Supposed tax-inefficiency of bonds in taxable/TDFs

[CarefullyCarele$$]

Hi everyone, I'm new to the sub and just recently set up my taxable situation with a low-fee robo advisor that utilities tax-free muni bonds, after learning about the tax-inefficiency of bonds within TDFs - I have them in all my tax-deferred accounts. I like having a mirrored allocation and being hands-off: which is true of both TDFs and robo-advisors.

However, I recently decided to run some numbers. I compared the after-distributions-tax performance of VT (which is where most TDFs start at 90%+ to equities) to Vanguard Target Retirement Income Fund (at 30/70, this is where most TDFs end). Looking at data over the last ~15 years, the tax-cost difference between them is only about 0.45-0.5%, which averages to about ~0.25%/yr over a target date fund's investing lifetime.

0.25%/yr: that's also coincidentally the cost of most robo-advisors. And genuinely doesn't sound like a life-changing difference, much closer to a rounding error. And this is at the very highest federal tax bracket, the actual cost for most of us is bound to be much lower.

So, I wanted to ask: Why is conventional wisdom against using bonds/TDFs in taxable if the extent of the inefficiency is so small? Especially compared to the simplicity and behavioural discipline that a TDF instills? (Let us disregard the issues with TDF and capital gains for a second, and purely look at tax-inefficiency of bonds in taxable)

[nalor511]

Robos have the robo cost, the fund cost, plus the tax cost of frequently trading. It's more than 0.25%

[donaldfair71]

Hi everyone, I'm new to the sub and just recently set up my taxable situation with a low-fee robo advisor that utilities tax-free muni bonds, after learning about the tax-inefficiency of bonds within TDFs - I have them in all my tax-deferred accounts. I like having a mirrored allocation and being hands-off: which is true of both TDFs and robo-advisors.

However, I recently decided to run some numbers. I compared the after-distributions-tax performance of VT (which is where most TDFs start at 90%+ to equities) to Vanguard Target Retirement Income Fund (at 30/70, this is where most TDFs end). Looking at data over the last ~15 years, the tax-cost difference between them is only about 0.45-0.5%, which averages to about ~0.25%/yr over a target date fund's investing lifetime.

0.25%/yr: that's also coincidentally the cost of most robo-advisors. And genuinely doesn't sound like a life-changing difference, much closer to a rounding error. And this is at the very highest federal tax bracket, the actual cost for most of us is bound to be much lower.

So, I wanted to ask: Why is conventional wisdom against using bonds/TDFs in taxable if the extent of the inefficiency is so small? Especially compared to the simplicity and behavioural discipline that a TDF instills? (Let us disregard the issues with TDF and capital gains for a second, and purely look at tax-inefficiency of bonds in taxable)

Keep in mind that the tax inefficiency of holding bonds that yield less than 1% will hit entirely different than the tax inefficiency of bonds yielding 4%.

The tax-friendly way to solve that is that now those bond funds in taxable (which I too believe weren’t as tax-inefficient as the reputation- at 1%) should be tax loss harvested, and I’m sure a robo will do this. But the new yield moving forward, and dividends taxed at ordinary income rates, will drag much more than they have been for several years.

[little_star]

In my opinion, the problem with target-date funds in a taxable account is that you cannot control the glide path if you decide later to have a different asset allocation without selling (and thus incurring capital gains tax). It is far better, in my opinion, to hold the individual funds, in a tax efficient manner (bonds in tax deferred, stocks in both tax-advantaged and taxable).

I would never use a robo-advisor. Investing does not have to be difficult. If you use a 3-fund portfolio, you should be able to manage your funds without paying a robot or a person to do so for you.

[CarefullyCarele$$]

Robos have the robo cost, the fund cost, plus the tax cost of frequently trading. It's more than 0.25%

I'm using Vanguard Digital Advisor which is 0.15% inclusive of ETF expense costs - that's only 0.07% more than their target date index funds. Most robos (including Betterment and Wealthfront) prioritise rebalancing with distributions and contributions, and only sell to rebalance if the allocations are wildly off, which is the same as DIY-ing. For a lot of new investors like me who are prone to tinkering, the 0.07% (or I'll dare say even the 0.25% you'd be charged with many other robos) is well worth it for the ability to be hands off, take advantage of automatic tax-loss harvesting and the behavioural discipline that comes with using them.

Most people will always tinker if you give them a chance to tinker, which is what manual rebalancing and management is: a chance to tinker. TDFs and robos by their very nature of being "automatic" and "advisors" - prevent tinkering which can have huge costs. It might not be worth it for everyone I'm sure, especially for more experienced and disciplined investors such as yourself and others on this forum - but for me and many others like me, that extra 0.07% is insurance to keep me from fiddling with my investments and stay the course.

]

Keep in mind that the tax inefficiency of holding bonds that yield less than 1% will hit entirely different than the tax inefficiency of bonds yielding 4%.

