Volatility and Sustainable Withdrawals

Volatility and Sustainable Withdrawals One of a series of papers on the Confident Retirement® approach

We will look at an overview of the various studies through the years, and update you on the current thinking with interviews with some of the thought leaders. *Bill Bengen as rocket scientist: “It Was 20 Years Ago Today”, Jonathan Guyton, Journal of Financial Planning, October 2014, volume 27, No. 10; pages 24-25.

By Craig Brimhall, CFP®, CRPC®, CFS® Vice President - Wealth Strategies Since Bill Bengen’s ground-breaking study of just over 20 years ago, there have been many studies on the issue of paying out sustainable income from a diversified portfolio in retirement.* There have been many variations on this theme, and sometimes the “sustainable distribution” numbers move a little higher or a little lower (see side bar). A good example of when this may happen is to incorporate the current environment of record low interest rates with the high likelihood of rates going up and bond values declining. Most historical illustrations, and even Monte Carlo asset return assumptions, show bonds with positive returns (in the neighborhood of 5%, for example), but there is a distinct possibility such returns may not materialize in a rising rate environment. (Incidentally, that has led to the argument that a bond ladder with individual bonds with specific bond maturities is a better fit for this potentially rising interest rate environment in 2015 and forward. Of course, credit risk is still an issue.) But, the gold standard of a 4% withdrawal from the total portfolio at retirement, with inflation adjustments to the spending, still holds water, with the important exception that careful review and cash flow management every year is a must. Few things work on a “set-it-andforget-it” basis. Drawing down a portfolio is definitely not one of them.

Sustainable withdrawal numbers may move a bit higher if time frames are shorter (e.g. shorter life expectancies), higher returns are anticipated or modelled, there is an ability to “flex” spending based on portfolio performance, or not expecting steady inflation adjustments in spending each year. Withdrawal amounts may show lower numbers if time frames are lengthened, lower returns on assets are anticipated or assumed, or cost-ofliving adjustments are built into the assumptions.

EXECUTIVE SUMMARY • Even after 20+ years of research, the old “4% sustainable

VOLATILITY MATTERS . . . . AS DOES TIMING It should be noted that the sustainable withdrawal rates are based on

withdrawal rate from a diversified portfolio over 30 years” still

“worse-case scenarios”, so it is a way to “stress test” your portfolio for

stands the test of time, however, with some important revisions.

survivability. Before Monte Carlo simulations, most tests were done

• Some studies suggest that, due to historically low interest rates right now, a 2.8% withdrawal rate may be more appropriate. • Others suggest that as long as there is “adaptive adjusting” to

using the historical record. Retiring before a bear market, or series of bad years (end of 1999 for example…see illustration below) looks totally different than retiring right before bull markets (the end of 1987,

payouts and not automatic COLAs (cost of living adjustments), that

as an example). The first set of examples assumes an all stock portfolio

number may be as high as 5%. (“Adaptive” or “dynamic” adjusting

of $1 million invested in the S&P 500 Index.

is simply adjusting the payout based on portfolio performance from time to time, thereby preventing the portfolio from too severe of a drawdown in bad market years.) • And taking longevity into consideration, older retirees with a shorter time frame can see their “sustainable” number higher than 5%. • There appears to be universal agreement that there is no logical “set-it-and-forget-it” payout methodology, even at lower rates of distribution. Whereas a low rate of distribution (3% or less) may be

Looking at the charts below…retiree “A” retired year end of 1999 and after he did so, the S&P 500 Total Return gave him a NEGATIVE .95% per year for the next decade, or cumulative -9.10%. Retiree “B” chose 12/31/1987 as her retirement date (right after “Black Monday” occurrence in October), and she saw the S&P return 18.04% per year for the next decade, or a cumulative 425.67%!!

sustainable with a COLA every year, the chance of underspending

This is the same index (not identical investments since the stocks in the

and leaving a very large “terminal value” or large estate then

index do change), and a similar time period; it just shows how

appears.

unpredictable the returns over time can be. (Illustrations run on

• The “right” combination, although this is definitely a decision that

Morningstar Advisor Workstation.)

each investor needs to make, is probably a high enough rate of distribution to be sustainable, with adaptive adjusting every year (or as needed) to keep overall portfolio balances appropriate, and aiming to whatever final estate amount the investor desires to leave to others.

ILLUSTRATION #1: TIMING MATTERS – RETIREE “A”

Page 2 of 21

Incidentally, retiree “A” (above) was probably feeling good about

the history of the stock market (October 19, 1987…the Dow Jones

retirement at first because the mood of the country was almost

Industrial Averages shed 508 points or 22.61% in one day), and a

euphoric (the “happy days are here again” feeling of the technology

recession was headed our way in the summer of 1990, not to be over

boom). But then came a recession. Actual returns were probably

until the spring of 1991. However, actual returns were surprisingly

shockingly disappointing.

good.

And retiree “B” (below) might have been a bit apprehensive since the

Feelings may not be reality.

country was just starting to climb out of the biggest one-day decline in

ILLUSTRATION #1: RETIREE “B”

Whereas the timing of one’s retirement may, or may not, be voluntary,

It has been said that “high risk means high returns.” That is a myth

one thing we can control is our asset mix.

since the first half of the sentence does not assure the second half.

Will being an aggressive investor save the day?

High risk means high volatility and higher probability of uncertainty. If you hit it right, you can reap great rewards, but if not, it may spell ruin.

Page 3 of 21

ASSET MIX MATTERS

ILLUSTRATION #2; chart 2: 12/31/1999—12/31/2014: 50% stocks

Let’s look at another scenario similar to illustration #1 above, but now

(S&P 500 TR) and 50% bonds (Barclay’s Aggregate) also with a 4%

looking at it from the perspective of two retirees during the SAME time

withdrawal rate and a 2% COLA. The portfolio rebalances if the 50/50

period, but with different asset mixes.

ratio is out of balance by 10% or more.

The charts below demonstrate the relative stability of a diversified

Here is a period of 15 years when assumptions about returns (stocks

portfolio versus the all stock portfolio.

should outperform bonds over the long term) did not hold up. And the

ILLUSTRATION #2; chart 1: 12/31/1999—12/31/2014: all stocks (S&P 500 TR…Total Return); with a 4% payout (based on the beginning balance at retirement) and a 2% COLA.

concern about volatility with stocks is showing up in spades. Although both portfolios are still alive (not depleted), the “balanced” portfolio still has close to its original value, whereas the “all stocks” portfolio is down by 50%.

ILLUSTRATION #2; chart 1: ASSET MIX AND WITHDRAWAL RATES MATTER TO PORTFOLIO SURVIVABILITY ALL STOCK PORTFOLIO (S&P 500 TR) 12/31/1999-12/31/2014

All stock portfolio has not run out of money (yet), but the account has lost half of its value. Would you have “stayed the course”?

Page 4 of 21

ILLUSTRATION #2; chart 2

A 50/50 mix of stocks and bonds paid out the same income as the all stock portfolio but has close to the same beginning balance it started with 15 years prior. During this time period, volatility was lower and returns were higher.

