due to deflation or rising interest rates. B. Taxes are not paid until the bonds are redeemed. But C. they pay significantly less nominal/fixed interest rates than TIPS


Treasuries are available with maturity dates from one month to thirty years and yields that currently vary from 1% to 3%. These bonds sell on the open markets and are not redeemable until maturity. So, if interest rates have risen and you need the money you will sell at a loss. Conversely you could sell at a gain but interest rates cannot drop much from present levels.

Treasury advantages over EE-bonds: You can invest in shorter maturity instruments – that may have significantly better yield (over the shorter lifetime) than EEs sold short of original maturity..


These inflation-protected securities are sold with maturities from 5 to 30 years and pay a nominal interest rate (currently ranging from 0.1% to 0.9%). The inflation protection increases the value of the bond as inflation is positive. The amount of interest you receive depends on the current value of the bond. You will owe taxes on the payments in the year in which you receive the payments. You will also owe taxes on the increased value of the bond in the year that it increases – but you will not receive that increased value until maturity. As with other treasuries you can sell them on the open markets at a loss or gain.

TIPS advantages/disadvantages vs I-bonds: A. TIPS pay better nominal interest rates. But B. TIPS can lose value with deflation and C. You may owe taxes on money not received.

Municipal Bonds

Munies generally pay tax-free interest and thus are of interest to many high income individuals. For most people they have little benefit, are subject to default, and are added into MAGI and can result in additional SSRB being taxable – thereby removing some of the tax free benefit of owning them.

I know little of their total characteristics and will make no judgment of their suitability for anyone’s portfolio. There are none in mine.

Bond Funds

Rather than buying individual bonds and “laddering” them yourself, you can invest in a bond fund. Bond funds come in a variety of forms: government, municipal, corporate, high risk or low risk, long duration or short duration, leveraged or not etc.

Many funds buy long term bonds and then sell them as they reach short maturity. This can (and generally does) increase the return of the bond fund vs. individual ownership. On the other hand, bond funds can (and do) lose value – particularly leveraged bond funds.

Bond Funds can make it easier, and give a better return at minimal additional risk.


Annuities probably give you the highest guaranteed return. Note however, that there is a small (very small) chance of default. The higher return does not come from higher risk but rather from “Mortality Credits”. Basically, some individuals will experience a short lifetime after purchasing an annuity while others will live forever (OK not really forever). The long-lived ones will take home higher monthly payments at the expense of the short-lived ones.

Finally, one other good thing about annuities, they will protect you from yourself late in life. For many of us, the time will come when we are not really able to fully comprehend and control our finances. If at a younger age you purchased a lifetime annuity, you will be hard pressed to give away your nest egg. In essence you have already given it to the insurance company. Sure, someone can come along and take the rent money today – but next month another tranch of money shows up. It makes you a much less valuable target and guarantees monthly income regardless of your faculties.

Michael Kitces gives a good description of Mortality Credits and why they can beat bond ladders at .

The downside to annuities – you give you give up complete control of your capital to the insurance company. If you change your plans or experience a significant loss (market, health, or natural disaster) you will not have capital to bail you out. That is not always the worst thing (giving up control), but it is real.


Retirement is neither simple nor easy. Early retirement has an especially difficult set of circumstances and probabilities with which to contend. However, retirement is not the unwieldy, complex and overwhelming problem that the FP community would have us believe. Boomers are retiring in record numbers – many of them happily without near the savings “needed” according to the “experts”.

The income needed in retirement is significantly less than your needs while working. It is critical to understand where your money goes while working in order to determine how much you will need in retirement. Doing a careful calculation to determine how much you spend today (SPENDING) to maintain your lifestyle is critical to determining A. how much you will want tomorrow in retirement and B. how much you will have to save to generate that income in retirement.

Estimating your SSRB accurately is also critical as this is likely to be half or more of your ATRI.

Taxes in retirement will, for many people and most average income folks, be negligible, as compared to those paid while working.

Use the 4% rule to determine how much savings you need and expect to invest in retirement. The 4% rule basically assumes that investments will make a mere 1% real rate of return. Additional savings can be used to “improve” the RR. Even with these conservative assumptions, the amount needed to fund retirement is nowhere near as high as many FPs contend.

