Managing Withdrawals

[Mustang]

I just read a very good article by Christine Benz talking about which withdrawal system might be best for you. www.morningstar.com/articles/1028793/the-first-question-you-should-ask-about-in-retirement-withdrawal-rates

She leaves out a couple of methods such as the income system (living off interest and dividends) and the bucket system (a complicated multi-fund approach) but for the most part it gives a good description of the two other primary withdrawal systems: a fixed or static system and a flexible or dynamic system. In their simplest forms a fixed or static method is based on real dollar withdrawals and is designed to protect the retiree's standard of living. The flexible or dynamic method is based on a percentage withdrawal and is designed to protect the portfolio's value by changing the amount of the withdrawal according to market performance. There are dozens of variations of each usually incorporating ceilings and floors.

There appears to be two methods of testing the effectiveness of the withdrawal systems: using historical data and using Monte Carlo simulations. Bengen used historical data for 51 30-year payout periods (1929-1979) to come up with the 4% Rule. A few years later the Trinity study verified his results using 55 30-year payout periods. In 2018 Wade Pfau updated the Trinity study and confirmed the results again. Even for the worst 30-year payout period in history a 50/50 portfolio did not run out of money for a 4% initial withdrawal with subsequent withdrawals adjusted for inflation.

Michael Kitces, going back to 1871 found that the maximum withdrawal varied from 4-10%, that 4% was the worst case scenario, and that most of the time the retiree ended the 30-year payout period with considerable wealth.

The Monte Carlo methods test the withdrawal strategy using 10,000 computer generated simulations. Some are pretty far out and most analysts using it say an 80% success rate is acceptable. Recently some say 50% is acceptable if used with a flexible withdrawal system.  Christine Benz references a March 1, 2021 article by Amy Arnott which says the 4% Rule has an 86% probability of success. Why isn’t that acceptable? www.morningstar.com/articles/1026956/will-the-real-retirement-income-number-please-stand-up

A few months earlier another Morningstar analyst, John Rikenthaler questioned the viability of the 4% Rule. (“The Math for Retirement Income Keeps Getting Worse: Revisiting the 4% Withdrawal Rule,” (October 8, 2020). Using Buffet’s forecasting formula he tested a portfolio of 50% S&P 500 stocks and 50% treasury bonds and showed that it survived a 30-year payout period. The worst he could say was that it didn’t have a high enough ending balance. www.morningstar.com/articles/1003885/the-math-for-retirement-income-keeps-getting-worse

In the article Christine Benz asks some important questions such as: Do you have other sources of stable income? Can you adapt to large year to year changes in income? Studies have shown that flexible or dynamic withdrawal systems result in less than planned income almost half the time.

The type of withdrawal system planned should be a major input into your portfolio’s design. For example, if you plan income only withdrawals then you have to invest heavily in investments that pay interest and dividends. You can also use more than one withdrawal system. I plan on using a dynamic system for my IRAs (RMDs) and a stable system (4% Rule) from my taxable accounts.

Benz mentioned variations that include ceilings and floors. There are other variation such as taking less than a full inflation increase each year or (because most of the time considerable wealth is accumulated) using Michael Kitces step up method which gives the retiree a raise every three years if the portfolio does well.

As I said, I liked the article and I liked the questions she asked.

[retiredat48]

Mustang...and others...here's my withdrawal strategy...been withdrawing for over 27 years now.  Please critique. 

As background, I submit that for most investors, their main goal is you do not want to be withdrawing monies for retirement purposes at the bottom of bear markets. Fair enough, most feel this way.

Also, many will be taking Req'd Min Distributions from their mostly untaxed portfolio (IRAs). I have done such IRA withdrawals for 27 years now (taking out money to live on), and found the following scheme to be my best method to reduce the risk of taking out during bear market bottoms, without holding excessive cash. Some call it "IMMUNIZING YOUR PORTFOLIO." Here it is...ponder...you can get up to two years of wait time without having to take an RMD or withdraw a penny:



------------------------------------------------------------

FWIW, here's what I do...



