Withdrawal Rates at 50-50 Stocks/Bonds

[sdfire]

My spouse's eyes glaze over when I start to talk about this stuff. Can I ask this group to help me check my math and assumptions? Thanks.

We are planning a 3% withdrawal rate over 50 years. Here are my assumptions (all figures present dollars before tax):

50%: Fixed income earns 3.5% (fixed income is in long-term munis at 5%, I and EE bonds at approximately 3.5%, taxable muni bond fund at 3.5%, total bonds in IRA at 2.5%, and "cash" at 1% or so.) Taking 3% leaves .5% for inflation.

50% Stocks: I need stocks to earn an extra 5.5% per year in order to give the stock half of my portfolio enough of a bounce to come up with an extra overall portfolio 3% for inflation each year. As a result, my stocks need to earn 3% for withdrawal, plus an extra 5.5% for a total of 8.5%.

The .5% extra from the 50% in fixed income plus the 5.5% extra from the 50% in stocks gives me 3% extra for the entire portfolio to cover inflation. Am I getting this right?

Needing to average 8.5% on stocks has me concerned. I probably shouldn't be concerned because our existing budget is just a hair above 2%. Therefore, a 2.5% budget would give us a nice cushion. At a 2.5% withdrawal rate, my math looks like this:

50% in fixed income earns 3.5% leaving an extra 1%.
50% in stocks needs 2.5% plus 5% for a total of 7.5%

I'm more comfortable knowing I've covered withdrawals and inflation at 2.5% with a 7.5% stock return.

On the other hand, I wonder if 2.5% is unnecessarily skimpy. If I just take out 3%/year which was my plan, I take out more after good years and less after bad. Because my budget is closer to 2%, absorbing the variations should present zero problems. In fact, I suspect that we'd actually have a hard time spending the full 3% anyway. Habits die hard.

Anyway, thanks for "listening" and hopefully not glazing over. Any thoughts or corrections most welcome.

SDFIRE

 

[brewer12345]

What does firecalc say?

 

[jebmke]

I do all my analysis in real dollars - makes it easier to keep things in constant purchasing power. For now I am assuming FI has 0-.5% real and equity is 3-4% real.

 

[HFWR]

What brewer and jeb said...

I'm planning, hopefully conservatively enough, for 4% real for stocks. Better would be good. BSV is currently yielding better than jeb's 0-0.5%, but worse than 3.5%...

The OP's plan looks relatively sound, but needs some FireCALC and Monte Carlo on the side. The straight line math can be deceiving...

 

[sdfire]

What does firecalc say?

I did several firecalc runs a few years ago using a 3% withdrawal rate adjusted for inflation. They showed a very high 90's percent success rate (its been long enough that I've forgotten the exact numbers).

But, I'm not planning on the typical 3% plus inflation withdrawal strategy handled by firecalc. I'm planning a straight 3% maximum withdrawal. By definition, I will have a 100% success rate. And, as I said in my original post, I suspect we will spend closer to 2.5%.

My question was asking for a confirmation on the math for the required returns to meet a 3% inflation rate. And, any other thoughts on the plan.

If we get any social security, that will be gravy on top of the plan.

Thanks.

 

[bo_knows]

I'm planning a straight 3% maximum withdrawal. By definition, I will have a 100% success rate.

Not to be the wet blanket, but I'm sure you're aware that the market has some big gyrations. If 3% = 40k, and you ran through the 2009 crash, would your "straight 3%" of say 28k (rough estimate of market drops for your AA) be enough to cover expenses?

Going with a historical-based calculator can tell you better if your 3% starting WR would survive.

 

[sdfire]

I do all my analysis in real dollars - makes it easier to keep things in constant purchasing power. For now I am assuming FI has 0-.5% real and equity is 3-4% real.

OK. I'm afraid that I get confused moving back and forth between nominal and real dollars. So, it is easier for me to look at nominal dollars/returns and back out inflation. Is it just a simple matter of reducing the expected return by the expected inflation rate? If so and if you assume 3% inflation, wouldn't that mean that you just subtract 3% from the returns I stated? So, my question restated using real dollars would be:

I am assuming .5% real return for fixed income.
And If I have a 3-4% real return on stocks, I'm afraid that I come up short on inflation protection. For a 50/50 stock/fixed income portfolio with 3% inflation and 3% withdrawals, I need a real total portfolio increase of 3% to keep up with inflation. If half of my portfolio gets 3-4% real and the other half only gets .5%, then, my overall portfolio increase is only 1.75-2.25% in real terms. Right??