Don’t forget to always use feedback when determining how much you need to save in order to reach your goal. Any additional savings made while working will decrease your SPENDING and savings goal. This decrease can be significant.


I am not a FP professional of any type. The opinions expressed here concerning the FPs in general are mine. I successfully retired at 48 on well under 20x my salary in savings – that hardly means that you can.

While my advice is free, and I would like to think well justified, it most certainly could be wrong. Again, just because it worked for me does not mean it will work for you. Only you can decide if/when/how to plan and finally retire. The cost of any mistakes made in planning or retiring will be borne by you, not me. So – be careful. Read, learn, study, ask questions.

Finally, LEARN. Remember, your fidelity guy, me, and even uncle Eddy are not the guys that are going to suffer from the (possibly) bad advice that you might act on. You, alone, are responsible. I am always happy to share my thoughts, and even the mechanics of how I did it. Times are different now and the appropriate solutions may not be the same.
Good luck, and let me know if I can assist (or share an opinion) on anything.


FP                Financial Planner

MFJ              Married Filing Jointly

FICA             Federal Income Contribution Act (Social Security and Medicare taxes)

PRGA           Primary Retirement Goal A — ensure a 95% probability of success

PRGB           Primary Retirement Goal B — maintain you pre-retirement lifestyle

TIPS             Treasury Inflation Protected Securities

ATRI             After Tax Retirement Income

SPENDING  What you are spending today, before retirement, to live

IRA                Individual Retirement Account

SS                  Social Security

ROI               Return on Investment

MAGI           Modified Adjusted Gross Income

FRA              Full Retirement Age

AIME           Average Indexed Monthly Income

GI                 Gross Income

SOR              Sequence of Return

SSRB             Social Security Retirement Benefits

Eliminating Inflation in Retirement Planning

Using a Real Rate of Return to Simplify Retirement Planning

Planning for retirement is an important exercise that has been made significantly more difficult then necessary by the inclusion of inflation as a variable. Many calculators and Financial Advisors (FAs) assume investment returns (ROI) and inflation separately. Therefore one must understand and estimate future inflation as well as future market returns to adequately assess their retirement savings and spending plans.

Either of these inputs (ROI or Inflation) is difficult (‘nee impossible) to know. Lack of knowledge of ROI and the possibility of poor ROI early in retirement is the sole cause of Sequence of Return (SOR) risk. And, obviously, inflation can cause a significant lack in purchasing power over time. So, despite their “unknowability” they seem to be important.

In this article I intend to show that combining the two into a “real” return (that is the ROI above inflation) significantly decreases the complexity of retirement calculations. At a minimum it reduces the number of variables. A real return (RR) is often a knowable variable in the real world investment behaviour of many retirees. And, finally, there are real world products that real retirees can and do purchase that correlate inflation with ROI to guarantee RR.

To start, inflation will be fixed between 0% and 9% per annum while the RR will be fixed at 1%. Savings accumulation will occur for 20 years with savings balances growing by RR plus inflation.

Spending will be taken following the 4% rule. That is after accumulating for 20 years the first year of “retirement will experience a withdrawal of 4% of the total savings balance. The annual withdrawal will increase at the rate of inflation. Total balances will continue to grow by RR plus inflaition.

The determining factor for comparison will be the number of years until the entire savings is spent – portfolio lifetime (PL). Subsequent analysis will be performed to determine the sensitivity of PL to inflation (at constant RR).

Example Setup
Setting up the initial example in a spreadsheet is quite simple. Increasing each year’s savings by inflation and increase the total savings by inflation plus RR and adding the current years savings. Then after 20 years, the savings rate is adjusted to negative 4% of the total savings after year 20 – and subsequently increased by inflation every year. The total savings calculation does not change – only instead of adding the years savings the years spending is subtracted.

Table 1. Shows the encoding (for an Excel spreadsheet) for the simple analysis and the inputs for the simplest example: RR=1% inflation = 0%. Table 2 shows the results.

