Semi-Fixed Minimum Withdrawal Starting at 3%

[sdfire]

I recently hit my savings goal which was to save enough to support a 3% withdrawal rate at my current budget. Woo Hoo! I will likely continue working for a while because I don't hate it and I'm being well paid.

But, I'm thinking through my plans so that I have the comfort of knowing that I can walk away at any time if I want. I am nearly 47 and my spouse is nearly 45. So, we need to make this money last for 50+ years. We are invested 50/50 stocks/bonds.

Given the 4% "rule," my thinking is that a 3% withdrawal rate is pretty conservative. But, to add a level of conservatism, we are happy with our budget based on the 3% withdrawal rate. As a result, even if the portfolio jumped dramatically, we would probably stick fairly close to our initial budget plus inflation. On the other side of the fence, my math tells me that if the stock market falls 50% sticking with our current budget would have us withdrawing at a 4% rate. And, knowing my spouse and myself, I have no doubt that we'd decrease spending a bit if we suffered a 50% stock loss. Our 3% based budget is generous enough that we could cut a bit without too much pain (keep our cars for another year or two, take one less vacation, etc.). So, the plan would be to start with 3%, go with 3% plus inflation in years with increases and go up to 4% in years with decreases (recognizing that we'd probably cut spending in years with big losses).

My primary question is whether there is a way to analyze this plan using the current withdrawal calculators? I tried with Firecalc, but couldn't figure out how to input exactly what I'm suggesting.

My other question is whether this seems conservative, aggressive or something in between. All thoughts welcome.

[ziggy29]

Quote:

Originally Posted by sdfire On the other side of the fence, my math tells me that if the stock market falls 50% sticking with our current budget would have us withdrawing at a 4% rate. And, knowing my spouse and myself, I have no doubt that we'd decrease spending a bit if we suffered a 50% stock loss. Our 3% based budget is generous enough that we could cut a bit without too much pain (keep our cars for another year or two, take one less vacation, etc.). So, the plan would be to start with 3%, go with 3% plus inflation in years with increases and go up to 4% in years with decreases (recognizing that we'd probably cut spending in years with big losses).

First of all, welcome to the board!

Secondly, the "4% rule" and its derivatives usually only set the *first* year's withdrawals and is adjusted for inflation annually. In subsequent years you wouldn't calculate a percentage of the portfolio to withdraw; you would simply take the previous year's withdrawal plus an inflation adjustment (if any). If you withdraw $50,000 in the first year and inflation was 3% over that year, you would withdraw $51,500 in the second year regardless of whether your portfolio was up or down.

[mmcc]

For 50 years, 3% is conservative by most standards.

FireCalc can do the calculations but the available data is not ideal for 50-year tests. FireCalc by default uses Shiller's data, which goes from 1871 to 2008. This is the most data available for long tests, but it does not support specifying a well-diversified stock portfolio so the results are on the very conservative/low side. Even with "low" results though, it shows a 100% success rate for 3% with a 50/50 allocation. If you use FireCalc's option for a mixed stock portfolio, then data is only available from 1927 to 2005 (last time I looked). This is shorter than desired for a solid 50 year test, but it shows you can withdraw up to around 4% for a 100% success rate (assuming a well-diversified stock portfolio).

Stepping back a bit though, it good to note FireCalc assumes annual rebalancing back to target values. There is strong evidence that annual rebalancing is not the best way to harvest incoming during retirement, although most of the financial industry does not seem aware of the research. If you are interested, there is a lot of information around on other withdraw strategies, including variable withdraw rates. (Unfortunately, with the other strategies a calculator like FireCalc is not available to compare, so its safer to avoid any that does not show solid backtesting.)

[samclem]

If you haven't already done so, I'd recommend ypou play around with FIRECalc a bit and see what you can learn.

A lot depends on what you mean by "4%" or "3%" Is this a withdrawal you'll make based on your ending portfolio balance each year, or an amount starting at 4% of portfolio and adjusted each year for inflation? FIRECalc will let you do things either way. Obviously, if you disregard your portfolio value and just adjust for inflation, your annual "take" will be very smooth, but you could run out of money.

