More Galton Goofiness

Look familiar? The decision tree above yields the distribution below. The distribution is a population of results but the destination of any individual ball is particular. If success is ending up on the right it is the result of enough rightward decisions and the sequence of rightward decisions to get you there. In the above there are 12 decisions applied to each ball before it reaches the mouth of its final bin. If the first 6 decisions are leftward it becomes increasingly unlikely you will wind up on the right. If the first 6 decisions are rightward it becomes increasingly unlikely you will end up left aka fail. This is the essential definition of sequence of return and the essential display of conditional or Bayesian probability. In your retirement spend down getting through the first half rightward has a big effect on eventual success.

Here is the Galton board taking a nap. In this case success is winding up on top.

This is a Monte Carlo of a 60 year 4% withdrawal on a 60/40 portfolio. It is basically a statistical creation of the Galton board placed on its side. Instead of the whole board being represented only the success is represented at the top of the graph, the “right” cohort of balls. After failure there isn’t much point in gilding the lily about how bad the failure is so instead the bottom graph represents how soon failure occurs. In the above 7869/10000 simulations succeeded, meaning about 25% of these retirements failed by the 60th year. I’ve come to believe this is the proper way to view retirement, not some rearward looking analysis. The rearward analysis says nothing about the future it only rehashes the past. I stopped by FIREcalc yesterday and ran my numbers, and I think the projections are optimistic. I was reading a forum discussing how closely FIREcalc can predict and what percent failure is a reasonable percentage of failure, in other words using FIREcalc as the mechanism to understand what is enough. That’s the real rub. Enough is based on what you will actually need from the start of portfolio deflation (retirement) until the end and that number, the most important number, is a single data point the projected WR based only on what would have worked in the past dating back to 1870 or something. Lincoln was shot in 1865 and I’m not sure 1870 data is relevant to the analysis.

The above analysis tells you likelihood of failure and when that is likely to occur. The probabilities are conditional. Here are the probabilities based on the first 3 years of return being the worst

Pretty stark only 2173 out of 10,000 simulations succeed. Here is the super safe 3% WR success rate on a 60 year portfolio with the worst SOR first.

Better, only about half fail instead of 78% but hardly super safe. In the Galton board example it took about half the choices (6/12) before the likelihood of failure to dramatically start to recede and the likelihood of success to overwhelm. 30 years is a long time to live on pins and needles. It’s actually not quite as gloomy as that but what’s the point of blogging if you can’t create a bit of dramatic tension. It’s during those initial 30 years however you have the best chance of affecting where your ball will reside at the end, that’s when the conditional probability is most active on the outcome. It definitely colors the perception of how much is enough.

The above discussion shows the risk involved in leveraging your future. The 4% portfolio is more leveraged than the 3% portfolio Here is 2%

9234/10,000 succeed after 60 years using a 2% leverage. Probably why Bernstein uses 2% as super safe. I get a lot of pushback on this stuff but knowledge of how to adjust the odds is power.

Authoring Part Deux

To author effectively you need tools. A couple tools I think are useful. One is AoF’s spreadsheet course over on Udemy. This course will teach you how to make an effective financial analysis tool using XL or another spreadsheet. I highly recommend. A spreadsheet allows actually granularity to be applied to the future and your projected need as opposed to the MMM 4th grade math approach. Granularity is how you move the ball rightward on the Galton Board. Granularity implies the application of a Bayesian understanding to reality.

The second tool is Jordan Peterson’s self authoring tool It’s not a financial tool per se but a means to create and add and record granularity to your narrative, and make a record of the thought process. Memory is a vital component of creating the future. If you can’t remember what you were thinking or doing accurately you can’t adjust the probabilities. Sometimes it becomes evident the scheme needs whole sale revision and future directed journaling is quite useful. It is a means to grow your tree of life and it can not be done wrong. Here is a 2 second example

“my time is my own so this course is to find how to spend my time. I want enough to live without the need to work or worry about finance. I want enough to be able to provide for catastrophic situations. I want enough to see my kids get a BA or BS degree the prestige of the school is not relevant, getting educated is relevant”

By writing down short statements needs and wants get specifically elucidated and a plan can be developed around each and later modified because you will be able to actually remember what you were thinking. The Peterson tool is specifically designed to create future narratives which can have a high probability of success since they are directed and not just sitting down with Notepad and creating pages of gobbledygook. If you want to pound in a stake you need a hammer. To pound in a BIG stake, BIG hammer

The journal provides the narrative and nthe spreadsheet provides the quantitative means to understand funding the narrative. Big Hammer

Authoring Your Life

Next week I’m heading for bypass surgery so I may be back or not. I may wind up dead or I may wind up stroked drooling in my lap going weedle weedle weedle. It’s a matter of odds. , so what are my odds?