First, I always try to own about a half years living expense worth of shorter term bond funds such that if the stock market crashes, can be tapped for spending/withdrawals (or added into the stock market!). It is part of my asset allocation. I do not own much "cash." These shorter term bond funds do not fluctuate much in total return. I opt for the higher interest, versus absolute safety of cash, MM Funds or CDs. Over time, this added interest adds up greatly.



Then each December, I assess my asset allocation and previous stock and bond fund yearly performances to date. Then, I decide, and select one or more stock funds to take some off the table, to give a second years short term cash holding, if strategically beneficial. This means, if the stock market zoomed upward in a given year, I likely will take some off the stocks table...a form of rebalancing.



For example, In Dec 2013, I sold some Energy Funds moving it to VFSTX for the second year withdrawals. Year 2014 I did not sell any that past December, but planned on selling Vanguard's Health Care Fund, for my RMDs throughout the year. This selling/taking RMDs was completed summer 2015. At the end of that year 2015, I further sold more from my Healthcare Fund to meet the upcoming 2016 year RMDs. This turned out to be a good sale also, as Healthcare declined a lot in 2016. So I have a two year cushion to market selling, as I can wait until end of 2017 to take 2017 RMDs.



 December 2019, I completely exited Energy Sector funds...as laggards, to build full cash for 2020 RMDs. With the stock "meltup" in January ...I sold more stock funds. With Covid, I sold more in early Febuary (All documented/posted real time, if anyone is interested). So I now have the full vaccination of two years worth of RMD distributions...IN CASH/SHORT TERM BOND FUNDS.



Note that if you do this selling in December, and the market goes into an immediate bear market, you are set for 24 months before any cash replenishment need is necessary to take required RMDs. That is, current year RMDs come from the first bucket, and the next year RMDs need not be taken until December of that year, giving 24 months worth of wait. Bear markets usually recover a lot in this time frame.



If money is absolutely needed monthly, live off a HELOC or equivalent for awhile...or from bond funds. Remember, in the vast majority of market history, bond funds go up when stock funds decline. Bond funds, part of ones allocation, are the usual source of spending monies in stock bear markets.



Lastly, many say you cannot time the market. OK...but investors taking RMDs must realize you have to time the market, generally EACH YEAR, if you live on distributions like I do. You must sell some stock funds at times to replenish your bond funds, if you are taking everything from bonds. There's no escaping this fact. Otherwise, your stock/bond allocation gets distorted. Better to do such timing strategically...with a plan.



With this December strategy, you immunize your portfolio withdrawals for up to 24 months...if you like. Few storms last longer than that.


---------------------------------------------------------


R48

[dgva]

Mustang, Thanks for this thread and your comments. We have been debating our retirement spending/cash flow plan for some time. Over a year ago, I read Christine Benz Bucket strategy. Initially, it sounded appealing, but as I started looking at operationalizing it, I decided it was not for us. Clearly, these are personal decisions that may be somewhat unique to ones circumstances and experience.

R48, appreciate your comments and will be thinking through your experience more. I have read your comments on many threads and have benefited from your experience.

We have been in full retirement about a year. Our immediate plan: We moved most of our TSP holdings to Fidelity, but left a residual in the G fund, planning to spend it out in the next 3 years. With that and TIAA Traditional IO, our spending will be covered until age 72. Afterwards, need about 2-2.5% from retirement funds (unless a personal health challenge/death occurs to one of us).

Irrespective of the need to withdraw, RMDs are looming. So, I have been thinking about methods to harvest RMDs while avoid selling in a bear. I can see the merit of keeping a couple year's RMDs in relatively low risk instruments. Of course, this does generate potential opportunity costs as the price for safety.

Interested in others experience. Thanks for the Thread, Mustang.

[yogibearbull]

Note that TSP G Fund and TIAA Traditional annuity are stable-value funds. TIAA Traditional offers a variety of options that include Interest-Only [IO] option. Both have limited/restricted availability.

[Mustang]

retiredat48 , Thank you for your response. Your approach seems a little like the bucket approach that Christine Benz writes about using short term bonds as cash. I'm still exploring possibilities. I set up RMDs before the age was extended to 72. Since I believe last minute changes tend to screw things up I let it ride. We are just paying the taxes and reinvesting what we receive.

