On page 37 of this month’s Money Magazine, Penelope Wang makes a strong suggestion that everyone should stop passing up on Roth 401Ks because they are better than traditional 401Ks. This was shocking news to me. I figured that if this is true, it should be “Front Page” material, not a buried in a one-page article with an extremely vague title: “Stop Passing Up This Great Deal.”
When I was writing this article originally, it wasn’t available online. Thus my research was limited to a spirited debate of it in the Boglehead forums. However, in the spirit of doing a little true journalism and receiving a lot of serendipity, I delayed the article a couple weeks. The delay means that a version of this article is now online (not the exact article, but very much the gist of it), sporting a more descriptive title of: “The Great Retirement Account You’re Not Using.”
Before I get into the article itself, I want to kick things off with my view of 401(k)s and Roth 401ks and explain why I believe one would choose one vs. the other. Spoiler: The biggest difference between the two is when you pay your taxes. The traditional 401(k) uses pretax money and is taxed when withdrawn. The Roth 401k uses post-tax money, but is not taxed when withdrawn.
Let me give you an example. Let’s say that I make $100,000 and I want to save 10% of that ($10,000) in some kind of 401(k) vehicle. Let’s say that I’m in a 25% tax bracket (I’m trying to keep the numbers simple.) If I pick the Roth 401(k), that $10,000 gets taxed, leaving me $7,500 in the account. If it grows 10% a year for 20 years, I’ll have $50,456 to withdraw tax-free. If I pick the traditional 401(k), that $10,000 goes in the account. If it grows 10% a year for 20 years, I’ll have $67,275 which will be taxed at withdrawal. After paying the same 25% in taxes, I have… $50,456 in spendable money.
So in this example, it doesn’t matter which one you do. They come out the same. The reason why someone would choose one over the other is the tax rate. If you are in low tax bracket and expect to be a higher one in the future, you want to pay those taxes now and get them out of the way. If you are in a high tax bracket now and expect to be in a lower on in the future, you want to pay those later at the lower tax rate.
Imagine my confusion when reading an article that seems to claim that the Roth 401k is almost always better. Penelope Wang says:
“Every dollar you save in a Roth 401(k) is worth more than a dollar you put in a pretax account. That’s because you’ll eventually pay income taxes on those pretax dollars while you get to keep every penny in a Roth. Granted, you get an upfront tax break by saving in a traditional 401(k), and you can invest that savings.”
Essentially she’s comparing a post-tax Roth 401(k) dollar with a pretax 401(k) dollar. It simply doesn’t make sense to do that kind of comparison. If you look at a Roth dollar without understanding that you’ve already paid tax on it, it is going to appear that you escaped the tax system altogether, which is simply untrue. If you downplay the upfront tax break by the traditional 401(k), you are missing the whole benefit of that 401(k).
In the aforementioned Bogleheads discussion on the topic, readers have found what appears to be the T. Rowe Price’s research.
Much of the Money Magazine article seems to say that the Roth 401(k) is better than the traditional 401(k) because of that T. Rowe Price research. I find that a less than satisfying explanation. I’d rather read exactly what the benefits are in terms of the applicable tax law.
It was at this point where the serendipity kicked in. I got an email about with all the people going to Fincon this year and spotted T. Rowe Price‘s Stuart Ritter, the person credited for the research. After some phone tag, we connected and I got a much more satisfying explanation. I hate to create a cliff-hanger (just kidding, I love it!), but that explanation is worthy of its own article, so tune in tomorrow.
I want to leave by saying that I’ve probably been too harsh on Ms. Wang. Often, editors create the titles in magazines, so the lack of descriptive was not likely her doing. In addition, writing words online is cheap and easy. I don’t have specific space requirements like print magazine writers do. I can afford to give a much more detailed explanation in an entirely separate article. I don’t think anyone could have done the topic justice in the space provided. It’s a shame it wasn’t a two or three page feature.
“Additionally, a 25% tax is subtracted annually from the taxable account during the years leading up to retirement and then was taxed at the same rate as their income during retirement.”
So they’re assuming you put that money in something that generates ordinary income? That seems like an odd assumption.
