Why do you think the Vanguard Target Fund is so tax inefficient

[gavinsiu]

So in morningstar, you can examine the price tab and get the tax efficiency. If we examine the tax efficiency of Target 2050

Vanguard Target 2050 VFIFX

The tax cost ratio is listed as 1.37%. According to morningstar, this mean you return is reduced by 1.37% if you are at the highest tax bracket. If we look at a bond fund like total bond market.

Vanguard Total Bond Market VBTLX

We see that Total bond Market has a tax cost ratio of 1.21%. Wait, this mean Target fund is worse than bond in tax efficiency. Let's look at a competitor

Fidelity Freedom Index 2025

The tax ratio is only 0.59%, this is comparable to a total stock market nidex which is at about 0.37%. The question is why the vanguard tax ratio is so high. Do you think this has to do with Vanguard's lawsuit with Target funds? Did that occur within the last 3 years?

[rkhusky]

There was a large cap gain for VFIFX in 2021. Since then, cap gains are back to normal.

[bonesly]

The question is why the vanguard tax ratio is so high. Do you think this has to do with Vanguard's lawsuit with Target funds? Did that occur within the last 3 years?

The lawsuit isn't the root cause; it's a by-product of the internal fund churn caused by Van introducing a lower-cost institutional class of their TDFs, so when all the big-balance institutional investors jumped ship to the new lower class, it caused churn in the investor class, which in turn created a huge tax-efficiency issue (large enough for a lawsuit to proceed). Until that one-time churn event is off the radar of Morningstar's look-back period for calculating tax-efficiency of funds, it's going to look like Van's funds are not that tax-efficient.

-----
Edit: That churn was in 2020, so shouldn't be on Morningstar's 3-year tax look-back period.

Van 2050 is 1.37 but the category average is 1.75. When we look at it is 0.59, so I think the reference to Fido 2025 is just a type (that is 2050).

It's still puzzing because both the Van 2050 and the Fido 2050 have an 18% "Potential Capital Gains Exposure," so you'd think that by M*'s description they'd be the same unless Van has had higher capital appreciation in the 3y window than Fido (but then if that were true, Van's PCGE would be higher than 18%, not the same as Fido 2050).

Their formula is Tax Ratio = 1 - (1+After-Tax Return) / (1+Pre-Tax Return) per their white paper.

M* has been known to post data that doesn't match the data from the brokerage sites, so it's worth looking at the actual distributions from the brokerage sites (without selling shares) over 3y for each fund and verifying, but an exercise for another day.
-----

As an aside, I would stick with similar TDFs so rather than Van 2050, I'd suggest Van 2025 to compare to Fido 2025 (probably doesn't change the conclusion, but it's an inconsistency in analysis that's easily fixed).

I'm also not sure how VBTLX has a cost ratio of only 1.21% when there's no expected price appreciation and all of the distributions are taxable in a Taxable account (which is the only type of account that tax-efficiency matters in).

[henry]

The question is why the vanguard tax ratio is so high. Do you think this has to do with Vanguard's lawsuit with Target funds? Did that occur within the last 3 years?

Could be. Those large distributions came at the end of 2021. I suppose that tax cost ratio is based on the last 3 full years available which would be 2021, 2022, and 2023, although I don't know for sure.

[Geologist]

If you search on this board, you will find threads that describe an important weakness of the Morningstar tax cost analysis is that it often doesn't recognize that funds are paying qualified dividends (that is, it treats all the dividends as non-qualified and therefore having a higher tax cost). I'm not sure how easy it is to determine how Morningstar is doing this for any particular fund.

[gavinsiu]

I'm also not sure how VBTLX has a cost ratio of only 1.21% when there's no expected price appreciation and all of the distributions are taxable in a Taxable account (which is the only type of account that tax-efficiency matters in).

I am curious, too. but the fidelity total index is even lower.

[gavinsiu]

OK, I figure out the reason why the tax ratio for bonds are so low. The tax ratio is subtracted from the return, so let's say you say the tax ratio is 1%, you subtract that from your return, so if you fund return 10%, then a 1% tax ratio means your return is reduce to 9%.

This mean you can't compare tax between different asset classes. Let's say you have a 1% on a bond fund that return 4%, hat 1% just removed 25% of your return. Let's say it's 1% on a stock fund that return 7%, then 1% is 14% of return.

It is good for comparing similar funds. for example fidelity 2050 is pretty smilar to Vanguard target 2050, so you can use the tax ratio to compare the two funds.

[bonesly]

...let's say you say the tax ratio is 1%, you subtract that from your return, so if you fund return 10%, then a 1% tax ratio means your return is reduce to 9%.

