Is Tax loss harvesting worth it?

[alfaspider]

Of course, but $3k ordinary deduction for capital losses makes it better than a pure delay. Plus, money has time value. A 20-year delay in tax is equivalent to a reduction of more than half (depending on what your discount rate is). TLH is especially useful for accumulators who are using a taxable account for tax-advantaged overflow. They may not realize the gains for more than 20 years, and by managing tax lots plus avoiding spending your taxable account to zero, you may be able to defer forever and let your heirs get the basis step up.

Those are valid points. But if your money is invested in the stock market "the time value of money" is based on the fact that stocks generally go up over time. By tax loss harvesting and resetting your cost basis to a lower level that same mechanism is working against you. It means that eventually (assuming you sell at some point) your taxable capital gain is greater than it would have been if you had not tax loss harvested. Figuring out the math on whether it's worth it would require precise knowledge of what is going to happen to the value of your stocks between the time you TLH and the time you sell those shares. In other words, you would have to accurately predict the future of the stock market.

No, the time value of money is working for you because you now have money for additional investments that would otherwise just be sitting in your tax basis. Let's put some numbers on it:

You have $10,000 invested in year 1. In year 2, you TLH when the value drops to $7,000. Assume for simplicity your marginal tax rate is 33%. For the $3,000 tax loss, you get $1,000 cash benefit in year 2, which you put into the same investment. Now, you effectively have $8,000 compounding instead of $7,000.

Now, assume the market has returned roughly historical amounts after the initial fall and your investment doubled from year 2 to year 10. With TLH, you now have $16,000 with $7,000 basis. If you did not TLH, you only have $14,000 with $10,000 basis. Now suppose you sold in year 10. With TLH, you recognize $9,000 of gain. At 20% capital gains rate, you have $14,200 after tax. Without TLH, you have $4,000 in gain leaving you with $13,200 after tax.

So under this scenario, you come out $1,000 ahead after selling even though your basis went down when you did the TLH transaction. The difference becomes more substantial if you are talking 20 or 30 years. It doesn't really matter how much gain you have for this to work, but more gain equals more benefit. I suppose you come out behind if you have long-term losses, but if you are looking at losses over 20 or 30 years, taxes are probably not your primary problem and you would have been better in cash the whole time.

[Artsdoctor]

This is very common. If you take a look at national statistics, the majority of investors have their investment portfolio in their retirement accounts.

All of that said, there are some distinct advantages of having at least a small amount of investments in a taxable account if possible. It can make charitable giving extremely tax-efficient, and you can use carryover losses even if the amount is relatively small.

I have a good amount in taxable, but most of it is long term gains. I also continue making taxable contributions but they’re smaller than my 401k

TLH depends heavily on new contributions, the longer you are invested.

Just highlighting that TLH generally tends to require a fairly significant correction for me to take advantage of.

Totally agree. It can truly be hard to find losing assets if you've been growing them for many years and nothing new has been added. From a tax perspective, it's usually more efficient to rebalance in a tax-advantaged account although when the opportunity arises in a taxable account, I'll take it. All of that said, tax-loss harvesting in the early stages of taxable investing can be carried forward for many years, even decades.

[LilyFleur]

It is so individual.

A few years back, I sold a position in Verizon and tax loss harvested almost $3,000 exactly. It was a year that income increased significantly because of a Roth conversion. It helped on my tax bill, but not hugely.

For many investors, Michael Kitces is not wrong.

If you take a tax loss of $3,000 but you sell a house in San Francisco and are taxed on $500,000 of capital gains over the exemption, $3,000 is not going to move the dial much.

It all really depends, and I have found in retirement that my tax strategy varies a lot depending on the year.

It is a great tool to have in your back pocket, I think.

[toddthebod]

If you take a tax loss of $3,000 but you sell a house in San Francisco and are taxed on $500,000 of capital gains over the exemption, $3,000 is not going to move the dial much.

That's still $600-$700 in tax savings today that you won't have to pay back for decades if ever for about 5 minutes of effort. Is there a better $/hr investment on the planet?

