As mentioned last time, perhaps the most important variable affecting retirement outcomes and expectations is the cost of living during retirement. For decades, it has become nearly gospel that retirees will need to live on about 80% or 85% of their before-tax, pre-retirement income during their retirement. (Alternatively, some calculators – such as the one I talked about last time from TIAA-CREF - use 100% of the after-tax income.) So according to these calculators, if an individual is living on $50,000 a year before retirement, she'll need about $40,000 a year during retirement. That income need will need to be met through one or more sources such as a pension, social security, dividends or by liquidation of financial assets.
This 80% or so retirement income estimation has been so well established for so long it is hardly considered whether or not it reflects reality for the majority of retirees. And it doesn't. In the graph below from Advisor Perspectives we see that that ratio of household income for those over 65 compared to those in the 45-54 year-old age group is barely 50%. In fact, only in the last few years has this ratio gone above 50%. This ratio has been below half for most of the last 50 years.
(As noted last time, the reason I am not comparing retiree incomes over 65 to those in the next younger age bracket – 55-64 – is because that age bracket already has a large number of retirees (the average retirement age is about 63), pushing average incomes for this age group lower. Using ten year increments, the 45-54 age group is likely to have the highest percentage of individual still in the work force, and represent the “best-fit” comparable before-retirement to after-retirement age group. In fact, since the 45-54 age group includes younger 45-50 year-olds (generally lower-earning compared to those a decade older), it is likely that average incomes in the 50-60 age group or 55-60 age group would be even higher, showing an even greater drop off between pre-retiree income and income after retirement.)
The graph below compares the same information in a somewhat different way. Comparing by U.S. state, it shows the ratio of pre-retiree to retiree income. In this case however, it is comparing the entire 45-64 age group to those over 65. As noted, since those near 60 or in their early 60s are more likely to be retired than those in the younger 45-54 age group, including this older segment of (generally) pre-retirees is reducing the average income for the entire 45-64 age group. (If we were able to compare the average income of those in younger and older age groups who were actually retired we would see an even greater drop off in average incomes and more accurately reflect the actual decline in income after retirement.) Despite this upside bias to the pre-retirement income group in this graph, we see in the caption that the average 65 and older household still had just 57% of the average income compared to those 45-64.
The downside of upside bias
Last time I looked at TIAA-CREF's retirement calculator. This time I'll look at Fidelity's. Using the same figures as last time, age 55, $60,000 yearly income, saving $500 a month, $200,000 in current assets, and a “balanced” portfolio, Fidelity, not surprisingly, again found that I don't have and am not saving nearly enough to retire. With Fidelity's calculator I am given three choices as to spending expectations: will I spend less, more or the same in retirement? If I predict spending less, the calculator will estimate 15% less, if I predict higher spending in retirement, then it will calculate 15% more. Those are the choices. If I believe that I might live like the millions of current retirees who live on 40% or 50% less lower pre-retirement income, I am not given any such option. Even if I predict “lower” (15%) spending in retirement, it will most likely be far above what I and my fellow retirees will experience. But the calculator will also have the effect – perhaps the main effect – of scaring users like me into saving more and retiring later than I otherwise would.
This is the general bias throughout the financial services and investment industry – to save as much as possible and work as long as possible. On its face, such “conservative” recommendations sounds good and will help ensure potential retirees have enough for retirement. Since we are often told (as I have also done) that many workers aren't saving enough for retirement, it would seem that such advice is helping to solve that problem. However, when we see such a pattern in their assumptions, advice, and projections that are often based on falsehood, it pays to be skeptical about the motivations of financial institutions when looking for an unbiased, accurate assessment of individual and unique retirement needs.
Financial companies make money by managing assets. That is of course their primary goal, to manage as many financial assets as possible and the bias in their advice reflects their motivation. Financial companies do not want workers retiring at a young age, living modestly during retirement, spending down their assets. They want pre-retirees to stay pre-retirees, saving and investing large amounts of money and delaying retirement – and any possible liquidation of those assets - as long as possible. Pre-retirees are often told they need to save 10%, 15% or even 20% of their pre-retirement paycheck in order to have a realistic chance at meeting these accumulation or retirement income figures. (I always find the recommendations for very high savings rates for pre-retires somewhat amusing. If a pre-retiree is able to save 10%, 15%, or 20% before retirement – that is, able to live on a much lower income now - and despite the intrinsically higher pre-retirement cost-of-living expenses – commuting, raising children, mortgages, taxes, college education, etc., then that essentially proves the pre-retiree can live on much less during retirement.)
The amount of assets required for retirement based on these calculators is often shocking, and discouraging, to potential retirees. Here's a calculator by Merrill Lynch. I again used similar hypothetical information: age 55, $200,000 in assets, $60,000 in yearly income, $500 additional investments per month, a “Moderate” investment style, retiring at age 67, and using the default 85% of pre-retirement income rule-of-thumb suggested by Merrill Lynch. Based on these inputs, the program estimated that I would need about $1.3 million in financial assets to retire. (Estimated life expectancy with Merrill Lynch's calculator, as with some others I've seen, is 93, the 75 percentile in terms of life expectancy for a 67 year-old. One in four will live longer.) According to Merrill Lynch, if I simply continue investing my existing $200,000 but am unable to save any more before retirement, I won't be able to retire until somewhere between age 74 and 81, depending on market conditions.
Compare this estimation of necessary retirement financial assets to the real world. Here is a calculator from DQYDJ.com, using 2016 Federal Reserve data, showing median net worth by age group in the U.S. If I search for median net worth for those in the 60-64 age group, not including home equity (and most retirement calculators also ignore home equity) I find that the median (50 percentile) net worth in this age group is about $106,000. In the 65-69 age group, the median net worth is even lower, at about $95,000, and in the 70-74 age group lower still, at about $78,000.