My genetics are CAD. The genetics in my case is based on metabolic syndrome which is a constellation of disease process including obesity HTN, T2DM. lipid abnormalities and possibly neuro degenerative disease. Each of these “sub diseases” in my opinion is a phenotypic expression of the main disease, metabolic syndrome. The sine qua non of metabolic syndrome is insulin resistance and insulin resistance is pretty much strictly diet based. The culprit is carbohydrate, all of it. Insulin is a master anabolic hormone and it will take precedence in hormone competition. Excess blood glucose causes insulin to be secreted and the goal of insulin is to store this excess energy and the end organ is to store the excess glucose as fat in the fat cell. Constantly bombarding the tissue with glucose, constantly bombards the tissue with insulin, which constantly bombards the tissue with an anabolic stimulation. The normal is to cycle. The cyclic hormone is glucagon which is catabolic. In the absence of hyperinsulinemia, glucagon undoes what insulin does. Glucagon promotes turning stored fat into free fatty acids available for energy and it promotes the conversion of glycerol into glucose through neogenesis in the liver. You can eat only protein, some fat and no carbs and you will have a normal blood glucose and adequate stores of glycogen based on the liver’s ability to make glucose from glucerol and amino acids. The body knows what to do. It does not involve a lot of hocus pocus or vegan schmegan or vitamin schmitiman it involves turning off the constant anabolic insult of tissue by hyperinsulinemia which is directly caused by carbohydrate consumption. The most extreme form of this diet is Zero Carb in which you live on meat some fat and water and let the enzyme system of the body normalize and heal.

I’ve had CAD for decades based on this poorly understood metabolic syndrome scenario and I’ve treated the sub diseases accordingly under supervision. I have managed to avoid actual muscle damage through a cardiac stent and exercise. but my genetics for the vascular inflammatory response to hyperinulinemia and my coronary anatomy is my problem. The heart has 2 main arteries right and left. My heart has a vestigial right artery. not diseased just barely there. My left is patent and the only artery feeding my heart. The left immediately subdivides to 2 arteries the circumflex and the LAD and those are diseased. The circ has a stent and the LAD was too small to stent. That is my anatomy PERIOD. You have no control over your anatomy, that is the hand you are dealt. My course after the stent was to slow the progression but stenting doesn’t change your genetics nor your anatomy. I exercise 4 hours per week and can raise my heart rate to 180 for a sustained period (bordering on anaerobic metabolism) and can recover to a normal heart rate in less than 2 minutes and my blood pressure response to extreme exercise is non hypertensive which means my vasculature responds to correctly to the metabolic load of exercise. It wasn’t always so. In previous years, though I exercised, the exercise was suboptimal and my heart responded in a decompensated fashion as opposed to it’s present proper response. I got to proper response through proper training. Since I retired I didn’t have a 60 hour a week job to get in the way. By changing my diet to zero carb, I lost weight and I lost inflammation and I lost hypertension and diabetes which are the killer combo of the CAD HTN DM Obesity quartet of metabolic syndrome. My arterial disease and anatomy remain.

During my cath I had them shoot my carotids and my c’arotids are clean. No disease. I also had the leg arteries doplared and no PVD. I don’t smoke which means my cilia work correctly and my lungs are not scarred. I do not have pulmonary HTN based on cath data and my CVP normally runs zero to +1 which means no component of failure. I got to this state through effective training not half assed suggestions from the Mayo clinic. I didn’t know if I would be able to be bypassed, my cardiologist felt I was not a surgical candidate and would be doomed to stents, a substandard solution given my anatomy. Given my rate of decline I figured I had 2-3 years left before the disease moved from salvageable to failure. I consulted with a surgeon had he said he could do a LIMA to the LAD and a vein graft to the circ, a 10-15 prognosis improvement. Because I spent a year preparing for surgery, getting all my ducks in a row even though not a surgical candidate according to my cardiologist. My acute prognosis post op is

This would have been my scores a year ago before I went into effective training

The point of this article is you have the power to author your life, but only if you do so effectively. Doing something is not enough, doing what is necessary is what is required.

If I had followed conventional treatment, I would still be living the slide 2 risk profile. I was fortunate to hit upon a combination of protocols that changed the odds, and it’s all about changing the odds. Hitting that combination was not random. It took all the experience of my 67 years. There is a properly productive way to play and an ineffective way that is actually pseudo random. If you are doing ineffective stuff you are basically doing random stuff. The process is Bayesian in that an action then effects the probability of all subsequent probabilities. This idea is best described by a Galton Board.

The video describes a normal distribution of events, and is based on the random 50/50 odds of a ball tracking it’s way to its final bin. Bayesian statistics suggests that although the populations behavior is normative the fate of any individual ball is not random. If you wantto end up on the “right” side of the board you have to bounce as much as possible rightward, and your rightness therefore can be affected by making choices with rightward probabilities and risks.

Bayes theorem is described in this video on the Kahn site. By going into training I affected my odds moving them considerably rightward, and by a combination of correct weighting of choices the odds improved even more. Conceptually this in my opinion is the best way to approach retirement as well as life. It’s about growing and pruning trees and affecting the probabilities which can be calculated. The above calculator is a medical example. Monte Carlo in the financial realm is another example.

