There are all kind of popular formula as to retirement. 4% x 25, 3.3% x 30, 3% x 33, and so on. These formula describe withdrawal rate and are based on longevity. So what’s the risk? WR in these formula is all about performance. No one really expresses risk.
Here is a picture (monte carlo simulation) of a $1,000,000 portfolio, in a Bogelhead 3 portfolio, with a 40,000K/yr withdrawal projected 30 years, and no sequence of return bias and using historical inflation.
The program runs 10,000 simulations and survives 8,679 times. If I shorten retirement to 20 years survival improves to 9575 times. 15 years takes us to 9915 survivals. 50 years is 7331 survivals.
Here is with a 2 year bad SORR in the first 2 years, 4% over 30 years normal inflation. This is how your portfolio would perform if you retired into a recession, or were laid off because of a recession and decided to retire.
You survive 5,914 times out of 10,000
Here is 50 years because you decided to retire early or couldn’t find a job.
You survive for 50 years only 2827 times.
You say BUT I WOULD CUT BACK TO SUPER SAFE 3%! Here is 50 years at 3%
Better, you survive 6263 Times. How about if this was a 30 year retirement instead of 50 still retiring into a recession?
Better still! Your up to 8435 using “super safe” 3% and 2 bad years of initial SORR. Did you ever read the refrain how FI gives you independence? You call these last few independence? Notice the down tick at the beginning of the graph. This IS a graphical representation of SORR
How about a Bogelhead 3 @ super safe 3% WR over 30 years with normal SORR?
9535 certainly safer but super safe?
How about 50 years super safe 3% normal SORR?
8808 successes over 50 years! That means you fail 1 out of 8 times.
What about 2% Hell nobody retires on only 2%!
First over 30 years:
Now you’re talking! 9929 successes over 30 years at 2% WR! What about 2% at 50 years?
9678, almost 97% survival for 50 years at 2% WR normal inflation, normal SORR. This is why Bernstein says 2% over 30 years is basically bullet proof
These graphs were generated using the Monte Carlo simulator in Portfolio Visualizer. Monte Carlo does a forward looking simulation of 10,000 probable futures and then statistically orders them in terms of likelihood. You can decide, but the risk analysis is quantitative forward looking as opposed to “rule of thumb” or historical . I base my retirement on Monte Carlo.
No pressure in this post. Retire at 40? What the hell!
FIRE types use formula to determine retirement. It goes “take your retirement amount multiply x 25 and withdraw 4%. Or multiply x 33 and take 3%. People think this is “safe” based on looking in the rear view mirror. A 1998 study looked at 30 year aliquots of time from before 1900 to 30 years ago and judged nobody ran out of money so it must be safe. The 30 years was expected to cover from say age 60 to 90. This some how was generalized to save 25 or 33 take 4% or 3% and you’ll never run out of money so once you hit the “number” pull the trigger. Some problems: 1. Past performance does not equal future result. 2. It doesn’t really make mathematical sense. Retirement is based on both Risk and Reward, not just projected Reward. Here is a graph of Human Capital:
The red line represents what your potential earnings capacity is from birth to about age 70. After 70 consider yourself burned out.
This graph represents a “Mr. normal retirement” The orange line is saving and investment from ages 20 to 60, and the blue line is deflation from retirement at age 60 till death at 90. He gave up a dab of red line earnings retiring at 60 but he accumulated over 40 years and so he has more than enough to cover 30. The area under the triangle represents the money you made during accumulation or the money you spend in retirement. You generated 1.33 times more money in accumulation than you spent in retirement. He will die with money in the bank and a cushion in case of bad times. He could guarantee his 1x in TIPS and put .33 in riskier assets and likely leave his kids a nice chunk. His risk of running out of money is extremely low.
This is Mr. FIRE. He accumulates from 20 to 38, 18 years and then deflates for 52 years. He’s hard charger with a great job and lives in his mother’s basement so he basically saves it all. He gives up a huge chunk of earnings potential (ages 38 to 70, 32 years potential) and he needs to generate twice as much money during retirement as he generated in accumulation to satisfy 52 years. To generate twice as much money he needs to be in fairly aggressive assets with high risk. As risk goes up failure goes up. He waves his hands a lot about side gigs and such but hand waving may or may not generate that extra needed to cover till death do us part. This guy is essentially retiring in debt. He is not covered. He presumes it will work out.
