Tax loss harvesting - wash sale rule

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
iceport
Posts: 3952
Joined: Sat Apr 07, 2007 4:29 pm

Re: Tax loss harvesting - wash sale rule

Post by iceport » Fri Sep 14, 2018 3:11 pm

Nate79 wrote:
Fri Sep 14, 2018 2:21 pm
iceport wrote:
Fri Sep 14, 2018 1:21 pm
Nate79 wrote:
Fri Sep 14, 2018 9:56 am
I think the IRS originally got it right when they said two different mutual funds can not be substantially identical. They should have left that guidance in the documents and left it at that.

To broadly say to mutual funds that track the same index are substantially identical is false and we know this to be false. For those that actually read the case law review that was posted recently (I think in another thread) where it went thru the case law of SI for bonds there are clear examples that show how there could be examples of index funds tracking the same index that are not SI.

Consider these two examples:

1). Two S&P500 index funds but one has an ER of 0.04% and one has an ER of X% (let's say 0.5% or let's say 1%). There total returns are different. Are they SI?
2). Two S&P500 index funds, one pays their dividends quarterly and one pays dividends annually. Are they SI?
Wasn't that old guidance written before index funds were in common use?

Anyway, taking the letter of the law so far away from the spirit of the law in the two examples offered keeps me questioning why folks think they wrote the wash sale rules in the first place. If they wanted to close a tax haven for those holding the securities of the day — i.e. not mutual funds or ETFs — it wouldn't make sense to make it so easy to open it back up again.

This is a coupon-clipping interpretation of wash sales: Scan a free coupon at the cash register and you save money. Use your promo code and save. Exchange your losing index fund from Fund Company A for an index fund tracking exactly the same index from Fund Company B and, voila, with a few extra keystrokes the wash sale rules don't apply to you. (Except in the case of wash sales, there is no reason to promote tax avoidance.)

Maybe the 1921 equivalent would be if they let investors by-pass the wash sale rules by using one stock broker to sell the losing stock and a different stock broker to buy it back again.
No.
IRS publication 564, 2009. The spirit of the law was to prevent buying the exact same security to get a tax benefit.
Then the use of the qualifying term "substantially" (as in "substantially identical") is curious.
Nate79 wrote:
Fri Sep 14, 2018 2:21 pm
Did you read the case law review which was for bonds?
I'm not sure. Do you have a link?

I did find the following legal review and Kitces article (that you referenced before) to be quite informative in providing historic perspective.

New Twists on an Old Plot: Investors Look to Avoid the Wash Sale Rule by Harvesting Tax Losses with Exchange-Traded Funds

Does Tax Loss Harvesting “Almost” Substantially Identical Mutual Funds And ETFs Trigger A Wash Sale Problem?
"Discipline matters more than allocation.” ─William Bernstein

User avatar
fortyofforty
Posts: 1089
Joined: Wed Mar 31, 2010 12:33 pm

Re: Tax loss harvesting - wash sale rule

Post by fortyofforty » Fri Sep 14, 2018 3:12 pm

We should drop all the insults, no matter how softly phrased, since they are insults, nonetheless.

Here is a simple example. Imagine an index that tracks exactly one share each of exactly ten companies. It's the One-Ten Index.

We have Company A that charges .0501% fees, and Company B that charges .0500% fees, for their version of their One-Ten Index Fund.

Do you imagine you could buy 100 shares of Company A's One-Ten Index Fund, and 100 shares of Company B's One-Ten Index Fund, then, if the price drops, immediately sell all 200 shares, just swapping them out for each other, enjoying the tax loss without fear of the wash-sale rule?

If you say "no", then why not? Their expense ratios are different. Their total returns are different. Their ticker symbols are different. Their mailing addresses are different.

If you say "yes", then why? Are the underlying holdings not nearly (substantially) the same (identical)?

