Tax Efficiency of Treasuries

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
JamesSFO
Posts: 3404
Joined: Thu Apr 26, 2012 10:16 pm

Re: Tax Efficiency of Treasuries

Post by JamesSFO »

Call_Me_Op wrote:I am not sure why there is so much confusion about this. Let's say I have $10,000 and I want to decide whether to put it in a tax-deferred account (say an after-tax IRA) or keep it in taxable. In both cases, I plan to invest in a 4% treasury bond. And, to eliminate the apparent source of confusion, in BOTH cases I will pay taxes out of a cash account that is yielding nothing. I will reinvest at 4% for simplicity.
Ok you changed the hypothetical from the original now. When originally presented there were taxes coming out of case 1 from the investment and out of magic cash for case 2. As fnk notes, you've now shown that 39% tax rate > 33% tax rate.
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Tax Efficiency of Treasuries

Post by grabiner »

Call_Me_Op wrote:More importantly, I have shown that if you pay out of cash, tax-deferral in and of itself is essentially useless. (There is a minor issue of interest on the cash, which I eliminated by eliminated cash interest rate.)
But this isn't a minor issue; you aren't leaving extra cash around today as cash to pay your tax bill in 20 years. If you are maxing out your retirement plans, you could use the extra cash to buy long-term municipal bonds and have a low-risk return well above 0%, or I-bonds or TIPS and have a return which tracks inflation; alternatively, you could buy stock and expect even higher returns. If you aren't maxing out your retirement plans, you could add the money to your retirement plans where it will grow tax-deferred or tax-free for even better returns.

That is the benefit of tax deferral, whether in a tax-deferred account or just in the tax deferral of capital gains from stocks. (And the additional advantage of most tax-deferred accounts is the tax deduction on contributions.)
Wiki David Grabiner
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Tax Efficiency of Treasuries

Post by Call_Me_Op »

JamesSFO wrote:
Call_Me_Op wrote:I am not sure why there is so much confusion about this. Let's say I have $10,000 and I want to decide whether to put it in a tax-deferred account (say an after-tax IRA) or keep it in taxable. In both cases, I plan to invest in a 4% treasury bond. And, to eliminate the apparent source of confusion, in BOTH cases I will pay taxes out of a cash account that is yielding nothing. I will reinvest at 4% for simplicity.
Ok you changed the hypothetical from the original now. When originally presented there were taxes coming out of case 1 from the investment and out of magic cash for case 2. As fnk notes, you've now shown that 39% tax rate > 33% tax rate.
No. There was no change in the analysis. I just couched it a bit differently. Originally, I did not say where I got the cash to pay the taxes.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Tax Efficiency of Treasuries

Post by Call_Me_Op »

grabiner wrote:
Call_Me_Op wrote:More importantly, I have shown that if you pay out of cash, tax-deferral in and of itself is essentially useless. (There is a minor issue of interest on the cash, which I eliminated by eliminated cash interest rate.)
But this isn't a minor issue; you aren't leaving extra cash around today as cash to pay your tax bill in 20 years. If you are maxing out your retirement plans, you could use the extra cash to buy long-term municipal bonds and have a low-risk return well above 0%, or I-bonds or TIPS and have a return which tracks inflation; alternatively, you could buy stock and expect even higher returns. If you aren't maxing out your retirement plans, you could add the money to your retirement plans where it will grow tax-deferred or tax-free for even better returns.

That is the benefit of tax deferral, whether in a tax-deferred account or just in the tax deferral of capital gains from stocks. (And the additional advantage of most tax-deferred accounts is the tax deduction on contributions.)
I always keep cash on hand, and always pay my taxes out of cash. I think the effect you are discussing exists, but it is not that significant if someone always keeps a cash allocation.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
User avatar
FNK
Posts: 1360
Joined: Tue May 17, 2011 7:01 pm

Re: Tax Efficiency of Treasuries

Post by FNK »

Call_Me_Op wrote:I always keep cash on hand, and always pay my taxes out of cash. I think the effect you are discussing exists, but it is not that significant if someone always keeps a cash allocation.
There's a big difference between holding an emergency fund that has a lot of uses and locking up cash for 20 years with no interest to make a point. In the second case, if you have cash today that you know will need to be used in 20 years, you should buy an EE bond with it. Throw in a doubling of that cash and see how your math works out.

