Fund Expense Ratio Vs. Tax Expenses

[galacticb]

Hello everyone,

This may sound like a stupid question, so please let me apologize in advance if that is the case. While working on my overall portfolio strategy, I became curious about quantifying the relative trade-off between expense ratios and keeping funds in a tax-deffered account. Obviously, the ideal situation is to adjust your allocation so that you get the benefits of both worlds. However, suppose that you encounter a situation in which this is not possible.

Is there a way to quantify the added expense of holding a fund in a taxable account in a way that is comparable to an expense ratio? I'd like to be able to do something like the following example:

Fund A (Tax-Deffered Account)
Expense Ratio = 0.40%
Tax Penalty = 0.00%
Effective Expense Ratio = 0.40%

Fund B (Taxable Account)
Expense Ratio = 0.30%
Tax Penalty = 0.20%
Effective Expense Ratio = 0.50%

I didn't know if there was a general strategy that could be used based on publically available information such as a fund's turnover rate or annual dividend distributions.

Of course, this might be oversimplifying things, but I figured that it couldn't hurt to ask.


Sincerely,
GalacticB

[dbr]

You can go to Vanguard on a page like this and scroll down to "after tax returns"

https://personal.vanguard.com/us/funds/ ... st=tab%3A1

and you can read the note there:

https://personal.vanguard.com/us/conten ... rnsJSP.jsp

Or you can go to an M* page like this one and also read the data and the notes:

http://quicktake.morningstar.com/FundNe ... ountry=USA

[Rager1]

Hello everyone,

This may sound like a stupid question, so please let me apologize in advance if that is the case. While working on my overall portfolio strategy, I became curious about quantifying the relative trade-off between expense ratios and keeping funds in a tax-deffered account. Obviously, the ideal situation is to adjust your allocation so that you get the benefits of both worlds. However, suppose that you encounter a situation in which this is not possible.

Is there a way to quantify the added expense of holding a fund in a taxable account in a way that is comparable to an expense ratio? I'd like to be able to do something like the following example:

Fund A (Tax-Deffered Account)
Expense Ratio = 0.40%
Tax Penalty = 0.00%
Effective Expense Ratio = 0.40%

Fund B (Taxable Account)
Expense Ratio = 0.30%
Tax Penalty = 0.20%
Effective Expense Ratio = 0.50%

I didn't know if there was a general strategy that could be used based on publically available information such as a fund's turnover rate or annual dividend distributions.

Of course, this might be oversimplifying things, but I figured that it couldn't hurt to ask.


Sincerely,
GalacticB

Here's how I do it.

At the end of each year, I calculate our portfolio weighted expense ratio by multiplying each account's value at year-end by the fund's latest reported expense ratio. Summing the expense ratio cost for each fund, divided by the portfolio value at year-end, gives a weighted expense ratio for the portfolio at year-end.

For the tax calculation, after completing our tax returns, I delete all schedule B and schedule D entries and save that return with another name. By comparing the difference in taxes for federal and state between the 2 returns, I can see the tax cost associated with the portfolio for that year. Dividing the difference in tax cost by the portfolio value at year-end gives the tax cost ratio. (Hint: the larger number is always the tax cost ratio).

Ed

[galacticb]

Thanks for the useful links and analysis!

~ GalacticB

[xerty24]

But computing your realized taxes isn't the correct comparison.

Fund A, high-yield dividend fund that pays 5% per year, will cause you to pay a lot of taxes on those dividends (especially next year!). On the other hand, Fund B is a no-dividend fund that pays 0% per year and retains all those payments as unrealized capital gains. Let's suppose that both funds have the same total return pre-tax. If it helps, think about both funds earning 5% where A pays it out each year while B does not.

You need to compare the tax-deferral associated with the higher unrealized capital gains in fund B, vs the taxes paid (and not deferred) in fund A. This will depend on your dividend-vs-capital gain tax rates, your holding period, the relative dividend yield (and more generally turnover-inducing capital gains), etc.

[galacticb]

You need to compare the tax-deferral associated with the higher unrealized capital gains in fund B, vs the taxes paid (and not deferred) in fund A. This will depend on your dividend-vs-capital gain tax rates, your holding period, the relative dividend yield (and more generally turnover-inducing capital gains), etc.