For me, I will not sell and buy any fund or ETF that tracks the same index until waiting the necessary time period, to avoid the wash-sale rule. I don't care what company manages it. I don't care what the expense ratio is. I don't care what it's called. If it tracks Index X, it is "substantially identical" in my mind, and I will conduct my financial transactions accordingly.
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

User avatar
iceport
Posts: 3952
Joined: Sat Apr 07, 2007 4:29 pm

Re: Tax loss harvesting - wash sale rule

Post by iceport » Fri Sep 14, 2018 3:20 pm

Well — and this is absolutely true! — I was just notified by a recorded voice message that the IRS has issued a warrant for my arrest and my property is being monitored. Crime just doesn't pay... :wink:
"Discipline matters more than allocation.” ─William Bernstein

Nate79
Posts: 3325
Joined: Thu Aug 11, 2016 6:24 pm
Location: Portland, OR

Re: Tax loss harvesting - wash sale rule

Post by Nate79 » Fri Sep 14, 2018 4:25 pm

iceport wrote:
Fri Sep 14, 2018 3:11 pm
Nate79 wrote:
Fri Sep 14, 2018 2:21 pm
iceport wrote:
Fri Sep 14, 2018 1:21 pm
Nate79 wrote:
Fri Sep 14, 2018 9:56 am
I think the IRS originally got it right when they said two different mutual funds can not be substantially identical. They should have left that guidance in the documents and left it at that.

To broadly say to mutual funds that track the same index are substantially identical is false and we know this to be false. For those that actually read the case law review that was posted recently (I think in another thread) where it went thru the case law of SI for bonds there are clear examples that show how there could be examples of index funds tracking the same index that are not SI.

Consider these two examples:

1). Two S&P500 index funds but one has an ER of 0.04% and one has an ER of X% (let's say 0.5% or let's say 1%). There total returns are different. Are they SI?
2). Two S&P500 index funds, one pays their dividends quarterly and one pays dividends annually. Are they SI?
Wasn't that old guidance written before index funds were in common use?

Anyway, taking the letter of the law so far away from the spirit of the law in the two examples offered keeps me questioning why folks think they wrote the wash sale rules in the first place. If they wanted to close a tax haven for those holding the securities of the day — i.e. not mutual funds or ETFs — it wouldn't make sense to make it so easy to open it back up again.

This is a coupon-clipping interpretation of wash sales: Scan a free coupon at the cash register and you save money. Use your promo code and save. Exchange your losing index fund from Fund Company A for an index fund tracking exactly the same index from Fund Company B and, voila, with a few extra keystrokes the wash sale rules don't apply to you. (Except in the case of wash sales, there is no reason to promote tax avoidance.)

Maybe the 1921 equivalent would be if they let investors by-pass the wash sale rules by using one stock broker to sell the losing stock and a different stock broker to buy it back again.
No.
IRS publication 564, 2009. The spirit of the law was to prevent buying the exact same security to get a tax benefit.
Then the use of the qualifying term "substantially" (as in "substantially identical") is curious.
Nate79 wrote:
Fri Sep 14, 2018 2:21 pm
Did you read the case law review which was for bonds?
I'm not sure. Do you have a link?

I did find the following legal review and Kitces article (that you referenced before) to be quite informative in providing historic perspective.

New Twists on an Old Plot: Investors Look to Avoid the Wash Sale Rule by Harvesting Tax Losses with Exchange-Traded Funds

Does Tax Loss Harvesting “Almost” Substantially Identical Mutual Funds And ETFs Trigger A Wash Sale Problem?
No, this one:
http://g2ft.com/papers/G2FTAdvancedTopi ... lBonds.pdf

marcopolo
Posts: 1050
Joined: Sat Dec 03, 2016 10:22 am

Re: Tax loss harvesting - wash sale rule

Post by marcopolo » Fri Sep 14, 2018 5:09 pm

fortyofforty wrote:
Fri Sep 14, 2018 3:12 pm
We should drop all the insults, no matter how softly phrased, since they are insults, nonetheless.

Here is a simple example. Imagine an index that tracks exactly one share each of exactly ten companies. It's the One-Ten Index.

We have Company A that charges .0501% fees, and Company B that charges .0500% fees, for their version of their One-Ten Index Fund.

Do you imagine you could buy 100 shares of Company A's One-Ten Index Fund, and 100 shares of Company B's One-Ten Index Fund, then, if the price drops, immediately sell all 200 shares, just swapping them out for each other, enjoying the tax loss without fear of the wash-sale rule?

If you say "no", then why not? Their expense ratios are different. Their total returns are different. Their ticker symbols are different. Their mailing addresses are different.

If you say "yes", then why? Are the underlying holdings not nearly (substantially) the same (identical)?

For me, I will not sell and buy any fund or ETF that tracks the same index until waiting the necessary time period, to avoid the wash-sale rule. I don't care what company manages it. I don't care what the expense ratio is. I don't care what it's called. If it tracks Index X, it is "substantially identical" in my mind, and I will conduct my financial transactions accordingly.
I am not sure your analogy leads to your conclusion.

