All posts by Tom Canfield

Why are you taxed at 18.5% in the 10% marginal tax bracket?

When you are retired and receiving Social Security, the 10% and 12% marginal tax brackets on ordinary income are a myth. Most all of your distributions from your traditional IRAs – such as your RMD – are taxed at 18.5% or 22.2%. Very little is taxed at 10% or 12%. The culprit is the calculation that increases the amount of Social Security that is taxed. For most all of the 10% and 12% brackets, each added $1,000 of ordinary income increases the amount of taxable Social Security income by $850. Each $1,000 from your traditional IRA results in $1,850 of taxable income. That’s 85% more tax than you’ve probably assumed.

 

 

This effect holds true until you’ve added enough ordinary income to reach the maximum of 85% of Social Security that is taxed. The way this exactly works will be different for you than it is for Patti and me, but I think our graph is representative of how it would look for you: for us, the first $52,000 of distributions from our traditional IRAs is taxed at an average of 20.7%. Distributions more than that are then taxed at 22% until the 24% marginal rate and/or an IRMAA tripwire; those points are at much greater distributions than shown on the graph. (You can see the worksheet that I used to calculate taxes here.)

 

None of our distributions from our IRAs is taxed at 10% and just $3,000 of the first $52,000 of distributions from our IRAs is taxed at 12%. We pay an effective rate of 20.7% on the first $52,000 of distributions.

 

Details:

 

Social Security is taxed from 0% to a maximum of 85%. A somewhat complex formula calculates the percentage that is taxed. Here’s my spreadsheet. You can input the numbers from your tax return for a normal year and calculate the percentage of Social Security that is taxed for each increment of distributions from your traditional IRA.

 

== Does this matter? ==

 

• Current retirees can’t do anything about this. Traditional swamps Roth for most all of us: Roth didn’t exist until 1997 and employers didn’t start offering it as part of their 401k plans until 2005. It turns out, though, that traditional IRAs were the right choice. Marginal tax rates we avoided were perhaps 16 to 23 percentage points greater in some years than the 22% tax that we pay today.

 

 

 

An IRA gives a basic ~20% advantage relative to the same investment in a taxable account. The math of avoiding 38% tax then and paying 22% now means you pick up an added 26%. You just aren’t getting another 15% to 20% benefit on top of that if your distributions were really taxed at 10% or 12%.

 

 

• Younger folks in the save and invest phase can benefit from this information: emphasize Roth. You really don’t make a mistake by picking Roth if you are in the 22% marginal tax bracket, and I think I’d argue that Roth is just fine if you are in the 24% bracket.

 

If you assumed ALL your distributions for desired spending were from after-tax Roth, a 1040 tax return would be simple, and you’d be in a very in a low tax bracket. For Patti and me our 1040 with no taxable distributions from IRAs gives this result:

 

– 32% percent of our Social Security is taxed.

 

– That taxable amount + all other ordinary income (primarily money market dividends for us) keeps us in the 10% marginal tax bracket for ordinary income.

 

– We’d pay NO TAX on dividends qualified for capital gains tax and NO TAX on our normal amount of capital gains from our sales of securities for our spending. We don’t get close to the threshold of taxable income that triggers the 15% capital gains taxes.

 

 

Conclusion. Most folks think that they’ll be in a lower tax bracket when they are retired. That’s basically a myth. I found that Patti and I effectively pay ~21% tax on the first $53,000 of distributions from our traditional IRAs and then 22% thereafter (until the distant 24% marginal bracket).

 

The culprit is the way the taxable portion of your Social Security benefit is taxed: each $1,000 of income from distributions from our traditional IRAs triggers an added $850 of taxable Social Security income. In the 10% bracket, we are really paying 18.5% on each added $1,000; for most all the 12% bracket, we are paying 22.2% on each added $1,000.

 

Traditional (pre -tax) IRAs were the right choice for current retirees. Marginal tax brackets were much higher in the past: in some years, retirees contributed to their traditional IRAs and avoided paying perhaps 38% marginal tax then. Paying 22% now means they picked up another 26% on top of the basic 20% advantage of a retirement account.