Would Morningstar's Tax Cost Ratio effectively deal with this difference? (Obviously, you are going to have to pay taxes on the net gains posted by the fund whenever you liquidate your account. However, does this metric capture the extra loss that is experienced by having to realize capital gains and dividends earlier in the investment cycle?)

~ GalacticB

[dbr]

You need to compare the tax-deferral associated with the higher unrealized capital gains in fund B, vs the taxes paid (and not deferred) in fund A. This will depend on your dividend-vs-capital gain tax rates, your holding period, the relative dividend yield (and more generally turnover-inducing capital gains), etc.

Would Morningstar's Tax Cost Ratio effectively deal with this difference? (Obviously, you are going to have to pay taxes on the net gains posted by the fund whenever you liquidate your account. However, does this metric capture the extra loss that is experienced by having to realize capital gains and dividends earlier in the investment cycle?)

~ GalacticB

After tax cost and/or tax cost ratio are simplistic metrics for describing the tax properties of an investment. Such metrics don't even begin to address the issue of building a lifelong tax optimization model, should you imagine that such a thing is even practical.

There are investment and cash flow models such as i-orp that do include tax consequences of different financial planning scenarios. You could look there for some insight.

[livesoft]

One should also note that the M* tax analysis assumes the highest tax bracket. You really need to understand your personal situation before you go overboard with the data which may not apply to you.

[galacticb]

One should also note that the M* tax analysis assumes the highest tax bracket. You really need to understand your personal situation before you go overboard with the data which may not apply to you.

It sounds like the phrase "enough knowledge to be dangerous" applies here. I will fall back to a more personal analysis of my tax situation.

Thanks again for the replies.

~ GalacticB

[Shawn]

What I do is compute an "effective yield" for different scenario's. This yield is obtained by considering all factors that affect the total return after a given number of years (e.g., 30 years, as a long-term investor). These factors include expense ratios and the various taxes (income, dividends, capital gains) that reduce the actual return. Obviously, most of the factors can be only estimated since one doesn't know future tax rates or market conditions.

For example, if an employee receives $1 in 2010, one can compute how this dollar behaves from the day it is received to the day it becomes available for spending in 2040. One might compute that this dollar will grow to $10 if put in a tax differed 401k (effective yield 8.0%), or to $8.75 if put in a taxable investment account (effective yield 7.5%). In this case, it's better to put the dollar in the retirement plan. On the other hand, perhaps the costs/fees are very high in the retirement plan and the calculations show that the $1 will grow to only $7.60 after 30 years (effective yield 7.0%). In this case, it's better to put it in the taxable investment. I use effective yield as opposed to total return because it is more intuitive (IMO), and it's easier to compare alternate scenarios (e.g., different investment timescales).

I've found this to be the best way to do "apples to apples" comparisons since it essentially tracks the return of a dollar from when it is received (e.g., as income) to when it is available for spending (e.g., as retirement income). (Note that the first thing that happens to a dollar invested in a taxable account is that it is taxed as income, and so initially it goes from being worth $1 to being worth perhaps $0.75 - initially the return and yield are actually negative.) While the actual returns can be only estimated due to unknown future conditions, this still allows one to test different scenarios such as your Fund A and Fund B examples. Unfortunately, while the equations are not overly complex, many of them are not trivial (e.g., changing basis due to annual taxes on reinvested dividends in taxable accounts).

Nitpick. In your example for Fund A, I don't believe it is correct to say that the Tax Penalty is 0.00%. There is a tax penalty, and it is likely to be significant. This tax penalty is differed, but it can't be ignored. This is why it is important to consider the entire lifespan of the funds while they are in your control.

[galacticb]

Nitpick. In your example for Fund A, I don't believe it is correct to say that the Tax Penalty is 0.00%. There is a tax penalty, and it is likely to be significant. This tax penalty is differed, but it can't be ignored. This is why it is important to consider the entire lifespan of the funds while they are in your control.

Hi Shawn, thanks for the reply.