You'd probably also find that Bob Clyatt's book "Work Less, Live More" is a good read. He discussed a number of strategies, including a "95% rule" that dampens the downswings in withdrawals from a down market, and which can be modelled in FIRECalc.

The ER Board Retirement FAQ page has some threads you'll be interested in.

I also enjoyed "Gummy's'" discussion of withdrawal strategies.

Welcome to the board

[sdfire]

Thanks very much for all of the comments. I've been monitoring these boards all the way back to the Motley Fool days. I've read over and over that a 4% withdrawal rate with annual inflation adjustments is very, very safe for 30 year withdrawal periods. Therefore, I started aiming for assets that would allow 4% withdrawals. As I came close to 4%, I realized that I wasn't totally comfortable with "barely making it" and with a 50 year need I decided to shoot for 3%. Now that I've hit 3%, I realize that I need to get comfortable that 3% plus inflation is conservative enough that it really will last for 50 years. On that score, I'd love to hear more reassurance.

I will check into the annual rebalancing research. As I am still in the accumulation phase, I essentially rebalance with every piece of income. I had planned on annual rebalancing in retirement so I need to understand those issues. I will say that as I've accumulated more assets, I've been surprised to find that I am more risk averse than before. I guess I feel like I've finally achieved the goal I set and I don't want to see it quickly vanish.

An alternative to 3% plus inflation that I considered was simply going with 3-4% of the portfolio every year. By definition that never depletes the portfolio. Honestly, with a fairly generous budget at 3% I think we could handle some decreases if we got hit early with a stock downturn. My thinking is that a 50% stock drop only puts me at a 4% withdrawal rate with a 50/50 portfolio. On the flip side, if the portfolio had big increases, we'd probably stick with 3% plus inflation withdrawals. But, again, I'm just not sure how to model that in Firecalc.

[audreyh1]

Quote:

Originally Posted by sdfire Thanks very much for all of the comments. I've been monitoring these boards all the way back to the Motley Fool days. I've read over and over that a 4% withdrawal rate with annual inflation adjustments is very, very safe for 30 year withdrawal periods. Therefore, I started aiming for assets that would allow 4% withdrawals. As I came close to 4%, I realized that I wasn't totally comfortable with "barely making it" and with a 50 year need I decided to shoot for 3%. Now that I've hit 3%, I realize that I need to get comfortable that 3% plus inflation is conservative enough that it really will last for 50 years. On that score, I'd love to hear more reassurance. I will check into the annual rebalancing research. As I am still in the accumulation phase, I essentially rebalance with every piece of income. I had planned on annual rebalancing in retirement so I need to understand those issues. I will say that as I've accumulated more assets, I've been surprised to find that I am more risk averse than before. I guess I feel like I've finally achieved the goal I set and I don't want to see it quickly vanish. An alternative to 3% plus inflation that I considered was simply going with 3-4% of the portfolio every year. By definition that never depletes the portfolio. Honestly, with a fairly generous budget at 3% I think we could handle some decreases if we got hit early with a stock downturn. My thinking is that a 50% stock drop only puts me at a 4% withdrawal rate with a 50/50 portfolio. On the flip side, if the portfolio had big increases, we'd probably stick with 3% plus inflation withdrawals. But, again, I'm just not sure how to model that in Firecalc.

That last approach also has the advantage of keeping up with the portfolio growth. A good way to handle the lean years when you cut back, is to set aside a little extra from the good years. That way you can smooth things out a bit more than a straight percentage would have you do.

Audrey

[Ed_The_Gypsy]

IIRC, the 4% rule (annually adjusted for inflation) applies to 60/40 portfolios.

Galeno had a slightly different approach. See below from 2002.