In the past couple months I’ve become somewhat disillusioned with FIRE. It’s all taken on the sheen of bullshit. I see all kinds of cute schemes which are likely ineffective and random in the main. For example people invest in low cost index funds. Investing in low cost index funds guarantees you won’t do better than market return and in fact that is the point of this way of investing, yet people persist in trying to guild the lily and outsmart the market with this scheme or tat scheme, this hustle or that hustle and no way of determining the effectiveness of the hustle. A t some point it becomes about marketing and not knowledge. A t some point it becomes Amway aka multievel marketing, and I think the point has been reached. The FIRE Blogoland is not any less susceptible to Galton board statistics than anything else since its human endeavor raised to the level of religion. I see much in the middle and no one trying to Bayesian their way to the right, rather opting for clicks over ideas. Too bad really, all those wasted words selling soap instead of ideas.

Next three months are gonna be a bitch for me, and not sure what my FIRE apatite will be post op. Depends on whether my squash will still work at speed or I get pump brain, time will tell. In the mean time keep effectively biasing your decisions rightward so you wind up on the right side of the mean. That is how to author your life, by creating truly effective rightward narratives and avoid the click bait.

More On the Concept of Yield Shield and 4 x 25

I’ve been thoroughly enjoying Big ERN’s series on yield shield 29, 30, and 31. In a very even handed way Karsten pretty much demolishes the illusion of SOR “safety” associated with yield based retirements. I think the illusion comes from the way FIRE is sold by the bogglehead crew and others like Mustache. The shockingly simple math is the math of a passbook savings account, not a equity investment account, and given the passbook savings rate even that is far from safe. 1M @ 4% with a 4% WR works fine till the first 4% goes to 3 % or inflation goes to 3.5%.

I was reading another article by a young physician who has 1/2M in the bank and “decided” 3.3M would give him a FI retirement, but now it’s up to 5M. If you actually inflation adjust to just nominal inflation, $150K/yr added to 500K takes 19 yrs to get to 5M. Imagine at high inflation how long it would take. At 3.5% inflation for the same rate of return it now would take 22 years to make that 5M nut and every year after you pulled the trigger your effective WR and “safety” would be less. I’m just playing with shockingly simple 4th grade math here, not calculus or differential equations. Look at these numbers. A cut in the rate of return from 4% to 3%, with a 4% WR wipes you out in 47 years. An increase in inflation from 2% to 3.5% blows you up! These small changes represent SOR. They are what constitute SOR. SOR giveth and SOR taketh away. The yield shield does not take into account SOR because the calculation is based on a constant as in it’s not a variable but life is not based on a constant. SOR by it’s very definition is a variable. It’s a sequence not a constant. Where is the sequence variable in 4 * 25? In 3 * 33? Yet people blindly plan their lives around these variable-less formula and behave as if there is some assurance of success and get all pissed off at Suze O when she dare challenge the illusion.

Lets take the illusion of success and turn it on it’s head using a simple FV calc No fancy Monte Carlo statistics but iterative 4th grade math.

This is a 1M 4% WR 30 year retirement @ 6% return. 2.25% inflation and a 12% tax load. In 30 years your 1M at 6% would generate 2.748M. Taxes, inflation and WR would eat 2.000M. Your have 1.030M of buying power left

Here is the scoop if you make 4% on your 1M instead of 6%. After taxes inflation and WR your 1M has 398K of buying power left.

Here is the scenario at 2% return all else the same. 64K of buying power left. Better hurry up and die! You will notice what we did literally was vary the sequence of return. We turned the constant into a variable, with a pretty dramatic result! THIS is what Suze O knows! She intuitively understands there is a variable in there somewhere to worry about. Ignoring the variable is the face of retirement failure. This retirement is levered to at least 2% rate of return to satisfy the assumptions.

Lets say you get a diagnosis where instead of 40K WR you need 52K WR. For example the drug Repatha costs about $1000/mo

You ran out of money a long time ago. Your Repatha bill killed your portfolio off at 23 years. “Varying” the WR is part of SOR.

Here is a 60 year Mustache retirement. Mustache retired at 30 with shockingly simple math so this calculation takes him to 90.

At 6% Mustache still has 1.8M left in the bank at age 90.

At 4% return he better hurry up and die! Sucka is flat out of dough. 2 variables changed in this scenario SOR and longevity and this portfolio is leveraged to a 4% rate of return for success. This is the problem when you create scenarios like the Yield Shield. If your money is coming from yield your are ignoring growth and growth is your protection against inflation. It’s also the the problem with formula like 4 x 25 where you just pull numbers out of thin air and call it good enough. If Mustache gets a bad diagnosis at 55 or 60, with the above 4% plan the guy is screwed.

Skinning Cats and Fleshing Out Disaster Funding.