This is Mr. middle of the road. He retires in his 50’s, say 52 after working 32 years of saving his ass off plus working a second job and spends his 32 years of accumulation on 38 years of retirement. His retirement money is invested in safe assets which return an extra 6 years of free retirement. His failure risk is low. Mr. middle of the road retires a tiny bit in debt but his potential compounding over 38 years should easily make up that difference. No Walmart jobs starting at age 70 for this joker he has fishing on his agenda. Remember we consider age 70 as burned out regarding Human Capital.
3 retirements, 3 pictures of risk. Every picture tells a story don’t it.
I seem to be obsessed with Roth conversion. I’ve analyzed it 68 ways to Sunday, and written articles on that. My take away was to develop a scheme of the most efficient conversion. What I discovered is to convert at best efficiency and you have a fair pile to convert you need to be living on cash and it works best if you are married.
Love and Conversion
The tax law is such that being single kills conversion. A married couple filing jointly can convert much more at a lower rate
Here is a list of maxed out contributions for married and the tax on an equivalent contribution for a single. Notice the tax bracket creep. Being married helps to the tune tens of thousands or more. ALWAYS CHECK THE SINGLE V MARRIED CALCULATION! In this calculation the deductible is the sum of my over 65 and my wife’s under 65 rates. If you die your wife will be taxed at single rates and have a single deductible which means her taxes will be far higher than while married for a similar income. You have to plan your portfolio deflation accordingly because Uncle Sam will happily take her dough. In addition after your death her SS will decrease so taxes go up and SS goes down meaning a bigger strain on portfolio deflation. Filling only to the 12% level doesn’t convert enough money to dent the tax burden of bigger TIRA.
The Time of the season
Till 2025 taxes are favorable, but the tax cuts are temporary! If you can, make hay while the sun shines. I spent months optimizing the most efficient conversion strategy. It turns out best conversion is in the years just prior to RMD, or when you can live off cash and are no longer funding IRA’s. If you planned for conversion, you can free up some cash to live on while using all of your tax bracket to convert. The number of years and amount you need is based on IRA size. If your going to convert the whole shootin’ match of say a 1.7M IRA your best conversion is 340K over 5 years. Your tax bill will be $321,000. Past the top of the 24% bracket things start getting drained by the tax man pretty quickly. If you plan to live on 100K per year you will need 821000 in available cash to effect your conversion. You would need 500K anyway to live on so the extra risk is only 321K or an extra 64K per year. You plan for that during accumulation, and if you do it right part of that 321K will be from interest,
If you put 10K per year away for 20 years at 4% that’s 320K at the end of 20 years, 34% of which is interest. But wait there’s more! If you start at age 30 and put an extra 14K per year away for 20 years at 4% that will virtually pay for the entire 5 year Roth conversion including 100K/year living expense and 64K taxes/year starting at age 65. 14k/yr for 20 years @ 4% grows to 450K. 34% of that money is interest. 450K for an additional 15 years @ 4% with no added money is 820K meaning your Roth conversion would only cost 280K, (20 x 14k) including 5 years of living expense. If that isn’t magic, nothing is! Essentially your 820K Roth conversion would be paid with 2/3 interest and 1/3 principal, and your non Roth conversion portfolio would continue for 5 more years of compounding untouched. As long as you’re married filing jointly you can pull 100K/yr out of a taxable account with no cap gains tax. In the mean time since you don’t need SS because you saved and have all this interest to spend, you can let SS ride to age 70 when you will get a bigger payout. I hear people call SS gravy. Not this daddyo, I call it red meat. A larger SS means my therefore smaller WR becomes bullet proof.
You can fund over longer or shorter periods but it’s hard to live on cash for say 10 years while doing funding. Longer periods of funding tends to be in smaller amounts per year so you do save some taxes. Part of the reason I retired was so I would have adequate time to do my Roth conversions efficiently. Best choice ever!
Discovery is in the Details
My Original plan was to convert 100%. I consider those taxes already belonging to Uncle Sam. The problem with waiting is Uncle Sam controls the tax rate and RMD forces you to withdraw. If you’re forced to withdraw A LOT and taxes go up you pay A LOT in taxes. Taxes are low till 2025 so my inclination is to get them out of the way.
This is an image of how RMD works the red line is ROI with RMD. The black line is ROI with Roth conversion. RMD continually increases and therefore taxes continually increase causing the slope of the red line to fall off as time goes on. The black line however shows what happens when you pay a chunk to start and then let the ROI recover. Eventually black beats red. It takes some time but if you live long enough it pays, and when it pays is significant. If you don’t live long enough you die with less than if you didn’t do the conversion, but your wife will reap that black line benefit. It starts paying just when you might need assisted living or some significant medical expense. This is a kind of sustainable self insurance. Yes you may need to forgo yet another trip around the world to pay for the conversion but when you need memory care that trip around the world will be long forgotten. (it turns out the black and red lines are also a pictorial of what mildly bad SORR looks like). The above pic dramatizes the differences between early SORR (black) and late SORR. In my own portfolio I ran actual quantitative curves, holding all other return risk stresses constant, the early SORR portfolio (black) outperformed the late SORR portfolio (red) by close to 1M over a 30 year horizon but it took something like 15 years for the lines to cross.