The tax-friendly way to solve that is that now those bond funds in taxable (which I too believe weren’t as tax-inefficient as the reputation- at 1%) should be tax loss harvested, and I’m sure a robo will do this. But the new yield moving forward, and dividends taxed at ordinary income rates, will drag much more than they have been for several years.

Appreciate your response, thank you - this makes a lot of sense. But are you suggesting that using a taxable bond with tax-loss harvesting (which I'm glad my robo will do for me because I wouldn't know where to start) is actually better than using a tax-exempt municipal bond? (I'm in a high tax state, in a high income industry)

In my opinion, the problem with target-date funds in a taxable account is that you cannot control the glide path if you decide later to have a different asset allocation without selling (and thus incurring capital gains tax). It is far better, in my opinion, to hold the individual funds, in a tax efficient manner (bonds in tax deferred, stocks in both tax-advantaged and taxable).

I would never use a robo-advisor. Investing does not have to be difficult. If you use a 3-fund portfolio, you should be able to manage your funds without paying a robot or a person to do so for you.

I understand your perspective, but I use VG Digital Advisor which is just 0.07% more than their TDFs - well-worth it for someone like me for whom rebalancing is just a chance to tinker, and eliminating that is helpful.

The inflexibility of TDFs is ofc another issue, and I understand where you're coming from - but for me, handing my money to a TDF (or a robo) is much like handing my money to a low-fee trusted advisor, and I'm willing to accept it's not ideal, but it's good enough for me.

But my question is regarding this supposed tax-inefficiency of bonds being exaggerated. For the convenience of TDFs, 0.25% (or much less than 0.25% in most cases) is not life changing. I do think @donaldfair71's point of future yields being higher is something to think about tho.

[goingup]

You're smart to recognize your behavioral tendencies in investing. We really are our own worst enemies.

Using a robo advisor is fine. The only knock I've seen about them is cost, which is really not much of a knock because they are so low-cost compared to full service advisors. The other issue is that you'll likely end holding a lot more funds because of the buying/selling and tax-loss harvesting. Owning lots different funds doesn't bother some investors. (It would drive me bonkers because I want things simple.)

Don't hold a TDF in a taxable account. It's not just about bond distributions. It's about the changing glide path, capital gains distributions, as well as bond dividends, which will grow over time. If you like a TDF, keep it your 401K or other tax-advantaged accounts. https://www.bogleheads.org/wiki/Target_date_funds

As a high income person, keep broad market stock funds and muni bond funds in taxable. This has been the conventional wisdom, and even more so now that bond fund yields are so high.

[NiceUnparticularMan]

You are encountering the truism that many people here seem to devote a lot of energy to avoiding taxes that are not likely to make much difference to their overall outcomes, particularly when you net out everything on a lifetime basis.

I have analogized this before to those "hypermiling" forums where they swap ideas for squeezing out small fractions of additional miles per gallon. Of course it is fine to think broadly, say in terms of buying an appropriately fuel-efficient car (I'd say this is equivalent to making sure to fully use tax-protected accounts when possible). But all the stuff these hypermilers do beyond making reasonable vehicle choices has less and less marginal effect.

And yes, sometimes it is actually dangerous, in ways you might think outweigh any plausible fuel cost savings.

And yes, I think sometimes people here in the name of "hypermiling" their marginal taxes are really taking on additional behavioral risks that outweigh any plausible tax cost savings.

So my two cents is you can consider whatever someone suggests, but you can also ignore the more dogmatic people who say things like "You should not do X because of such and such tax concerns." Like, maybe, but often maybe not, and you can reasonably decide whatever tax concern they are referencing is outweighed by other considerations.

And using Target funds for retirement income savings held in taxable accounts to carry out a mirroring strategy is one of those things. Generally speaking, the lifetime tax benefits of not doing that are way smaller and less certain than some people seem to argue. And it is true that may not be an ideal approach if you anticipate wanting to make a lot of significant plan changes, but for some people that is a feature not a bug! Meaning the behavioral benefits of just having something you never tinker with, that in fact it would be bad to tinker with, could be seen as a positive in the greater scheme.

As a final thought--my personal rule of thumb is that somewhere between 0.5 and 1% of additional estimated cost (annualized), I would at least start thinking seriously about whether saving those costs would be worth some additional behavioral risk. But even in that range, it can make sense. And below 0.5%--eh, lower is better, of course, but I would think behavioral risks can easily trump estimated cost differences in that range.

But that's just me.

[CarefullyCarele$$]

You're smart to recognize your behavioral tendencies in investing. We really are our own worst enemies.

Using a robo advisor is fine. The only knock I've seen about them is cost, which is really not much of a knock because they are so low-cost compared to full service advisors. The other issue is that you'll likely end holding a lot more funds because of the buying/selling and tax-loss harvesting. Owning lots different funds doesn't bother some investors. (It would drive me bonkers because I want things simple.)