Table 1: Excel encoding for first few rows of the simplest example
1 Year Savings Total
2 1 1000 1000 RR= 1%
3 2 =B2*(1+E$3) =C2*(1+E$2+E$3)+B3 Inflation = 0%
4 3 =B3*(1+E$3) =C3*(1+E$2+E$3)+B4
21 20 =B20*(1+E$3) =C20*(1+E$2+E$3)+B21
22 21 =-0.04*C21 =C21*(1+E$2+E$3)+B22
23 =B22*(1+E$3) =C22*(1+E$2+E$3)+B23
50 49 =B49*(1+E$3) =C49*(1+E$2+E$3)+B23

Table 2. Results for the simplest example
1 Year Savings Total
2 1 1000 1000 RR = 1%
3 2 1000 2010 Inflation = 0%
4 3 1000 3030.1

20 19 1000 20810.9
21 20 1000 22019
22 21 -880.76 21358.43
23 22 -880.76 20691.26

50 49 -880.76 -77.3226

Saving $1000 a year in a zero inflation environment with RR equal to 1% yields total savings at the end of 20 years of $22019. The first year’s withdrawal (and all subsequent years as inflation is zero) is $880.76 and the account goes belly up in year 49.

Once the spreadsheet is set up it is now easy to adjust the inflation rate (cell E3) and one can quickly notice that increasing the inflation, while maintaining RR results in a longer PL. Increasing inflation by a mere 1% increases the PL to 50 years. An inflation rate of 3% takes it out to 50 years and higher inflation rates have a minimal effect on PL.

Sensitivity Analysis
Table 3 shows the sensitivity of the PL to input changes in both RR and inflation. Using a baseline case (RR=0% and inflation = 0%), which gives a PL of 45 years. Table3 gives the PL increases for various RR and inflation rates.

Table3 Lifetime increases (years increase in PL from the baseline case of 45 years)
Inflation      0.5%    1%    2%    3%
0%         2        4      10      22
1.5%        3        5      11      23
3%          3        5      12      24
6%          4        6      14      25

It is possible that the increase in PL as inflation increases is an artifact of the timing – when is the deposit made vs. interest earned and funds withdrawn etc. But, it is clear that the rate of inflation has very little impact on the PL so long as the RR is held constant. The inflation rate can be made to vary during both accumulation and spending without any significant impact on the results – I have not included the details of that analysis here.

The RR as expected has a significant impact on the PL. The apparent increasing rate of change for PL as RR increases from 1, to 2, or 3% is real and to be expected as the PL for RR equals 4% and inflation equals 0% is infinite as the 4% payout is equal to the 4% gain in the total savings. Increasing inflation in that scenario results in an ever-increasing portfolio value.

Real expectations for RR
Many retirees border on the conservative side of investments. This may be particularly true for early retirees trying to avoid SOR risk and intend to follow a reverse glide path later in retirement. To the extent that a portfolio is invested conservatively RR becomes more knowable. If in fact much of the portfolio is invested in inflation-protected securities, RR can be estimated with considerable accuracy.

In the 1990s the US treasury issued Treasury Inflation Protected Securities (TIPS). The initial rates (RR) for these issuances were in the range of 3.5%. A good buy by any standard. Today 30-year TIPS offer inflation protection and a yield (RR) of just under 1%. Savings bonds also offer an inflation-protected version – I-Bonds. The RR of I-bonds is essentially zero.

Investing in tips or even I-bonds is simple and secure way for ordinary retirees to protect themselves from inflation and guarantee a real return – albeit a small one. Investing in inflation-protected securities protects retirees (especially early retirees) from sequence of return risk. For many middle class retirees that have saved “enough” but not too much, securing a retirement with inflation-protected assets is a major goal. Inflation protected securities go a long ways towards making that goal safely and easily achieved.

Many FAs and retirees opt for market based returns rather than the safety of government bonds. Obviously, if the retiree invests in markets then there is a good chance for better returns. While working and saving market downturns can be easily absorbed. There is sufficient time to allow for market recovery and there is always the option of working for additional years to cover the losses. However, when retired, the SOR risk can overwhelm the best of retirement plans.

Generally speaking, I should think that any retirement plan should contain a basic plan that is based on an RR of 1% in retirement. The time for risk is before retiring. Additional funds not needed for retirement funding can – and probably should – be invested in higher risk assets. The bottom line however is that if you cannot afford to retire on an RR of 1% you should consider that you cannot afford to retire.