I have a well documented plan for post retirement risk management, which includes varied accounts, pre-tax, taxable brokerage, and Roth. My plan is to annuitze the pre-tax accounts through RMD after whittling down their present size through Roth conversion to something that won’t force me into higher tax brackets too quickly. It’s a perverse race between taxes and death. My Roth is primarily disaster money, in case of emergency break glass. The Roth is the most valuable money I own since I own it tax free, while the other two are still taxable, as ordinary income in the case of the pre-tax, and as capital gains in the case of the brokerage.

My pre-tax RMD has included a tax discount because a certain percentage of it was funded with post tax money and the government only taxes you once on pre-tax money. As I was working through my taxes it occurred to me there may be a different format for disaster spending, by spending down the pre-tax money first.

This is a picture of a $600K RMD amortized on the government’s schedule over 45 years @ 3% return. The green is the yearly disbursement and blue the remaining portfolio. It occurred to me after the 7.5% “standard write-off deduction” is met, in the case of a medical emergency you could basically extract extra money from the account tax free, never paying taxes once you meet the threshold. This means a 600K @ 3% pretax (mostly bonds) could provide about 8 years of of catastrophic care @ 85K/yr essentially tax free or at a very low rate. That extra 8 years of growth would be tacked onto the Roth and Brokerage. A 1M Roth would be worth 1.5M if it were risked in a portfolio returning 6% for 8 years with no withdrawal.

Let’s say you make 40K/yr on SS of which 85% or 34K is taxable lets say you generate a $50K bill fighting your cancer diagnosis. The 7.5% threshold is $2,550 so you should be able to pull $47K out of your pre-tax account, tax free to pay your medical bills, and leave the Roth to grow and pay your living expense later if you survive. Your married filing jointly tax bill would be less than $1K/yr on $90K of spendable money You can pull 50K/yr from a 3% 600K portfolio for 15 years before the stream runs dry. In the meantime the Roth grows to 2.3M. How do you spell relief? I spell it “well thought plan”.

A nice strategy to have up your sleeve if the time comes.

Yield Shield Schmield

There is a raging little argument going on over at Millennial Revolution. It’s over living off yield vs living off growth. Wanderer loves his “yield shield” and “cash cushion” and Big ERN is a growther through and through and claims living off yield is a mirage. I’m with ERN on the yield argument but I think there is merit to a tweaked cash cushion approach as well. My cushion tweak is to use an efficient frontier tangent portfolio instead of cash, so it grows in growth years efficiently but does it’s job in down years of protecting the portfolio against early SORR. My tweak also is once the tangent is out of money do not refill it just move to start living off the higher risk portfolio.

Here is 30 years of S&P 500 returns starting in 1988, 5.1% plus 2.245% yield for a net reinvested inflation adjusted 7.35% annualized return!

Here is 20 years of S&P starting in 1998 1.7% growth, 1.9% yield for a total of 3.626/yr total return.

here is a 2008 to 2018 look see, 9.366 growth and 2.281% yield for a whopping 10 year reinvested growth of 11.647% Reinvested yield is a supercharger! $1,000,000 at 11.647 becomes OVER $3,000,000 in 10 years. $1,000,000 at 9.366 becomes $2,440,000 in 10 years plus
281K of compounded dividend interest for a net 2,721,000. If you lived off the yield you would have on the average 28K/yr to live on. If you lived off the growth 3M-2.4M = 600K over 10 years = 60K/yr. Both portfolios after 10 years would be 2.4M

In the 20 year case 1M would grow to 2.04M over 20 years @ 3.636. If you lived off the yield you would live on 23,000/yr and the end portfolio value would be 1.18M. If the reinvest guy winds up with a 1.18M portfolio after spending some WR he could could live on 43K/yr. Both portfolio’s after 20 years are worth 1.18M . If the reinvest guy lived on 23K/yr he would have a portfolio worth 1.38M , 200 grand more than the yield living dude.

It is what it is. Reinvesting the yield is like a dollar cost averaging a yearly cash infusion. The good thing about yield is it tends to be non correlated with S&P growth, so you DCA pretty much the same amount each year whether the S&P is high or low, so some years you buy low, some high but you’re always buying and allowing that dividend money to be exposed to growth. On really good years your growth on the dividend way out performs your WR so only some % of the dividend is extracted and the rest just continues to grow, and grow and grow. In the down year you would prefer not to sell so have some low risk tangent frontier portfolio around to sell instead. I once was enamored with dividend stock scenarios till I did the math. You can goose yourself into using “high dividend stocks” but those can be dangerous and tend to be in concentrated sectors. GE paid a good yield till it didn’t. In the meantime it lost nearly ALL of it’s value. A yield shield doesn’t shield anything of the company goes into negative growth.

I’m with Big ERN except I include the ability to use tangent funds on years the market is down (YTD less than zero) to live on, and still reinvest the dividends when the stock price is low. Do that for a few down years especially in the first half of retirement an u gone b a winner!