Quote:

Originally Posted by galeno Yes I am following the same mechanical withdrawal plan. In short, it goes like this: FI = 25% 2y living expenses in MMF 2y living expenses in 2y CD (maturing in 1y) 2y living expenses in 2y CD (maturing in 2y) Stocks = 75% At the end of the year I sell 4% of the value of my stock portfolio and buy a 2y CD. I let half of the maturing 2y CDs go to MMF and I buy another 2y CD with the other half. I then divide the entire FI balance by 72 and that's my monthly draw for the year. At year end, I repeat the process. Even with this three year bear market, my monthly FIRE income fluctuates very little thanks to my FI buffer. It's a similar approach to intercst's inflation adjusted withdrawals but slightly different in that I let the long term growth of my stock portfolio indirectly take care of any inflation or deflation in the economy.

[Duke of Sands]

Ed, any idea how this has worked out since 2002?

[MichaelB]

Quote:

Originally Posted by sdfire Thanks very much for all of the comments. I've been monitoring these boards all the way back to the Motley Fool days. I've read over and over that a 4% withdrawal rate with annual inflation adjustments is very, very safe for 30 year withdrawal periods. Therefore, I started aiming for assets that would allow 4% withdrawals. As I came close to 4%, I realized that I wasn't totally comfortable with "barely making it" and with a 50 year need I decided to shoot for 3%. Now that I've hit 3%, I realize that I need to get comfortable that 3% plus inflation is conservative enough that it really will last for 50 years. On that score, I'd love to hear more reassurance. .

A 3% withdrawal rate for 50 years means your portfolio needs to keeps its current real value over time – at least for 20 years. IOW, you need 3% real return.

With a 50/50 split between equities and fixed income, if you got 1.5% real return in fixed income you would need 4.5% real return in equities. Right now 10 year TIPS are yielding 1.7%, so the real question is do you think you will get 4.5% real return in equities?

This is within the range of the more bearish expectations of investors like Grantham and Hussman. Not safe but reasonable. For my own planning I think 3% is entirely reasonable and would focus on other aspects and risks of ER.

[sdfire]

Quote:

Originally Posted by MichaelB A 3% withdrawal rate for 50 years means your portfolio needs to keeps its current real value over time – at least for 20 years. IOW, you need 3% real return. With a 50/50 split between equities and fixed income, if you got 1.5% real return in fixed income you would need 4.5% real return in equities. Right now 10 year TIPS are yielding 1.7%, so the real question is do you think you will get 4.5% real return in equities? This is within the range of the more bearish expectations of investors like Grantham and Hussman. Not safe but reasonable. For my own planning I think 3% is entirely reasonable and would focus on other aspects and risks of ER.

Thanks for these thoughts. They actually made me feel more confident. My FI should return better than 1.5% as I have a significant percentage in I-bonds with a base return of 2-3.3%. And, even if I need a 4.5% real return in equities, I think that is very realistic especially over the long haul.

[ziggy29]

Quote:

Originally Posted by sdfire And, even if I need a 4.5% real return in equities, I think that is very realistic especially over the long haul.

My "assumptions" for equities are 4% real over the long term, but this isn't out of the ballpark, either -- and it's not like my crystal ball isn't a little wonky.

In any event, in order to get the best real return in equities in terms of generating retirement income, one key is to keep enough in cash and/or fixed income on hand so you are never *forced* to sell stocks at bargain prices into a serious bear market. In this recent cycle, perhaps, you would have not sold equities in 2008 or much of 2009, but by late 2009 or early 2010 (once stocks recovered to reasonable valuations) you could have sold some to replenish the cash and bonds you sold for income in the previous year or two. It sort of feels like market timing, but it's really also a play on the tendencies for asset classes to mean-revert over time -- a lot like rebalancing an asset allocation.

[sdfire]

Quote:

Originally Posted by Ed_The_Gypsy IIRC, the 4% rule (annually adjusted for inflation) applies to 60/40 portfolios. Galeno had a slightly different approach. See below from 2002.

This is intriguing. I had originally thought about an approach where I did something very similar. Any idea how to model what would happen if you:

Took 3%, 3.5%, or 4% from portfolio every year? How bad would your income swings be? By definition you would never exhaust the portfolio.

What to do with the non-stock portion of the portfolio is a puzzle I'm still working on, but I don't think it is key to this question.

Thanks!

[teejayevans]

Quote:

Originally Posted by sdfire My other question is whether this seems conservative, aggressive or something in between. All thoughts welcome.