Bad timing and a volatile portfolio can work against a retiree and lead to

not), and (2) the assumed increase or adjustment in withdrawals taken

a depletion of assets.

in subsequent years (the “inflation adjustment” or COLA…cost of living

Many of the tests that have been run are either projections of future returns, or “rolling periods” of actual past returns, or other similar methodology. They almost all universally show that income should be sustainable at about the 4% distribution level (assuming 30 years of retirement and roughly a 50/50 stock/bond asset mix based on the portfolio balance at the time of retirement).* However, of the many factors that may affect the outcome, the two biggest appear to be: (1) the portfolio asset allocation (and whether it rebalances every year, or

adjustment…makes a big difference to survivability of a portfolio). Built in to the assumption of any asset allocation are assumptions about volatility. Even a higher-returning asset (in the long run) can cause problems in the shorter-term if volatility is high. Downside protection is a major concern for most retirees. (It should be noted that the 4% rule has been challenged in light of the historically low interest rates and relatively high valuation for stocks at this particular point in time, in the summer of 2015.)

* one notable exception being a paper titled: “Low Bond Yields and Safe Portfolio Withdrawal Rates” by Blanchett, Finke, and Pfau

Page 5 of 21

“SEQUENCE OF RETURNS” RISK

identical withdrawal rates. The ONLY difference in the two scenarios is

Although much has been written about “sequence of returns” risk, it

the start date of distribution. Since the first retiree, John, was lucky and

may be news to many investors that the order of returns is one of the

retired at the beginning of a long bull market, we decided to “break the

most critical factors in how long money may last in the distribution

rule” and have withdrawals way above the accepted or prudent amount

phase of life. It is also a factor during accumulation if an investor is

of 4%...we went for 6% AND a 4% COLA. The point will be, as you will

dollar-cost-averaging, but, it makes absolutely no difference in the

see…hit the “retirement day lottery” right, and you could go well above

resulting ending balances if you have invested a lump sum.

what might be considered “safe” or prudent. Be unlucky enough to

Consider this interesting math:

retire right before a bear market, well…you’ll see what happens in the illustration.

Take a lump sum and apply the returns in following order, two scenarios: (1)

+20%, +10%, 5%, -10%, and -15% As an example, in this first scenario, a $250,000 lump sum ends up with $265,073, or a compounded annual growth rate (CAGR) of 1.177% Now reverse the order of returns:

(2)

JOHN: Start date: Withdrawal rate: COLA (cost of living adjustment) Beginning balance: Total withdrawals: Ending balance on 02/28/2015: Average annual return:

12/31/1990 6% 4% $1 million $2,498,754 $1.75 million 9.97%

BOB: Start date: Withdrawal rate: COLA (cost of living adjustment) Beginning balance: Total withdrawals: Ending balance on 02/28/2015: Average annual return:

12/31/1999 6% 4% $1 million $1,309,472 $51,982 3.80%

-15%, -10%, 5%, +10%, and +20% $250,000 in this second scenario ends up at $265,073; CAGR = 1.177%

However, if you were to take the numbers above and start to withdraw money from a portfolio with those returns, the first scenario will survive a 4% annual distribution ($1,000,000 taking out $40,000 at the beginning of each year (no COLA); those returns will end the 5th year with a balance of $885,805. This is also its lowest balance; the high balance was over $1.2 million). In the second scenario, identical returns but in the opposite order, the investment survives, but the final balance at the end of year 5 is $811,742 after falling to a low balance of $691,320. Not only is the final number $74,063 lower, but would someone stick with a plan that was down $308,680 at one point (-30+%)? This is another major factor in the discussion of a portfolio’s “survivability.” From a mathematical standpoint, it may survive, but from an investor behavioral standpoint, the ship may have been abandoned along the way (usually at the low point).

What makes these illustrations even more remarkable is that John has had 9 more years of distributions (and COLA raises) and is still going strong. Bob had 9 fewer years in his retirement and is just about ready to run out of money (and his portfolio has paid out only a little more than one half of John’s). Keep in mind that John and Bob suffered through the same bad years. However, for John, they happened later in retirement so did not affect him as much. Bad market years early in retirement are why portfolios

When we look at a series of returns historically, we have the benefit of

should be “stress tested” for prudent withdrawal rates. If an investor is

knowing the end result. An investor does not have that luxury since

lucky enough to hit the jackpot and start withdrawals in a year like a

returns are in the future. So to personalize it a bit more: would you

1990, then more money should be available. If an investor draws the

stay with a portfolio and payout plan if you were down 30% or more?

short straw and enters retirement into a bear market, then starting with

Risk tolerance is that ability to hang tight in tough times.

a low withdrawal rate (until markets recover), or “adaptively adjusting”

AN EXAMPLE OF SEQUENCE OF RETURNS RISK IN A RETIREMENT SITUATION

withdrawals based on portfolio performance, seem to be the wise ways to help make the money last.

We will look at two potential retirees in the following illustration: they have identical portfolios (a 50/50 mix of stocks and bonds) and

Page 6 of 21

Not knowing what type of market environment you may face as a retiree has been called: “the retirement day lottery.” Your ticket may have you retiring right before a raging bull market (like John) or your retirement day might just have you right at the beginning of a bad bear (like Bob). Or, of course, you may experience something in between. It’s because it is a big unknown that we would suggest you start with a prudent withdrawal rate and be willing to make adaptive adjustments to that income each year, depending on portfolio performance and circumstances.

ILLUSTRATION #3: PUTTING IT TOGETHER: JOHN AND BOB…AND THE LUCK OF THE DRAW; JOHN’S SCENARIO 12/31/1990 to 02/28/2015

John starts withdrawals 12/31/1990 at 6% of the beginning balance plus 4% cost of living adjustments each year. He got lucky as the markets entered a long bull market.

Page 7 of 21

ILLUSTRATION #3: PUTTING IT TOGETHER: JOHN AND BOB…AND THE LUCK OF THE DRAW; BOB’S SCENARIO

Bob starts withdrawals 12/31/1999 at 6% of the beginning balance plus 4% COLA each year. He was not so lucky as the market entered a bear market.

BACK TESTING VS FORWARD-LOOKING

(and, of course, an average of all of the scenarios). This modelling is

“Back testing” is using the historical record to test how investments or

where the “probability of success” comes from. A retirement portfolio

portfolios may have done during a specific period of time. This is what

with a stipulated payout might have an 80% “probability of success and

is used in each of the illustrations used so far in this paper. One of the

a 20% probability of failure.” Of all of the stochastic (Monte Carlo)

weaknesses of using the historical record is that is it highly unlikely, if

models we examined, most show that a 4% withdrawal, with some

not impossible, that history will literally repeat itself with the same or

variation on a COLA adjustment each year, is survivable for a 30 year

similar events, in the same order and magnitude as it has in the past.

retirement. HOWEVER, when adjustments are made to the modelling

But as Mark Twain put it: “History doesn’t repeat itself, but it does

assumptions, such as the current historically low interest rates and high

rhyme.” And that is why it may be useful to at least look at past years to

equity P/E ratios (thereby lowering return assumptions for both stocks

see how portfolios performed. They may not repeat exactly, but the

and bonds), the 4% withdrawal may be too high.

“rhyme” or rhythm of the economy and market cycles do have a somewhat familiar pattern. The historical record can give us a good glimpse of what might have happened to a retirement portfolio if a

“Success” by many of the models means that you have at least one dollar left at the end of the time period chosen to run the simulation

person had been retired and lived through a particular period.