If so, that is my concern. 1.75-2.25% doesn't keep up with 3% inflation. Over 50 years (I'm 50 now), I could be in trouble even with my 3% withdrawal regardless of ups and downs plan.

But, on the other hand, given the firecalc results, I feel like I'm being very conservative with 3-4% real returns.

Thoughts?

 

[sdfire]

Not to be the wet blanket, but I'm sure you're aware that the market has some big gyrations. If 3% = 40k, and you ran through the 2009 crash, would your "straight 3%" of say 28k (rough estimate of market drops for your AA) be enough to cover expenses?

Going with a historical-based calculator can tell you better if your 3% starting WR would survive.

Our current budget is just a hair over 2% of assets. Therefore, if stocks took a huge dive and we lost 50% of 1/2 of our portfolio, taking a straight 3% of the remaining 3/4's of the original portfolio would still be higher than our current budget. Using round numbers divisible by 4 for ease: If we had a portfolio of $8 million an initial 3% withdrawal would be $240K. If we lost $2 million in the stock 1/2 of the portfolio, the portfolio would be $6 million and a 3% withdrawal would be $180K. That's still higher than a 2% withdrawal or $160K of the original portfolio. Make sense?


That's one of the reasons that the straight 3% appeals to me. I think I can deal with the gyrations and I'm comforted by knowing that I won't continue to pull high amounts out when my portfolio tanks periodically. I would take out more after good years and less after bad years. And, the reality is that I probably won't take out the max in the good years. Its just not our personality type.

 

[brewer12345]

SDfire,are you familiar with ESRBob's 4%/95% withdrawal methodology? You might want to take a look at it as a comparator for how absurdly conservative your plans are.

 

[rayvt]

We are planning a 3% withdrawal rate over 50 years. Here are my assumptions ... As a result, my stocks need to earn 3% for withdrawal, plus an extra 5.5% for a total of 8.5%

Are you under the impression that stocks will deliver the return that you "need"? They don't.
A quote: "Wanting or needing x% return doesn’t cause it to be available in the market."

Don't overthink it. It's not just your spouse that's doing MEGO -- it's some of us, too. Your plan implicitly assumes that the market (stocks & bonds) will do as you plan for them to do. It doesn't work that way. The market does whatever it feels like doing -- your task is to deal with what it does.

That implict assumption is embedded in all your posts in this thread, with wordings like "required returns" & "I am assuming .5% real return for fixed income".

Use a standard, simple asset allocation, rebalance periodically, take withdrawals from the different asset classes to keep them balanced. Use Firecalc or any of the other hundreds of retirement calculators to see what they say.

 

[Midpack]

I did several firecalc runs a few years ago using a 3% withdrawal rate adjusted for inflation. They showed a very high 90's percent success rate (its been long enough that I've forgotten the exact numbers).

But, I'm not planning on the typical 3% plus inflation withdrawal strategy handled by firecalc. I'm planning a straight 3% maximum withdrawal. By definition, I will have a 100% success rate. And, as I said in my original post, I suspect we will spend closer to 2.5%.

FIRECALC has several spending choices, not just constant dollar/inflation adjusted. Sounds like you want to model 3% of remaining portfolio, you can do that with FIRECALC as well (see pic below). FIRECALC can give you more objective answers than anything a member here can. What's right for you and what's right for me can be considerably different even if we have the same portfolio, spending and retirement duration. There are members here who sleep fine retiring with a 75% success rate, and others who need 200% to pull the trigger.

Beyond average returns, are you familiar with the impact sequence of returns has on a portfolio? It's relatively negligible during accumulation, but highly significant once contributions end and distribution/withdrawal begins. I didn't see any acknowledgment in your OP.

The fundamental question for retirement withdrawal planning is what actual returns and sequence of returns will be. Assuming 8.5% or any other constant return is guaranteed to be wrong, you can be sure it won't happen like that. Looking at market history, Monte Carlo or some variation of returns will do more to help you understand how much is right for your risk tolerance and time frame.

 

[jebmke]

What brewer and jeb said...

I'm planning, hopefully conservatively enough, for 4% real for stocks. Better would be good. BSV is currently yielding better than jeb's 0-0.5%, but worse than 3.5%...

The OP's plan looks relatively sound, but needs some FireCALC and Monte Carlo on the side. The straight line math can be deceiving...