 

Younger folks in the save and invest phase of life who are in the 22% bracket really can’t go wrong with 100% going to Roth. A traditional IRA is ~never better than Roth. Retired folks who could use Roth IRAs solely for their spending will likely be in the 10% tax bracket. They may not even pay any tax on dividends and gains that qualify for capital gains tax rates.

Inflation was a hot topic in the news this week: pointing to +4% annual rate.

Inflation was a hot topic the last two days. The last three months are running at an annual rate at more than 4% inflation. This post shows six graphs I use to follow the inflation trends. We get the last report for inflation for March at the end of this month: Personal Consumption Expenditures (PCE) – the one that the Fed favors for decision-making on interest rates. The one good graph is the fifth graph on Producer’s Price Index: the increase for March was slight.

 

Going deeper: below I display a table and the same six graphs that I’ve use to follow the trends in inflation. This table has PCE through February, not March.

 

PCE is for month-ending February. Inflation for March is published at the end of April.

 

== Inflation trends ==

 

The two most widely-reported measures of inflation are Seasonally-adjusted inflation and Core inflation. These and most all other measures of inflation are reported at about the two-week point in the following month.

 

Seasonally-adjusted inflation increased by 0.38% in March. The rate over the last six months aims at an annual rate of 3.1%.

 

 

Core inflation excludes volatile energy and food components. This is similar to the measure favored by the Federal Reserve. Inflation increased by .36% in March. The last six months aim at an annual rate of 3.9%. The last three aim at 4.5% inflation.

 

 

Personal Consumption Expenditures (PCE) excluding Food and Energy (February) is the measure of inflation that the Federal Reserve Board favors. The last six months aim at an annual rate of 2.9% and the last two aim at 4.3% inflation.

 

Data through February. Data for March is published at the end of this month.

 

== History of 12-month inflation ==

 

Full-year inflation measured by CPI-U shows that inflation for the last 12 months has been 3.5%.

 

 

== Producer’s Price Index ==

 

The change in producer prices will impact consumer inflation. PPI over the last six months is at a -2.9%. annual rate. The increase for March was slight and mu lower than the 1.47% increase in February.

 

 

== Services ==

 

The last six months aim at an annual rate of 5.7%.

 

 

 

Conclusion: Inflation for the last three months is running at 4.5% or greater.

Is it best to hold international stocks in a taxable or tax deferred account?

This was an interesting article at Morningstar: Should You Keep Foreign Stocks Out of Your IRA? Tax considerations suggest yes…. There is an added tax wrinkle with international stocks or funds, but I conclude the opposite: the tax considerations CLEARLY say it’s preferable to hold your international funds in a Roth or Traditional IRA account. You earn about 6.0% after-tax return when international is held in your taxable account, and you earn about 6.8% after-tax return when it’s held in your IRA.

 

• Details of two tax wrinkles:

 

== Foreign taxes paid ==

 

In your taxable account, you pay a 15% tax on all dividends you receive from your taxable stock funds (FSKAX and VXUS in our case). Some of the taxes that you pay are to foreign countries. So that you are not double-taxed, you get a direct tax credit (deduction) on US taxes for “foreign taxes paid” shown in box 7 of your 1099-DIV.

 

You don’t pay tax on dividends for securities held in your IRA, but YOU DO NOT GET BACK the taxes you’ve paid foreign countries. The total return (price + the effect of dividends reinvested) is LESS in an IRA than a taxable account because of the lower net dividends you have to reinvest. I calculate the effect is about 0.3% less in total return per year.

 

The 8% tax rate on foreign dividends is about right for most international stock funds.

 

== High dividend rate on International ==

 

The dividend yield on international stocks is about 3.1%. You are paying taxes each year on dividends in your taxable account. You are compounding a slightly smaller top-line amount each year than you would if you paid no taxes on the dividend. The lower top-line growth is not offset by your higher cost basis. You net less over time, and you net less from a greater dividend rate.

 

It requires a spreadsheet to figure out the effect of paying taxes on dividends: for 7.1% real return and 1.3% dividends year, an investment held 25 years in a retirement account results in about 16% more than an investment in a taxable account. I show here the first few years of my spreadsheet that tracks the difference in return for an investment in a taxable account and a retirement account.