(e.g. 30 years). Some people would not consider it a “success” if they

If you look at how a portfolio may have done in a period that had very

planning. “Failure” doesn’t necessarily mean you would run out of

negative markets right after retirement, that is what we referred to as

money. It simply means that IF the investments performed the way the

stress testing a retirement portfolio. Of course, if you plan for the worst

modelling shows and IF no course corrections were made along the

and hope for best, and if you retire into more of a bull market than a

way, the retiree might run out of money. And, there are lots of things

bear, everything may turn out better than expected, and you may have

that can be done year by year to prevent the disaster from happening.

nice “raises” along the way. What can we surmise about the future?

ended their life with a dollar, while others might say that was perfect

It is important to note that while the simulations are “scientific

That is where the models called “Monte Carlo simulations” come into

guesswork” and the probabilities are mathematically sound, they are

play. They attempt to be forward looking. Certain assumptions are

still assumptions, and tiny tweaks in the assumptions (such as asset

made about asset performance, inflation, interest rates, economic and

returns, inflation, etc.) can quickly change the outcomes. The

market conditions and a host of factors, and then a portfolio scenario is

calculations may be objective, but the inputs that make up those

run though thousands of possible outcomes. The results create a

calculations, unlike physics, are subjective.

“best” outcome, a “worst” outcome” and many scenarios in between

Page 8 of 21

We make those comments only to emphasize the point that

But the issue of the timing of your retirement distributions, as shown in

summarizes this paper, and the findings and interviews you are about

the illustration of John and Bob, can greatly affect the amount of

to read: the biggest factors to having a portfolio last a lifetime, and

withdrawals that will be sustainable. A declining portfolio while

longer, is the withdrawal or spending rate and the ability to make

withdrawing funds accelerates the drawdown. This is especially

changes in spending as the portfolio values change. The total return on

problematic during the early years of retirement. If the decline occurs

the portfolio is very important, and volatility control is a big issue, and

later in retirement when the portfolio presumably has grown sufficiently,

the “sequence of returns” risk must be managed, but the biggest factor

it is less critical for a couple of reasons: (1) hopefully a higher balance

the consumer directly controls is the spending level being taken from

when compared to the income that needs to be withdrawn, and (2) a

the portfolio.

shorter time frame left in which the withdrawals must be made.

It is interesting to note that automatic dollar-cost-averaging into a highly volatile portfolio actually can work very well for an accumulator, but doing the same during distribution can become “dollar cost ravaging” since a portfolio can experience a significant drawdown from which it may not recover. The math of periodic contributions and distributions are 180 degrees apart. During accumulation, you can buy more shares with a fixed dollar amount if the price falls, so an eventual recovery can significantly lift your portfolio. During distribution, you must sell more shares at lower prices to generate a fixed dollar amount, and an eventual recovery may not prevent your portfolio from failing, so a much more thoughtful approach is needed for those in the payout phase of life versus the accumulation phase.

THE EFFECT OF VOLATILITY ON THE MATH OF SUSTAINABLE WITHDRAWAL RATES

of success” lower for all equity portfolios than for more balanced

A higher returning portfolio (in the long run) might end up also being a

investors to stay the course if they see their portfolios down a

higher volatility portfolio; but that does not mean it will last longer. Two

significant amount. Therefore, seeking to find a lower volatility portfolio

reasons: (1) high volatility may mean that withdrawals that occur in

with consistency may end up being more appropriate and lasting longer

really bad market years may draw down the portfolio enough that

than the “home run” portfolio with higher volatility. As it has wisely

recovery becomes difficult and that is why you may see the “probability

been said of baseball players and investment portfolios:

portfolios; (2) behavioral finance reasons – it may be very difficult for

“It is better to be consistently good than occasionally great”

Page 9 of 21

The chart below is taken from one of the most well-known studies

bonds; (2) the highest success rates, IF the rate of withdrawal was in

completed right after Bill Bengen’s seminal work of 1994. It was

the 4-5% range, seems to be somewhere near a 50/50 mix, with a

published in 1998 by professors at Trinity University, San Antonio, and

slight bias to bonds in the “no COLA” example, and slight bias to stocks

is now referred to as “the Trinity Study”. Its official name is:

in the “COLA” example, and (3) higher weighting of stock above a 50-

Retirement Savings: Choosing a Withdrawal Rate that is Sustainable.

75% concentration did not seem to improve the probability of

It is authored by Philip L. Cooley, Carl M. Hubbard, and Daniel T. Walz.

success…it actually decreased it slightly. However, not shown here is the interesting phenomenon that occurred if the withdrawal rate was

These are only two lines from a larger chart in their work, but some

ABOVE 5%, and then a higher concentration of equities, even though

interesting things come to light when the numbers are examined: (1)

failure rate was increasing, still had a better chance of survival than did

the lowest success rate happened to portfolios that were too heavy in

the other mixes with a higher percentage of bond holdings.

30-year retirement horizon: no COLA; data from 1926-1995. Probabilities of success – withdrawal rate 5%

100% bond

75% bond

50/50

75% stock

100% stock

51%

100%

100%

98%

95%

20%

71%

95%

98%

95%

with COLA 4%

"Probabilities of success" means having some money in the portfolio at the end of the period. NO COLA data: Numbers rounded to the nearest whole percentage. The number of overlapping 15-year payout periods from 1926-1995, inclusively, is 56; 20-year periods, 51; 15-year periods, 46; 30-year periods, 41. Stocks are represented by the Standard and Poor’s 500 Index, and bonds are represented by long-term high-grade corporates. Data source: author’s calculations based on data from Ibbotson Associates. With COLA: same data as first footnote; however, the withdrawal amount is increased by inflation (or decreased due to deflation) as measured by the Consumer Price Index (CPI).

In summary, it appears as though adding more potential return with

age), Pfau and Kitces comment that their research has “significantly

higher volatility (standard deviation) to the mix does not add to the

disruptive implications.”

probability of success, but may possibly detract from it slightly. This issue of changing the mix of investment DURING retirement has received quite a bit of attention lately, with significant work being done by Wade Pfau and Michael Kitces (January 2014: Reducing Retirement

However, they go on to comment that “The success of a retirement scenario is heavily influenced by the sequence of returns.” Dr. Pfau had examined that issue in a paper of September 2013 titled:

Risk with a Rising Equity Glidepath).

The Lifetime Sequence of Returns: A Retirement Planning Conundrum.

In that article, after a thorough discussion of literature review on the

He examined both the accumulation period up to retirement, and the

subject, Pfau and Kitces ran 10,000 Monte Carlo simulations for 121 various glidepaths (asset allocations that change over time) and made these conclusions: “Across all time horizons and withdrawal rates, when examining the approach which provides the highest sustainable withdrawal rates…the results consistently show support for rising equity glidepath portfolios. Generally, depending on the underlying assumptions, the optimal starting equity exposures are around 20 – 40% and they finish at around 40 – 80%.” It is interesting to note that since this conclusion appears to be so contrary to the “accepted wisdom” (one decreases equity exposure with

distribution period after retirement, and came to these conclusions: “…the compounded return in the last 15 years (prior to retirement) explains 65% of the wealth accumulation.” Further, “the compounded return in the first 10 years of retirement can actually explain 77% of the final retirement outcome.” This would explain why it makes so much sense, mathematically, to reduce equity exposure during the most critical time right before and right after retirement, and then build it up later. Pfau goes on to explain that the very first year of retirement accounts for more than 14% of the final outcome for all of retirement cash flows,

Page 10 of 21

and the first 5 years account for over half of the probabilities of

(1)

“success” or “failure.”

withdrawals):

Some have referred to this period right before, and right after, retirement as “the red zone,”* and Pfau concludes his comments with: “Sustainable withdrawal rates are disproportionately explained by what happens in the early part of retirement.”