Also, try the kind of tax deferral people do use, either deductible Trad or a Roth conversion of your non-deductible contribution. That's the kind of advantaged accounts I was talking about.
Bob's not my name
Posts: 7417
Joined: Sun Nov 15, 2009 8:24 am

Re: Tax Efficiency of Treasuries

Post by Bob's not my name »

Call_Me_Op wrote:More importantly, I have shown that if you pay out of cash, tax-deferral in and of itself is essentially useless. All that really matters is the relative tax rates.

As far as throwing way all the advantages of tax-deferred accounts, that's the whole point. There is no fundamental advantage (not counting pre-tax contribution) unless your tax rate in retirement will be lower
You haven't shown this. Tax-deferral has a quantifiable advantage, which you have taken off the table by allowing tax-deferred compounding in the taxable account, which pays no taxes.

Your model, though flawed, still confirms the conventional wisdom that a non-deductible TIRA may not be advantageous for a tax-advantaged asset. Your tax rate on Treasury interest now is 33% s. 39% on IRA withdrawals later.

Even that is improbable. Only a very narrow slice of taxpayers are in the 33% bracket without having AGI above $200,000 (single) or $250,000 (married), which means their tax rate on Treasury interest will be 36.8% due to the ACA tax on interest, while their tax rate on IRA withdrawals in the same nominal bracket will be 33% plus state tax (which may be deductible against federal, as grabiner pointed out, depending on AMT effects).
Last edited by Bob's not my name on Wed Dec 26, 2012 6:29 am, edited 1 time in total.
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Tax Efficiency of Treasuries

Post by grabiner »

Call_Me_Op wrote:I always keep cash on hand, and always pay my taxes out of cash. I think the effect you are discussing exists, but it is not that significant if someone always keeps a cash allocation.
You may pay tax out of your cash allocation, but you presumably keep a cash allocation which is relatively constant except for short-term expenses. If you are planning to contribute $5500 to a non-deductible IRA next week and invest it in bonds, you have presumably left an extra $5500 in cash now rather than investing it in your taxable account, but you aren't also keeping an extra $2000 in cash because you expect to owe that much tax when you withdraw the money in 2032. Rather, you will withdraw an extra $2000 (from some account) in 2032 to pay your tax, and that $2000 might be the earnings on $1200 you invested this year.
Wiki David Grabiner
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

getting more explicit about cash flows

Post by afan »

Taxable scenario: Each year you receive taxable interest. Your after tax rate of return is 0.67 x 4 =2.68%. This is your return whether you take the money out of this account, or take it from a separate account. If you take it from a separate account, the economic effect is that you reduced the amount in the treasury fund to pay taxes, then replenished it by adding more to the T fund by taking it from the cash account. In other words, your total investment is no longer $10,000. It is now $10,000 + tax on the interest.

Consider the end of year 1. You collect $400 interest in your initial investment. You owe $132 in federal tax. You take this out, but then replace the $132 from your cash account. Your total investment is no longer 10,000. It is $10,132. You now collect interest for 19 more years. To do a fair comparison you would have to also invest another $132 in the tax deferred scenario. Let's do that, put it in the same taxable T fund at 4%. Again, the after tax return is 2.68%. By the end of the term this would be worth $219, after tax. You need to add this to the total in the tax deferred scenario. At the end of year 2 you again compute the tax in the taxable account, pay that amount in tax, and on the deferred scenario put that into a side account at 4%, 2.68 after tax. Compound at 2.68% for 18 years and add the total to the proceeds from the tax deferred, plus the proceeds from the year 1 investment. Repeat thus every year. You have now investwd $10,000 at the start and smaller, but growing, amounts each year as you pay taxes (taxable case), or move money into the side fund (deferred case). At the end of 20 years you now have the same amounts invested in the taxable and deferred scenarios, and it would be meaningful to compare the totals.