Don't hold a TDF in a taxable account. It's not just about bond distributions. It's about the changing glide path, capital gains distributions, as well as bond dividends, which will grow over time. If you like a TDF, keep it your 401K or other tax-advantaged accounts. https://www.bogleheads.org/wiki/Target_date_funds

As a high income person, keep broad market stock funds and muni bond funds in taxable. This has been the conventional wisdom, and even more so now that bond fund yields are so high.

Thank you! Robo-advisors seem to get a lot of flak around here, and I totally get why. But for people like me, they've essentially solved portfolio construction and management for a very low fee. I think in the long run, their fees which are 0.25% or lower, are well worth it. Appreciate you recognising that.

To your point about too many funds: VG, Betterment, WF all either put you by default in a simple 3-fund (VG), or you can choose to be in one (Betterment, WF). So portfolio construction wise, it's really not that different. Ofc, the tax-loss harvesting would generate parallel funds (essentially doubling the number), but wouldn't that be the case with manual TLH too? Just curious.

Appreciate you detailing out other concerns with TDFs, definitely gives me a lot to think about. I'm currently in a combination of VTI/VXUS/VTEB (managed by VG Digital Advisor), and I'm glad robos exist.

You are encountering the truism that many people here seem to devote a lot of energy to avoiding taxes that are not likely to make much difference to their overall outcomes, particularly when you net out everything on a lifetime basis.

I have analogized this before to those "hypermiling" forums where they swap ideas for squeezing out small fractions of additional miles per gallon. Of course it is fine to think broadly, say in terms of buying an appropriately fuel-efficient car (I'd say this is equivalent to making sure to fully use tax-protected accounts when possible). But all the stuff these hypermilers do beyond making reasonable vehicle choices has less and less marginal effect.

And yes, sometimes it is actually dangerous, in ways you might think outweigh any plausible fuel cost savings.

And yes, I think sometimes people here in the name of "hypermiling" their marginal taxes are really taking on additional behavioral risks that outweigh any plausible tax cost savings.

So my two cents is you can consider whatever someone suggests, but you can also ignore the more dogmatic people who say things like "You should not do X because of such and such tax concerns." Like, maybe, but often maybe not, and you can reasonably decide whatever tax concern they are referencing is outweighed by other considerations.

And using Target funds for retirement income savings held in taxable accounts to carry out a mirroring strategy is one of those things. Generally speaking, the lifetime tax benefits of not doing that are way smaller and less certain than some people seem to argue. And it is true that may not be an ideal approach if you anticipate wanting to make a lot of significant plan changes, but for some people that is a feature not a bug! Meaning the behavioral benefits of just having something you never tinker with, that in fact it would be bad to tinker with, could be seen as a positive in the greater scheme.

As a final thought--my personal rule of thumb is that somewhere between 0.5 and 1% of additional estimated cost (annualized), I would at least start thinking seriously about whether saving those costs would be worth some additional behavioral risk. But even in that range, it can make sense. And below 0.5%--eh, lower is better, of course, but I would think behavioral risks can easily trump estimated cost differences in that range.

But that's just me.

Thank you, I feel seen! Really helpful and puts things in perspective. Keeping yourself from tinkering with your own investments is an underrated benefit.

I'm happy with using a robo for now, since they have basically the same benefits as TDFs, along with slight advantages like automatic TLH and being able to use tax-exempt bonds.

But TDFs were the original robos in a lot of ways, so it was interesting to discover that most robos today are actually priced at around the supposed tax-inefficiency cost of TDFs, which is ~0.25%.

Once again, really appreciate your thoughtful response.

[aristotelian]

Even if the tax cost of the bond fund in taxable is in itself negligible, you are in effect increasing the stock allocation of your 401k, causing 401k gains to be taxed as income rather than long term capital gains. Thus you could have a muni bond in taxable with zero tax cost that still reduces your overall tax efficiency.

Also have you considered the Vanguard target date fiasco? https://www.morningstar.com/stocks/less ... s-surprise

[CarefullyCarele$$]

Even if the tax cost of the bond fund in taxable is in itself negligible, you are in effect increasing the stock allocation of your 401k, causing 401k gains to be taxed as income rather than long term capital gains. Thus you could have a muni bond in taxable with zero tax cost that still reduces your overall tax efficiency.

Also have you considered the Vanguard target date fiasco? https://www.morningstar.com/stocks/less ... s-surprise

I'm sorry but could you please elaborate? In a more ELI5 way?

I did, and it seems more a minor tax inefficiency than anything else. Here's a video that gets to my point: https://youtu.be/n8X1vZTjZS8

[retiredjg]

But my question is regarding this supposed tax-inefficiency of bonds being exaggerated.

I see it this way.

Bonds are taxed at your marginal rate. Let's say that is 28% or 32% which I think will also trigger the NITT at an additional 3.8%. The bond dividends are taxed each year - you cannot pick the year to take the income and you cannot pick when to pay the tax.

Stocks are taxed at the cap gains rate which would be 15% (and maybe the NITT 3.8%). Stock cap gains are mostly paid when/if you choose to sell shares.
.