 

== I compare the two ==

 

I use my spreadsheet to compare the after-tax return from international fund held in a Taxable vs. IRA account, but I don’t need a spreadsheet to reach the conclusion that an investment in an IRA is better than one held in a taxable account.

 

• Taxable effectively knocks off 15% of the growth of an investment from taxes on dividends and on the ultimate gain when sold. The 15% isn’t a direct reduction in the return rate, but it’s simplest to think that you are lowering a 7.1% pre-tax return rate about 6.0% after-tax return rate.

 

• The investment in an international fund has about 0.3% knocked off a 7.1% annual return rate because of the “lost” foreign taxes: net of about 6.8% after tax return rate.

 

My spreadsheet shows a dollar advantage to holding your investment in a retirement account in all years.  The relative dollar  benefit increases with time.

 

 

 

Conclusion: Some suggest you should hold an investment in an international fund, like VXUS, in a taxable account and not in a retirement account. The argument is that your net return in a Retirement Account will be less because you lose foreign taxes paid.

 

The logic is straightforward, though. It clearly makes sense to hold your international stocks in your retirment account.

 

• Your after-tax return rate on a taxable investment is less because of the 15% gains tax. A 7.1% pre-tax return rate is about 6.0% after-tax rate.

 

•  Your after-tax return rate on international stocks in your IRA is about 0.3% less from the lost foreign taxes paid. It’s about a 6.8% after-tax return rate.

 

The dollar benefit is small in early years but grows over time.

Should we lose faith in bonds as valuable insurance to protect our portfolio?

I view the bonds we retirees hold as a form of insurance – downside protection. We need to sell from our portfolio each year for our spending. We don’t want to sell stocks when they’ve taken a deep dive. Selling stocks then for our spending just magnifies the decline. We want to give stocks time to recover and repair our portfolio. We want to sell something else that has performed much better. Most everyone agrees that the “something else” is bonds.

 

2022 could challenge our faith in bonds as valuable insurance: the real return for stocks cratered and bonds were worse – the worst in history.

 

This post shows that 2022 was the outlier. Stocks have declined in 30 of the last 98 years. Bonds have outperformed stocks in 27 of those 30 years and by an average of 15 percentage points. They haven’t been perfect, but one has to conclude that bonds give us excellent downside protection when stocks crater. Don’t change the mental image of bonds as insurance: we are wise to hold bonds that we can sell for our spending when stocks crater.

 

 

DETAIL:

 

== Stocks generally swamp bonds ==

 

When stocks are good, they are very good relative to bonds. Real stock returns have been positive in 68 of the last 98 years – about seven of ten years. In those years, stocks returned more than bonds in 62 of the 68 – better than nine of ten years – and outperformed by an average of 16 percentage points per year: stocks averaged +19% real return per year and bonds averaged +3% in the 68 years.

 

Stocks greater than 0% real return in 68 of the 98 years 1926 through 2023. Yes, stock returns were +50% annual return two times: 1933 and 1954.

 

== Bonds swamp stocks when stocks decline ==

 

When stocks are bad, bonds are very good relative to stocks. Real stock returns have been below 0% in about three of ten years – 30 of the past 98 years. In those years, bonds returned more than stocks in nine of ten years – 27 of the 30 – and outperformed by an average of 15 percentage points: bonds averaged 1% real return and stocks averaged -14% real return in the 30 years.

 

Stocks less than 0% real return in 30 of the the 98 years 1926 through 2023.

 

The outlier for poor performance for bonds relative to stocks was 2022; the return for bonds was the worst in history; 9 percentage points worse than the second-worst year in history (1946).

 

== Bonds even better when stocks worst ==

 

Bonds have been even better relative to stocks when stocks were their worst. Bond returns have been better in 19 of the 20 worst years for stocks – years when the real return for stocks was roughly -9% or worse. Bonds outperformed by 20 percentage on average.

 

2022 is the bad outlier: the only time bonds have preformed worse than stocks when they cratered. 2008 is the good outlier: bond returns were 55 percentage points better than stocks.