I Took The Plunge

I’ve been sitting in quite a bit of cash for a year, largely in my brokerage account. I have 3 brokerages all 3 pay bupkus on the money market accounts. I think it how they generate those “free” mutual funds. They shave points off the money market returns to pay for the mutual funds. Individual investors are notorious fo leaving money un-invested and therefore un-risked and all that un-invested money just keeps the light on plus a tidy profit for the brokerage. With the FED taking it’s boot off the banks you can actually get a tiny bit of interest income these days. I considered putting the dough into a CD ladder but CD ladders tie up the money and create taxes. I did put the money in a short term muni bond fund which “should” have paid me a little tax free while waiting to spend down the balance over the next 4 years, but with interest rates backing up the Muni fund went down and I promptly lost 5K so I decided to take the loss and make another plan.

I could ladder a CD for about 3% average yearly return having trounces of the ladder come due across the period in time to pay for hamburgers and taxes, but the non-liquidity about that plan bugged me. It tied up the money in a way that would generate a penalty if I needed some cash fast for some reason. It turns out there is now online banking which pays pretty high interest, compounded daily, perfectly liquid, FDIC insured to 250K. So I decided to stuff that money into a couple high interest savings accounts. The account yields 2.45% annually, compounded daily paid monthly. The interest is taxable, so given my Roth conversion at the 24% bracket for every dollar of interest I make I have to send Sam 24 cents, but keeping 76 cents ain’t such a bad deal seeing as how presently I’m making 0.1%. If I had a CD ladder I would have a similar tax drag.

The logistics were easy. I have my accounts linked so I merely validated the internal and external accounts and started transferring money. Accounting for the transfer was virtually over night but I’m sure setting the account will take a couple days. The savings account does not have checking or any other “features”. It houses money, compounds money and allows transfer in and out. Unlimited transfer, limited transfer out is limited to 6 electronic transfers per period, or by check PRN. I generally transfer twice a year, enough to cover hamburgers and taxes and pay bills through a checking account. Should generate enough over the next 4 years for 3 free trips to Europe or maybe 2 to Asia, after taxes.

Progressive Tax Code




Here is a little graphic representation of the first 4 tax brackets. The taxes you pay are the area under the curve. Blue is 10% 10% ends where Green or 12% starts at about 20K. I extended each graph to show the progression. 12% is from 20K to about 80K so it’s progression ends when Red starts and it’s 3 times bigger than 10%, but it’s slope increases only a little bit. This is the middle class. Red is 22%. It extends to about 160K 80 to 160K is as long as from 0 to 80K but notice the slope virtually doubles. I put a little black piece at the end to give a feeling of what 24% looks like, 24% extends to about 320K or double again from 0 to 160K. A progressively increasing slope up to 37% plus the 3.8% surtax. This is why it pays to Roth convert a big TIRA. The top of the 12% including standard married deductions extends to about 104K/year and every dollar after that is sloped up as the government eats an extra dime from every additional dollar. If you can avoid the green to red transition the tax saving soon enough becomes NON TRIVIAL. You avoid that by cleaning out the TIRA to a point where it takes a long time for RMD to push you past the green to red transition.

It’s interesting to note once past 600K the graph is no longer progressive but becomes linear at a 37% slope plus the sur tax. What that means is if you are pulling 4% WR in ordinary income on 15M dollars you pay the same “tax rate” as someone who is 4% on 30M or 300M.

Toward a Quantitative Understanding of Retirement

I’m going to define retirement as age 65. If you want to call it something else be my guest but that’s my definition. The society is defined around age 65 retirement in terms of social services. The problem with FIRE is there is no definition and you can’t “quantitative” a non definition, you can just blubber about simple math and bullet points. How you get to 65 is your business. How you fund beyond 65 is your business. Understanding the post 65 landscape however greatly informs how to get to 65, so MY analysis starts aimed at the end not aimed at the middle like most FIRE mumbo jumbo. Once you have a clear working model of what happens after 65 then you can start screwing around with the precedents. Included in the model are typical retirement amounts one might acquire by 65 in the upper middle class married brackets. I chose the numbers because they point out places of optimization. Whether and how you optimize is your business. I’m interested in an understanding and not a precise how to guide, but enough detail that you can make choices.

Roth Conversion

My choice is to Roth convert at least some of my TIRA money. There are several cliffs in the tax code as well as progressiveness in the tax code and they are on goingly cliffy and progressive. In addition everybody dies. It’s a known known. When you die is a specific expression of your genetic phenotype and is a known unknown and that introduces the risk of longevity, since you need enough money and a well controlled enough burn rate that your money will out live you. Forget about dying the richest person in the cemetery. Richest, poorest, the point is to die with the security and dignity money buys you and not forced to live a substandard life. This model is about the relationships of portfolio survival not about the fantasy of yacht ownership. Since we have a known unknown you should insure that unknown. Some people buy SPIA which I think is a waste, you wind up paying the insurance company for what you can do yourself. In my analysis the Roth provides that insurance against a high risk and catastrophic situation, as well as some buffer to ongoing unforeseen WR issues as well as some tax relief in the long term and a divorce from the government from control of your retirement accounts, since Roth’s don’t RMD. To own a Roth you have to fund a Roth and funding a Roth can be expensive but there are means to optimize that funding.