Finally, for planning purposes, the RR of market returns is generally not much better known than ROI and inflation. Using RR merely removes two unknowns and replaces them with one. That process probably adds nothing to the accuracy of retirement modeling. The advantage of using RR is probably specific to the case where the retiree intends to use guaranteed return investments that make knowing the expected RR a reality.


No one knows the forward going rate of return on equities or bonds. In addition it is not possible to really know the future of inflation. However, knowing both of these numbers is important to many retirement planning models. Modeling retirement scenarios with “guesses” about these important numbers is unlikely to yield good results.

Conflating the market ROI and inflation rates to determine the real rate of return (RR) however eliminates two unknowable variables into one variable. Sensitivity analysis shows that PLs with constant RR are not sensitive to inflation

The value of this one variable – The Real Rate of Return – can be more easily and accurately determined. The current yield on 30 year TIPS (1% at present) provides a floor for the value of RR in retirement modeling.

I suggest that all plans should proceed using RR of 1% as a baseline. While modeling success will be largely dependent on the actual RR used, using a known and achievable rate like the current 1% eliminates two unknowns from retirement planning in favor of one known.

It is hard to fathom why the FA community at large continues to use two unknown and unknowable variables as inputs to their retirement models when one known variable can adequately replace both of the unknowns. Using the knowledge of the RR needed for a successful retirement gives a benchmark to the retiree for determining if his investment model is working.

Bond Duration: How and why bond lifetimes relate to interest rate sensitivity

Bond duration

Consider a simple two-year zero coupon bond in zero interest environment. Obviously the value of that bond is its face value – 100 dollars.

Now, if the interest rates climb to 1% the value of the bond falls to 100/(1+1%)^2 = 98.03, since an investment of $98.03 will yield $100 dollars at maturity if the YTM is 1%.

Notice that the Bond value has fallen by just about 2% — The numerical value 2, not coincidentally, is the time in years to maturity. In the case of zero coupon bonds, the decrease in value is the duration multiplied by the interest rate change.

The question that might come to mind is “Why does the bond value fall a percentage that is equal to the lifetime (duration)?” The answer is in a double – first order analysis of the problem. First, a first order expansion of the exponential and then a first order expansion of the quotient.

  1. (1+x)^n = 1 + nx + n(n-1/)x^2 /2 + … ~ 1+nx for small x
  2. 1/(1+nx) = 1 – nx – (nx)^2 …   ~ 1 – nx for small x
  3. Which combined give us 1/(1+x)^n… ~ 1 – nx
    1. Alternatively, 1/(1+x)^n = (1+x)^(-n)… ~ 1 – nx
  4. For the example being used, 100/(1+1%)^2 ~ 100(1 – 2*1%) = 100(1 – .02) = 0.98

And so, we can see how the duration (the bond lifetime in the case of a zero coupon) reflects the percentage change in bond value for a 1% change in interest rates. The bond value changes a percentage about equal to the product of the interest rate change and the duration.

The duration is the weighted average maturity of cash flows. That is, it is the sum of the present value of the bond payments times the time to those bond payments divided by the sum of the present values of the time payments.

ti = time to the ith payment

PVi = the present value of the ith payment

The fact that it is about equal to the percentage change in bond value for a 1% change, while not coincidental, has nothing to do with its actual value, definition, or calculation.

What about bonds other than zero coupon

Let’s go back to a 2 year not quite zero coupon – but still in a zero interest rate environment. Consider a two year bond that pays x dollars of face value after 1 year and (1-x) dollars after the second year. Given a 0% interest rate – this bond also has a value of 100. But, the duration is no longer 2 years.

Duration = (x*1 + (100-x) *2)/(100) = (200-x)/100 years = 2-.01x

We can readily see that the duration of the bond varies from 2 years (x=0) to 1 year (x=100).

Now if we raise the interest rate to 1% we can recalculate the bond value as

x/(1+1%)^1 + (100-x)/(1+1%)^2 ~ x(1- 1*1%) + (100-x)*(1-2*1%) = 100 – ( 2 – .01x )

and we can easily see that the decrease in value is just about equal to the duration.

Retirement Savings: Tax Deferred or Tax Free?