[Artsdoctor]

It is so individual.

A few years back, I sold a position in Verizon and tax loss harvested almost $3,000 exactly. It was a year that income increased significantly because of a Roth conversion. It helped on my tax bill, but not hugely.

For many investors, Michael Kitces is not wrong.

If you take a tax loss of $3,000 but you sell a house in San Francisco and are taxed on $500,000 of capital gains over the exemption, $3,000 is not going to move the dial much.

It all really depends, and I have found in retirement that my tax strategy varies a lot depending on the year.

It is a great tool to have in your back pocket, I think.

This is where people fall off the cliff in recognizing the value of TLH: the house sale. The $250,000/$500,000 exemption has not been indexed to inflation. So when it was originally instituted, few people thought they'd go over the exemption. It's true that a lot of people move every 7-10 years; however, there's still quite a few people who stay in their houses for 20-30+ years. That's a lot of equity, especially in high cost of living areas. If you've been diligently TLHing during bear markets, you might have carryover losses which are substantial. And in California, those losses can be carried over from year to year, unlike some states. Those losses can really make a huge difference when you ultimately sell a house with substantial gains because those gains could easily be taxed at a marginal rate of over 30% or so (federal and state).

So if you really think you're going to be staying in your house for a "very long time" and real estate prices are moderate to high, keep those renovation receipts and tax-loss harvest when appropriate.

[LilyFleur]

It is so individual.

A few years back, I sold a position in Verizon and tax loss harvested almost $3,000 exactly. It was a year that income increased significantly because of a Roth conversion. It helped on my tax bill, but not hugely.

For many investors, Michael Kitces is not wrong.

If you take a tax loss of $3,000 but you sell a house in San Francisco and are taxed on $500,000 of capital gains over the exemption, $3,000 is not going to move the dial much.

It all really depends, and I have found in retirement that my tax strategy varies a lot depending on the year.

It is a great tool to have in your back pocket, I think.

This is where people fall off the cliff in recognizing the value of TLH: the house sale. The $250,000/$500,000 exemption has not been indexed to inflation. So when it was originally instituted, few people thought they'd go over the exemption. It's true that a lot of people move every 7-10 years; however, there's still quite a few people who stay in their houses for 20-30+ years. That's a lot of equity, especially in high cost of living areas. If you've been diligently TLHing during bear markets, you might have carryover losses which are substantial. And in California, those losses can be carried over from year to year, unlike some states. Those losses can really make a huge difference when you ultimately sell a house with substantial gains because those gains could easily be taxed at a marginal rate of over 30% or so (federal and state).

So if you really think you're going to be staying in your house for a "very long time" and real estate prices are moderate to high, keep those renovation receipts and tax-loss harvest when appropriate.

So let's say I tax loss harvest and have $3,000 a year of losses to put on my yearly tax returns, for ten years. In year 5 of this process, I sell a home in California and have $100,000 in taxable capital gains (after subtracting the cost basis, improvements, selling costs, capital gains exemption, etc.) Don't the capital gains taxes have to be paid the year that the home is sold? So, I could take $3,000 off my taxable income in the year I am taxed on capital gains of $100,000 for the sale of my home?

In California, all capital gains are taxed at the regular income tax rate by the state, unlike federal taxes which are taxed at the capital gains rate. I suppose selling a home in the same year as incurring significant medical deductions for assisted living or a nursing home would help on those taxes.

These numbers are totally made up, by the way.

[bling]

This is where people fall off the cliff in recognizing the value of TLH: the house sale. The $250,000/$500,000 exemption has not been indexed to inflation. So when it was originally instituted, few people thought they'd go over the exemption. It's true that a lot of people move every 7-10 years; however, there's still quite a few people who stay in their houses for 20-30+ years. That's a lot of equity, especially in high cost of living areas. If you've been diligently TLHing during bear markets, you might have carryover losses which are substantial. And in California, those losses can be carried over from year to year, unlike some states. Those losses can really make a huge difference when you ultimately sell a house with substantial gains because those gains could easily be taxed at a marginal rate of over 30% or so (federal and state).