I chose a TIRA value at 65 of 1.4M and I analyzed what is the “best” ongoing outcome post conversion. A common maneuver would be to convert to the top of the 24% tax bracket which would virtually eliminate the TIRA. This maximizes future tax relief at the expense of high early cost of conversion. A 340K “top of the 24%” conversion will cost you 65K in taxes per conversion year. A conversion of 250K will cost about 35K so for 90K more conversion you pay nearly twice as much in taxes. In addition income passed 250K for marrieds starts to pile up excess penalties like a 3.8% tax surcharge and double or triple Medicare costs beyond the excessive taxation, it truly is about soak the rich. My conclusion was media via, middle way. For a 1.4M TIRA a plan aimed at 4x 250K conversion was about most efficient in my estimation. You could go to more years but each year has considerable expense on it’s own that needs to be funded. I further analyzed the best conversion time is likely between 65 and 70 because of Medicare eligibility which is likely cheaper than the commercial alternative. You also don’t have to convert equally, so you could convert biggly early and smaller later, but it’s good to convert the most earliest to get the tax free compounding going.

The Cost of Conversion

The cost of conversion is determined by the cost of living plus the excess taxes. The cost of living is not different that non conversion cost of living so if you live on $9000 a month your 4 year cost of living is $432,000 the taxes on 4 years of equal conversion is about 140,000 for a total of $572000. The taxes on the 432,000 cost of living over 4 years for the non converted life, if you are married is 40,000. So in the end your conversion costs an additional 100K in taxes over and above just the cost of living. If you convert 1M, you basically pay a dime in taxes per dollar and you are done with taxes on that money forever. If you convert to the top of the 24% you convert 1.36M and you pay 256,000 for the privileged or a net of 216,000 in extra taxes for an average tax rate of 15 cents on the dollar. You go AH HA I could convert less and save even more! While true the point is to pay the taxes now while living on cash not later when you have other income streams like SS and RMD generating taxes. Also the point is to get at least about 60% out of the TIRA to reduce the RMD and it’s tax burden to a better level. This is how to do the analysis, I’m not telling you what to do. This is what I did. I made a spread sheet to explain conversion. Using 4% growth and 2% to adjust for inflation as the constraints

My conversion is over 4 years starting at 66. The income stream includes Roth conversion investment income and SS income both gross and taxable. I actually chose 245K since I wasn’t 100% sure of my investment income and Roth now has a rule that you can’t re-characterize the contribution, Once contributed you owe those taxes. SS is tax advantaged in that you pay on only 85% of the benefit. SS is also inflation adjusted but I forgot to include that in the example. More on SS to come. You can see at 4% growth there is still a lot of money left in the TIRA even with conversion. My goal is around 500K-600K in the TIRA at the time of age 70 RMD. More in the TIRA generates a bigger tax bill forever, less in the TIRA generates a bigger tax bill sooner (10 cents v 15 cents). Remember this is a concept piece not a how to guide.

This is the schedule of my conversion the numbers are slightly less because my conversion amount was reduced by SS and investment income, but still nearly 1M is in the Roth by age 70 which was the goal. I had about 15K in my Roth to start. From this point the Roth just grows unmolested unless needed. You have a million bucks in the bank just in case. A spare million IS true financial independence.

SS concerns.

My wife worked early in our careers but then became a SAHM and raised my daughters. We home schooled so I assure you she was employed just uncompensated. Given her earning record at age 62 she is eligible for 800/mo and I am eligible for half of her 800 or 400 for a net 1200/mo or about 14K/yr. An extra 14K pays half the tax cost of conversion for a year and that money is tax advantaged so I can use tax advantaged money to pay for conversion and save my cash for other uses. My SS plus her SS at age 70 will generate 55,000/yr of which 46750 is taxable. Max payout for SS is 59000 for a family. So “our” earning record taken as a whole is 93% of max SS. SS is not a trivial benefit especially because of the tax advantage. Depending on your tax bracket SS essentially pays it’s own taxes up to the 15% average tax rate at which point the marginal rate over takes the advantage. People blow off SS in their analysis but what it allows for is a lower WR early in retirement because it replaces some of the WR burden on the portfolio. This is anti-SORR and desirable for portfolio longevity. SS has therefore more subtle utility on the end game than doing the “bent finger” analysis might suggest.