(2)

a.

7.56% aatr (average annual total return)

b.

4.23% standard deviation (10-year annual average)

c.

30.46% best year (10/81 – 09/82)

d.

– 4.46% worst year (11/93 – 10/94)

S&P 500 TR since 01/31/1973 (assumes no withdrawals): a.

If we look at the math of a decline, it is easy to see why it is so important to avoid downdrafts. Although true during accumulation, it is magnified in importance during distribution since money is coming out at the same time values are declining.

Barclay’s Aggregate Bond since 01/31/1973 (assumes no

(3)

10.43% aatr

b.

14.76% standard deviation

c.

61.18% best year (07/82 -- 06/83)

d.

-43.32% worst year (03/08 – 02/09)

50/50 mix of the two indecies

Notice that if a portfolio declines by 10%, it only needs slightly more

a.

than an 11% return the next year to get back to break even. $100,000

b.

7.02% standard deviation

X 90% (a 10% loss) = $90,000 X 1.1111 (an 11.11% return) =

c.

40.93% best year (07/82 – 06/83)

$100,000. But a portfolio that declines 25% needs a 33% return the

d.

-17.96% worst year (03/08 – 02/09)

next year to get back to even! $100,000 x 75% (a 25% loss) = $75,000 X 1.3333 (a 33.33% return) = $100,000. Low volatility portfolios are very important for two reasons: the math, and investor behavior.

THE RATE OF WITHDRAWAL DOES AFFECT YOUR INVESTMENT RETURN The average annual return of an investment, or an asset allocation, will change as soon as you change from a lump sum investment to making additional deposits or make various withdrawals. Small amounts may change the final calculated AATR (average annual total return), but larger withdrawals may change it substantially, and the more volatile the investment, the more likely that will happen.

9.78% aatr

Mixing in bonds (or some other portfolio stabilizers) may reduce overall return, but greatly reduces volatility…and volatility matters. Notice that the standard deviation falls in half (7.02% vs. 14.76%) in a 50/50 portfolio compared to all stocks. But the return was only .65% per year less. A word of caution about any lessons to be learned from the previous paragraphs: they are hypotheticals looking back at history (sometimes referred to as “back testing”). And, bonds in particular have a very interesting anomaly to consider: interest rates peaked in 1981 and have been in fairly consistent decline until very recently when they bottomed at all-time lows.** Bond values increase as interest rates

The reason is quite straightforward: additions or withdrawals mean you

fall; therefore, we just went through a 30-year bull market in bonds that

are buying (or selling) shares at differing prices than the original lump

most likely cannot be repeated anytime soon since rates cannot

sum and the time period measured for the AATR is now also different

continue to fall lest they become negative. Therefore, in imagining

for those additional transactions.

returns in the future, different assumptions probably need to be built

Here are some examples of that using Morningstar: • A portfolio that is a 50/50 mix of S&P 500 TR (Total Return) and Barlcays Aggregate Bond Index from 01/31/1973 until 02/28/2015 had an average annual return of 9.78%/year.

into simulators since the future bond returns probably can’t replicate the recent past (at least the past 3 decades). The Barclay’s Aggregate (bond index) from 12/31/1980 to 02/28/2015 averaged 7.93% AATR (average annual total return), and this is probably not a reasonable assumption to make in 2015 looking forward.

• The same portfolio and time frame, with a 4% withdrawal rate and 2% COLAs shows an AATR (average annual total return) of 9.60%. • The S&P 500 TR from 01/02/1973 to 02/28/2015 had an AATR of 10.37% • The same index during the same time period with 4% withdrawals and a 2% COLA each year shows an AATR of 9.65%. Volatility matters. Volatility can affect the “probability of success” just as total return can. Listed below are statistics from Morningstar Advisor Workstation for (1)

*”Retirement Red Zone” is a registered trademark of Prudential Insurance Company of America. **federalreserve.gov: selected interest rates/historical data

the S&P 500 TR, (2) the Barclay’s Aggregate Bond Index, and (3) a 50/50 portfolio. Take a close look at the standard deviation numbers (volatility) and the “best year”, “worst year” statistics.

Page 11 of 21

“SAFE” (SUSTAINABLE) WITHDRAWAL RATES…REVISITED

Today’s retirees will be more strained to spend principal to achieve a

On January 21, 2013, Morningstar Investment Management published

4% sustainable withdrawal rate. If we assume that the historical risk

a paper by David Blanchett, CFA, CFP® (Head of Investment Research

premium for stocks and other asset characteristics remains the same,

for Morningstar) and Michael Finke, Ph.D., CFP® (Professor at the

but we adjust the average return on stocks and bonds downward by

Department of Personal Financial Planning at Texas Tech University)

2.76% to reflect today’s lower bond yields, we obtain higher failure

and Wade D. Pfau, Ph.D., CFA (Professor of Retirement Income at the

rates for the 4% rule.

American College).* They make the argument that “the average annualized bond total return over the next 10 years is expected to be 1.8%...this return is almost 4% less that the 5.5% average annual return on the Ibbotson Intermediate-Term Government Bond Index from 1930-

Figure 2 repeats Figure 1 but includes portfolio success rates for Monte Carlo simulations based on today’s low yields. The ability to make these adjustments is unique to Monte Carlo simulations. These

2011.”

adjustments question the safety of 4% as a sustainable withdrawal

And incidentally, it is in this paper which has the now-famous statement

allocations between 80% and 100%, and the success rate is only 73%

about a 2.8% withdrawal rate being the new “safe” withdrawal rate,

in these cases. With low bond yields, it is extremely difficult for a bond-

replacing the 4% standard.

heavy portfolio to sustain 4% withdrawals, and success rates fall as low

The paper goes on to make convincing arguments that we should “adjust our expectations” about asset returns in the upcoming years since we are starting from record low yields (and therefore

strategy. The highest portfolio success rates are seen with stock

as 10% for an all-bond strategy (based on bond funds, not laddered individual bonds).” (The chart on page 13…Figure 2…is from the same article)

unsustainably high prices for bonds). And price/earnings ratios (P/E) for stocks are elevated as well, when compared to the historical averages. Currently, (the summer of 2015), bond yields are still wallowing near record lows** and stock P/E ratios (whether you looked at trailing P/E, estimated forward P/E, or Robert Shiller’s adjusted P/E) are at elevated levels, it may be time to adjust return expectations downward to some degree and then run Monte Carlo simulations and look at the results. In a similar vein, below are a few sentences pulled directly from a recent article in Advisor Perspectives (September 2, 2014) titled: “Can Returns Still Use a 4% Withdrawal Rate?” “Today we are dealing with a situation in which Shiller’s cyclically adjusted price-to-earnings ratio (CAPE) is well above historical averages, while bond yields are at historical lows. Retirees are also exposed to sequence-of-returns risk, as the returns experienced in early retirement will have a disproportionate impact on their final retirement outcomes. When spending down their principal in the meantime, it will be unfortunate even if interest rates and market valuations “normalize” in the next 5-10 years. Today’s high-valuation/low-yield situation has been quite rare in U.S. history, indicating we are in uncharted territory when trying to determine if the 4% rule will remain a safe strategy. Historical simulations don’t analyze this possibility, but with Monte Carlo simulations we can adjust our capital-market expectations to better account for the types of returns that are more likely to be experienced in the future. Here is an example to demonstrate the vulnerability of the 4% rule. On Aug. 15, five-year Treasury Inflation-Protected Securities were yielding negative 0.28%. This is 2.76 percentage points less than the historical

* Low Bond Yields and Safe Portfolio Withdrawal Rates, by David Blanchett, CFA, CFP, Michael Finke, Ph.D., CFP, and Wade Pfau, Ph.D., CFA; Morningstar Investment Management, January 21, 2013. ** federalreserve.gov: selected interest rates/historical data

2.48% real return experienced by intermediate-government bonds.