While working, many people are faced with a choice between saving in tax deferred account (like a Traditional IRA or 401(k) ) or a tax free account (like a Roth IRA)

The traditional IRA costs less to fund today, as the money saved is usually not taxed today. but… when the savings are withdrawn in retirement the withdrawals are considered ordinary income and taxable as such.

The Roth IRA costs more today, but in retirement, withdrawals are tax free. Roth withdrawals do not even count in calculating Modified Adjusted Gross Income (MAGI) – the important number in determining taxation of social Security (SS) benefits.

So, the pertinent question is which is tax favorable. Let me state right up front; tax deferred is preferable. There may be situations when paying your taxes today beats paying them later – but those circumstances are almost certainly rare and applicable to a small minority of workers.

First Question

The first question is – Given equal tax rates, is it better to save tax free or tax deferred? If everything were equal before and after retirement it would make no difference. Sure, one of these accounts grows tax free while the other grows taxable income. But the one that grows taxable income starts with more money – because it is cheaper to fund.

Let’s compare the results for a worker in the 25% marginal tax bracket and assume the money will be taxed at 25% when withdrawn. Let us further assume that it returns 2% a year above inflation. If the worker has $100 of income to fund a Roth or a traditional IRA, his initial balance is Roth – 75$ or Traditional – $100. The Roth is lower because he must pay the government $25 of tax on the $100 of income. The traditional saver incurs no tax liability the year that he saved it. The savings grow and are then withdrawn. Tax is withheld from the traditional savings and not from the Roth. Table 1 shows the results.

Table 1: Comparison of Roth vs. Traditional IRA Assuming Tax Rates Are Constant.

Roth Traditional inflation
income $100.00 $100.00 2%
tax $25.00 $0.00 real rate of return
initial saving $75.00 $100.00 2%
Year Roth Traditional
0 $75.00 $100.00
1 $78.00 $104.00
2 $81.12 $108.16
3 $84.36 $112.49
4 $87.74 $116.99
5 $91.25 $121.67
6 $94.90 $126.53
7 $98.69 $131.59
8 $102.64 $136.86
9 $106.75 $142.33
10 $111.02 $148.02
11 $115.46 $153.95
12 $120.08 $160.10
13 $124.88 $166.51
14 $129.88 $173.17
15 $135.07 $180.09
16 $140.47 $187.30
17 $146.09 $194.79
18 $151.94 $202.58
19 $158.01 $210.68
20 $164.33 $219.11
withdrawl $164.33 $219.11
tax $0.00 $54.78
value $164.33 $164.33

Assuming that the savings are taxed identically, it matters not whether the money is taxed before (in a Roth) or after (in a Traditional) growing in a tax advantaged account. The only question as to which is best is which has the lower tax rate.

Second Question

Will tax rates be identical before and after retirement? That answer is certainly – NO.

For a couple earning in the neighborhood of 100K per year, the marginal rate is 25%. The marginal rate might be 15% for all/some of the funds for lower incomes. The funds invested into a Traditional IRA decrease the taxes at the marginal tax rate of the year in which they were invested. Contributions to a Roth generate no tax savings. The important – and harder to answer question – is “What tax rate will apply to withdrawals from the Traditional IRA?”.

First, we do not know whether tax rates will go up or down in the future. Surely, as has been said for the past 20 years, tax rates must go up in the future to help pay off the debts incurred over the past 30 year. However, that ignores the facts – tax rates have not gone up significantly and have actually fallen. Let’s just assume that we have no crystal ball and taxes won’t change much.

What is the tax rate for retired folks? Well that depends much on Adjusted Gross Income (AGI) and Modified AGI (MAGI). Are we talking marginal rates or average rates? Am I drawing SS benefits or not?

Again considering that couple making about 100K while working. They might have saved 600K over 30 years for their retirement in addition to SS benefits. This savings could easily come from a 10% yearly contribution to a 401(k) type account with a 5% employer match earning a real 2% return for thirty years.