That doesn't seem to match up with the definition in M*'s white paper...
Their formula is Tax Ratio = 1 - (1+After-Tax Return) / (1+Pre-Tax Return) per their white paper.

That formula has raw return in both the numerator and denominator so different asset classes should work.

If VTSAX has an expected return of 11.5% (historical avg from 1928-2017) and taxable LTCG distribution of 1.2%, which is taxed at 15% bracket, then the pre-tax return is 11.5% and the after-tax return is 11.5% - (1.25% x 15%) = 11.32%. Remember, no shares were sold... it's tax-efficiency of just holding it without selling.

So M*'s Tax Ratio = 1 - (1+11.32%) / (1+11.5%) = 1.44%

If the VBTLX has an expected return of 5.2% (historical avg for 10y T-Notes from 1928-2017) and, for argument, 100% of the return is bond interest that's taxable at 22%, then the pre-tax return is 5.2% and the after-tax return is 5.2% x (1 - 22%) = 4.06%.

So M*s' Tax Ratio = 1 - (1+4.06%) / (1+5.2%) = 18.39% (i.e., it's not very tax-efficient in a Taxable account).

That's why I'm so puzzled that they say this value is 1.21% for VBTLX and not some thing in the 15-20% range (Tax Ratio would be almost 31% for the 37% tax bracket compared to 22% example).

[gavinsiu]

...let's say you say the tax ratio is 1%, you subtract that from your return, so if you fund return 10%, then a 1% tax ratio means your return is reduce to 9%.

That doesn't seem to match up with the definition in M*'s white paper...
Their formula is Tax Ratio = 1 - (1+After-Tax Return) / (1+Pre-Tax Return) per their white paper.

That formula has raw return in both the numerator and denominator so different asset classes should work.

If VTSAX has an expected return of 11.5% (historical avg from 1928-2017) and taxable LTCG distribution of 1.2%, which is taxed at 15% bracket, then the pre-tax return is 11.5% and the after-tax return is 11.5% - (1.25% x 15%) = 11.32%. Remember, no shares were sold... it's tax-efficiency of just holding it without selling.

So M*'s Tax Ratio = 1 - (1+11.32%) / (1+11.5%) = 1.44%

If the VBTLX has an expected return of 5.2% (historical avg for 10y T-Notes from 1928-2017) and, for argument, 100% of the return is bond interest that's taxable at 22%, then the pre-tax return is 5.2% and the after-tax return is 5.2% x (1 - 22%) = 4.06%.

So M*s' Tax Ratio = 1 - (1+4.06%) / (1+5.2%) = 18.39% (i.e., it's not very tax-efficient in a Taxable account).

That's why I'm so puzzled that they say this value is 1.21% for VBTLX and not some thing in the 15-20% range (Tax Ratio would be almost 31% for the 37% tax bracket compared to 22% example).

The explanation give was something like this;

For example, if a fund had a 2% tax cost ratio for the three-year time period, it
means that on average each year, investors in that fund lost 2% of their assets to
taxes. If the fund had a three-year annualized pre-tax return of 10%, an investor in
the fund took home about 8% on an after-tax basis. (Because the returns are
compounded, the after-tax return is actually 7.8%.)

I didn't work out the formular like you did.

[grabiner]

OK, I figure out the reason why the tax ratio for bonds are so low. The tax ratio is subtracted from the return, so let's say you say the tax ratio is 1%, you subtract that from your return, so if you fund return 10%, then a 1% tax ratio means your return is reduce to 9%.

This mean you can't compare tax between different asset classes. Let's say you have a 1% on a bond fund that return 4%, hat 1% just removed 25% of your return. Let's say it's 1% on a stock fund that return 7%, then 1% is 14% of return.

It is good for comparing similar funds. for example fidelity 2050 is pretty smilar to Vanguard target 2050, so you can use the tax ratio to compare the two funds.

However, it's just as useful for comparing two funds in a portfolio. If you hold equal amounts in these two funds, one in tax-deferred and one in taxable, you will lose the same amount to taxes regardless of which fund is in your taxable account. (In that case, it would be better to have the bond fund in the taxable account, because the stock fund will have an additional tax cost for capital-gains tax when you sell the shares.)

A more important issue is that the tax costs are computed assuming the highest federal tax and no state tax. This makes the tax cost on bond funds unrealistic, because investors in the top tax bracket would hold muni funds, not taxable bond funds, in a taxable account. (My rule of thumb is that the effective tax cost of a muni fund is 1/3 of the yield; that is, a muni fund yielding 3% has about the same risk as a taxable fund yielding 4%, so you should prefer the muni fund if your tax rate on bond interest is more than 25%.)

[afan]

Big picture, I assume it is tax inefficient because it has been created to hold in tax favored accounts. As people found out, holding it in taxable accounts can be a mistake.