In your example, i would say it is a wash sale because the two funds do hold the same exact securities in the same exact proportions.
But, index funds do not work like. They typically sub-sample the index to get something that is "representative" of the index.
For example FSTVX says:
Normally investing at least 80% of assets in common stocks included in the Dow Jones U.S. Total Stock Market Index, which represents the performance of a broad range of U.S. stocks.
Another firm doing the same would be tracking the same index, but quite likely have a different set oif stock holdings, in different proportions, and still be representative of the index.

Even the funds themselves only claim to be representative, and not substantially identical to the index. Can two things (funds) that are only representative of, not identical to, a 3rd thing (index) somehow become substantially identical to each other?

You can argue about whether the fact that they have high correlation indicates they are "substantially identical", but i don't see how it could be argued that the underlying holdings are substantially identical.
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
fortyofforty
Posts: 1089
Joined: Wed Mar 31, 2010 12:33 pm

Re: Tax loss harvesting - wash sale rule

Post by fortyofforty » Fri Sep 14, 2018 7:24 pm

marcopolo wrote:
Fri Sep 14, 2018 5:09 pm
fortyofforty wrote:
Fri Sep 14, 2018 3:12 pm
We should drop all the insults, no matter how softly phrased, since they are insults, nonetheless.

Here is a simple example. Imagine an index that tracks exactly one share each of exactly ten companies. It's the One-Ten Index.

We have Company A that charges .0501% fees, and Company B that charges .0500% fees, for their version of their One-Ten Index Fund.

Do you imagine you could buy 100 shares of Company A's One-Ten Index Fund, and 100 shares of Company B's One-Ten Index Fund, then, if the price drops, immediately sell all 200 shares, just swapping them out for each other, enjoying the tax loss without fear of the wash-sale rule?

If you say "no", then why not? Their expense ratios are different. Their total returns are different. Their ticker symbols are different. Their mailing addresses are different.

If you say "yes", then why? Are the underlying holdings not nearly (substantially) the same (identical)?

For me, I will not sell and buy any fund or ETF that tracks the same index until waiting the necessary time period, to avoid the wash-sale rule. I don't care what company manages it. I don't care what the expense ratio is. I don't care what it's called. If it tracks Index X, it is "substantially identical" in my mind, and I will conduct my financial transactions accordingly.
I am not sure your analogy leads to your conclusion.

In your example, i would say it is a wash sale because the two funds do hold the same exact securities in the same exact proportions.
But, index funds do not work like. They typically sub-sample the index to get something that is "representative" of the index.
For example FSTVX says:
Normally investing at least 80% of assets in common stocks included in the Dow Jones U.S. Total Stock Market Index, which represents the performance of a broad range of U.S. stocks.
Another firm doing the same would be tracking the same index, but quite likely have a different set oif stock holdings, in different proportions, and still be representative of the index.

Even the funds themselves only claim to be representative, and not substantially identical to the index. Can two things (funds) that are only representative of, not identical to, a 3rd thing (index) somehow become substantially identical to each other?

You can argue about whether the fact that they have high correlation indicates they are "substantially identical", but i don't see how it could be argued that the underlying holdings are substantially identical.
The example most used here concerns S&P 500 Index Funds. The rule does not require "identical", only substantially so. Vanguard and Fidelity can and do hold every one of the stocks specified by Standard & Poors. The proportions might be off ever so slightly, but they are, for all intents and purposes, substantially identical. Something "less than precise correspondence" is required. The argument is easily made because it is true. Demonstrably, unarguably true, in fact.

I suggest you refer to Hanlin, in which the court ruled:
The wash sales provision is designed to eliminate fictitious losses. As losses are a matter of economics, so the fiction lies in the lack of any change in economic position on the part of the taxpayer.
Part of the difficulty with this issue is that the court ruling requires that you look inward for answers. If you are claiming "fictitious losses" by selling funds from one security that holds a bundle of underlying stocks and buying another that holds a substantially identical bundle of underlying stocks within 30 days, you are violating the wash sale rule. The rule doesn't say "do whatever you think you can get away with." Nor does it say "do whatever you want, because chances are you won't be caught."
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

marcopolo
Posts: 1050
Joined: Sat Dec 03, 2016 10:22 am

Re: Tax loss harvesting - wash sale rule

Post by marcopolo » Fri Sep 14, 2018 8:30 pm

fortyofforty wrote:
Fri Sep 14, 2018 7:24 pm
marcopolo wrote:
Fri Sep 14, 2018 5:09 pm
fortyofforty wrote:
Fri Sep 14, 2018 3:12 pm
We should drop all the insults, no matter how softly phrased, since they are insults, nonetheless.