With bonds you make a little money and pay tax at a high rate. With stocks you make more money (over the long run) and pay a lower rate of tax.

Taxable bonds in a taxable account do not sound very efficient to me.

[NiceUnparticularMan]

I'm happy with using a robo for now, since they have basically the same benefits as TDFs, along with slight advantages like automatic TLH and being able to use tax-exempt bonds.

But TDFs were the original robos in a lot of ways, so it was interesting to discover that most robos today are actually priced at around the supposed tax-inefficiency cost of TDFs, which is ~0.25%.

Funny how that works! Almost like this is a competitive market driving down marginal costs for substitutable products . . . .

Anyway, sort of a side note, but back in the day that 1% upper bound in my rule of thumb was motivated by the fact it could reasonably cost up to 1% to have a competent fiduciary manage an account per standard Bogleheadish principles, and I thought in many cases that could be money well spent if necessary.

But lower costs are definitely better, and it appears that robos are now basically driving down that cost, at least for the sorts of basic services most Bogleheads would be interested in.

So that's great! And who knows--maybe robo costs will get so low they actually compete directly with Target funds even in tax-protected accounts. That would be cool.

[aristotelian]

Even if the tax cost of the bond fund in taxable is in itself negligible, you are in effect increasing the stock allocation of your 401k, causing 401k gains to be taxed as income rather than long term capital gains. Thus you could have a muni bond in taxable with zero tax cost that still reduces your overall tax efficiency.

Also have you considered the Vanguard target date fiasco? https://www.morningstar.com/stocks/less ... s-surprise

I'm sorry but could you please elaborate? In a more ELI5 way?

I did, and it seems more a minor tax inefficiency than anything else. Here's a video that gets to my point: https://youtu.be/n8X1vZTjZS8

If you are keeping your overall allocation constant, more bonds (less stocks) in taxable means more stocks (fewer bonds) in your 401k.

Stock dividends and gains in taxable are taxed favorably using long term capital gains rates, which is lower than tax on ordinary income. 401k withdrawals are taxed as ordinary income.

Therefore, holding bonds in taxable causes more of your stock allocation gains to be taxed as income rather than capital gains.

Yes, it is a relatively minor optimization and you can ignore it if you prefer bonds in taxable for other reasons.

[CarefullyCarele$$]

But my question is regarding this supposed tax-inefficiency of bonds being exaggerated.

I see it this way.

Bonds are taxed at your marginal rate. Let's say that is 28% or 32% which I think will also trigger the NITT at an additional 3.8%. The bond dividends are taxed each year - you cannot pick the year to take the income and you cannot pick when to pay the tax.

Stocks are taxed at the cap gains rate which would be 15% (and maybe the NITT 3.8%). Stock cap gains are mostly paid when/if you choose to sell shares.
.

With bonds you make a little money and pay tax at a high rate. With stocks you make more money (over the long run) and pay a lower rate of tax.

Taxable bonds in a taxable account do not sound very efficient to me.

Hi again , thank you for your input. I get where you're coming from, but I'm not saying bonds in taxable (as part of TDFs) are tax efficient. All I'm saying is the cost of this tax-inefficiency seems exaggerated when looking at the historical data (not more than 0.25% over a TDF's lifetime). So it might be worth it for some people considering the behavioral discipline and simplicity embedded in a TDF.

Of course, now that there are robos out there which do basically the same thing but using tax-exempt bonds instead and with automatic TLH, my point is kinda moot. But there seem to be many who have qualms with robo-advisors, for them TDFs might still be a great solution despite the (very slight) tax-inefficiency - which is not the conventional wisdom.

I'm happy with using a robo for now, since they have basically the same benefits as TDFs, along with slight advantages like automatic TLH and being able to use tax-exempt bonds.

But TDFs were the original robos in a lot of ways, so it was interesting to discover that most robos today are actually priced at around the supposed tax-inefficiency cost of TDFs, which is ~0.25%.

Funny how that works! Almost like this is a competitive market driving down marginal costs for substitutable products . . . .

Anyway, sort of a side note, but back in the day that 1% upper bound in my rule of thumb was motivated by the fact it could reasonably cost up to 1% to have a competent fiduciary manage an account per standard Bogleheadish principles, and I thought in many cases that could be money well spent if necessary.

But lower costs are definitely better, and it appears that robos are now basically driving down that cost, at least for the sorts of basic services most Bogleheads would be interested in.

So that's great! And who knows--maybe robo costs will get so low they actually compete directly with Target funds even in tax-protected accounts. That would be cool.

Absolutely - totally see your point! The upstart robos (Betterment, Wealthfront, etc) have already basically forced the legacy providers to come out with their own low-cost robos (Vanguard, Fidelity, Schwab, etc). Now that the genie's out of the bottle, it's impossible for them to go back. Plus: most of them are by default very Boglehead in their portfolio construction. Most people who use these robos are Bogleheads without even knowing it - which is insane when you think about it, but how lucky for young people like me who are just starting out on their investing journey.