Of course, if you assume your annual "investments" of taxes on the taxable account earn 4%, but your side fund on the deferred scenario earns 0%, then the deferral is less appealing. But I think the explicit cash flows make it clear how arbitrary that 4% vs 0% assumption is. In effect you have assumed a 4% taxable rate for your annual investments into the taxable account, but a 0% rate when you put money into the side fund. You could do that, but it would not make much sense.


Of course, this assumes you actually liquidate the deferred account at year 20, which you don't have to do.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Tax Efficiency of Treasuries

Post by Call_Me_Op »

FNK wrote:
Call_Me_Op wrote: Also, try the kind of tax deferral people do use, either deductible Trad or a Roth conversion of your non-deductible contribution. That's the kind of advantaged accounts I was talking about.
Perhaps - but that's not the type I was talking about.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: Tax Efficiency of Treasuries

Post by Call_Me_Op »

Bob's not my name wrote:Tax-deferral has a quantifiable advantage, which you have taken off the table by allowing tax-deferred compounding in the taxable account, which pays no taxes.
That's the whole point. My analysis is absolutely correct for the assumptions stated - and that is the case I set-out to analyze. Different assumptions will, of course, modify the results.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
Topic Author
Call_Me_Op
Posts: 9881
Joined: Mon Sep 07, 2009 2:57 pm
Location: Milky Way

Re: getting more explicit about cash flows

Post by Call_Me_Op »

afan wrote: Of course, if you assume your annual "investments" of taxes on the taxable account earn 4%, but your side fund on the deferred scenario earns 0%, then the deferral is less appealing. But I think the explicit cash flows make it clear how arbitrary that 4% vs 0% assumption is. In effect you have assumed a 4% taxable rate for your annual investments into the taxable account, but a 0% rate when you put money into the side fund. You could do that, but it would not make much sense.
It makes sense if there is a relatively large part of your AA devoted to cash - which applies to my taxable AA. There are many reasons to hold an allocation to cash, and if it is there for those reasons, it's no big deal to pay taxes out of there. The Permanent Portfolio, for example, holds 25% in cash.
Best regards, -Op | | "In the middle of difficulty lies opportunity." Einstein
User avatar
JamesSFO
Posts: 3404
Joined: Thu Apr 26, 2012 10:16 pm

Re: Tax Efficiency of Treasuries

Post by JamesSFO »

Call_Me_Op wrote:
Bob's not my name wrote:Tax-deferral has a quantifiable advantage, which you have taken off the table by allowing tax-deferred compounding in the taxable account, which pays no taxes.
That's the whole point. My analysis is absolutely correct for the assumptions stated - and that is the case I set-out to analyze. Different assumptions will, of course, modify the results.
I think the revised set of assumptions where you read the whole thread through and realize what you are actually talking about the punchline is that you are showing a lower tax rate produces a higher return. Not sure why you (a) don't edit the first post to clarify your assumptions (tax rates, type of IRA, where cash is coming from) and (b) why you think this is particularly interesting or should change the conventional wisdom about where to put funds...
FinancialDave
Posts: 1819
Joined: Thu May 26, 2011 9:36 pm

Re: Tax Efficiency of Treasuries

Post by FinancialDave »

The main conclusion for me on reading this was I just wasted a lot of time since I live in Washington State.

Of course much like the analysis assumes future tax rates, I could assume a future state income tax (which is not out of the question.)

:D

One piece of advice is to remember that not all taxes are "marginal" and secondly that in retirement a larger portion of your income could turn out to be non-taxable, especially if it is being pulled from a taxable account where most of the tax has already been paid, such as your example.
I love simulated data. It turns the impossible into the possible!
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Re: Tax Efficiency of Treasuries

Post by afan »

The concern is not the portion of AA held in cash. The concern lies in the interpretation of the different rates of return as attributable to tax treatment, as opposed to different amounts, timing, and interest rates on investments. I ask not why you have a given percent in cash, but why you assume a long term T rate for the taxable account-both for initial and subsequent investments long term T rates for the deferred account, but cash rates for subsequent investments in the side fund.