I recognize that there is certainly a tax penalty in both scenarios. What I really was trying to quantify was the *additional* penalty that will be incurred specifically by the inability to defer earnings (e.g. dividends and capital gains distributions) of the taxable fund until the money is withdrawn. I am even willing to ignore the benefits gained from not paying tax on either investment principle (Traditional 401k/IRA) or withdrawls (Roth 401k/IRA), as I would hold my investment ratio between retirement and taxable accounts constant.

Maybe a slightly different (albeit hypothetical) example would help to clarify. Suppose that I plan to invest in both a Total Stock Market fund and an REIT fund. Furthermore, assume that my total contribution exceeds my maximum 401k contribution. For simplicity, pretend that IRAs don't exist and that my 401k plan provides two funds:

Total Stock Market Index Fund (expense ratio 0.1%)
REIT Index Fund (expense ratio 2.0%)

The expense ratios clearly suggest that the TSM fund is the better option for the 401k. However, REIT funds are less tax efficient than TSM funds. I was simply trying to find a way to identify at what point the expense ratio outweighed the tax consequences of holding a less tax-efficient fund in a taxable account.

It was more of a curiosity question than anything. Sorry for all of the confusion.

~ GalacticB

[Shawn]

The expense ratios clearly suggest that the TSM fund is the better option for the 401k. However, REIT funds are less tax efficient than TSM funds. I was simply trying to find a way to identify at what point the expense ratio outweighed the tax consequences of holding a less tax-efficient fund in a taxable account.

It was more of a curiosity question than anything. Sorry for all of the confusion.

Yes, you were clear in your original post. I guess my point was that to fully assess the trade-off between tax consequences and expense ratios, it is often necessary to consider the complete life-cycle of the funds. Depending on the specific circumstances, the result may not be linear. I don't believe there is a simple rule of thumb, although in general I believe a reduction in ER can be more beneficial than modest tax-efficiency.

I haven't looked at the scenario you describe above (TSM vs REIT) but I have considered a related situation. Assume that the desired allocation is 100% TSM and there is equal space in the taxable and tax-differed accounts. Also assume that the dividend yield on TSM is 2%, and that the ER is 0.1 in both accounts. The easiest thing to do would be to put TSM in taxable and TSM in tax-differed and be done with it. The 2% dividends in the taxable account will be taxed on an annual basis but these same dividends will be tax-differed in the retirement account.

An alternate approach would be to put TSM Growth (1% dividend yield) in taxable and TSM Value (3% dividend yield) in tax-differed. The combined allocation still would be TSM (Growth + Value = TSM). The reason for doing this would be to take advantage of the fact that Value will be somewhat less tax efficient than Growth and hence will "do better" in a tax-differed account. Using this strategy, one will pay annual taxes on the 1% dividend yield whereas the 3% dividend yield will be tax-differed. If the ER of the Growth and Value funds is the same as TSM (0.1), then this half-and-half approach likely will be better assuming a relatively long investment horizon and similar current and future tax rates. (If applicable, one must also consider the impact of qualified dividends in taxable vs fully taxed dividends in the tax-differed retirement account).

However, ER's on funds that represent the Growth and Value sides of TSM are likely to be higher than a pure TSM fund. The question is, at what point does the higher ER's exceed the tax benefit of this half-and-half investment approach? It turns out that for the scenario of interest to me (my tax rates; 30 year investment period), if the ER's of Growth and Value are only slightly higher (~0.1) than the ER of TSM, then the increased ER's overwhelm the tax efficiency of the half-and-half approach.

In other words, a relatively small change in ER (e.g., 0.1) can easily wipe out the benefits of a tax efficient strategy. Of course, this is only one example where the tax-efficiency of one option isn't that much greater than the other option. It would take a significantly greater ER difference to wipe out the tax-efficiency of having TSM in taxable and REIT in tax-differed (but I haven't quantified this). Still, the conclusion is that seeking low ER's may be more beneficial than trying to use an alternate strategy that is more complex but only moderately more tax efficient.

Sorry for the long explanation. I can be verbose.

[galacticb]

Thanks for the reply Shawn. I appreciate you taking the time to type out that explanation. It was a great example! I guess the moral of the story is that tax trade-offs are very situation-specific and depend on one's personal investment objectives.

Thanks again for everyone's help.

~ GalacticB