Page 12 of 21

(End of direct quote and chart from Advisor Perspectives.)

Notice that the green “probability of success” line above in the chart

Although Monte Carlo simulations had been in use for some time (the

(which is built from assumptions about the current market environment)

original “Monte Carlo” dates back to 1946 and the name…it is

is appreciably lower than the more commonly used assumptions and

rumored…comes from the methodology being used to assist in

historical record of the lines (blue and red) shown near the top of the

estimating how to win at certain casino games in, of course, Monte

graph. The point? If regression to the mean is true, and we are at

Carlo).

historically low yields in bonds and relatively high valuations in stocks, then we may need to change our assumptions and expectations about returns in the upcoming years.

Simulations of asset returns and retirement income projections had not come into the fore at the time Bill did his seminal work, so he used the historical records and the now-famous “4% rule” assumed static

Does this mean the 4% rule won’t work? Not necessarily…but it may

automatic increases in payouts. (For example, a withdrawal of $40,000

not be quite as “safe” as some have assumed. Probably more

on $1,000,000 (4%) would go up the next year by a specified cost-of-

important than the initial “rate” of withdrawal are the annual

living increase; e.g. $41,200 is a 3% increase over the $40,000 starting

adjustments to that withdrawal.

figure). The goal was to keep retiree’s spending at the same “real”

Bill Bengen first developed the idea of a 4% sustainable withdrawal back in October of 1994 when his work was first published in the Journal of Financial Planning. Mr. Bengen, in addition to being a CFP (Certified Financial Planner), also has a degree in aeronautics and astronautics from MIT. Among his many published works, he coauthored “Topics in Advanced Model Rocketry” back in 1973 (Journal of FP October 2014, volume 27, No. 10). So we believe it may be safe to say that finding an appropriate withdrawal rate isn’t rocket science. Mr. Bengen’s work relied on the historical records of asset returns.

(inflation adjusted) level. Modern analysis of Bill’s work has made three major adjustments. • First is to use Monte Carlo simulators that “project” possible future scenarios and see if a payout rate is sustainable. • Second is adjusting the payout period based on remaining life expectancy rather than a standard period (and a variation on that theme is the IRS table for Required Minimum Distributions, or RMDs)

Page 13 of 21

• Third, and probably the biggest, is the adoption of withdrawals being “dynamic” or “adaptive” to how your portfolio is performing and what the markets and economy are doing.

Annuities (RiverSource Life Insurance Company). Below are summaries of those conversations. As evidenced in their biographies (located in the endnotes), these

Concerning the second point about life expectancies, the IRS Table I in

thought leaders have made major contributions to the issue of safe or

Publication 590 can be used as a divisor (denominator) since the idea

sustainable withdrawal rates in retirement.

is that dividing an account balance by the number corresponding to age (and the divisor changes every year as the person ages) should liquidate the account by the end of life expectancy. The table terminates at age 111. Assuming you are certain you would make it that long, you could use this as a guide and maybe “round up” a little bit, but it is interesting to note that the factor for age 60 is 25.2 and for

THE INTERVIEWS, AND THE ANSWER: START WITH A “PRUDENT” PAYOUT AMOUNT, AND MAKE YOUR PAYOUTS “ADAPTIVE”: BILL BENGEN

age 65, it’s 21. Take an account balance and divide by 25.2 and you

Mr. Bengen, who is now happily retired from his financial planning

will see that the distribution percentage is just under 4% for someone

practice in Southern California, talked about what he had learned over

age 60, and it’s just a bit over 4.7% for someone age 65. The

the years in helping so many retire, and how he felt about his own

withdrawal percentage is not necessarily static, but as a person ages,

withdrawal rate in retirement. He commented how important portfolio

he or she should be able to take slightly higher percentages of the

reviews are, at least annually, and that Monte Carlo simulations should

remaining balance since life expectancy (the period of years remaining)

be rerun each year as well. He said he starts getting nervous when the

is getting shorter.

“probability of success” drops below 90% and heads toward 75% or

For JOINT life expectancies, the factor will be larger, making the withdrawal percentage lower. But that is the point…joint life

80%. The former rocket scientist said that whereas the laws of physics are predictable; financial markets are not.

expectancies are longer than either individual, so the time period for

He directed his retiring clientele to consider four possible steps to deal

paying out income will be longer; therefore the “safe” withdrawal rate

with the sustainable withdrawal issues:

will be lower, assuming a married couple wants income for life for both.

(1) use more actively-managed funds over passive strategies (his

After many years of testing, studying, calculating, researching and

point about “matching” the S&P 500, which would be a passive

publishing, what are the conclusions? Does the old “4% rule” created

strategy: “Who cares? Do you want to ‘match’ negative 38% like

by Mr. Bill Bengen over 20 years ago still hold water? In essence, yes it

2008?”),

does, but with thoughtful revisions. Keep in mind that “variations on a theme” might have to do with one or more of the following: (1) assumptions about capital market (asset) returns and volatility in the future, and inflation, (2) the time period to take those withdrawals, and (3) whether or not you want to assume inflation-adjusted withdrawals, or a more “dynamic or adaptive” approach to withdrawals (some years you may take more, some less, and some no raises, etc.).

Although there are many articles and studies published on this issue, one we would recommend is from The Journal of Financial Planning,

(2) raise cash, (3) look at hedging strategies against equity volatility, and (4) using annuities can be very helpful to hedge longevity risk. He had a very thought-provoking comment when he said he has a derivation of Warren Buffet’s saying about never losing money: “Rule #1: ‘never have a big loss’, and Rule #2: ‘never forget rule #1.’” Additionally, Bengen recommends that all investors should keep these three things in mind as they enter retirement: (1) lower your expectations about market returns; (2) be prepared to adjust your spending through the years; and

October 2012 (Vol. 25, No. 10). Its cover article is entitled: “Safe Withdrawal Rates: What Do We Really Know?” It features a retirement

(3) Expect the unexpected – protect your capital.

roundtable of very well-known professionals who publish a lot about these issues: Bill Bengen, Jonathan Guyton, Wade Pfau, and Michael Kitces.