This is equivalent to taking money from the taxable T fund to pay taxes, creating a side fund to accept the replenish payments, but holding that side fund in cash at 0%interest. If you want to hold it as cash, then to do a comparison of the value of deferral you have to hold at 0% in both the taxable and deferred cases. Otherwise you attribute different performance to tax treatment rather than pretax investment returns.

And, if you invest more in the taxable account, then any comparison of final values with the deferred account is meaningless.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Net present value

Post by afan »

The most robust way to handle the situation where you really want to invest different amounts of money at different times into the two scenarios is to do net present value calculations. You will need to specify a discount rate. It would seem logical to use the same discount rate for both the taxable and the tax deferred case. If you really want to accumulate cash at 0% interest in the deferred case, then, having eliminated interest rate risk, it would seem appropriate to use a lower discount rate.'

Note- you cannot use a 0% nominal interest rate for the cash investment in the taxable case, at least not as you have defined it. You are putting the cash into the T fund. If you were taking cash out of the T fund, and replenishing that amount into a 0% interest cash account then you would be using a 0% interest rate in the taxable example. As it is, you are using 2.68% after tax.

So my suggestion is to use 4% pretax as the interest rate for both the taxable and the deferred accounts, create a taxable side fund to the deferred account, putting the amounts into it each year that you use to pay income taxes from the taxable account, and use the same discount rate for the taxable and the deferred scenarios.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Re: Tax Efficiency of Treasuries

Post by afan »

Since the exercise of paying the taxes out of a separate fund leads to increases in investment in the taxable account, it distorts any attempt to analyze the returns. The after tax return on the taxable account under this assumption- 4% pretax return, 33% tax rate- is 0.67 x 4=2.68%. That is the return, no matter how you pay the taxes.

Taxable:
invest 10,000 at 2.68% for 20 years. The final value, after tax would be $17,265.

Tax deferred: invest 10,000 at 4% for 20 years. At the end, cash out, pay tax and keep the rest.

value before tax is 21,911 including 11,911 of gain.
at 33% tax rate, the after tax value is 17,980. More than the taxable account.
at 39% tax rate, the after tax value is 17,265. Same as the taxable account.

If the tax rate for the gain in the deferred account were greater than 39%, while the rate on the taxable remained 33%, then the taxable account would be a better deal.

Now assume you defer for 40 years, instead of 20. The numbers change:

Taxable:
invest 10,000 at 2.68% for 40 years. The final value, after tax would be $28,803.

Tax deferred: invest 10,000 at 4% for 40 years. At the end, cash out, pay tax and keep the rest.

value before tax is 48,010 including 38,010 of gain.
at 33% tax rate, the after tax value is 35,466. More than the taxable account.
at 39% tax rate, the after tax value is 33,186. More than the taxable account.

Not only is deferral a better deal at 20 years at the same tax rate, and at a tax rate up to 6% higher for the deferred investment, but the superiority of the deferred approach increases as the time horizon increases.

If you start putting money into an after tax IRA at age 21, you have 50 years of deferral before you are required to start taking some out. But then you do not need to take it all out. So the deferral continues to work in your favor.

Finally, remember that in the recent past one has been able to make charitable donations with IRA funds without realizing the income, or paying tax on it. If one were to do this, then the effective tax rate on the IRA would become zero. You do not have this option with a taxable bond investment. Another advantage of deferring income.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Tax Efficiency of Treasuries

Post by grabiner »

afan wrote:Finally, remember that in the recent past one has been able to make charitable donations with IRA funds without realizing the income, or paying tax on it. If one were to do this, then the effective tax rate on the IRA would become zero.
This is not a tax rate of zero, because you lose the tax deduction; the only advantage is that the income never appears, so it doesn't affect other income-based limitations. For example, if you withdraw $10,000 from an IRA, then write a $10,000 check to charity, you have $10,000 of income and $10,000 of deductions, with no net tax effect, just as if you made a direct charitable distribution.

However, if you are in the phase-in range for Social Security taxation, then withdrawing $10,000 from the IRA (raising your adjusted gross income by $10,000) might make $8500 of Social Security taxable at 15%, leading to a tax cost of $1275. The direct charitable distribution (which expired in 2012 but might be reinstated) avoids this tax.