 

 

Conclusion. The very poor performance for bonds relative to stocks in 2022 could shake our mental image of stocks as valuable insurance that we use to protect our portfolio. When we look at the 30 years that stocks have declined over the past 98, the advantage for us retirees to hold bonds is clear, even including 2022. Bonds return much more than stocks – on average 15 percentage points more – when stocks crater. We are wise when we disproportionately sell or totally sell bonds and limit the damage to our portfolio when stocks crater. We buy time for stocks to recover and repair our portfolio.

How do the last two years compare to the worst sequences of return in history?

I’ve been encouraged about returns to date for this year and the outlook for our economy, but 2022 was a really bad year. Stocks and bonds combined for the roughly the fifth worst portfolio return in history. We rebounded in 2023 and are doing well in 2024 so far. I’m encouraged that we may not be riding a sequence of return similar to the most harmful in history.

 

This post gives a snapshot of where we stood at the one-year and two-year marks compared to the two most harmful return sequences in history. At the two-year mark, our 2022+2023 portfolio was better off compared to those two periods. And we’re clearly in positive territory so far in 2024.

 

== 2022: really bad ==

 

The combination of the sixth-worst stock returns and the worst bond returns in history combined for a portfolio return that was not better than the fifth-worst portfolio return in history. (For someone with a portfolio mix of 60% stocks and 40% bonds – which I hate – 2022 was the worst portfolio return in history.)

 

2022 was VERY unusual. It’s rare that bonds don’t outperform stocks when they decline. Bonds have outperformed 27 of the 30 times stocks have declined. And 2022 was the only time that bonds were worse than stocks when stocks declined by more than 10% real return.

 

Stocks have declined in 30 of the last 98 years, and bonds have been better in 27 of those years. 2022 was the only year when bonds had worse return than stocks when stocks cratered.

 

== Most harmful sequences ==

 

We’ve had three horrible start-points for a retirement portfolio in the last 98 years: withdrawals combined with horrible returns means a portfolio depletes. The three most harmful sequences for a portfolio started in 1929, 1969 and 2000.

 

Three most harmful periods for a retirement portfolio: sequences that started in 1929, 1969 and 2000. 1969 was the worst: steep deline for stocks and longest period of 0% cumulative return for both stocks and bonds.

 

The sequence that started in 1969 was the most harmful for retirees withdrawing each year from their portfolio for their spending. A poor four-year start was followed by the second worst two-year return in history. It’s the worst six-year return period by a long shot. The 1969 sequence is the one with the potential to deplete a portfolio to the point that it no long supports a full withdrawal for spending. It’s the return sequence intersects zero the quickest in the displays at FIRECalc.

 

I built a spreadsheet that shows the detail for one example: a portfolio lasts 20 years if it rode the 1969 sequence. Assuming that sequence is the worst we could expect in the future, 4.5% is a safe withdrawal rate for essentially zero chance of depletion in 20 years.

 

The second-most harmful sequence started in 2000. You can see a .pdf of a spreadsheet with the same inputs.

 

== Compare ==

 

I compare 2022 with the start of those two: 1969 and 2000. The portfolio return rate at 85% Stock and 15% bonds (my portfolio mix) was TWICE AS BAD of a start. That conclusion holds for all reasonable mixes of stocks vs. bonds. Has to: bonds were worse than stocks.

 

 

2023 was a strong rebound, and at the end of two years, the 2022+2023 result was better than the two by 4% to 10%. That’s not staggeringly better, but it is better.

 

 

We are not out of the woods, but the point at the end of 2023 and the trend so far in 2024 encourages me to think that the 2022 sequence won’t turn out to be one of the most harmful in history.

 

 

Conclusion: 2002 stock and bond returns resulted in about the fifth worst portfolio return in the last 98 years. Stock returns were the sixth worst and bond returns were the worst.

 

The return rate for a portfolio was more than TWICE AS BAD as the first year of the two most harmful return sequences for a retirement portfolio in history. Alarm bells went off, or should have gone off.

 

2023 was a strong rebound and was far better than the second-year return for the most harmful sequences of return in history. Our 2022+2023 portfolio value is GREATER than if we would have been riding the two most harmful sequences in history. Not by a whole lot, though. And we’ve continued to improve so far in 2024. I’m encouraged and think that that the horrible 2022 may not be the start of return sequence similar to the most harmful in history.