I once looked closely at the bucket approach. Howard Evensky who pioneered the approach said in a 2010 interview the sensible number of buckets for a do-it-yourself investor was two. He is said to have simply bolted on a cash account to his total return portfolios. Distributions are only from growth. Benz has called it pruning back the investment. Modern advocates have complicated it. It is an approach that is far more complicated than I want and managing multiple funds such as Health funds, energy funds, etc. isn't anything I could teach my wife.

But, as Evensky pointed out, it doesn't have to be that complicated. Christine Benz of Morningstar tested a simplified bucket approach. She attached a cash bucket to Vanguard Balanced Index Fund (60% stock and 40% bonds). For the period 2000-2017 she compared it to her multi-fund approach. She said she preferred the multi-fund approach to the single fund approach because when rebalancing once per year to replenish cash there were always funds that needed pruned. She said using Vanguard Balanced Index, the cash account was sometimes depleted and funds had to be sold early. She said the simplified approach finished the test having a balance that was $400,000 less than the multi-fund approach.

If I read your post correctly you use a HELOC to cover such instances.

I had a problem with Benz's test. Why did she pick Vanguard Balanced Index? And why an 18-year period starting in 2000 instead of a 20-year period starting in 1998? There was a big correction in 2000. Vanguard Balanced Index suffered a sequence-of-return loss losing money in 2000, 2001, and 2002 and it took a huge 22.2% loss in 2008. A reader pointed out that if she had used Vanguard Wellesley Income Fund instead of Balanced Index, the single fund approach would have had a ending value greater than her multi-fund portfolio. Wellesley Income lost money in 1999 but had solid returns 2000-2002. And, it only lost 9.8% in 2008.

I think the bucket approach is a great approach for people who enjoy managing their funds daily.  It just isn't anything I could teach my wife.

Even the way you take RMDs requires a hands on analytic approach. Timing it so that you can go 24 months without incurring a penalty.  My RMDs are automated.  The fund company calculates them takes out the taxes and sends us a monthly check.  As the article in the first post said, it a trade off between sustainability and livability.  The flexible or dynamic withdrawal methods (of which RMDs are one) are designed to protect the portfolio.  While not totally ideal if the market drops and the portfolio loses value less is taken.

[Mustang]

I came across another article on withdrawals. This time it isn't about choosing a fixed or dynamic method. The author uses the 4% rule in all tests. This time he was looking at various methods to determine which to withdraw the money from different asset classes, stocks or bonds. His target allocation was 50/50/ and the only transaction made each year was the withdrawal. He did not change funds or force re-balancing. money.com/best-retirement-withdrawal-strategies/

He used historical data from 1928 he tested 58 30-year payout periods. He said the CAPE Median withdrawal strategy performed the best. Equal withdraws was next. And third was withdrawing from stocks or bonds based upon last year's performance. Finishing last was the rebalancing method, 3-year moving average method and 7-year moving average method.

It was an interesting reading.  And according to his test it made an average of up to $4 million in the ending balance (starting with $1 million). I'm going to see if I can find more on the subject.

[Tibbles]

money.com/best-retirement-withdrawal-strategies/

Mustang, The results reported in the above article, posted by Mustang, are interesting and seem relevant regardless of whether one prefers a fixed or a dynamic withdrawal strategy. I was surprised that the simple strategy of rebalancing to fixed targets was so greatly inferior to the CAPE-based strategy. I wouldn't be sure just how to apply the latter, since I don't know where to find the current long-term median, and wouldn't be confident that being guided by the long-term medium will work well in the future, since the CAPE has moved upward a lot over the last 40 years (possibly for good reasons). But the results do seem to justify what I think most of us are tempted to do: to take withdrawals mainly from assets that seem the most expensive, rather than simply rebalancing to fixed targets. Also, the results square with the thesis defended (I think persuasively) in a lengthy Boglehead thread titled: CAPE: A much stronger predictor of stock returns than many think.