MICHAEL KITCES

In an effort to answer these questions brought up in this paper, in

Mr. Kitces, a nationally recognized and frequently published academic

February of 2015, we decided to set up phone interviews with all of the gentlemen mentioned in the paragraph above, plus David Blanchett of

in this field of retirement planning, had very similar discussion points as

Morningstar, and Doug Dunning and Steven Wolfraith of RiverSource

Page 14 of 21

Bill. His greatest emphasis is about adjusting withdrawals as the

Take a portfolio balance and divide by 31 and notice what happens:

portfolio changes.

$1,000,000 / 31 = $32,258, or a 3.2% distribution rate. Keeping in

He noted that spending rates determine which is more important – total return or volatility. His research indicates that a lower rate of

mind that the table does go out past age 100, and you may want to “adjust up” just a little bit.

withdrawal will be most affected by the overall return of a portfolio, but

Pfau’s point is that life expectancy tables are a factor and the “RMD

a higher withdrawal rate can be more affected by the volatility of that

methodology” is not a bad course of action, but with some revision.

portfolio. Recently, he and Wade Pfau completed a study that lead

Your payouts would adjust annually since your portfolio balance would

them to conclude that a “U”-shaped glide path concerning equities

change and your IRS table factor based on age would change.

(stocks) may make the most sense (gradually reduce equity exposure right before retirement, and gradually take on more exposure after retirement). This concept combines the sequence of returns risk with the desire for the best overall total return long term. His point is: the last decade before retirement is the most important in getting assets accumulated prior to drawing income (savings should begin much earlier, but the compounding effect becomes much more pronounced in the later years), and the first decade after retirement is

He also emphasized the importance of the “red zone”* -- the last few years leading up to retirement and first few years of taking distributions after retirement. He believes that retirees should “reduce portfolio volatility when they are the most vulnerable.” In addition to his research (along with Mr. Kitces) into the “U” stock allocation glide path, he is currently researching “valuation-based asset allocation for retirement”, which infers that you skew your allocation in

the most critical in dealing with possible negative sequence of returns.

favor of stocks when the P/E ratios are below average and you would

With this in mind, it is his contention that reducing equity exposure right

theory is predicated on the fact that financial assets should “regress to

before retirement should help to avoid any potential downdrafts at your

a mean.”

time of great decisions. Then slowly build it back up as you age in retirement. Down markets during the first few years of retirement have a more negative effect on a portfolio than negative years later on,

reduce your stock allocation when P/E ratios are above average. This

Pfau said that retirement should be thought of in the same way a pension plan treats it: streams of cash flows that need to be funded. In

assuming balances have grown and withdrawal rates are in line.

other words, each person’s retirement has a “funding ratio,” which

Kitces, like Bengen, had a similar threshold with his clientele: “I get

sufficient assets, in other words, you are “fully funded” or over funded,

nervous if the probability of success gets down in the 75%-80% range; if

you have the luxury of doing other things (spend more, gift more, etc.).

it’s way above 90%, we want discuss this as well…maybe do some

However, if you are underfunded, then that suggests other choices,

gifting? Or spend some more? Or do a ‘set aside’ for something later,

such as: add more money to your accounts if at all possible, delay

just in case.”

retirement, spend less, make wise Social Security or pension choices,

The best piece of advice from the interview was his take on how a

would be the present value of the needed income streams. If you have

do part-time work, consider a reverse mortgage, etc.

conversation should occur if an investor sees the probability of success

Pfau believes that in this era of potentially rising interest rates, using an

(or failure) at an unacceptable level.

income annuity as a replacement for bonds makes sense in some

Concerning the probability of failure, Michael said, “We probably should stop calling it that. To prevent failure, it is really just a ‘probability of adjustment.’ If the retiree is willing to, and does, adjust his or her

cases. *”Retirement Red Zone” is a registered trademark of Prudential Insurance Company of America.

spending along the way, then that should just about eliminate the probability of failure.”

WADE PFAU Dr. Wade Pfau, a noted academic in retirement planning, has some additional variations on this theme of sustainable retirement income. He brought up the required minimum distribution (RMD) table published by the IRS, which calculates a divisor that should liquidate a balance over the remaining life expectancy of an American male or female. This number changes for each year you age. As a quick example, the IRS table in Publication 590 (used for determining Required Minimum Distributions from IRAs, for example), shows the factor for someone age 65 to be 31.

JONATHAN GUYTON Mr. Guyton is a planning practitioner and published researcher. Mr. Bengen, Mr. Kitces, Dr. Pfau, and Mr. Guyton are not only contemporaries, but have also shared ideas and research over the years. As an active financial planner, Mr. Guyton has some applicable comments regarding financial behavior. Since expectations drive much of our behavior, he believes that advisors need to emphasize the importance of setting proper expectations. Guyton says people have a tendency to “do the right thing when in withdrawal mode, if they understand the consequences. When things get bad, we get cautious;

Page 15 of 21

when things are great, we spend more.” He expresses how important it

Modern portfolio theory (MPT) statistics use terms like “alpha” (excess

is to NOT “set-it-and-forget-it” but to plan on adjusting withdrawals

returns above expectations) and “beta” (volatility compared to a

regularly.* Guyton’s “Big 3” rules are: (1) Set reasonable expectations, (2) Review consistently, (3) Adjust withdrawals appropriately by following the strategy you’ve chosen.” In an article in the Journal of Financial Planning,** Guyton advocated the creation of a “set aside” or slush fund with some extra funds (not counted as part of the total portfolio on which withdrawals are based). His justification is that unforeseen events or a change of mind could mean unanticipated spending. A set-aside fund is specifically designed so retirees will not interrupt their withdrawal plan from the core portfolio. They can “raid” the slush fund instead.

benchmark), but Mr. Blanchett introduced a concept called “gamma,” which is a more nuanced idea. Gamma is defined as a great financial planning relationship which adds value above and beyond the possible alpha generated by a well run portfolio. Gamma is a combination of ideas such as proper asset allocation, annuity allocation, tax-efficient investing and drawdown ideas, and dynamic withdrawal concepts which adds additional value and wealth through smart techniques.

DOUG DUNNING AND STEVE WOLFRATH: Dunning and Wolfrath have developed a tool for financial advisors to use in helping clients monitor their spending compared to their assets. By applying the previously discussed ideas, they created the “Income Guide.” Dunning, a 34-year veteran of financial services, and Wolfrath, a 20year veteran, are both Fellows of the Society of Actuaries and members of the American Academy of Actuaries. Wolfrath is also a Chartered Life

* Sequence of returns: “Sequence-of-Returns Risk: Gorilla or Boogeyman?” Jonathan Guyton, Journal of Financial Planning October 2013.

Underwriter (CLU). A sentence on the literature for the Income Guide program says it is:

**Discretionary or “slush” fund: “When an Ounce of Discretion(ary) Is Worth a

“an innovative approach that can help clients monitor the projected

Pound of Core”, Jonathan Guyton, Journal of Financial Planning, February 2013.

sustainability of their lifestyle income in retirement.” In order to achieve this, the program sets a “prudent income amount” that, based on age, applies to the annual withdrawals which could be taken from the

DAVID BLANCHETT

portfolio and have that portfolio last a lifetime.