Their income (67K) in retirement comprises SS (~ 42K) and Retirement income (~ 25K). At these levels, about 7K of SS benefits are taxable, for total AGI of 32K. Which gives, after deductions, a taxable income of 9K and total tax of 900 dollars. $900 dollars tax on withdrawals of $25,000 is a total tax rate under 4%. The marginal tax rate (18.5%) comes from the 10% bracket plus an additional $0.85 increase in income due to the SS tax torpedo. But, this 18.5% marginal rate has little to do with how much tax is collected from their retirement savings withdrawal. Attributing all of their taxes to their retirement withdrawals (which seems fair, after all without these withdrawals there would be no tax liability) we can see that savings were made at the 25% rate and withdrawn at 4%.

Even if the savings were made at the 15% rate, the tax rate in retirement would be significantly lower – and that is not even considering that if their tax rate while working was lower their SS benefit while retired would be less and their tax bill would consequently also decrease.

It is also worth noting that this couple has 67K of income and a tax liability under one thousand dollars. And the elderly complain about their onerous tax liabilities.

Other Considerations

There are other considerations that may help to impact your choice. I mentioned that there are methods to avoid the tax torpedo. Delaying SS benefits while living on taxable withdrawals, rolling over to a roth after retirement but before claiming SS, and rolling over to a Roth in a year when you have a significant hit to income – large itemized deductions (due to health or catastrophe) or capital losses.

Generally speaking delaying taxes just gives you a lot more flexibility as to when to pay than just paying the high rate while working. Yes, there are possibilities that you will pay more – but it is hard to envision the scenario.


Saving in a tax deferred account while working is almost certainly better than saving in a tax free retirement account. Taxes saved for contributions to a Traditional IRA are generally saved at the highest marginal rate (15% or 25% for most folks) while withdrawals are taxed at an average rate that is much lower; even while the marginal rate might be higher. The flexibility gained to time withdrawals to minimize taxes and the effects of the tax torpedo also contribute to the idea that deferring taxes is better than paying them.

On savings and retirement: part 2


In the first part of this article we saw how to estimate how much you need/want to spend in retirement. The basic rule is that you will want to replace all that you spend now – which is something in the range of 35 – 70% of your income.

In this section we will discuss how to estimate how much income you can expect from Social Security (SS), IRAs, 401ks and other savings accounts. We will start with SS as it  is the most knowable and the easiest to estimate. Knowing how much you can expect from SS will help determine how much you need from retirement savings and consequently how much you need to save.

I will also discuss, very briefly,  the effect of taxes both before and after retirement with a special discussion on taxes and SS payments.

Finally, I will close with examples of savings plans for couples that have incomes in the 30 to 120K a year range.

Social Security

When discussing how much you can expect from SS, I will be using average annual income. Social Security indexes your income over the years to account for salary inflation. If you have not had a significant change in salary for many years (small raises and COLAs only) then your final salary is probably not a whole lot higher than your SS index adjusted average. For an accurate determination of your expected benefit go to the social security web site ( and use one of the calculators – or just ask for a benefit analysis.

To determine your benefit – you need to determine your Average Indexed Monthly earnings (AIME). Again, for the sake of simplicity, I have just used  your final monthly salary. These examples are accurate as of 2017. You can expect to receive 90% of the first 885 dollars of AIME. In addition you can expect to receive 32% of the AIME above 885 but below 5336. Finally you will get 15% of your AIME that exceed 5336.

Table 1. Expected SS benefit and income replacement percentage
Income 30000 40000 50000 60000 70000 80000
SS benefit 15750 18950 22150 25350 27550 29050
Replacement % 53% 47% 44% 42% 39% 36%

Despite the progressive nature of the benefits, those at the lower end of the income scale will probably find it hard to live on half their income. Those at the higher end may very well find that 38% of their income is nearly sufficient to their needs.

There is, for married couples another tremendous benefit. Whether the spouse has worked or not, he is entitled to a SS benefit that is no less than 50% of the working spouses benefit. If we add that additional benefit we find that low income couples replacement percentage would rise to 75%. while the high end climbs to 57%.

One thing is immediately clear – SS replacements percentages for married couples are already in the 50-70% of income range. This explains how so many people actually do retire despite not having sufficient savings (according to financial experts) to do so.

401(k)s, IRAs, etc.

If you’ve been following through you now have an estimate of how much income you need in retirement and how much you should count on coming from social security. Obviously the difference needs to come from your personal retirement savings, pensions (which I am not considering), work in retirement, or lifestyle changes.