So if you really think you're going to be staying in your house for a "very long time" and real estate prices are moderate to high, keep those renovation receipts and tax-loss harvest when appropriate.

So let's say I tax loss harvest and have $3,000 a year of losses to put on my yearly tax returns, for ten years. In year 5 of this process, I sell a home in California and have $100,000 in taxable capital gains (after subtracting the cost basis, improvements, selling costs, capital gains exemption, etc.) Don't the capital gains taxes have to be paid the year that the home is sold? So, I could take $3,000 off my taxable income in the year I am taxed on capital gains of $100,000 for the sale of my home?

In California, all capital gains are taxed at the regular income tax rate by the state, unlike federal taxes which are taxed at the capital gains rate. I suppose selling a home in the same year as incurring significant medical deductions for assisted living or a nursing home would help on those taxes.

These numbers are totally made up, by the way.

you can carry any unused losses forward indefinitely though. so yes, you pay taxes on capital gains of the year you realize them, but you can offset those gains from any unused losses you've had over the years.

[toddthebod]

This is where people fall off the cliff in recognizing the value of TLH: the house sale. The $250,000/$500,000 exemption has not been indexed to inflation. So when it was originally instituted, few people thought they'd go over the exemption. It's true that a lot of people move every 7-10 years; however, there's still quite a few people who stay in their houses for 20-30+ years. That's a lot of equity, especially in high cost of living areas. If you've been diligently TLHing during bear markets, you might have carryover losses which are substantial. And in California, those losses can be carried over from year to year, unlike some states. Those losses can really make a huge difference when you ultimately sell a house with substantial gains because those gains could easily be taxed at a marginal rate of over 30% or so (federal and state).

So if you really think you're going to be staying in your house for a "very long time" and real estate prices are moderate to high, keep those renovation receipts and tax-loss harvest when appropriate.

So let's say I tax loss harvest and have $3,000 a year of losses to put on my yearly tax returns, for ten years. In year 5 of this process, I sell a home in California and have $100,000 in taxable capital gains (after subtracting the cost basis, improvements, selling costs, capital gains exemption, etc.) Don't the capital gains taxes have to be paid the year that the home is sold? So, I could take $3,000 off my taxable income in the year I am taxed on capital gains of $100,000 for the sale of my home?

In California, all capital gains are taxed at the regular income tax rate by the state, unlike federal taxes which are taxed at the capital gains rate. I suppose selling a home in the same year as incurring significant medical deductions for assisted living or a nursing home would help on those taxes.

These numbers are totally made up, by the way.

In year 1 you realized $30,000 of losses, and you have no capital gains and will have none for the remainder of this example except the house sale.
Year 1: Deduct $3,000 from income.
Year 2: Deduct $3,000 from income.
Year 3: Deduct $3,000 from income.
Year 4: Deduct $3,000 from income.
Year 5: Sell house, realize $100,000 taxable gain. Subtract $18,000 of carryover losses. Pay taxes on only $82,000 of gains.
You have no further losses to carryover to years 6 and beyond. You also have no losses to deduct from income in year 5, as they are used up by the realized gains first.

[Nathan Drake]

Another issue with Tax Loss Harvesting - Bid/Ask Spread, Fund Performance Drift, etc.

You are incurring a cost with selling assets through the Bid/Ask spread. Similarly, if you sell VTI to by VOO, it's possible that the two could deviate quite a bit in just 30 days to the tune of 1% or more. At what volume of stock lot does it make sense to sell?

Analyzing Transaction / Opportunity Costs:
A $20,000 sell order to generate a $3,000 loss could potentially drift 1%, or $200 after 30 days vs. your target fund. A bid-ask spread cost may be 3 basis points. Multiply that by 4 transactions (Sell-Buy day 1, Sell-Buy Day 31) and that's 12 basis points. Or $24. Not terrible, but if you have more exotic funds like AVUV -> DFSV, that bid-ask spread could be 10-20 basis points per transaction, or potentially 80 basis points for $168.