We’re really within 3% of the prior peak for stocks in November 2021.

We’re within 3% of the inflation-adjusted peak for stocks in November 2021. This is a sharp change from my post at the two-year mark, when I concluded we had to see a real increase of 25% to get back to that peak. Wow. I’m encouraged that we may put that -23% real decline in 2023 behind us. That was the sixth worst calendar-year decline since 1926 and the third worst in my lifetime. If we put that peak from 2 1/3 years ago behind us, this will be one of the fastest recoveries for stocks from a year of steep decline.

 

== Details of the 4% ==

 

I use S&P Global’s index for Total US Stocks. I use the index for Total Return. I use the peak on November 8, 2021. The index, without adjusting for inflation is now more than 9% above that. But we’ve had ~12% inflation. When I adjust, we are less than 3% below the peak, and we have to have a real increase of less than 3% to surpass it.

 

 

== Implications for our SSA ==

 

We might calculate to a real increase in our Safe Spending Amount when we recalculate at the end of this year (SSA, Chapter 2, Nest Egg Care). I previously thought this would be IMPOSSIBLE. Real stock returns were more than +8% in November, and US stocks are up another than 13% from December.

 

I calculate our real portfolio return since December 1 at 8.7%. Patti and I need a ~19% real portfolio return for our calculation year (December 1 – November 30) to calculate to a real increase in our SSA next Nov 30 or Dec 1. (You can follow my calculation to find the ~19% that we need here.) +10 percentage points more from where we are now is not impossible.

 

Your SSR%s are different than ours and your mix of stocks and bonds may be different from ours. But my conclusion is that you also will be close to calculating to a real increase in your SSA at the end of this year if stock returns are +10% over the next nine months.

 

 

Conclusion. We’ve had a steep climb for stocks since early November. The outlook for our portfolio is very different than my look then. We are now just 3% in real return from surpassing the prior market peak in November 2021. If we can pass it and put it behind us, this will be one of the fastest recoveries from a steep annual decline in stocks.

 

The performance over the last four months gives me some hope of calculating to a real increase in our Safe Spending Amount the next time I calculate on December 1. Patti and I need less than 10% real portfolio return for the balance of this year to calculate to a real increase. 10% will put you in the ball park of calculating to a real increase in your SSA.

The latest inflation release confirms the high rate in January

Two months ago, it looked like we were clearly headed to 2% inflation. November inflation, using the measure favored by the Federal Reserve was the lowest in four years. The inflation rate for January was issued yesterday and confirmed a reading earlier in the month. Inflation of 0.4% was four times that of November and the highest in the last 12 months. That one month is an annual rate of 4.8%, but the measure of inflation the Federal Reserve favors tracks to 2.3% annual rate over the past eight months.

 

Going deeper: below I display a table and the same six graphs that I’ve use to follow the trends in inflation.

 

 

====

 

The two most widely-reported measures of inflation are Seasonally-adjusted inflation and Core inflation. These and most all other measures of inflation are reported at about the two-week point in the following month.

 

Seasonally-adjusted inflation increased by 0.38% in January. The rate over the last six months aims at an annual rate of 3.5%.

 

 

Core inflation excludes volatile energy and food components. This is similar to the measure favored by the Federal Reserve. Inflation increased by .39% in January. The last six months aim at an annual rate of 3.7% and the last eight aim at 3.2% inflation.

 

 

Personal Consumption Expenditures (PCE) excluding Food and Energy is the measure of inflation that the Federal Reserve Board favors. Inflation of 0.42% in January was less than the prior January but is greater than any of the last 12 months. The last six months aim at an annual rate of 2.5% and the last eight aim at 2.3% inflation.

 

 

== History of 12-month inflation ==

 

Full-year inflation measured by CPI-U shows that inflation for the last 12 months has been 3.9%.

 

 

== Producer’s Price Index ==

 

The change in producer prices will impact consumer inflation. PPI over the last six months is at a -2.5%. annual rate.

 

== Services ==

The last six months aim at an annual rate of 5.7%.