Budget

I’ll include my budget analysis more as an example of how to think about it than suggestive. When I retired I wasn’t sure what to expect. I read all the dope on “being happy” money and the “80% rule” and bla bla bla. None of it means anything. Your budget has a lower limit and it’s not the same lower limit as your W2 budget. It’s not the same because the W2 budget has part of the actual cost of living subsidized by your employer in the form of benefits. This is not a trivial benefit. It includes insurances, paid vacation, fees, sick days etc as well as retirement benefits and management of your life in terms of filing taxes. Your 100K in W2 terms may be 125K or more in terms of the risk that is paid for. When you quit all that extra risk befalls you. In addition W2 tends to give one a sensation of having “time”. Need a new roof? You got some “time” to cover that with the W2 and shuttle some dough toward the roof. Need a new roof when retired? Get out the check book and you better do it pronto because ignoring it only costs more. I just mention that since while you have the W2 you have some luxury to save up for the know known one off expenses like a roof or A/C or car etc.

In my case I had all of my early retirement “one offs” planned for in a separate account. I settled on 10K/mo as my “target”, did my planning around that and initially it worked out to be a rational choice. After 1 1/2 years 9K is what it averages out to be. That 9K covers all the additional risk of living beside the bottom line cost. I left the estimate at 10K and use that extra 12K/yr as pin money for splurges. The extra 12K pretty much means I never have to say no or feel constrained. I retired with 2 kids in college. One is launching, the other still has 2 years but #1 may go to grad school, such is life. If I need more I’m adequately funded but I like having a plan to use as a governor. In my working life I never budgeted. Once retired I strictly budgeted till I understood the moving parts using Mint. Now I check Mint twice a month and export the month data into a spreadsheet for further analysis.

The Post RMD Plan

Given that I now understand quantitatively and relatively precisely my actual needs and cash flow, I wrote a spread sheet which incorporates the plan including RMD, SS income, investment income, and taxes. By mapping out the plan year by year I get a precise and predictable picture. Notice the analysis has not thus far even engaged “net worth”, only “net need” and “net available cash flow” namely SS, investment income, RMD and self insurance. This analysis is a quantified picture of my future with contingencies in built, not shockingly stupid math.

This is not my exact plan but close. At age 70 SS provides 55K/yr and RMD provides 25,547 of income. I just set investment income as a constant 25K because it’s there but unknowable, a known unknown but over the past decade that’s proven to be a reasonable estimate so it’s better than a guess. You can read through the columns and see what I anticipate including what I anticipate my tax load to be. If the laws change I merely redo the estimated taxes. The RED part is what happens when I die. My wife is younger and her family is long lived. In my family all the men are dead by 80. I could last longer than that but who knows? If I last longer I’ll make a new projection, in the mean time the Roth continues to grow unmolested so I have back up. When I die she goes from married jointly with 2 deductions to less SS (she will take the survivor benefit when I die) and she will have a dramatically larger tax percentage. So the inflation adjusted gross income of 137843 drops to 103069 a 34775 loss in income but the taxes effectively go up a bit. The average tax rate increases at 80 from 19.7% to 37.2% at age 81. So mama is paying more taxes on less income! This needs to be planned for. In addition there is another income drop when SS is scheduled to be cut 25% in 2034. This is another 9825 hit to mama’s income and also needs to be planned for, else mama goes from living a comfortable life style to having to make moves to reduce expenses because you followed some dumb assed simple math and a few bullet points.

The plan of course is to have enough! Now we get to the portfolio aspect

This is a schedule of what it costs to generate 9000/mo inflation adjusted, given the proceeding RMD SS investment income and taxes burden. In this plan mama still get’s her inflation adjusted 9K/mo. I didn’t run this out 30 years since that’s your job. After tax income is compared to inflation adjusted spending and a deficit signal is generated. The deficit IS the portfolio WR. In this example SS is treated as an annuity, RMD is likewise treated as annuity. In my case I reinvest my investment income yearly. That dollar cost averages money into the portfolio yearly and I (or mama) extracts money late after it’s had time to compound. Notice the deficit is small for the first decade. Maybe 18K. If my investment income is 25K that’s a net positive dollar cost average of 7K into the portfolio. At my death the WR changes dramatically, but by then the size of the portfolio has grown also. The changes also include the 2034 cut to SS. Not to worry however! The burn rate is mostly covered by the post tax portfolio but note it’s starting to decline, which is fine. Mama is now 80 and there is plenty of money left in the portfolio. At her age 80 the TIRA still has 450K left in the account. I set the TIRA to a conservative 3% which pretty well assures its longevity as an annuity and controls the tax bite. RMD is considered separately therefore from WR. If absolutely necessary you can extract more than RMD from the TIRA whenever you want. RMD is required minimum. The taxable continues to provide hamburgers and is the real source of WR since that deficit is what you need to withdraw. How foreign to the usual boggle head boiler plate an actual variable WR FREAK OUT! If you need some money look at what’s sitting next to post tax money why it’s old Mr Roth! Mr Roth is there in case mama starts to run short or has the big whammy like a cancer diagnosis or something.

HOW MUCH?