Suddenly, your total cost using a more exotic fund is $200 + $168 = $368. Your tax savings on $3,000 loss @ 24% bracket would be about $720.

So $720 - $368 = $352 now. Seems much more minor, but you'd never really know this impact because the costs are all basically hidden and harder to accurately track.

Of course this doesn't even consider potentially executing a trade sub-optimally. At least at Fidelity, I can't immediately "Sell to Buy" an ETF with extremely quick precision. So having to do two separate transactions may take a minute or two, and by that time the market has potentially moved against you. The less large the position you are selling is, the higher the benefit. $10,000 in sales for a $3k loss is much more efficient than $20,000 in sales for a $2k loss, but the former situation is much more rare.

That's why it almost seems like TLH only makes sense during a 20% drop scenario or more.

[grabiner]

Another issue with Tax Loss Harvesting - Bid/Ask Spread, Fund Performance Drift, etc.

You are incurring a cost with selling assets through the Bid/Ask spread. Similarly, if you sell VTI to by VOO, it's possible that the two could deviate quite a bit in just 30 days to the tune of 1% or more. At what volume of stock lot does it make sense to sell?

Analyzing Transaction / Opportunity Costs:
A $20,000 sell order to generate a $3,000 loss could potentially drift 1%, or $200 after 30 days vs. your target fund.

The drift is a random factor; you are just as likely to gain $200 as to lose $200 for being in a different fund.

A bid-ask spread cost may be 3 basis points. Multiply that by 4 transactions (Sell-Buy day 1, Sell-Buy Day 31) and that's 12 basis points. Or $24.

The cost should be halved here; if you buy at the ask and sell at the bid, you have lost only one bid-ask spread, not two. Also, 3 basis points is high for the most popular funds among Bogleheads: Vanguard's page on bid-ask spreads lists 0.01%-0.02% for funds such as VTI, VOO, and VXUS.

Not terrible, but if you have more exotic funds like AVUV -> DFSV, that bid-ask spread could be 10-20 basis points per transaction, or potentially 80 basis points for $168.

Here, the cost is more significant. I estimated in With high spreads, wait for large ETF losses before harvesting that it is worth harvesting an ETF loss if the amount of the loss is 10 times the spread (or the combined spread of both ETFs if you plan to switch back). If you already have significant carryover losses, it should be 20 times the spread.

I made that post in 2020, when low-volume ETFs had 0.5% spreads, and thus it wasn't worth harvesting unless the loss was over 10%. But at 0.1% spreads, even a 2% loss gives a small benefit from harvesting. (I still wouldn't harvest a 2% loss on a low-volume ETF, because the value of my time is more than the tax savings.)

[JayB]

I recently sold some Agency bonds at a loss and replaced them with safer Treasurys of similar maturities. As a result, I will have capital loss carryovers for many years because I can only claim a net $3K loss per year and do not usually have any offsetting capital gains. However, when I did some spreadsheet modeling covering the next decade plus, I found an unexpected result: I was projected to avoid Medicare's IRMAA Part B surcharges for a year or two for spouse and self and realize some noteworthy future savings as a result. Of course, my actual numbers won't exactly follow the projections, but this was nevertheless a pleasant surprise.

[Nathan Drake]

Another issue with Tax Loss Harvesting - Bid/Ask Spread, Fund Performance Drift, etc.

You are incurring a cost with selling assets through the Bid/Ask spread. Similarly, if you sell VTI to by VOO, it's possible that the two could deviate quite a bit in just 30 days to the tune of 1% or more. At what volume of stock lot does it make sense to sell?

Analyzing Transaction / Opportunity Costs:
A $20,000 sell order to generate a $3,000 loss could potentially drift 1%, or $200 after 30 days vs. your target fund.

The drift is a random factor; you are just as likely to gain $200 as to lose $200 for being in a different fund.