If you intend to use savings for all of the difference, you will need to save approximately 25 times the difference between your SS benefit and your total spending. This allows for the 4% withdrawal rule with your annual withdrawals increasing by inflation every year.  See Wade Pfau blogs for a discussion of the 4% rule.

Now, that you have figured out how much you need to save, you can begin to establish a savings plan to accomplish that goal. This is where your plan gets feedback. If you decide that you need to save 10% of your salary to reach your goal, you will find that your goal has shrunk. If you are saving an additional 10% of your income while working that is an additional 10% that you do not spend and therefore do not need to replace.

Tax considerations

First off you need to consider saving in a pre-tax vehicle (e. g. traditional IRA) or post -tax (Roth IRA). Obviously saving pre-tax is easier now (it reduces your income tax) but then the savings are taxed as ordinary income when withdrawn.

There are many considerations when it comes to traditional vs. Roth IRAs or 401(k)s, but, briefly, remember that your contribution to a traditional IRA gets a tax break generally at your highest marginal tax rate. For many working couples this is often 25%. Since SS is generally not taxed (see below there are huge exceptions) and an elderly married couple has 23K personal and standard deductions, total adjusted gross income (AGI) up to 42K remains in the 10% tax bracket.

I personally favor traditional IRAs because retirement savings are taxed at a rate much closer to your average tax rate when withdrawn (as opposed to your marginal tax rate). For example, if you save 30K at the 25% margin while working you avoid $7500 of tax. If you withdraw 30K in retirement (even while having 15K of SS benefit) your total tax bill would be about $1000. Despite the fact that you are in a 15% marginal tax bracket (see below) your 30K of income has generated a mere 1K in taxes – a 3.33% tax rate on your retirement withdrawal.

SS benefits can be taxable and the marginal tax rate for retired individuals can be pretty high (as much as 46%). This is known as the SS tax torpedo. The critical number is your  “Modified Adjusted gross Income” (MAGI). This is half your social security and all earned, unearned, and tax free interest income plus taxable retirement account withdrawals. It does not include Roth distributions.

So, to be a bit more accurate with marginal tax rates, most retired couples experience a 15% marginal tax rate – even when they are in the 10% tax bracket income range due to the Social Security tax torpedo. If the couple finds themselves in the 15% tax bracket – their marginal tax rate may be least 22.5%.

For single filers a combined income greater than 25K allows the next dollar of MAGI to cause 50 cents of SS benefit to be taxable. If you are in the 15% marginal tax bracket – you have suddenly moved into the 22.5% bracket. When your combined income hits 34K, 85 cents of your SS benefit becomes taxable on the next dollar. If you are at the 25% marginal rate you have now moved into the 46% tax bracket. The limits for Married Filing Joint are 32K and 44K.

I won’t even pretend to optimize retirement income from various pre-/post- tax strategies. I will show that retirement savings can be done without imposing draconian savings goals on working couples. Optimizing their various contributions to different savings vehicles will probably change for nearly every income level, residence locality, retirement age, earned vs. unearned income,  employee matching in 401(k), expected returns, and who knows what else.

Some Examples

I have run some examples for married couples of approximately the same age, retiring at full retirement age, that might have saved for 30 years before retiring. If a 401(k) was available, I assumed that the employer would match it up to 5%. SS taxes are collected on all Earned income. I assumed that IRAs and 401(k) accounts would return 2% above inflation every year  The initial withdrawal from savings was 4% of the account balance. Local and state taxes were set at 3-8% pre-retirement and assumed to be negligible in retirement. Federal taxes were paid using only standard deductions.

These are simplistic examples. Their goal is not to set about a proper savings plan for these couples but merely to show the approximate levels of spending available while working given a savings plan that can be expected to replace that spending with after tax income in retirement.