The real tax advantage is for stocks in a taxable account, because you can donate appreciated stock to a charity and not pay tax on the capital gains. If you paid $2000 for stock which is now worth $10,000, and you have held the stock more than a year, you can get $8800 out of the stock for your own use, paying $1200 tax. But if you donate that stock to a charity, the charity gets the full $10,000, and you get a $10,000 deduction.
Wiki David Grabiner
User avatar
FNK
Posts: 1360
Joined: Tue May 17, 2011 7:01 pm

Re: Tax Efficiency of Treasuries

Post by FNK »

grabiner wrote:The real tax advantage is for stocks in a taxable account, because you can donate appreciated stock to a charity and not pay tax on the capital gains.
That's right. You get to double-dip.
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Tax rate zero

Post by afan »

Under current law a charitable contribution from an IRA would be treated as income recognition and a deduction. I was referring to the recently expired and perhaps to be resurrected provision that allowed on to contribute to the charity without recognizing the income. Then this hypothetical $8,000 appreciation on a $2,000 investment would produce a $10,000 charitable contribution, no reported income, no deduction, and no tax. Hence, a tax rate of zero.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Tax rate zero

Post by grabiner »

afan wrote:Under current law a charitable contribution from an IRA would be treated as income recognition and a deduction. I was referring to the recently expired and perhaps to be resurrected provision that allowed on to contribute to the charity without recognizing the income. Then this hypothetical $8,000 appreciation on a $2,000 investment would produce a $10,000 charitable contribution, no reported income, no deduction, and no tax. Hence, a tax rate of zero.
And my point is that the effective tax rate is not zero, even in this situation. If you donate $10,000 in cash (or, still better, appreciated stock) to a charity, you get a $10,000 tax deduction, so you pay only $6300 out of pocket in the 37% combined tax bracket. If you donate $10,000 from an IRA, under the expired rule, you pay no tax and it costs $10,000, and under the current rule, you realize $10,000 in income and deduct $10,000, so you also pay no tax unless you hit a phase-out. Thus $10,000 withdrawn from the IRA is only worth $6300 whether you spend it or donate it to charity.
Wiki David Grabiner
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Re: Tax Efficiency of Treasuries

Post by afan »

Perhaps we need to define the term "tax rate."

If I withdraw $10,000 from an account, deliver all $10,000 to a charity, incur no tax liability, and pay nothing in taxes, then I define the "tax rate" as zero.

amount contributed to charity = (amount withdrawn)- tax paid

tax rate=(tax paid)/(amount withdrawn) or, equivalently in this case tax rate=(tax paid)/(taxable amount withdrawn)

Since the numerator is zero, the denominator does not matter.

since "tax paid"= 0, the tax rate=0.

If I withdraw $10,000 from an account, deliver all $10,000 to a charity, incur no tax liability, and pay nothing in taxes, and the tax rate is something other than zero, then what is it? More importantly, who paid the tax, and to whom was it paid? I paid nothing, the IRS collected nothing. So how do you calculate the tax rate?

If the Tea Party got its way and eliminated the income tax altogether, then the transaction would be the same- 10,000 withdrawn, 10,000 to charity, no tax. Would you say that in the absence of any income tax at all the tax rate would still be something other than zero?

The old provision differed in that the money was never recognized as income, so it never triggered any income related provisions of the Code.

Please explain how you calculate the tax rate.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Tax Efficiency of Treasuries

Post by grabiner »

afan wrote:Perhaps we need to define the term "tax rate."

Please explain how you calculate the tax rate.
The tax rate is zero on a direct charitable contribution from an IRA, but the tax rate is minus 37% on a direct charitable contribution from anywhere else, so the IRA withdrawal still has a relatively higher tax cost. I prefer to look at the $3700 forfeited tax reduction as an effective tax cost.
Wiki David Grabiner
afan
Posts: 8193
Joined: Sun Jul 25, 2010 4:01 pm

Re: Tax Efficiency of Treasuries

Post by afan »

It may be possible to find a better deal than zero, in the form of a negative rate, but the rate under discussion must be zero.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Post Reply