A bid-ask spread cost may be 3 basis points. Multiply that by 4 transactions (Sell-Buy day 1, Sell-Buy Day 31) and that's 12 basis points. Or $24.

The cost should be halved here; if you buy at the ask and sell at the bid, you have lost only one bid-ask spread, not two. Also, 3 basis points is high for the most popular funds among Bogleheads: Vanguard's page on bid-ask spreads lists 0.01%-0.02% for funds such as VTI, VOO, and VXUS.

Not terrible, but if you have more exotic funds like AVUV -> DFSV, that bid-ask spread could be 10-20 basis points per transaction, or potentially 80 basis points for $168.

Here, the cost is more significant. I estimated in With high spreads, wait for large ETF losses before harvesting that it is worth harvesting an ETF loss if the amount of the loss is 10 times the spread (or the combined spread of both ETFs if you plan to switch back). If you already have significant carryover losses, it should be 20 times the spread.

I made that post in 2020, when low-volume ETFs had 0.5% spreads, and thus it wasn't worth harvesting unless the loss was over 10%. But at 0.1% spreads, even a 2% loss gives a small benefit from harvesting. (I still wouldn't harvest a 2% loss on a low-volume ETF, because the value of my time is more than the tax savings.)

Thanks for the clarification on quantifying the costs. That helps guide my decisions more. While drift may be less in MCW funds, the possibility is higher in factor funds. DFSV loads less on value, so would be expected to drift negatively vs. AVUV. Perhaps not 1% in a month on average, but that roll of the dice still can't quite be quantified.

I think TLH may be quite a bit easier for MCW funds than Factor funds in terms of lessened costs. Less drift. Less spreads.

Good rule of thumb for the loss = 20X spread.

Another aspect of this is that usually when you are rebalancing with DCA, you are buying the asset that is down the most. So you may have regular contributions that would cause a Wash Sale due to the "30 day before AND after a tax loss sale". So almost seems like the opportunity would present itself if you have a multi-fund portfolio that all universally drop significantly.


I'm not sure if you use Fidelity, but are there any ways to set up trades so that they can transact simultaneously?

[LilyFleur]

So let's say I tax loss harvest and have $3,000 a year of losses to put on my yearly tax returns, for ten years. In year 5 of this process, I sell a home in California and have $100,000 in taxable capital gains (after subtracting the cost basis, improvements, selling costs, capital gains exemption, etc.) Don't the capital gains taxes have to be paid the year that the home is sold? So, I could take $3,000 off my taxable income in the year I am taxed on capital gains of $100,000 for the sale of my home?

In California, all capital gains are taxed at the regular income tax rate by the state, unlike federal taxes which are taxed at the capital gains rate. I suppose selling a home in the same year as incurring significant medical deductions for assisted living or a nursing home would help on those taxes.

These numbers are totally made up, by the way.

In year 1 you realized $30,000 of losses, and you have no capital gains and will have none for the remainder of this example except the house sale.
Year 1: Deduct $3,000 from income.
Year 2: Deduct $3,000 from income.
Year 3: Deduct $3,000 from income.
Year 4: Deduct $3,000 from income.
Year 5: Sell house, realize $100,000 taxable gain. Subtract $18,000 of carryover losses. Pay taxes on only $82,000 of gains.
You have no further losses to carryover to years 6 and beyond. You also have no losses to deduct from income in year 5, as they are used up by the realized gains first.

Thank you.
I did not understand that you could take a greater than $3000 loss for capital gains (not just regular income).

[bling]

Of course this doesn't even consider potentially executing a trade sub-optimally. At least at Fidelity, I can't immediately "Sell to Buy" an ETF with extremely quick precision. So having to do two separate transactions may take a minute or two, and by that time the market has potentially moved against you. The less large the position you are selling is, the higher the benefit. $10,000 in sales for a $3k loss is much more efficient than $20,000 in sales for a $2k loss, but the former situation is much more rare.

you can queue up multiple trades using ActiveTrader on Fidelity to make it instantaneous.