The point of all this is to consider how much you need. At age 65 in this scenario you need 1M post tax, you need 1.4M TIRA, you need about 600K in cash to do the conversion. Which is 3M AT 65 for a 30 year retirement taking mama to about 95. Judiciously executed you have well over age 95 money available. I Monte Carlo’d the stuff as best I could and got about a 98% probability of success with these assumptions. What you do to survive prior to age 65 is a separate problem. If you borrow from the 3M portfolio to survive before age 65, it’s like taking out a second mortgage on your house, and the bigger the second mortgage the greater the chance of loss. You can retire at 65 with less than 3M or you can quit and screw up your % max of SS or you can fail to optimize taxes and pay a bunch when you are old. You can “max out your retirement accounts” and then wind up with no post tax account. Tons of alternate futures to play with This in my opinion is the proper way of thinking about FI and RE. When you quit you are taking out a loan on your future. If it’s a big high interest loan the odds are worse than if it’s a cheap small low interest loan. Oh I forgot you’re an investor not a speculator. The boggleheads told you so!

HR Block and the New Tax Code Rocks

I’ve had 49 W2’s over the years or the self employed equivalent. Over that time, once I entered medicine I started using tax accounting software DOS based on a floppy to start my journey. The software was TaxCut, which evolved to Kiplinger and finally was assumed by HR Block. I just stayed with the same package since filling in last years data was automatic and ongoing totals like 8606 post tax money in the IRA was preserved. This turned out to be a boon to me this year since I did my first IRA withdrawals for my Roth conversions. Because of the post tax basis in the IRA, I was able to reduce my taxes paid on the conversion, by 33% in my wife’s case and by 6% in my case. Her IRA’s are now completely converted. Mine are ongoing and my small IRA RMD withdrawal will last decades. The pro-rata rule will mean for every TIRA withdrawal I make I will get a 6% tax break till kingdom come. It’s not really a break as I already paid those taxes once, but it does reduce my AGI from what I will RMD when I’m 70, so my TIRA and my SS will have an ongoing tax adjustment for the better.

The software acting as a track keeper really helped. I kept the shoebox (actually bankers boxes) of old paper and sifted through to pull out the 8606’s from years gone by but some were missed and I completely missed my and my wife’s SEP IRA pro-rata contribution but the software didn’t miss any of it. When I finally got to those screens my pro-rata was filled in and significantly more than what I anticipated aka if found me some free money. The interview process made working through the Roth conversion and 1099-R info a breeze as well. I was kind of dreading that whole process.

The software now allows one to download account data directly from the brokerage server which saved hours and hours of sometimes confusing data entry. 1099-B was not implemented correctly so I had to enter that data by hand but I only had a dozen transactions last year so it only took about half an hour to enter and check. 1009-INT and 1099-DIV import worked fine. All three brokerages I use coughed up the goods but I had to enter the 1099 from my bank by hand. I was not able to get a 1099-INT directly into the program All of this data is also available in downloadable PDF format as well. I opened up the PDF for each account in it’s own browser tab and put that in the left monitor while I entered and massaged data in the right.

The new tax law knocks the hell out of deductions. You can wheedle the deductions all you want but unless your charitable is truly huge and you spent the year generating cancer treatment kinds of medical bills, it’s going to be hard as a retiree to beat the standard deduction. My plan worked perfectly. I underestimated my conversion a little because I wasn’t exactly sure of my revenue and cap gains distributions on my taxable accounts by my assumptions were almost exactly on point. I wrote off the cap gain with some Tax Loss Harvest from a 2018 harvest plus from some previous years. The TLH should hold through all of Roth conversion, keeping the tax bill low till my ordinary income returns to 12% bracket levels when I RMD and take SS. I did overpay my estimated but since I Roth converted late last year and I will be Roth converting in the next week or so for 2019 I decided to shmush the 2 years together applying the excess from 2018 to 2019 to try and avoid any penalties. I was able to dodge the 3.8% surtax by making a smaller conversion and through the pro-rata breaks.

I consider this a dry run since the whole 1099-R thing is new to me, but my tax life has simplified considerably with the incorporation of online data importing, well designed software with good longevity built in, and the whole code being streamlined. The bracket breaks don’t hurt either. I converted well into the 24% bracket but paid under 16% average. I’ll take it! My taxes were about 2 thousand below what I anticipated, nice surprise. some of that will go away next year as a child tax credit for my oldest will evaporate but by March I will have over half of my Roth conversion completed and safely tucked away growing tax free forever. The software double checked and triple checked and forced me to fill in a little missing data like Vanguard’s zip code on the my wife’s 1099-R. About 10 of those typo kind of mistakes to correct plus an analysis of low probability of audit and I was good to go.

I was all happy as I submitted online owing zero dollars and in a couple hours got the notification I was REJECTED! Not my problem but a goofed up form on HR Block’s end and a fix due out by the 22nd. It’s all right, I’ve moved on! Now I’m scheming on what to transfer for 2019. Usually taxes for me was a three day immersion experience, this year more like 3 hours . Now that I know what I’m doing with a pretty much guaranteed standard deduction next year should be more like an hour. The year after that my wife will claim SS and I will claim spousal benefits so I’ll get to learn about that and a few years after that I’ll pull the trigger on my SS and start RMD.