And to your point about robos being low-cost enough for tax-protected accounts: Vanguard's robo Digital Advisor is already almost as cheap as a TDF, at just 0.15% inclusive of fees.

Even if the tax cost of the bond fund in taxable is in itself negligible, you are in effect increasing the stock allocation of your 401k, causing 401k gains to be taxed as income rather than long term capital gains. Thus you could have a muni bond in taxable with zero tax cost that still reduces your overall tax efficiency.

Also have you considered the Vanguard target date fiasco? https://www.morningstar.com/stocks/less ... s-surprise

I'm sorry but could you please elaborate? In a more ELI5 way?

I did, and it seems more a minor tax inefficiency than anything else. Here's a video that gets to my point: https://youtu.be/n8X1vZTjZS8

If you are keeping your overall allocation constant, more bonds (less stocks) in taxable means more stocks (fewer bonds) in your 401k.

Stock dividends and gains in taxable are taxed favorably using long term capital gains rates, which is lower than tax on ordinary income. 401k withdrawals are taxed as ordinary income.

Therefore, holding bonds in taxable causes more of your stock allocation gains to be taxed as income rather than capital gains.

Yes, it is a relatively minor optimization and you can ignore it if you prefer bonds in taxable for other reasons.

Ah, I see. Thanks for explaining - definitely gives me even more reasons to stick to a robo/DIY in taxable vs TDFs.

[NiceUnparticularMan]

And to your point about robos being low-cost enough for tax-protected accounts: Vanguard's robo Digital Advisor is already almost as cheap as a TDF, at just 0.15% inclusive of fees.

Yeah, that is already rounding error territory to me.

Just another little historical note. Not even that long ago, the way these companies talked about their Target funds, they would typically describe them as good starting points for the average retirement savings investor, but that of course customization could make sense, and you could contact their advisory services to get such customization. To be clear, there was nothing nefarious about that, but it was a significant leap up in annual expenses to go from a low-cost Target fund to a personal advisor.

OK, but now robos are allowing you to customize, but without the need to pay an advisor. Or not pay them a recurring fee, meaning you could maybe pay a few small fixed fees to get help programming and reprogramming your robot periodically, but most years you could then leave the robot to implement your customized plan.

And from that perspective, this is bridging that gap between low-cost but uncustomized Target funds, and customized by significantly higher-cost advisories.

Again, pretty cool.

[CarefullyCarele$$]

And to your point about robos being low-cost enough for tax-protected accounts: Vanguard's robo Digital Advisor is already almost as cheap as a TDF, at just 0.15% inclusive of fees.

Yeah, that is already rounding error territory to me.

Just another little historical note. Not even that long ago, the way these companies talked about their Target funds, they would typically describe them as good starting points for the average retirement savings investor, but that of course customization could make sense, and you could contact their advisory services to get such customization. To be clear, there was nothing nefarious about that, but it was a significant leap up in annual expenses to go from a low-cost Target fund to a personal advisor.

OK, but now robos are allowing you to customize, but without the need to pay an advisor. Or not pay them a recurring fee, meaning you could maybe pay a few small fixed fees to get help programming and reprogramming your robot periodically, but most years you could then leave the robot to implement your customized plan.

And from that perspective, this is bridging that gap between low-cost but uncustomized Target funds, and customized by significantly higher-cost advisories.

Again, pretty cool.

Such a good point, totally agree. Very cool indeed!

[NiceUnparticularMan]

Just to throw another high level thought out.

So in the typical lifecycle of a Target fund, it doesn't really matter much if in a taxable account you use such a fund or instead a couple stock funds deep in accumulation, because in that range your Target fund will typically be almost all stocks anyway. It might be doing a little rebalancing between stock funds, but so might you be. It will also "miss" tax loss harvesting "opportunities", but these are not complete free lunches since they increase your basis. And eventually if you end up selling almost all these accumulated stocks during your life--a reasonable goal for retirement income savings--then you won't get a stepped up basis.

Of course if you are saving not for retirement income, but instead say to give to charities or leave as a legacy or so on, THEN you might want to just use stock funds. But if your Target fund is in fact part of your retirement savings strictly defined, this becomes a much smaller concern.

And also on the back end! Meaning once you are actually in retirement and using some mix of stocks and bonds in a taxable account to provide spendable income, then you are not experiencing tax drag due to reinvestment of taxed income. And actually, it gets complicated comparing that to, say, using a traditional IRA or 401K for that purpose. Roths are the best, of course, but then you might not want to actually use those prematurely to generate income . . . .

So the "bad" period is really just when you are in the transition between mostly stock deep accumulation portfolios and active retirement income portfolios, which a lot of Target funds start around 20 years before expected retirement.

And that's a thing, but really a smaller thing than some seem to think. And again, one might suggest that is precisely the period in which avoiding behavioral error is hardest AND most important. So . . . maybe not such a bad idea, understanding robos might be a viable alternative too.

[CarefullyCarele$$]

Just to throw another high level thought out.