BSV is showing a nominal .88% yield which translates to negative real yield. My numbers were real 0-.5%, not nominal. Intermediate term bonds are running in the rough range of 2.5-3% nominal right now depending on the credit risk.

 

[HFWR]

BSV is showing a nominal .88% yield which translates to negative real yield. My numbers were real 0-.5%, not nominal. Intermediate term bonds are running in the rough range of 2.5-3% nominal right now depending on the credit risk.

Yep, my spreadsheet says 0.83% YTD, so negative real of around 0.67%, though that is a misleading number, i.e. income so far divided by total allocation. Should be slightly better by year end.

Nothing to write home about, for sure, but that portion of the portfolio is designed to reduce volatility. Slightly better than the coffee can buried in the back yard! I sold HYG and LQD, prematurely in hindsight, as they've not gone down as rates have gone up, albeit ever so slightly.

I may move up the yield curve at some point, just not yet...

 

[sdfire]

Inflation Is My Primary Question

It appears that I'm not clearly asking my question. I've run all the firecalc options and the results say that for a 3% withdrawal adjusted for inflation with a 50-50 asset mix, I have a 100% chance of success over 50 years. And, of course, my plan to simply take a fixed 3% out each year is 100% guaranteed not to run out of money. And, yes, I understand that the market goes up and down and may go down for an extended period just when I retire.

The question I was trying to ask is whether I had my math right for the required average return (recognizing that returns go up and down) required to take a 3% withdrawal for 50 years and still keep up with inflation.

I take comfort in running firecalc. But, firecalc can only give you responses based on the assumptions you put into it regarding expected returns from fixed income, etc. Getting "behind" the firecalc numbers has me looking at the required returns to take 3% and still keep up with 3% inflation.

I started with the opinion that a fixed 3% withdrawal rate was very conservative (I assume that is what the comment saying that I was absurdly conservative addresses). But, my math says that based on the assumptions in my original post that the market needs to average 8.5% for my plan to keep up with inflation. Given historic returns, that still shows 100% success in firecalc. But, 8.5% is higher than I would have guessed if you just told me that someone planned to take out 3% forever. I would have guessed that 3% fixed would work easily. That brings me back to my bottom line question. Do my figures make sense?

 

[Animorph]

It appears that I'm not clearly asking my question. I've run all the firecalc options and the results say that for a 3% withdrawal adjusted for inflation with a 50-50 asset mix, I have a 100% chance of success over 50 years. And, of course, my plan to simply take a fixed 3% out each year is 100% guaranteed not to run out of money. And, yes, I understand that the market goes up and down and may go down for an extended period just when I retire.

The question I was trying to ask is whether I had my math right for the required average return (recognizing that returns go up and down) required to take a 3% withdrawal for 50 years and still keep up with inflation.

I take comfort in running firecalc. But, firecalc can only give you responses based on the assumptions you put into it regarding expected returns from fixed income, etc. Getting "behind" the firecalc numbers has me looking at the required returns to take 3% and still keep up with 3% inflation.

I started with the opinion that a fixed 3% withdrawal rate was very conservative (I assume that is what the comment saying that I was absurdly conservative addresses). But, my math says that based on the assumptions in my original post that the market needs to average 8.5% for my plan to keep up with inflation. Given historic returns, that still shows 100% success in firecalc. But, 8.5% is higher than I would have guessed if you just told me that someone planned to take out 3% forever. I would have guessed that 3% fixed would work easily. That brings me back to my bottom line question. Do my figures make sense?

FIRECalc's "success" definition is that you don't run out of money. However, you may indeed spend it down to $0 by the end of 30 years. For a 30 year period, a real return of 1.5% gives you a 4% of starting portfolio plus inflation withdrawal rate for example. Your numbers would ensure that your portfolio's value in real dollars never decreased. That's a tougher prospect, and it's the reason your return numbers look kind of high. Put a required final portfolio value equal to your starting value into FIRECalc and it won't look so optimistic.

Also understand that with FIRECalc's historical data there are not enough years to get a good handle on a 50 year period. The last 50 year period it can run has to start in 1962 or so. It is common for a 40 year or longer period to have fewer failures than a 30 year period.

The US geometric average real return since 1900 is 6.26% for stocks and 2.01% for bonds.