 

 

 

Conclusion: Monthly inflation in January was 0.4%. The rate for the Fed’s favored measure was the highest in the last 12 months. That’s obviously the wrong direction to the goal of 2% inflation, but the last eight months aim at a 2.3% annual rate.

How much more do you keep from investments in your retirement accounts?

I calculate that you gain at least 20% on the money you invest in retirement accounts relative to the same investment in a taxable investment account. I used a 25-year holding-period for your investment: if you are in the Save and Invest phase of life, that’s how long I assume you will hold your contribution this month to your retirement plan before you sell it for your spending. When I look back in time, that seems about right: Patti and I held the securities we sell for our spending for at least 25 years. Two components contribute to the +20% more: tax-free growth and avoiding tax on inflation. And you need time: those two provide a small benefit in ten years. It takes 25 years to reach that 20% benefit.

 

== Traditional = Roth ==

 

We want to compare the results from an investment in a taxable account to an investment in an IRA. To best understand, we need to start with the fact that a traditional (pre-tax) IRA gives the same after-tax return as a Roth (after tax) IRA; this assumes the marginal tax bracket at the time of contribution is the same as at time of sale of securities and withdrawal from the retirement account. In this example, I assume the marginal tax bracket is 22% at the time of contribution and at the time of withdrawal.

 

 

== Example: 14% more from tax-free growth ==

 

Once we understand that traditional = Roth, it’s a lot easier to understand when we compare an investment in a taxable account with the same investment in a Roth account. You already paid tax on the gross amount to invest in your taxable investment account or to invest within a Roth IRA.

 

• If you keep $1,000 in your taxable investment account, you will pay 15% capital gains tax on annual dividends and on the final gain at the time of final sale. (You may pay tax on capital gains distributions for some mutual funds, but let’s ignore that.)

 

• If you keep $1,000 in your Roth, you never pay these taxes: it’s obvious you will ALWAYS have more from Roth than from your taxable accounts. (The obvious caveat is that you do not violate the early withdrawal penalties from your retirement accounts.) You are not paying 15% tax on the growth over time. You are keeping 17.6% more of the growth for any holding-period.

 

 

The impact on the total that you keep depends on the amount growth vs. the initial investment. For the first few years, the total growth is small relative to the initial investment and the benefit of a retirement account compared to a taxable account is small. In the simple example below for 25 years, the growth portion is 5X that of the initial investment.

 

 

The simple example assumes all the gains are from price, not the combination of dividends reinvested and price. You only pay the capital gain tax at the end of 25 years. You have ~14% more from your Roth than you would from your taxable account.

 

== Perhaps 10 percentage points more ==

 

Three other factors make the benefit of a retirement account greater than 14% – perhaps 10 percentage points more, but I’ll just safely settle that an IRA results in 20% more than a taxable account.

 

 

• +3 percentage points: The benefit is really ~three percentage points greater when one calculates the impact of paying taxes on dividends on a taxable investment. A detailed calculation (It’s a big spreadsheet.) shows that you lose an increment of growth when you pay taxes earlier than in the simple example. This is a small effect for 10 years, but for 25 years the effect cumulates to 6% lower total growth and 3% lower net return.

 

• +2 percentage points: The benefit is two percentage points more because an IRA escapes the capital gains tax on inflation. The cost basis of an investment is not adjusted for inflation. The gain you calculate includes a portion that is just inflation. I calculate that the real capital gains tax rate is about  17% for the example of 25 years and 6X real growth.

 

• +4 percentage points: The benefit could be four percentage greater when one includes state taxes on capital gains. State taxes on gains vary. Nine states do not tax dividends or capital gains. Other states tax gains at 2.5% to over 10%. My state, PA, is on the low end and taxes dividends and capital gains at 3.1% and does not tax withdrawals from traditional or Roth IRAs. That would make the total capital gains tax in the simple case 18.1%. That raises the 14% benefit of an IRA to nearly 18%.

 

 

Conclusions: You benefit by about 20% greater after-tax return when you invest in a retirement account – either a traditional IRA or a Roth IRA – as compared to an investment in a taxable account. Obviously, this is for money you are investing for retirement, and I think 25 years is a proper holding-period for comparison.