As I was working through conversion it became obvious the highest risked assets needed to be converted first, and lower risked assets each succeeding year. In other words convert the AMZN before you convert VBMFX. VBMFX is the last to go. I have virtually all of my bonds at Vanguard in IRAs except for some LTIPS in a FIDO IRA. As I was reading through the tax code I found out there is a 3.8% surtax attached to income above 250K jointly. Homey don’t like surtax so I re-evaluated max conversion. In addition there is an increase in monthly medicare expense. Regular folks (170K jointly income) pay 134/mo. 170K to 214K folks pays 187.50/mo, I pay 348/mo because I made too much money in 2015 (267K to 320K range). I can save over $1500 per year in medicare expense. Over 5 years that’s 7500. In addition by limiting my conversion to 250K, I save the 3.8% surtax and quite a bit of taxes upfront from not converting at 340K/yr. What I discovered is by converting 250K x 2 and then 214K x 3 I can convert about 1.1M of risky assets, take advantage of medicare savings the last few years before RMD and avoid a 3.8% surtax. My total tax bill drops about 150K. What I am left with is 600K of bonds in the IRA. If bonds pay 3% my RMD is about 25K/yr on the average. The Roth will continue to compound during conversion so with 1.1M converted over 5years as described at 4% return I will have 1.6M in the Roth when I turn 70. I have the rest in a taxable account with a big chunk of LT cap loss to pair against cap gains.
Our combined SS will be a little over 50K and my RMD about 25K, My living expense averages 109K/yr so I only need tap my taxable account for 35K per year plus taxes. Taxes look like they will be $4600 and once I get this rigmarole done medicare will drop back to 134/mo. I will never have need to touch the Roth so it will just sit in the background peculating, and my 150K tax savings also sits in my taxable account gathering interest.
One other small detail. I don’t recommend using “timing” but if the market happens to be down, it’s an excellent time to convert. A smaller portfolio generates fewer taxes, and the conversion can recover in the Roth as easily as if left in the TIRA. Roth conversion can no longer be re-characterized so once you do it, it be done.
Angels (not devils) are in the details
This post demonstrates what happens on the other side of accumulation. Your goal becomes all about efficiency and what I call parsimony, making the best value out of what you have. You don’t want to let the air out of the portfolio balloon too soon and you certainly don’t want Uncle Sam letting it out for you! This is my present Roth conversion strategy. Convert the stocks, RMD the bonds, compound the tax savings from not converting the bonds, avoid extra taxes and Medicare penalties as much as possible. Efficient conversion however depends on planning during accumulation. If you can make it happen, that accrued interest can’t be beat.
I just installed WordPress and it’s a completely different experience than my old blogging software. More intricate, more powerful, less boilerplate. The learning curve is moderately steep. My old blogging software was simple. As I work through the jot’s and tittles it occurs to me the experience is not unlike retirement. Pre-retirement is simple. Spend less than you make, put what’s left in some type of investment account, sit back and let the deadly brilliance of corporate America compound your money for you. You control the investment amount, you control the timing. The market controls the rest. Retirement is like WordPress. Lotta moving parts. How long is retirement? What about return? What about risk? Do you have enough? What about taxes? Medical cost? The rent? How much does all that cost? Which brings us back to do you have enough?
I’m past accumulation and into deflation (a word I prefer to distribution). I did my part, pulled the train, and now it’s time for discovery. I read a lot about plans and projections, but not that much about the rubber meeting the road. I read the pipe dreams of the W2 types pulling numbers out of thin air, but I don’t read all that much by actual retired folks living off their squirreled away acorns. I read about side gig this and scheme me that but not so much about what that pays. I never read about failure, just projected success. Failure is what happens when projected success doesn’t work out. Study of failure is how you assure success. You study failure and then don’t do that. Success and not projected success is what pays the rent and buys the hamburgers. I read an article analyzing a report this morning that looked at retiring middle class Americans .
I’m one of them. Hopefully I planned well. I haven’t read the study because it’s not referenced but it gives me pause to read 75% of workers plan to work past retirement age. If this the face of failure?
I’m a little jazzed at getting this thing to post and it looks pretty good!