So in the typical lifecycle of a Target fund, it doesn't really matter much if in a taxable account you use such a fund or instead a couple stock funds deep in accumulation, because in that range your Target fund will typically be almost all stocks anyway. It might be doing a little rebalancing between stock funds, but so might you be. It will also "miss" tax loss harvesting "opportunities", but these are not complete free lunches since they increase your basis. And eventually if you end up selling almost all these accumulated stocks during your life--a reasonable goal for retirement income savings--then you won't get a stepped up basis.

Of course if you are saving not for retirement income, but instead say to give to charities or leave as a legacy or so on, THEN you might want to just use stock funds. But if your Target fund is in fact part of your retirement savings strictly defined, this becomes a much smaller concern.

And also on the back end! Meaning once you are actually in retirement and using some mix of stocks and bonds in a taxable account to provide spendable income, then you are not experiencing tax drag due to reinvestment of taxed income. And actually, it gets complicated comparing that to, say, using a traditional IRA or 401K for that purpose. Roths are the best, of course, but then you might not want to actually use those prematurely to generate income . . . .

So the "bad" period is really just when you are in the transition between mostly stock deep accumulation portfolios and active retirement income portfolios, which a lot of Target funds start around 20 years before expected retirement.

And that's a thing, but really a smaller thing than some seem to think. And again, one might suggest that is precisely the period in which avoiding behavioral error is hardest AND most important. So . . . maybe not such a bad idea, understanding robos might be a viable alternative too.

Very well put.

I personally think that robos have solved this and are basically just TDFs minus their concerns (flexibility, automatic TLH, tax-efficiency), and that 90% of all people (even somewhat experienced investors) are better off using a low-cost robo. But I know quite a few people who are skeptical of robos (as with any new technology, I'm sure TDFs elicited a similar response at first) and for them: TDFs are a great option indeed.

[donaldfair71]

Robos have the robo cost, the fund cost, plus the tax cost of frequently trading. It's more than 0.25%

I'm using Vanguard Digital Advisor which is 0.15% inclusive of ETF expense costs - that's only 0.07% more than their target date index funds. Most robos (including Betterment and Wealthfront) prioritise rebalancing with distributions and contributions, and only sell to rebalance if the allocations are wildly off, which is the same as DIY-ing. For a lot of new investors like me who are prone to tinkering, the 0.07% (or I'll dare say even the 0.25% you'd be charged with many other robos) is well worth it for the ability to be hands off, take advantage of automatic tax-loss harvesting and the behavioural discipline that comes with using them.

Most people will always tinker if you give them a chance to tinker, which is what manual rebalancing and management is: a chance to tinker. TDFs and robos by their very nature of being "automatic" and "advisors" - prevent tinkering which can have huge costs. It might not be worth it for everyone I'm sure, especially for more experienced and disciplined investors such as yourself and others on this forum - but for me and many others like me, that extra 0.07% is insurance to keep me from fiddling with my investments and stay the course.

]

Keep in mind that the tax inefficiency of holding bonds that yield less than 1% will hit entirely different than the tax inefficiency of bonds yielding 4%.

The tax-friendly way to solve that is that now those bond funds in taxable (which I too believe weren’t as tax-inefficient as the reputation- at 1%) should be tax loss harvested, and I’m sure a robo will do this. But the new yield moving forward, and dividends taxed at ordinary income rates, will drag much more than they have been for several years.

Appreciate your response, thank you - this makes a lot of sense. But are you suggesting that using a taxable bond with tax-loss harvesting (which I'm glad my robo will do for me because I wouldn't know where to start) is actually better than using a tax-exempt municipal bond? (I'm in a high tax state, in a high income industry)

So it's not that one is better than the other- one could be better than the other. Dependent on a lot things that would deserve their own topic/thread.

You can loss harvest both, earn very positive yields with both, etc.

But either direction you go, the higher the yield is on whatever taxable bond you use, the more holding in taxable impacts versus holding in tax-protected. This is because more of the total return derives from dividends (taxed at marginal tax rates) rather than capital appreciation (taxed at normal capital gains rates) as yields rise. So a bigger chunk of returns get taxed at higher rates, and the tax impact increases enormously.

[CarefullyCarele$$]

Robos have the robo cost, the fund cost, plus the tax cost of frequently trading. It's more than 0.25%

I'm using Vanguard Digital Advisor which is 0.15% inclusive of ETF expense costs - that's only 0.07% more than their target date index funds. Most robos (including Betterment and Wealthfront) prioritise rebalancing with distributions and contributions, and only sell to rebalance if the allocations are wildly off, which is the same as DIY-ing. For a lot of new investors like me who are prone to tinkering, the 0.07% (or I'll dare say even the 0.25% you'd be charged with many other robos) is well worth it for the ability to be hands off, take advantage of automatic tax-loss harvesting and the behavioural discipline that comes with using them.