Mr. Blanchett, head of research at Morningstar, commented on the

Unique to the Income Guide and “prudent income amount” tool is daily

difference between historical “look backs” and forward-assuming

monitoring and comparing the income being taken to the portfolio

Monte Carlo simulations. One of his main points: be cautious about

market value (also valued daily). If the PIA moves above OR below the

what you assume your returns may be.

suggested amount by certain percentages, the client gets a communication suggesting it may be time for a review of their

Concerning “sustainable withdrawal rates”, he said the range from 2%-

distribution amount.

5% is probably a reasonable starting point but emphasized that there may be a large variance in those possible starting numbers depending

If the portfolio values increase to the degree that the PIA could be 25%

on this issue: “How comfortable are you with making changes in your

higher (or more) than it is currently, a notice will be generated indicating

spending, over time?” If you are not, start conservatively. If you are,

that it might be possible to take more from the account. On the other

you could start higher, but be willing to adjust spending downward, if

hand, if the portfolio has decreased enough that the PIA would have to

needed.

decrease by 10-20%, then a suggestion is made that it may be time to think about adjusting withdrawals downward a bit. And if the PIA would

His other points: (1) a balanced portfolio generally seems to do best over time due to more predictability and less volatility; (2) annuities make sense as part of the mix to manage longevity risk; and (3) like Michael Kitces, he thinks “success” and “failure” in the simulation vocabulary needs to change. It should be more nuanced, such as “how much left over at the end,” or the opposite

have to decrease by more than 20%, then a communication is sent that it may be time to decrease or stop withdrawals for a time to allow portfolio recovery. If withdrawals, when compared to the portfolio value, stay within a range that does not suggest a reduction, or taking more, then they are considered to be “on track.” The notices are not rules, but suggestions for a review with their advisor for a discussion on the issue. Below are the “prudent income percentages” from RiverSource Life Insurance Company as of 01/2015.

issue: “risk of adjusting spending” (to avoid depletion).

Page 16 of 21

The withdrawal rate (prudent income amount) is based on 90%

III.

Sequence of returns, or the “retirement day lottery,” can

probability of success to age 95; a 50/50 mix between stocks and

make or break sustainable retirement income cash

bonds. The assumptions are prospective, with bonds assuming to return

flows.

4% and grading up to 6% returns in 10 years, and the equity side at 9%, with a standard deviation (volatility measurement) of the portfolio at

IV.

The biggest factor to long-term sustainable income success is most likely going to be the withdrawal amount and the “adaptive” or “adjustable” withdrawals that are

9%. Dunning and Wolfrath believe both clients and financial advisors should consider that every portfolio is unique and may not look like the model portfolio that is used for these projections. The series of returns a

determined each year. The importance of adjusting the withdrawal rate based on portfolio performance cannot be overemphasized.

retiree will experience will be higher or lower than assumed in the model; so reviews become critically important to insure that withdrawals are “in range” to keep the portfolio in proper balance. Dunning put it this way: “We don’t change investors’ payouts; we provide information to clients and their advisors to make intelligent decisions.”

POST SCRIPT: In most cases, fees and taxes are not a consideration in the calculations of income need or withdrawal rate. It is assumed that a “gross need” has been calculated, out of which taxes and fees will be

“We have tried to keep it simple, straightforward, and user-friendly. With all of the possible variables, calculations could go on forever. The most important factor is frequent review, and adjustments if needed, along the way.”

deducted. Keep in mind that repositioning a “non-qualified” portfolio (one subject to taxation every year) for retirement income may involve capital gains and/or losses. Studies are continuing in the area of “taxefficient withdrawals.” Our Tax-efficient Drawdown of Assets in Retirement whitepaper was published in April 2015 and may provide additional considerations

CONCLUSIONS: I.

A good starting point to creating sustainable retirement income is a prudent withdrawal amount. There may be a variance in opinion as to whether that starting point is 3% or less, or maybe as high as 5%, but if you “split the difference,” we’re right back to the 4% number.

II.

Total return makes a difference, and so does volatility…so we need to be careful to balance these two issues. Consistency of returns is more important.

regarding this area of study.

ENDNOTES: Retirement Roundtable with Bill Bengen, Jon Guyton, Wade Pfau, and Michael Kitces; Journal of Financial Planning, October 2012/ Volume 25; No. 10; page 36 ff Prudent Income Amount: from RiverSource Annuities Income Guide literature Probability of Adjustment: from “Renaming The Outcomes Of A Monte Carlo Retirement Projection”: by Michael Kitces on his blog, August 14, 2013. 2.8% withdrawal rate: from “New research article: The 4% Rule is Not Safe in a Low-Yield World”: by Wade Pfau on his Retirement Researcher Blog, January 15, 2013 AND “Low Bond Yields and Safe Portfolio Withdrawal Rates”, by David Blanchett, CFA, CFP, Michael Finke, Ph.D, CFP, and Wade Pfau, Ph.D, CFA; Morningstar Investment Management, January 21, 2013. To wit: (page 7 of the Morningstar report) “Bengen (1994) recommends a 4 percent initial withdrawal rate from a portfolio made up of 50 percent stocks and 50 percent intermediate term treasuries, is sustainable for a minimum of 33 years…” “Later research by Cooley, Hubbard, and Waltz (1998), often called the Trinity study, generally confirmed Bengen's findings…that if a retiree seeks a 75 percent probability of success, a 4-to-5 percent initial withdrawal would be a good place to start…” “These findings have been affirmed by other research as Milevsky, Ho, and Roberson (1997) and Jarrett and Stringfellow (2000), among others.” However, page 11 concludes: “This paper introduced a model that takes into account current bond yields when determining the probability of success…a retiree who wants a 90% probability of achieving a retirement income goal with a 30-year time horizon and (has) a 40% equity portfolio would only have an initial withdrawal rate of 2.8%.” “Asset allocation can significantly affect a portfolio’s ability to sustain a given cash flow during retirement”. Research by Ervin, Filer, and Smolita (2005), Tezel (2004), Cooley et al (2003), and Kaplan (2005) demonstrate that portfolios with lower equity allocations tend to generate higher probabilities of ruin…” (note to reader: this research was done just prior to the “great recession” of 2008-2009). In an additional note sent by Michael Kitces, he wrote: “On the importance of total return vs volatility: with very low withdrawal rates, the reality is that neither matters (because) you are spending too little to fail in any scenario. As the withdrawal rate rises (>3% to 6%), you need SOME return, but the volatility of that return begins to matter significantly as well. At very high withdrawal rates (6% or more), volatility actually matters less again and the driver is just total return – basically, you’re ‘swinging for the fences’ and you’ll either strike out or hit a home run. There’s no room for anything in between if you’re spending that aggressively.” Other readings: > “Retirement Withdrawal Rates and Portfolio Success Rates: What Can the Historical Record Teach Us?” Wade Pfau, fall 2011 edition of Retirement Management Journal > “Determining Optimal Withdrawal Rates: An Economic Approach” D. Williams and Michael Finke, fall 2011 Retirement Management Journal > “The 4 Percent Rule is Not Safe in a Low-Yield World” Michael Finke, Wade Pfau, and David Blanchett, Journal of Financial Planning > “Spending Flexibility and Safe Withdrawal Rates.” Michael Finke, Wade Pfau, and D. Williams, Journal of Financial Planning, 2012, volume 25(3) pages 44-51 > “Withdrawal Rates, Savings Rates, and Valuation-based Asset Allocation.” Wade Pfau; Journal of Financial Planning, April, 2012 > “Retirement Savings: Choosing a Withdrawal Rate That is Sustainable.” Philip Cooley, Carl Hubbard, and Daniel Walz (Trinity University, Texas), AAII Journal, February 1998 > “Renaming the Outcomes of a Monte Carlo Retirement Projection.” Michael Kitces blog posting of August 14, 2013. > “Financial Independence in Lieu of Retirement, and Other Phrases That Should be Banished from Retirement Planning.” Michael Kitces blog posting of October 22, 2014 > “Why 4% May Be a Sustainable Withdrawal Rate.” Gregg S. Fisher, CFA, CFP February 14, 2014 Financial Advisor magazine > “Capital Markets Expectations, Asset Allocation, and Safe Withdrawal Rates.” Wade Pfau, Journal of Financial Planning (January, 2012)