Table 2. Some examples of the required savings rate to maintain spending in retirement
salary husband 30000 30000
salary wife 0 30000
employer contribution husband 0.0% 0 0.0% 0
employer contribution wife 0.0% 0 0.0% 0
retirement contribution husband 0.0% 0 0.0% 0
retirement contribution wife 0.0% 0 0.0% 0
Roth ira contribution husband 2.0% 600 9.7% 2915
Roth ira contribution wife 0.0% 0 9.7% 2915
Social security 7.65% 2300 7.7% 4590
state and local 3.0% 900 8.0% 4800
Federal 900 5108
spendable income before retirement 25300 39672
expected SS benefit hers 7880 15105
expected SS benefit his 15760 15105
4% of retirement savngs 4% 0 4.0% 0
4% ira savings 4% 1450 4.0% 9461
less taxes – average rate 0% 0 0.0% 0
spendable income after retirement 25100 39672
total retirement savings needed 36250 236532
salary husband 60000 60000
salary wife 0 30000
employer contribution husband 3.2% 1909 5.0% 3000
employer contribution wife 0.0% 0 0.0% 0
retirement contribution husband 3.2% 1909 5.0% 3000
retirement contribution wife 0.0% 0 0.0% 0
Roth ira contribution husband 0.0% 0 5.8% 3506
Roth ira contribution wife 0.0% 0 5.8% 1753
Social security 7.7% 4590 7.7% 6885
state and local 8.0% 4800 8.0% 7200
Federal 4821 9158
spendable income before retirement 43880 58499
expected SS benefit hers 12561 15105
expected SS benefit his 25123 25123
4% of retirement savngs 4.0% 6195 4.0% 9736
4% roth 4.0% 0 4.0% 8534
less taxes – average rate 0.0% 0 0.0% 0
spendable income after retirement 43880 58499
total retirement savings needed 154886 456755
salary husband 60000 90000
salary wife 60000 30000
employer contribution husband 5.0% 3000 5.0% 4500
employer contribution wife 5.0% 3000 0.0% 0
retirement contribution husband 10.0% 6000 10.0% 9000
retirement contribution wife 10.0% 6000 0.0% 0
Roth ira contribution husband 0.0% 0 0.0% 0
Roth ira contribution wife 0.0% 0 11.5% 3447
Social security 7.7% 9180 7.7% 9180
state and local 8.0% 9600 8.0% 9600
Federal 13858 14608
spendable income before retirement 75363 74166
expected SS benefit hers 25123 15885
expected SS benefit his 25123 31770
4% of retirement savings 4.0% 29209 4.0% 21907
4% roth 4.0% 0 4.0% 5593
less taxes – average rate 12.5% -2781 4.8% -988
spendable income after retirement 76674 74166
total retirement savings needed 730225 687492

The lowest income individuals spend and therefore need to replace nearly 80% of their salary. This is the highest necessary replacement percentage shown. This would of course make perfect sense as their tax rate while working would be the lowest and they probably would not be saving much, if anything. In fact the taxes shown here for state and local are probably too high and he needs to replace even more of his salary.

Fortunately for these low income couples, SS can be counted on to replace nearly all of their spending. The difference between what SS would give them and what they need would be about 1500 dollars a year. That difference can be “made up”  fairly easily. Contributing 600/year to a traditional IRA would do it. But, let’s be honest, living on 30K a year this couple is unlikely to save anything. More likely they will work in retirement or adapt some lifestyle changes in retirement. It is unlikely that these would be onerous.

At the other end of the scale Couples earning 120K would need to save a considerable sum. The table shows that if they contribute 10% –  11.5% to a 401(k) or a Roth IRA then their spendable income, after taxes and savings  is about 60% of their gross salary. Saving 10% of your salary  at these income levels should not be overly difficult.


Savings goals for retirement are  generally vastly overstated by most advisors. Real retirement goals can be achieved that can easily maintain ones lifestyle without onerous savings loads. Even at income levels of 120K savings rates of 10% are often adequate to replace all of the pre-retirement spending with after tax post retirement income. Lower income levels have lower savings requirements to accomplish the same.

Many people read about how much they need to save (25 times their income ) and think “there is no way I can save that much”. At that point there is a tendency to just give up. The 120K income couples would need to save 1.75 – 3 million by this standard. I think that many people just give up. I calculate that 700K would be sufficient. An achievable goal. A goal that does not dishearten the saver.

Despite the general lack of savings throughout the country, people continue to retire. They have not saved nearly enough – according to the experts – to retire. How do they manage? Simple, the savings required are not nearly what the experts think. Small changes in lifestyle, part time work for very low income people, and retirement savings account that are a fraction of what general consensus deems necessary supplement social security payments and are generally sufficient.