 

Traditional (“Pre-tax”) IRA and Roth (“After-tax”) IRAs give you the same, basic after-tax return over time. You are gaining about 14% more in after-tax return primarily from tax-free growth of gains. You get a boost to 20% or more from escaping the tax on inflation inherent in the way gains taxes are calculated and from escaping state taxes on gains (This is varies according to your state).

Does it make sense to convert your IRA to Roth in Retirement?

I subscribe to a stock market newsletter. This month it had an article with the same title as this post, Does it make sense to convert your IRA to Roth in Retirement? I found the article confusing and incorrect. The decision on Roth conversion is pretty simple. I’ve written on this before (here and here), but I think this post gives a shorter, clearer explanation: it makes most sense to me to convert if you avoid paying taxes in a significantly higher marginal tax bracket in the future or if you can avoid a Medicare premium surcharge; you avoid paying taxes you do not need to pay; will have more after taxes than you otherwise would. This applies to few retirees.

 

If you don’t see that converting to Roth now avoid higher taxes in the future, you have no reason to convert. You get the same after tax return from a Roth IRA and a Traditional IRA. And it is painful now to convert: you are paying more in taxes; you’ll have less than you would otherwise to spend to enjoy now. You have to see real benefits to convert.

 

== It NEVER really hurts ==

 

It NEVER hurts to convert from traditional IRA to Roth. You could win with a conversion, but you can ever lose. I can’t construct a scenario where a retiree winds up with less after taxes from a conversion. (I’m not counting doing something illogical, like converting a large amount of traditional at a high marginal tax rate only to be in a lower marginal tax bracket later.)

 

== Details: three cases ==

 

It makes most sense to convert some traditional IRA to Roth in three instances for a small set retirees.

 

1) Convert if you are now in the 12% marginal tax bracket and think you will pay the 22% marginal tax bracket in the future. This might apply to a small set of retirees and near retirees, and it takes a lot of planning to benefit: folks in this situation may not have a lot of headroom to keep the conversion amount in the 12% marginal bracket: the top of the 12% tax bracket for a single taxpayer is about $63,000 gross ordinary income for 2024: double that for joint, married filers. If they’re in the 12% bracket, they can’t convert large amounts in any year and stay in the 12% bracket.

 

2) A subset of retires with high income – some of the top 10% of all retirees – might want to convert to have flexibility to avoid Medicare Premium surcharges. They can convert and lower future RMD; they an withdraw from their Roth for their spending in the future and not increase taxable income that can trigger a Medicare Premium surcharge.

 

3) Married, joint filers with high income might want to convert even at the 24% marginal tax bracket to avoid the 32% marginal tax bracket that the surviving spouse would pay as a single filer.

 

== Biggest benefit: about 12% more ==

 

The biggest benefit is that you would have about 10% more in the future from a conversion to Roth. The best case is that you are in the 12% tax bracket now, but think you will be consistently be in the 22% tax bracket in the future – perhaps after your RMD starts.

 

It makes sense to convert the amount from traditional to Roth that still keeps you in the 12% marginal bracket. You pay the 12% tax now and not the 22% tax later. You will keep about 13% more after taxes on the amount you converted. (If your heirs are in the 22% tax bracket when they withdraw from Roth, they’re getting this 13% benefit.)

 

 

== Other jumps in tax bracket ===

 

This same math applies to other jumps in tax brackets, but only one other jump gets close to the benefit of the 12% to 22% bracket. Folks in these brackets have very high income. You also have to think through: if you convert at 22% bracket and hold Roth at death, will your beneficiaries be in the 22% bracket when they withdraw?

 

 

== Two other cases that make sense ==

 

Those with high income should consider two cases where conversion makes sense: avoid paying taxes that you do not need to pay.

 

• Convert some to Roth to give yourself flexibility to avoid Medicare premium surcharges. (I consider the surcharges as added taxes.) If you are near a tripwire in income that triggers a surcharge, you’d use Roth for your spending and not added withdrawals from traditional IRAs or from security sales in your taxable account .