Page 18 of 21

William P. Bengen, CFP®, is a recently-retired financial advisor based in Chula Vista California. His seminal article, “Determining Withdrawal Rates Using Historical Data,” published in the Journal of Financial Planning in October of 1994 is the groundbreaking study for the now-famous “4% rule.” His published book, “Conserving Client Portfolios During Retirement” (FPA press, 2006) has inspired many others to build on his work.

Wade D. Pfau, Ph.D., CFA®, is a Professor of Retirement Income at The American College in Bryn Mawr, PA. His research article on “safe savings rates” won the inaugural Journal of Financial Planning Montgomery-Warschauer Editor’s Award, and his work on evaluating the outcomes of different retirement income strategies received an Academic Thought Leadership Award from the Retirement Income Industry Association. Mr. Pfau is much published and frequently quoted in the media.

Jonathan Guyton, CFP®, is principal of Cornerstone Wealth Advisors, Inc., a holistic financial planning and wealth management firm in Edina, Minnesota. He is also a researcher, author and frequent national speaker on retirement planning and asset distribution strategies, and a former winner of the Journal of Financial Planning Call for Papers competition.

Michael Kitces, CFP®, CLU®, ChFC®, RHU, REBC, is a partner and research director at Pinnacle Advisory Group in Columbia, Maryland. He publishes: “The Kitces Report” newsletter and “Nerd’s Eye View” blog. He is a practitioner editor of the Journal of Financial Planning.

Page 19 of 21

David Blanchett, CFA®, CFP®, is head of retirement research for the Morningstar Investment Management group. Blanchett’s research has been published in a variety of academic and industry journals. He has won awards for his research papers in 2007, 2012 and 2014. In addition to his CFA and CFP designations, he also holds a Chartered Life Underwriter (CLU), Chartered Financial Consultant (ChFC) and Accredited Investment Fiduciary Analyst™ designations.

Doug Dunning, Vice President, Annuity Product Development. Doug heads up annuity product development and management for variable, fixed and payout annuities at RiverSource Life Insurance Company (a subsidiary of Ameriprise Financial). He is a Fellow of the Society of Actuaries and a Member of the American Academy of Actuaries. He serves as a member of the LIMRA annuity committee.

Steve Wolfrath, FSA, MAAA, CLU®, is vice president of annuity development at RiverSource Life Insurance Company (a subsidiary of Ameriprise Financial Services, Inc.). In his role, Steve and his team are responsible for the design and pricing of variable, fixed, and payout annuities. He is a Fellow of the Society of Actuaries.

IMPORTANT DISCLOSURES The views expressed in this publication reflect the personal views of the author. Further, Ameriprise Financial Services, Inc. research department employee compensation is neither directly nor indirectly related to the specific recommendations or views contained in this publication. For important disclosures on securities mentioned in this analysis, please review available third party research reports and charts with applicable disclosures on our website at ameriprise.com, or through your financial advisor, or by submitting a written request to Ameriprise Financial Services, Inc., 1441 W. Long Lake Rd, Suite 250, Troy, MI, 48098. The S&P 500 is a basket of 500 stocks that are considered to be widely held. The S&P 500 index is weighted by market value (shares outstanding times share price), and its performance is thought to be representative of the stock market as a whole. The S&P 500 index was created in 1957 although it has been extrapolated backwards to several decades earlier for performance comparison purposes. This index provides a broad snapshot of the overall U.S. equity market. Over 70% of all U.S. equity value is tracked by the S&P 500. Inclusion in the index is determined by Standard & Poor’s and is based upon their market size, liquidity, and sector. The Barclays Capital U.S. Aggregate Index is an index comprised of approximately 6,000 publicly traded bonds including U.S. government, mortgagebacked, corporate and Yankee bonds with an average maturity of approximately 10 years. The index is weighted by the market value of the bonds included in the index. This index represents asset types which are subject to risk, including loss of principal. It is not possible to directly invest in an index.

Page 20 of 21

Annuities: Annuities are not federally or FDIC insured and are not a deposit of or guaranteed by any bank or any bank affiliate. All guarantees are based on the continued claims paying ability of the issuing company and do not apply to the performance of the variable subaccounts, which will vary with market conditions. Features and availability may vary by state. Variable annuities are insurance products that are complex, long-term investment vehicles that are subject to market risk, including the potential loss of principal invested. Clients should consider the investment objectives, risks, charges and expenses of the variable annuity and its underlying investment options carefully before investing. For a free copy of the annuity's prospectus and underlying investment's prospectus, send an email message to [email protected]. The prospectus should be read carefully before investing. Accessing policy cash value through loans and surrenders may cause a permanent reduction of policy cash values and death benefit and negate any guarantees against lapse. The amount that can be borrowed or surrendered will be affected by the surrender charges applicable to the policy. Loans may be subject to interest charges. Although loans are generally not taxable, there may be tax consequences if the policy lapses or is surrendered with a loan (even as part of a 1035 exchange). It is possible that the amount of taxable income generated at the lapse or surrender of a policy with a loan may exceed the actual amount of cash received. Surrenders are generally taxable to the extent they exceed basis in the policy. If the policy is a modified endowment contract (MEC), pre-death distributions, including loans, from the policy are taxed on an income-first basis, and there may also be a 10% federal income tax penalty for distributions prior to age 59 ½. RiverSource is part of Ameriprise Financial.

DISCLAIMERS Except for the historical information contained herein, certain matters in this report are forward-looking statements or projections that are dependent upon certain risks and uncertainties, including but not limited to, such factors and considerations as general market volatility, global economic and geopolitical impacts, fiscal and monetary policy, liquidity, the level of interest rates, and historical sector performance relationships as they relate to the business and economic cycle. This summary is based upon financial information and statistical data obtained from sources deemed reliable, but in no way is warranted by Ameriprise Financial Services, Inc. as to accuracy or completeness. This is not a solicitation by Ameriprise Financial Services, Inc. of any order to buy or sell securities. This Summary is based exclusively on an analysis of general current market conditions, rather than the suitability of a specific proposed securities transaction. We will not advise you as to any change in figures or our views. Past performance is no guarantee of future performance. Ameriprise Financial Services, Inc. Member FINRA and SIPC. Investment products are not federally or FDIC-insured, are not deposits or obligations of, or guaranteed by any financial institution, and involve investment risks including possible loss of principal and fluctuation in value. Neither Ameriprise Financial, nor any of its advisors or representatives, provides tax advice. © 2015 Ameriprise Financial, Inc. All rights reserved.

Page 21 of 21