Asset location is a fairly advanced topic in the first place. Many investors don’t pay any attention to it at all. However, what surprises me is that so many of those who actually do pay attention to it get it wrong.
Let’s be clear, of course–in the grand scheme of things, asset location doesn’t matter all that much. It’s much less important than your savings rate. It pales in comparison to your income. Asset allocation matters far more. Getting your investment expenses down into the realm of reasonable is also going to make a bigger difference. Heck, I don’t even think good asset location can make enough of a difference to justify hiring an advisor. But it can boost returns a little bit and so if you’re one of those willing to put in a little effort to bank that benefit, at least make sure you’re doing it right.
Pet Peeve #1 – Only Paying Attention to the Tax-Efficiency of the Asset Class
I see this one all the time- in blog posts, in internet forums, out of the mouths of financial advisors, and even in the books of people I have a great deal of respect for. But it’s wrong, wrong, wrong. When deciding whether to locate an asset in a tax-protected account or in a taxable account, you have to pay attention to BOTH the tax-efficiency AND the expected return of the asset. The problem with paying attention only to the tax-efficiency is that you don’t account for the benefit of improving your ratio of tax-protected to taxable. A quick example will demonstrate what I mean.
Consider two asset classes. The first has an expected return of 2% and all 2% is paid out every year and is taxable at your full marginal tax rate. We’ll call that 40%. It is maximally tax-inefficient but has a low expected return. The second has an expected return of 10% but pays out nothing in any given year and when it is sold in a taxable account, the gains are paid at the capital gains rate, let’s call that 20%.
Let’s say your asset allocation calls for 50% in each asset, and you have two accounts, a tax-free account (such as a Roth IRA) and a taxable account, which are equal in size, $10K a piece. You’re going to invest for 10 years, then pull it all out, pay any taxes due, and spend it. To keep things simple, we won’t rebalance, although it could be argued that you should be tax-adjusting the two accounts to account for the effect of the slowly increasing tax liability of the taxable account.
Let’s look at your two options:
Option 1: Put tax-inefficient, low-return asset class into the tax-free account and the tax-efficient, high-return asset class into the taxable account
After ten years, your tax-free account has grown to
=FV(2%,10,0,-10000) = $12,190
And your taxable account has grown to
=FV(10%,10,0,-10000) = $25,937
Of course, you must apply the appropriate tax rate to the taxable account.
The total of the two account is $35,736. Now let’s compare to the second option.
Option 2: Put tax-efficient, high-return asset class into the tax-free account and the tax-inefficient, low-return asset class into the taxable account
And your taxable account has grown to
=FV(2%,10,0,-10000) = $12,190
Again, you must apply the appropriate tax rate to the taxable account. Since the returns are taxed at 40% each year (not at the end), and 40% of 2% is 0.8%, then we can just change there return figure in the Future Value function from 2% to 1.2%.
=FV(1.2%,10,0,-10000) = $11,267
The total of the two accounts is $37,204. By putting your most tax-efficient asset class into the tax-protected account, you have earned an additional, after-tax $1,468. That’s equivalent to an additional 0.42% per year in return.
=RATE(10,0,-20000,35736) = 5.98%
=RATE(10,0,-20000,37204) = 6.40%
6.40% – 5.98% = 0.42%
Clearly, in at least some situations, the least tax-efficient asset class SHOULD NOT be put preferentially into the tax-protected account. I announced this in a post with the admittedly click-baity title Bonds Go In Taxable! and have caught lots of crap about it ever since by people who can’t do math (including a Boglehead who told me he doesn’t read anything I write since that post.)
Now, that’s all good and fine. All of you who have completed sixth-grade math now agree with me that there are times when theoretically it can make sense to preferentially put a tax-inefficient asset class into a taxable account. But here in the real world, we have to deal with the hand we are dealt, rather than perfectly tax-efficient asset classes returning a perfect 10% a year and perfectly tax-inefficient asset classes returning a perfect 2% a year.
When you add all of those other factors in, it can get a lot more complicated. In fact, it gets so complicated that to truly arrive at the right answer for you, you will have to make guesses about the unknowable future including tax rates, when you will die, future returns of asset classes, and what will happen to your assets when you die. Good luck with that. However, there are some general rules you can apply to your situation.
General Rule # 1
Tax-inefficient asset classes with a high expected return should go into a tax-protected account.
Tax-efficient asset classes with a low expected return should go into a taxable account (assuming your tax-protected accounts are full. If not, everything should go into tax-protected accounts if possible.)
This includes things like muni bond funds and any investment that has a high probability of going down in value (although I suppose if you knew that, you wouldn’t buy it in the first place.)
General Rule # 3
Don’t spend a lot of time or effort trying to decide where to put tax-inefficient, low returning asset classes or tax-efficient, high-returning asset classes.
General Rule # 4
When choosing between two asset classes, pay attention to the bigger difference. If there is a BIG difference in expected return (let’s say 5%+), go with the high-returning asset class in tax-protected. If the difference is small, (say <2%), go with the tax-inefficient asset class in tax-protected. If the expected return difference is between 2-5%, then you really shouldn’t worry about it. Your expense ratios will make a bigger difference.