Most people will always tinker if you give them a chance to tinker, which is what manual rebalancing and management is: a chance to tinker. TDFs and robos by their very nature of being "automatic" and "advisors" - prevent tinkering which can have huge costs. It might not be worth it for everyone I'm sure, especially for more experienced and disciplined investors such as yourself and others on this forum - but for me and many others like me, that extra 0.07% is insurance to keep me from fiddling with my investments and stay the course.

]

Keep in mind that the tax inefficiency of holding bonds that yield less than 1% will hit entirely different than the tax inefficiency of bonds yielding 4%.

The tax-friendly way to solve that is that now those bond funds in taxable (which I too believe weren’t as tax-inefficient as the reputation- at 1%) should be tax loss harvested, and I’m sure a robo will do this. But the new yield moving forward, and dividends taxed at ordinary income rates, will drag much more than they have been for several years.

Appreciate your response, thank you - this makes a lot of sense. But are you suggesting that using a taxable bond with tax-loss harvesting (which I'm glad my robo will do for me because I wouldn't know where to start) is actually better than using a tax-exempt municipal bond? (I'm in a high tax state, in a high income industry)

So it's not that one is better than the other- one could be better than the other. Dependent on a lot things that would deserve their own topic/thread.

You can loss harvest both, earn very positive yields with both, etc.

But either direction you go, the higher the yield is on whatever taxable bond you use, the more holding in taxable impacts versus holding in tax-protected. This is because more of the total return derives from dividends (taxed at marginal tax rates) rather than capital appreciation (taxed at normal capital gains rates) as yields rise. So a bigger chunk of returns get taxed at higher rates, and the tax impact increases enormously.

Ah, I see. That's very insightful, thank you!

[NiceUnparticularMan]

I personally think that robos have solved this and are basically just TDFs minus their concerns (flexibility, automatic TLH, tax-efficiency), and that 90% of all people (even somewhat experienced investors) are better off using a low-cost robo. But I know quite a few people who are skeptical of robos (as with any new technology, I'm sure TDFs elicited a similar response at first) and for them: TDFs are a great option indeed.

For that matter, despite seeing the potential, I am one of those people who have not yet used them because I want to see how it works out in practice for a bit longer first.

But my dream is having a system where the only financial decisions we have to make in retirement are how to spend/give our wealth.

And it is entirely possible that at some point in my personal future, robos will play a large role in getting us as close as possible to that dream.

[CarefullyCarele$$]

I personally think that robos have solved this and are basically just TDFs minus their concerns (flexibility, automatic TLH, tax-efficiency), and that 90% of all people (even somewhat experienced investors) are better off using a low-cost robo. But I know quite a few people who are skeptical of robos (as with any new technology, I'm sure TDFs elicited a similar response at first) and for them: TDFs are a great option indeed.

For that matter, despite seeing the potential, I am one of those people who have not yet used them because I want to see how it works out in practice for a bit longer first.

But my dream is having a system where the only financial decisions we have to make in retirement are how to spend/give our wealth.

And it is entirely possible that at some point in my personal future, robos will play a large role in getting us as close as possible to that dream.

Very fair, totally see your point. I can see how for someone who has spent a long time DIY investing, robos can suddenly seem like a radical shift. Thankfully for me, I'm just starting out, and robos seem a lot less daunting than DIY.

[grabiner]

But my question is regarding this supposed tax-inefficiency of bonds being exaggerated.

I see it this way.

Bonds are taxed at your marginal rate. Let's say that is 28% or 32% which I think will also trigger the NITT at an additional 3.8%. The bond dividends are taxed each year - you cannot pick the year to take the income and you cannot pick when to pay the tax.

Stocks are taxed at the cap gains rate which would be 15% (and maybe the NITT 3.8%). Stock cap gains are mostly paid when/if you choose to sell shares.
.

With bonds you make a little money and pay tax at a high rate. With stocks you make more money (over the long run) and pay a lower rate of tax.

Taxable bonds in a taxable account do not sound very efficient to me.

However, there is another factor working the other way. Stocks lose a smaller percentage of their gains to taxes, but they have higher expected gains which may lead to a higher total tax cost.

With bond yields as high as they are now, stocks in a taxable account are likely to be more tax-efficient in any tax bracket in which taxable bonds in a taxable account are reasonable.

In a very high bracket, it might be better to hold bonds in taxable, but to hold them as munis. For example, investors in the top tax bracket in CA pay 20% federal + 3.8% NIIT + 12.3% CA = 36.1% total tax on qualified dividends and long-term gains. At that tax rate, they would likely be better off with CA munis in taxable and stocks in tax-deferred, rather than stocks in taxable and taxable bonds in tax-deferred.

[muffins14]

Hi everyone, I'm new to the sub and just recently set up my taxable situation with a low-fee robo advisor that utilities tax-free muni bonds, after learning about the tax-inefficiency of bonds within TDFs - I have them in all my tax-deferred accounts. I like having a mirrored allocation and being hands-off: which is true of both TDFs and robo-advisors.