Here is a simple example. Imagine an index that tracks exactly one share each of exactly ten companies. It's the One-Ten Index.

We have Company A that charges .0501% fees, and Company B that charges .0500% fees, for their version of their One-Ten Index Fund.

Do you imagine you could buy 100 shares of Company A's One-Ten Index Fund, and 100 shares of Company B's One-Ten Index Fund, then, if the price drops, immediately sell all 200 shares, just swapping them out for each other, enjoying the tax loss without fear of the wash-sale rule?

If you say "no", then why not? Their expense ratios are different. Their total returns are different. Their ticker symbols are different. Their mailing addresses are different.

If you say "yes", then why? Are the underlying holdings not nearly (substantially) the same (identical)?

For me, I will not sell and buy any fund or ETF that tracks the same index until waiting the necessary time period, to avoid the wash-sale rule. I don't care what company manages it. I don't care what the expense ratio is. I don't care what it's called. If it tracks Index X, it is "substantially identical" in my mind, and I will conduct my financial transactions accordingly.
I am not sure your analogy leads to your conclusion.

In your example, i would say it is a wash sale because the two funds do hold the same exact securities in the same exact proportions.
But, index funds do not work like. They typically sub-sample the index to get something that is "representative" of the index.
For example FSTVX says:
Normally investing at least 80% of assets in common stocks included in the Dow Jones U.S. Total Stock Market Index, which represents the performance of a broad range of U.S. stocks.
Another firm doing the same would be tracking the same index, but quite likely have a different set oif stock holdings, in different proportions, and still be representative of the index.

Even the funds themselves only claim to be representative, and not substantially identical to the index. Can two things (funds) that are only representative of, not identical to, a 3rd thing (index) somehow become substantially identical to each other?

You can argue about whether the fact that they have high correlation indicates they are "substantially identical", but i don't see how it could be argued that the underlying holdings are substantially identical.
The example most used here concerns S&P 500 Index Funds. The rule does not require "identical", only substantially so. Vanguard and Fidelity can and do hold every one of the stocks specified by Standard & Poors. The proportions might be off ever so slightly, but they are, for all intents and purposes, substantially identical. Something "less than precise correspondence" is required. The argument is easily made because it is true. Demonstrably, unarguably true, in fact.

I suggest you refer to Hanlin, in which the court ruled:
The wash sales provision is designed to eliminate fictitious losses. As losses are a matter of economics, so the fiction lies in the lack of any change in economic position on the part of the taxpayer.
Part of the difficulty with this issue is that the court ruling requires that you look inward for answers. If you are claiming "fictitious losses" by selling funds from one security that holds a bundle of underlying stocks and buying another that holds a substantially identical bundle of underlying stocks within 30 days, you are violating the wash sale rule. The rule doesn't say "do whatever you think you can get away with." Nor does it say "do whatever you want, because chances are you won't be caught."
While some of the examples did center around S&P500, the original post was about a Total Bond Market funds, which i suspect would be even less like each other than a total stock market fund, just because the market is much bigger.

As for the intent, how far do you think that goes? If i sell a large pharma company and then buy another large pharma company because i want to maintain similar exposure to that sector of the market, would you say that would also constitute a wash sale? If not, than why would it be any different if i did so with mutual funds (that holds clearly different securities, like two TSM funds) to maintain similar exposure to the broader market?
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
fortyofforty
Posts: 1089
Joined: Wed Mar 31, 2010 12:33 pm

Re: Tax loss harvesting - wash sale rule

Post by fortyofforty » Sat Sep 15, 2018 6:54 am

marcopolo wrote:
Fri Sep 14, 2018 8:30 pm
fortyofforty wrote:
Fri Sep 14, 2018 7:24 pm
marcopolo wrote:
Fri Sep 14, 2018 5:09 pm
fortyofforty wrote:
Fri Sep 14, 2018 3:12 pm
We should drop all the insults, no matter how softly phrased, since they are insults, nonetheless.

Here is a simple example. Imagine an index that tracks exactly one share each of exactly ten companies. It's the One-Ten Index.

We have Company A that charges .0501% fees, and Company B that charges .0500% fees, for their version of their One-Ten Index Fund.