Comparing these real-world statistics to the more than $1 million estimations for necessary retirement assets based on the Merrill Lynch calculator shows a huge disconnect. In fact, using my example of $200,000 of current financial assets at age 55, I already have more financial assets than the average American in my age group and household income level. And yet, Merrill Lynch tells me that I might need to work another two decades or more to get through retirement. It would be nice to be a 67 year-old among the wealthiest 12% of Americans (the 88 percentile for those in the 65-69 age group), but apparently not essential for seven out of eight retirees who seem to be able to manage despite not having enough “necessary” assets.
Investment return assumptions are inflated
Adding to this mix of confusing financial advice are the average investment returns built into these calculators. It has long been frustrating to see the way that investors are misled regarding expected investment returns. This is the case generally, but particularly disturbing for potential or current retirees, who are more often looking for low-risk investment. Older Americans also usually have less room for error in case of a long bear market or when high returns don't materialize. Incredibly, and despite contrary evidence, there are still financial advisors and calulators telling those in or close to retirement to expect yearly long-term average returns of 8%, 9% or 10% per year for their portfolio. These projections have nothing to do with real-world experiences.
The simplistic way that these calculators (and some financial advisors) offer such projections is to look at historical averages for various asset groups, analyze the investor's projected asset mix and then projects those “expected” returns forward. There are a few problems with this method. First, there is no guarantee that past returns will be repeated in the future. In fact, the only guarantee is that future returns definitely won't be like those in the past, and even long-term returns are continually changing, as the markets swing from overvaluation to undervaluation and vice versa and long-term returns respond accordingly. For example, we can find 20-year stock market returns with negative real (after-inflation) returns and we can find 20-year returns that averaged over 10% per year.
Secondly, assuming market returns will equal investor returns disregards fees, trading costs and commissions, which often add up to 1%, 2%, or 3% of investor assets per year, directly reducing investor returns by the same amount. Those lost investor returns instead go to the multi-trillion dollar financial industry. Lastly, these investment return estimations ignore the propensity for average investors to buy assets that are relatively expensive and sell those that are relative inexpensive. Combining these last two factors results in massive average underperformance for investors. A study by John Bogle covering 25 years (1980-2005) of investor performance showed an average 5% yearly underperformance for stock mutual fund investors. Another study by Dalbar covering the years of 1988-2008 found an average underperformance of stock mutual fund investors of more than 6% per year (1.9% versus 8.4%). Factoring in the asset mix and lower returns of less volatile investments (cash, CDs, bonds, etc.), pushes the real world average returns even lower.
The financial industry starts their deception by using inflated income needs during retirement. They will develop calculators and use benchmarks not based on actual retiree cost-of-living, but will use inflated figures based on a very high ratio of pre-retirement income to post-retirement, oddly assuming that the two lifestyles and costs are the same. Using these inflated retirement income estimations in their calculations, savers and investors will then become convinced that their current savings and rates of saving are not enough. Many of these savers will be convinced to save more, boosting financial industry assets, while possibly leading to unnecessary complications and difficulties for the pre-retiree. Lastly, to boost financial industry assets even more, and to move more assets into higher expense (fees, commissions, trading fees, etc.) investments – namely, stock-based investments - the financial industry will advise the employee that she is not investing “aggressively” enough (even if she is in her 50s or 60s, or simply doesn't wish to invest aggressively). The more aggressive investment recommendations will be justified in order to accumulate the assets “necessary” for a satisfactory retirement.
The notion that everything should be as simple as possible, but not simpler, is generally attributed to Albert Einstein. This idea is critical in the area of retirement planning, which is not simple, even if a calculator or calculation makes it appear so. The assumptions used in retirement calculators are generally based on averages (such as inflation), with the underlying assumption that everyone and every time is average, which is of course not the case. Worse many of these benchmarks are be based on incorrect assumptions (average investment returns, yearly retiree cost-of-living) further widening the potential gap between projection and reality.
Because of all these variables affecting individual retirement needs, I suggest that instead of letting a calculator tell you how much you'll need to retire, that you actually figure it out. That means preparing a tentative budget for your retirement (or at least the first part of it - say 5 or 10 years). This personalized budget can then take into account individual cost-of-living changes in spending, such as housing, college costs, medical spending, etc. Pre-retirees that are saving much of their paycheck (for retirement, for example), can then back those out of income needs during retirement, further reducing necessary income expectations for retirement. Another positive surprise affecting many retirees (and often not accounted for by such retirement income rules-of-thumb) is how much lower taxes are after retirement. Besides generally higher income taxes before retirement, salary income includes 7.65% Social Security and Medicare tax, which pension income, Social Security, dividend or capital gain income does not. In addition, in many states retiree income is treated even more favorably, sometimes avoiding tax altogether. The result is that many retirees will pay much less taxes in retire, or even zero tax. Tax rate differences alone between a pre-retiree and his retirement can sometimes amount to 25% or 30% of income, even for a middle class household.
Because of the often very large changes in spending before retirement to after, many retirees are able to spend less than a third or even a quarter of what they were earning just before retirement, and do so comfortably. At the other end of the spectrum, many retirees (especially those in early retirement) actually spend more in retirement, sometimes twice as much in retirement than in the years before.
The bottom line is that a retirement calculator (often intrinsically flawed) should not be your guide to making one of the most important decisions of your life. If you're going to use a retirement calculator, or heed a recommendation based on one, make sure you know the underlying assumptions built into it and whether those assumptions will meet the reality test as well as your individual circumstances.