 

[Midpack]

If I follow how you arrived at 8.5% reading your OP, I'd question if "your figures make sense." You can't simply add and subtract percentages to arrive at a result, and it appears you've double counted inflation on the equity side. Your approach misses the effects of compounding and increasing real dollar withdrawals, an important factor.

But IF I understand what you're getting at, a 50/50 portfolio with 3% initial withdrawal and 3% annual increase for inflation thereafter for 50 years assuming no residual $ at end of plan and a 3.5% return on fixed income would require about a 4.8% composite return, or about 6.1% average return on the equity portion according to my simple two column spreadsheet. Not of much use IMO, but FWIW.

Again, assuming constant returns is less meaningful than FIRECALC results. Maybe your spouse is on to something...

 

[rayvt]

I think you're getting confused with the inflation question. As well as overthinking it.
You've got to be careful to apply inflation adjustments properly to every piece of the situation, and not accidently include it more than once in any piece.

If you take the standard approach of setting a SWR and annual increase for inflation, then that's all you need to do. The historical market returns already include inflation, so you don't need to include it again. If you subtract 3% for inflation, you are double-counting inflation, so your results will be overly pessimistic.

Realistically, there's not much need to run Firecalc over longer that ~30 years. A portfolio that survives 30 years is virtually certain to survive any longer number of years.
Nor does Firecalc predict the future. What it does is tell you what the portfolio would have done in every N year period since 1871. And not even every N year period--only those starting in January, not starting in any other month.

And I'm not sure about "my bottom line question. Do my figures make sense?"
First, what does it matter? If you say that you need the market to do 8.5% for your calculations to work. But.....so what? The market will do whatever it does, regardless of what calculations or assumptions you have made. This is a little bit like debating where are going to take J.Lo after the prom. IOW, an exercise in futility.

 

[Midpack]

Realistically, there's not much need to run Firecalc over longer that ~30 years. A portfolio that survives 30 years is virtually certain to survive any longer number of years.

Virtually certain? If 4% is the 30 year SWR, the indefinite duration SWR is closer to 3% based on every academic paper I've seen. That's a significant difference...

 

[sdfire]

If I follow how you arrived at 8.5% reading your OP, I'd question if "your figures make sense." You can't simply add percentages to arrive at a result, and it appears you've indeed double counted inflation on the equity side. Your approach also misses the effects of compounding and increasing real dollar withdrawals. Again, assuming constant returns is less meaningful than FIRECALC results.

But IF I understand what you're getting at, a 50/50 portfolio with 3% initial withdrawal and 3% annual increase for inflation thereafter for 50 years assuming no residual $ at end of plan and a 3.5% return on fixed income would require about a 4.8% composite return, or about 6.1% average return on the equity portion according to my simple two column spreadsheet.

Maybe your spouse is on to something...

Thank you for your response. Your spreadsheet is very interesting. I'm also very happy with your conclusion that the stock portion only needs to average 6.1%. But, I'll confess that I'm not understanding how that happens. I assume this has to do with compounding of both returns and withdrawals.

However, I believe that your spreadsheet is showing the results for a slightly different withdrawal approach than my plan. My plan is to use a straight 3% withdrawal regardless of the swings in the portfolio value. By definition I never run out of money. The questions that I've been playing with have to do with whether the portfolio will keep up with inflation using that approach. I'm not sure how to model that (I did run it on firecalc, is there a way to see the ending values as numbers instead of as a chart?).

Thanks again for your insights. No fair ganging up on me with my spouse . . ..

 

[samclem]

My plan is to use a straight 3% withdrawal regardless of the swings in the portfolio value. By definition I never run out of money.

So, you mean "I'll withdraw 3% of the ending value of the portfolio each year, regardless of the value, and that's what we'll live on" right?

I'm not sure what the difficulties are, as this is one of several normal ways of calculating withdrawals, and can be entered directly into FIRECalc without any need for special additional calculations. If you want to know if the portfolio (and, consequently, your withdrawals) will keep up with inflation, just specify that you want the results in "present year" dollars, then you can see it all instantly. Maybe I'm missing something.

 

[sdfire]

So, you mean "I'll withdraw 3% of the ending value of the portfolio each year, regardless of the value, and that's what we'll live on" right?

I'm not sure what the difficulties are, as this is one of several normal ways of calculating withdrawals, and can be entered directly into FIRECalc without any need for special additional calculations. Maybe I'm missing something.