Do you imagine you could buy 100 shares of Company A's One-Ten Index Fund, and 100 shares of Company B's One-Ten Index Fund, then, if the price drops, immediately sell all 200 shares, just swapping them out for each other, enjoying the tax loss without fear of the wash-sale rule?

If you say "no", then why not? Their expense ratios are different. Their total returns are different. Their ticker symbols are different. Their mailing addresses are different.

If you say "yes", then why? Are the underlying holdings not nearly (substantially) the same (identical)?

For me, I will not sell and buy any fund or ETF that tracks the same index until waiting the necessary time period, to avoid the wash-sale rule. I don't care what company manages it. I don't care what the expense ratio is. I don't care what it's called. If it tracks Index X, it is "substantially identical" in my mind, and I will conduct my financial transactions accordingly.
I am not sure your analogy leads to your conclusion.

In your example, i would say it is a wash sale because the two funds do hold the same exact securities in the same exact proportions.
But, index funds do not work like. They typically sub-sample the index to get something that is "representative" of the index.
For example FSTVX says:
Normally investing at least 80% of assets in common stocks included in the Dow Jones U.S. Total Stock Market Index, which represents the performance of a broad range of U.S. stocks.
Another firm doing the same would be tracking the same index, but quite likely have a different set oif stock holdings, in different proportions, and still be representative of the index.

Even the funds themselves only claim to be representative, and not substantially identical to the index. Can two things (funds) that are only representative of, not identical to, a 3rd thing (index) somehow become substantially identical to each other?

You can argue about whether the fact that they have high correlation indicates they are "substantially identical", but i don't see how it could be argued that the underlying holdings are substantially identical.
The example most used here concerns S&P 500 Index Funds. The rule does not require "identical", only substantially so. Vanguard and Fidelity can and do hold every one of the stocks specified by Standard & Poors. The proportions might be off ever so slightly, but they are, for all intents and purposes, substantially identical. Something "less than precise correspondence" is required. The argument is easily made because it is true. Demonstrably, unarguably true, in fact.

I suggest you refer to Hanlin, in which the court ruled:
The wash sales provision is designed to eliminate fictitious losses. As losses are a matter of economics, so the fiction lies in the lack of any change in economic position on the part of the taxpayer.
Part of the difficulty with this issue is that the court ruling requires that you look inward for answers. If you are claiming "fictitious losses" by selling funds from one security that holds a bundle of underlying stocks and buying another that holds a substantially identical bundle of underlying stocks within 30 days, you are violating the wash sale rule. The rule doesn't say "do whatever you think you can get away with." Nor does it say "do whatever you want, because chances are you won't be caught."
While some of the examples did center around S&P500, the original post was about a Total Bond Market funds, which i suspect would be even less like each other than a total stock market fund, just because the market is much bigger.

As for the intent, how far do you think that goes? If i sell a large pharma company and then buy another large pharma company because i want to maintain similar exposure to that sector of the market, would you say that would also constitute a wash sale? If not, than why would it be any different if i did so with mutual funds (that holds clearly different securities, like two TSM funds) to maintain similar exposure to the broader market?
As I wrote, you must look inward for answers.

If we are speaking of stock index funds, I provided hypothetical mutual funds for you to consider. It is helpful to do that, as a thought experiment, to better understand your perception of the underlying issues. Do you think tax loss harvesting by swapping shares between these two funds would be permissible or would that be a "wash sale"? To me, the answer is clear. If it is clear to you, too, then you have to begin working back from there. It is obviously not the "total return" difference alone, because that was covered. It is obviously not simply the expense ratios, because that, too, was covered. Is it merely that they own the same stocks, exactly? Maybe, but so do well-managed S&P 500 index funds, at least substantially. So, you are left with the totality of the features.

To explore your own example, if I sell one security and buy another security, those are by definition different. The companies have different market capitalizations, different product lines, different leadership, different debt structures, different products in development.... The list of differences is long. You cannot use a Preferred Stock and Common Stock, within certain parameters, if they are convertible (the court ruling goes into more detail). But, if you have a fund that owns 1999 stocks and another that owns 1999+1 stocks, with the last one being the lowest market capitalization, then those are certainly "substantially identical" because in almost every measurable respect they are the same.

To me, if I substitute one fund tracking Index X with another fund tracking the identical index, those are "substantially identical". If I want to capture a similar segment of equities, I will choose a different index. Simple.
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

Post Reply