 

• Married, joint filers with more than about $200,000 of ordinary income, should decide if they want to convert now to avoid the 32% marginal tax bracket when just one will be alive.

 

(You can read more here if you are at this level of income.)

 

 

Conclusion. Converting traditional to Roth NEVER hurts in retirement. I can’t construct a scenario where you would wind up with less after taxes if you convert. But it does hurt to pay taxes now rather than then. You only want to convert if you see real benefits.

 

Converting to Roth will give you more after taxes in the future in two cases:

 

1) if you can see that you (or the surviving spouse of the two of you) will be in a significantly higher marginal tax bracket in the future. (I’d say at least 8 percentage points higher.)

 

2) if you have high income subject to Medicare premium surcharges, Roth gives you flexibility in a future year to keep your income below a level that triggers a surcharge.

35 times my money

Every January track what happened to the $2,000 I contributed to my IRA many years ago. I imagine that I put my $2,000 contribution in a gift envelope and bought shares of a stock index fund. It sat there untouched for 35 years, accumulating shares from dividends reinvested; price per share also changed. Patti and I open the envelope each January and see how much is there: we know we can spend it all to ENJOY. This January, it had securities worth $69,800. I made ~35 times my money in dollars. That’s about 14 times my money when measured in real spending power. This post shows the details.

 

== Summary ==

 

I did four things right:

 

 • I saved every year at a much younger age;

 

 • I saved in retirement accounts with no taxes on the growth;

 

 • I invested solely in stocks;

 

 • I invested in each year and stuck with the same fund, equivalent to a very low-cost S&P 500 index fund.

 

Details:

 

I can use Morningstar to see the growth in value for any mutual fund over all the years its existed. Here is the history for the past ten “envelopes” from my $2,000 investments. I invested each in an S&P 500 index fund. My $69,800 this year is about 35 times my original $2,000.

 

 

That’s not really an apples-to-apples comparison. I have to adjust for inflation to find the multiple in real spending power. The inflation factor to apply is ~2.5. $2,000 then has the same spending power as $5,000 now. The multiple in real spending power for our latest envelope was ~14: $69,800/$5,000.

 

== How does this happen? ==

 

That real growth in spending power was from the compounding of the long-term real ~7.1% growth rate for stocks for many years. I like this semi-log graph that shows the cumulative real growth in value for stocks and bonds since 1926. If you start way back in 1926, $1 invested in stocks grew 900X and ~90X that for bonds.

 

 

At 7.1% real growth, I would expect stocks to ~double every ten years applying the Rule of 72: 35 years would be 3½ doublings. Three would be 8x (1 to 2; 2 to 4; 4 to 8) and the five years is half way to the next doubling, getting you to 12X. My multiples have been better than that: for example, a straight line from the start of 1989 to January 2024 is ever so slightly steeper than that 7.1% line.

 

 

Another way to see what happened is to plot the real return multiples from all 35-year sequences of return that started in 1871.  We have the inflation-adjusted monthly returns for stocks and bonds since then. (Data downloads from here.)

 

I plot the multiples of return for the rolling sequences of 35 years. The first sequence started January 1871 and ended December 1895. The second one started in February 1871 and ended January 1996. The last one ended in September 2023. (The data hasn’t been updated yet for the last months of 2023.) That’s a total of 1,441 complete 35-year sequences.

 

 

 

You can see from this that the return multiples of the most recent 35-year sequences are greater than a multiple of 12.

 

You also see the folly of holding bonds for a 35-year period. Bonds do not come CLOSE to stocks for any 35-year period.

 

 

Conclusion. Every January I look back to see what happened to the $2,000 that I contributed to my IRA. I started that in 1981 and contributed the maximum for decades. Each year I look back at the contribution that I made 35 years ago. I can see the growth of my investment – which never changed for 35 years – from the Morningstar site. My investment grew 35X in dollars. My multiple in real spending power was 14X. The multiples of my last ten contributions have been slightly better than one would expect from the long-run real return rate for stocks of ~7.1%.

 

The lesson is to save and invest when you are younger. Save in retirement accounts and benefit from no taxes on growth; invest solely in stocks; never vary from investing and holding a very low cost, broad-based stock index fund.