Yes, that is what I plan to do. I've played with firecalc and it shows 100% success. But, I don't know how to tell the actual portfolio values other than guessing from the chart. And, without the numbers, I can't tell if I'm keeping up with inflation. Firecalc calculates success at never running out of money. I'd like to see if my 3% withdrawals keep up with inflation.

 

[samclem]

Yes, that is what I plan to do. I've played with firecalc and it shows 100% success. But, I don't know how to tell the actual portfolio values other than guessing from the chart. And, without the numbers, I can't tell if I'm keeping up with inflation. Firecalc calculates success at never running out of money. I'd like to see if my 3% withdrawals keep up with inflation.

Okay, it's pretty easy.
1) On the very first screen there is a "spending" field. There's also a "Portfolio" field. Let's assume you are starting retirement today and your portfolio value is $750k and you want to withdraw 3% at the end of each year. You'd enter $750K here as your portfolio value and you'd enter $22,500 in "spending". Stay with me.
2) Click the "spending model" tab from the top and select "Percentage of Remaining Portfolio". That's all you have to do. You don't physically enter a percentage here, because FIRECalc is using the Portfolio" and "Spending" figures you already entered on the "Start Here" screen to calculate the %age you'll take every year.
3) Fill in the other info (Asset allocation, SS income, etc) as desired.
4) Notice the "FIRECalc Results" screen: All the numbers are in 2013 dollars. Because you selected "Percentage of Remaining Portfolio", this screen now looks different than what you saw before--there's no "Percent success" given. As you've noted, if you take a percent of each year's value, the portfolio can't "fail" (run out of money), but it might "fail" to keep up with inflation. So this results screen instead shows you if that happened when FIRECalc ran the data from historical returns. You can see instantly if your annual withdrawals and modeled portfolio values would have kept up with inflation (lines go "up" in constant 2013 dollars) or didn't (lines go down).

Does that show you what you want? It's a great program--definitely worth a contribution to the FIRECalc kitty.

 

[sdfire]

Yes, that is what I plan to do. I've played with firecalc and it shows 100% success. But, I don't know how to tell the actual portfolio values other than guessing from the chart. And, without the numbers, I can't tell if I'm keeping up with inflation. Firecalc calculates success at never running out of money. I'd like to see if my 3% withdrawals keep up with inflation.

OK. I am embarrassed to say that I just saw the little box that allows you to ask for a spreadsheet from firecalc. This very valuable tool allows me to see the value of my portfolio over time as well as the inflation adjusted portfolio value. Just exactly what I was looking for. Now, my only remaining question is how I can tell firecalc to assume a 3.5% fixed income return combined with a total market return for the stock portion using historic stock values?

 

[samclem]

Now, my only remaining question is how I can tell firecalc to assume a 3.5% fixed income return combined with a total market return for the stock portion using historic stock values?

I don't know how to do that, and I doubt it's possible.

As you may know, one (of many) good things about FIRECalc is that the historical data for each year is used in the sequence in which it actually occurred, and each year's data is used as a block with the other data from that year. That's realistic: you wouldn't want to have the 1972 inflation rate used in the same year as the 2012 ST interest rates: We're highly unlikely to have ST interest rates of 1% while inflation is at 10+% unless our economy and markets have come unglued. So, it probably makes more sense to use FIRECalc's historic CPI rate along with it's historic interest rate information rather than trying to enter your own interest rate.

In my opinion.

 

[sdfire]

I don't know how to do that, and I doubt it's possible.

As you may know, one (of many) good things about FIRECalc is that the historical data for each year is used in the sequence in which it actually occurred, and each year's data is used as a block with the other data from that year. That's realistic: you wouldn't want to have the 1972 inflation rate used in the same year as the 2012 ST interest rates: We're highly unlikely to have ST interest rates of 1% while inflation is at 10+% unless our economy and markets have come unglued. So, it probably makes more sense to use FIRECalc's historic CPI rate along with it's historic interest rate information rather than trying to enter your own interest rate.

In my opinion.

I follow the concept. But, how does firecalc include the costs of switching to higher interest rate products? For example, if you are currently holding total bond index fund, the NPV of the fund will decrease with every interest rate increase. Similarly, I currently hold a chunk of individual muni bonds, if interest rates skyrocketed, I could sell my muni bonds, but it would be at a significant loss. I may be better off holding to maturity. I don't want to get too off topic here with a discussion of fixed income options. But, I am wondering if firecalc somehow factors in the cost of switching fixed income products if interest rates rise?