For example, if you have a 1% CD and an emerging market index fund you expect 7% from in the long run, put the index fund in the tax-protected account. But if you have a bond fund paying 5% and you only expect 6% out of your total stock market index fund over your investment horizon, then put the bond fund in tax-protected. But if you have a fund that is slightly more tax-efficient and has a slightly lower expected return (think TSM vs SV), then it is a no-brainer to put it in taxable.
If you have a bond fund with a 3% yield and a stock fund that you expect 7% from…well, don’t worry about it. You’re not going to know which was the right decision for years and the consequences won’t be that large anyway.
If one asset class is only slightly more tax-efficient, but has a significantly higher expected return, put it in the tax-protected account. Likewise, if an asset has only a slightly lower expected return, but is much more tax-inefficient, put it preferentially in the tax-protected account.
If you’re not sure what the expected return is for your asset class, join the club, but Rick Ferri gives some reasonable guidelines with his crystal ball which seems to be somewhat clearer than mine. (If you care, he predicts 3.9% for 10 year treasury bonds and 7% for US Large Cap stocks over the next 30 years.)
Pet Peeve #2 – Thinking That Locating Higher Expected Return Asset Classes into Tax-free Accounts is A Free Lunch
Another mistake I often see otherwise intelligent people make is to try to give recommendations about what asset class should go into your tax-deferred account (think 401(k)) and what asset class should go into your tax-free account (such as a Roth IRA.)
The common recommendation is to put the higher expected return asset class into the tax-free account and the lower expected return asset class into the tax-deferred account. What these folks don’t seem to get is that a tax-deferred account is simply a tax-free account plus an account you are investing for the government’s benefit. The only part of that tax-deferred account you will ever get to spend is the tax-free portion. When you look at it that way, you really just have two tax-free accounts (plus a government account that goes along for the ride for a few decades.)
So when you put the asset with the higher-expected return asset class preferentially into the tax-free account without acknowledging that a significant chunk of the tax-deferred account doesn’t actually belong to you, you are really just changing your asset allocation to a riskier asset allocation. Yes, over the long run that is likely to give you a higher overall return, but only because you are taking on additional risk. It’s not the free lunch these complex asset location process gurus would like you to think it is.
Let’s do some math to illustrate. Let’s assume you have a $10,000 tax-deferred account and a $10,000 tax-free account and you have two asset classes- one of which you expect a 10% return from and one that you expect a 5% return from. We will also assume the effective tax rate at withdrawal from the tax-deferred account is 20%.
What you really have is:
- $10,000 tax-free account
- $8,000 tax-free account
- $2,000 government account
If you put the 10% expected return asset class into the larger tax-free account, and the 5% expected return asset class into the smaller tax-free account and the government account, and then let it ride for ten years, you end up with:
=FV(10%,10,0,-10000) + FV(5%,10,0,-8000) = $38,969 and the government ends up with =FV(5%,10,0,-2000) = $3,258
If you put the 5% expected return asset class into the larger tax-free account, and the 10% expected return asset class into the smaller tax-free account and the government account, and then let it ride for ten years, you end up with:
=FV(5%,10,0,-10000) + FV(10%,10,0,-8000) = $37,039 and the government ends up with =FV(10%,10,0,-2000) = $5,187
Basically, if you asset locate properly, you get $1,930 more and the government gets $1,930 less. That’s like getting a return that is 0.54% higher (6.90% to 6.36%) Cool, right? But it shouldn’t be surprising given what you now know. In the first situation, you had this after-tax asset allocation:
- $10,000/$18,000 = 56% asset class with 10% expected return
- $8,000/$18,000 = 44% asset class with 5% expected return
And in the second situation you had
- $8,000/$18,000 = 44% asset class with 10% expected return
- $10,000/$18,000 = 56% asset class with 5% expected return
The truly advanced at this point will point out that Roth IRAs aren’t subject to RMDs, providing another reason to go with the higher returning asset class in the tax-free account. That’s true, but it’s a rather minor point for most of us who actually plan to spend most of our money in retirement. Of course a 56/44 asset allocation is going to have a higher expected return than a 44/56 asset allocation. You didn’t do anything smart. You just took on additional risk.
Now, sometimes behavior trumps math, and if you somehow were able to fool yourself into taking on more risk than you thought you were, and that helped you stay the course with a more aggressive asset allocation, then maybe you did yourself a favor. But there’s no magic under the hood here. It’s just a bunch of hand-waving.
Back In the Real World
Of course, back here in the real world nobody is adjusting their asset allocation for taxes. Try it and you’ll see why. It’s hard enough to figure out what your effective tax rate on tax-deferred withdrawals will be. Now try to figure it out with your taxable account! That’s what I thought. Talked you out of it didn’t I?
So in the real world, throwing your higher expected return asset classes preferentially into Roth is a reasonable thing to do, even if it isn’t a free lunch. Just realize that if the Great Depression II occurs, you’ll wish you had done the opposite.
What do you think? What do you do with asset location? What has your biggest dilemma with it been? Do you try to put your highest expected return assets into your Roth IRA? Why or why not? Comment below!