I never heard anybody complain they have too much money.
TJ
p.s. you can always spend more...cruises, travel, new luxury vehicle
will eat up the extra no problem.

[Ed_The_Gypsy]

Quote:

Originally Posted by Duke of Sands Ed, any idea how this has worked out since 2002?

Good question, Duke. I checked it out. Short answer: depends.

I checked back 10 years with some simple choices. (My work is rough, so take it with a grain of salt.) If I knew how to post graphs, it would be easy to see the results.

If you used 75% S&P 500 + 24% TIPS fund (VIPSX--all funds are Vanguard) for the past 10 years, it appears that you would have experienced steadily declining annual income (eventually -40% of start) and wind up with only half of your starting pot. Ugh.

If you used a 50/50 mix of small cap value (VISVX) and emerging markets (VEIEX) instead of the S&P 500 (VFINX), the annual income would fluctuate a little (+4.5%, -6.5%) and the balance at the end of 10 years would be higher than you started.

A margarita portfolio (1/3 VISVX, 1/3 VEIEX, 1/3 VIPSX), taking out 4% per year and rebalanced every year produced a wider fluctuation in annual income (+40%,-15%) and 22% larger pot at the end.

Interestingly enough, a margarita portfolio with the standard 4% of initial, adjusted every year for an assumed 3.5% inflation wound up with a pot about 15% larger at the end of 10 years compared to the margarita less 4% per annum and the income stream was smooth.

A 50/50 VISVX/VEIEX with 4% indexed w/d wound up with +24% more in the pot at the end and of course the same smooth income stream.

The TIPS fund wound up being just a drag on returns in all cases. Interesting. I guess it is what you pay for insurance. I am uncertain what to do about this.

This supports a small-cap and value bias and a 50/50 US/international equity split (which is what I am doing personally). It tells me that it is hard to trust the large caps.

[REWahoo]

Quote:

Originally Posted by Ed_The_Gypsy If I knew how to post graphs, it would be easy to see the results.

The bunny can help you with that: How to post a screen shot

[IndependentlyPoor]

Yes, please post the graphs.

[Ed_The_Gypsy]

Quote:

Originally Posted by sdfire This is intriguing. I had originally thought about an approach where I did something very similar. Any idea how to model what would happen if you: Took 3%, 3.5%, or 4% from portfolio every year? How bad would your income swings be? By definition you would never exhaust the portfolio. What to do with the non-stock portion of the portfolio is a puzzle I'm still working on, but I don't think it is key to this question. Thanks!

sdfire, you get income swings, no matter how little you take out each year (I checked). See my points above. I was surprised at how little difference there was at the end of ten years between 4% initial indexed for inflation and Galeno's path. Both were still well ahead but one gave a lot smoother income stream. I wish I knew how to post graphs.

MichaelB's point is very good. It is the same as my point about TIPS dragging down the return. This suggests Les Antman's 4% per annum of a 50/50 US/foreign stock model. Les was quite willing to accept the income swings. I am thinking that a smooth income stream would be more desirable.

What do you think?

[Ed_The_Gypsy]

Thanks, REWahoo. I will try that now.

[Ed_The_Gypsy]

(1) 75% S&P 500 + 24% TIPS fund (VIPSX), Galeno w/d 4% pa
(2) 50/50 VISVX / VEIEX. Galeno w/d 4% pa
(3) margarita (1/3 VISVX, 1/3 VEIEX, 1/3 VIPSX), taking out 4% per year and rebalanced
(4) margarita portfolio, 4% indexed
(5) 50/50 VISVX / VEIEX, 4% indexed
Remember, this is very rough. Beware of my mistakes.
(Ouch! Sorry about the pix.)

[Duke of Sands]

Ed- thanks for all the work. Some surprises there for me. A question, my impression is that Galeno's used CDs and MM for 25 % of his portfolio, whereas you used TIPS fund for the FI portion??
Looks like a confirmation that 2000 was a crummy year to retire, assuming one's equities pot was S&P. Interesting that staying small value and international did that much better!