This bleeds into tomorrow’s article and is probably more relevant for that one. Nonetheless, I’ll bite… what would be a better assumption with that money? One could assume that they spend it on bacon or an iPad. One could assume that they bought and held for long-term capital gains. I vote running the numbers with the later, but I can see taking the middle ground between the two.
I was also confused by Ms. Wang’s article (my Money subscription ran out a few months ago so I didn’t see it until it hit the interwebs). I reached the same conclusion as Kosmo. While it’s true that they had to make some assumption about what was done with the tax savings, the assumption that is made is responsible for 100% of the difference they found between Roth and tax-deductible. Furthermore, the research used a $1000 contribution, meaning that in their own simulation, they could have just put the $250 into the tax-deductible 401(k) (thus causing the difference to disappear).
You’re immediately handcuffing the standard 401(k) option by making s worse choice with a chunk of the money than you are with any of the Roth money. Why not make the comparison based on the assumption that you make what is more likely to be the optimal strategy?
Assuming that you have no immediate need for the money, the only reason to make a short term investment with the money is if you can get a better yield than long term … and based on the other assumptions, that’s not the case.
Wouldn’t it be more realistic to invest the money in the exact same mix of securities as the 401(k) money (but without the tax advantage, of course)?
Yeah, they’re making this overly complex. Good point about the low contribution amount, Steve
In the real world, I have disposable income of 1333.33. I can make one of two choices with that money.
1) Invest the entire $1333.33 in a traditional
2) Pay the taxes on the money and put the remaining $1000 into the Roth
That’s it. Period. Why would you introduce any more complexity?
If you’re at the contribution limits, then you may have to put a portion into a taxable account, but if you think the 401(k) has the best investment vehicles, you’d still want to invest in comparable investment vehicles with the taxable money if you want a meaningful comparison. Otherwise, you’re comparing the tax difference AND the investment differences.
Here’s a hint at what changed my thinking on this a bit… again with more to follow tomorrow. In the real world, people don’t calculate a disposable income to come up with the $1333.33 number. They say, “Put 10% of my salary in there.”
“Here’s a hint at what changed my thinking on this a bit… again with more to follow tomorrow. In the real world, people don’t calculate a disposable income to come up with the $1333.33 number. They say, “Put 10% of my salary”
I used my hacksaw to pound in same nails today. Stupid thing did a horrible job.
Seems silly to blame the tool, doesn’t it?
What you’re describing isn’t worse performance by the traditional 401(k). If you use equivalent inputs, you get the same results with Roth or traditional.
The problem is that the tool is being used incorrectly. Instead of blaming the hacksaw, we need to learn to use the hacksaw for cutting and the hammer for pounding nails.
In other words, we need to teach people that if they are in the 25% marginal bracket, then a $1333.33 contribution to a standard 401(k) will cost them the same out of pocket as a $1000 Roth contribution. If we don’t do that and instead use bad information to tell them that the Roth is better in the vast majority of cases, then we are doing them a disservice. We are costing them money by having them invest in Roth when traditional would be better, because we’re giving this weird 1000 traditional + 250 taxable ordinary-income account option as the alternative.
I’d think that Money Magazine would want to promote financial literacy.
Here, I’ll help.
Roth / (1-tax_rate) = Traditional
1000 / (1 – tax_rate) = Traditional
1000 / .75 = 1333.33
Surely people are capable of this calculation, right? Teach a man to fish …
I mean, it’s not like we’re asking people do do calculus on a daily basis. We’re asking them to do ONE simply math problem perhaps once a year. I’d hope that most people could buckle down and tough their way through the problem.
I put the maximum in the regular 401K, where I get a deduction today. This saves to me an amount of taxes which is almost ( very close to) equal to the limit I can contribute to the ROTH IRA.So 17500 in tax deductible 401K, and a 30% tax, equals $5,250 that is then put in the Roth IRA.
I plan to retire early, and move a portion of my 401K into a Roth, by trying to avoid triggering any income taxes, When you have more money, you have more options on what to do, even if it initially looks slightly disadvantageous to a ROTH 401K.
If I had done the Roth 401K, I would only have $17,500, and no money for Roth IRA. But, I do think about my taxes disturbingly often