However, I recently decided to run some numbers. I compared the after-distributions-tax performance of VT (which is where most TDFs start at 90%+ to equities) to Vanguard Target Retirement Income Fund (at 30/70, this is where most TDFs end). Looking at data over the last ~15 years, the tax-cost difference between them is only about 0.45-0.5%, which averages to about ~0.25%/yr over a target date fund's investing lifetime.

0.25%/yr: that's also coincidentally the cost of most robo-advisors. And genuinely doesn't sound like a life-changing difference, much closer to a rounding error. And this is at the very highest federal tax bracket, the actual cost for most of us is bound to be much lower.

So, I wanted to ask: Why is conventional wisdom against using bonds/TDFs in taxable if the extent of the inefficiency is so small? Especially compared to the simplicity and behavioural discipline that a TDF instills? (Let us disregard the issues with TDF and capital gains for a second, and purely look at tax-inefficiency of bonds in taxable)

My investing philosophy is that "More money is more than less money".

If your bonds are taxed more when held in taxable than when they are held in tax-deferred, then you should hold them in tax-deferred.

Additionally, in retirement your withdrawals from pre-tax accounts are taxed at your marginal rate, whereas you get long-term capital gains treatment from selling in taxable. So, probably favors keeping lower-returning assets (bonds) in pre-tax vs post-tax.

even if it's 0.5% per year for the tax difference, that's a lot over 30 years

[retiredjg]

But my question is regarding this supposed tax-inefficiency of bonds being exaggerated.

I see it this way.

Bonds are taxed at your marginal rate. Let's say that is 28% or 32% which I think will also trigger the NITT at an additional 3.8%. The bond dividends are taxed each year - you cannot pick the year to take the income and you cannot pick when to pay the tax.

Stocks are taxed at the cap gains rate which would be 15% (and maybe the NITT 3.8%). Stock cap gains are mostly paid when/if you choose to sell shares.
.

With bonds you make a little money and pay tax at a high rate. With stocks you make more money (over the long run) and pay a lower rate of tax.

Taxable bonds in a taxable account do not sound very efficient to me.

However, there is another factor working the other way. Stocks lose a smaller percentage of their gains to taxes, but they have higher expected gains which may lead to a higher total tax cost.

True, but I'm not thinking in terms of total tax cost. To me, the "efficiency" part has to do with tax paid on a dollar of gain. Taxable bonds will trigger more tax on that dollar than LTCG.

In a very high bracket, it might be better to hold bonds in taxable, but to hold them as munis. For example, investors in the top tax bracket in CA pay 20% federal + 3.8% NIIT + 12.3% CA = 36.1% total tax on qualified dividends and long-term gains. At that tax rate, they would likely be better off with CA munis in taxable and stocks in tax-deferred, rather than stocks in taxable and taxable bonds in tax-deferred.

Agreed. You've been making this argument for quite awhile and have made a believer out of me for CA and maybe another state or two. However, I would not want all my bonds in munis even if more tax-efficient. When a choice is available, I like your "half muni, half of which is CA" idea quite a lot though.

[CarefullyCarele$$]

Hi everyone, I'm new to the sub and just recently set up my taxable situation with a low-fee robo advisor that utilities tax-free muni bonds, after learning about the tax-inefficiency of bonds within TDFs - I have them in all my tax-deferred accounts. I like having a mirrored allocation and being hands-off: which is true of both TDFs and robo-advisors.

However, I recently decided to run some numbers. I compared the after-distributions-tax performance of VT (which is where most TDFs start at 90%+ to equities) to Vanguard Target Retirement Income Fund (at 30/70, this is where most TDFs end). Looking at data over the last ~15 years, the tax-cost difference between them is only about 0.45-0.5%, which averages to about ~0.25%/yr over a target date fund's investing lifetime.

0.25%/yr: that's also coincidentally the cost of most robo-advisors. And genuinely doesn't sound like a life-changing difference, much closer to a rounding error. And this is at the very highest federal tax bracket, the actual cost for most of us is bound to be much lower.

So, I wanted to ask: Why is conventional wisdom against using bonds/TDFs in taxable if the extent of the inefficiency is so small? Especially compared to the simplicity and behavioural discipline that a TDF instills? (Let us disregard the issues with TDF and capital gains for a second, and purely look at tax-inefficiency of bonds in taxable)

My investing philosophy is that "More money is more than less money".

If your bonds are taxed more when held in taxable than when they are held in tax-deferred, then you should hold them in tax-deferred.

Additionally, in retirement your withdrawals from pre-tax accounts are taxed at your marginal rate, whereas you get long-term capital gains treatment from selling in taxable. So, probably favors keeping lower-returning assets (bonds) in pre-tax vs post-tax.

even if it's 0.5% per year for the tax difference, that's a lot over 30 years

Sure, my only argument would be that complexity has a very tangible cost for most people. The discipline that TDFs/Robos instill just by the very nature of their simplicity, can easily save much more than their cost in the long run. Simplicity is more money.