Betterment $50k "Safety Net"?

[milesdividendmd]

Yes, to exploit harvesting opportunities, it's better to hold the individual funds directly rather than a "fund of funds."

Depending on when your investments were made, a harvest of 1.6% of your principal this year would have been pretty easy to achieve without much work or monitoring.  You paid Betterment's computers to do it for you, which is fine but less consistent with the mustachian ethos, but admittedly it is also only true looking back with 20/20 hindsight -- in periods of low volatility, Betterment's computers will be able to tax loss harvest when we mere humans cannot.

I'll take 3.2%/year in annualized tax loss harvesting any year I can.

Can you honestly claim that you have done as well with your taxable accounts this year, Brooklyn?

If so you are a much more meticulous invester than I am. Which is entirely possible, I guess. (I have my doubts)

[Dodge]

In fact, as much as I'd like to call it a "safety net" or "rainy day" funds, I'd probably just leave it open and flexible to even become a home down-payment or car fund.

To answer your original question, it really doesn't make sense to use Tax Loss Harvesting in this situation.  Tax Loss Harvesting isn't free money, whatever you deduct is owed back when you sell.  If you earn $100,000 a year (28% tax bracket), and get the full $3000 deductible, you'll get $840 back in taxes this year.  However, if you plan on cashing out your investments in 5 years to buy a house/car, while you're still in the 28% tax bracket, you will owe the government an additional $840 in taxes that year.  In other words, while tax loss harvesting doesn’t necessarily produce any net tax savings, it is the equivalent of getting an interest-free loan from the Federal government.



If you invest that $840 back into the Betterment, and get an 11.5% gain for the year, your $840 will have grown to $936, for a profit of $96 from your interest free loan.  Not bad!  And how much will Betterment charge you for this service?  $255.  How much would Vanguard have charged you for a fully automatic "set it and forget it" 60/40 stock/bond fund that handles rebalancing across 10,752 unique assets?  $56

Don't pay $199 to earn $96.  Keep it simple.

The problem with your analysis is that when you "cash out" the investments in 5 years, they will be long-term holdings, so you will pay the long-term capital gains tax rate of 15%, not your earned income marginal tax rate of 28%. In other words, by deferring the tax for 5 years, Betterment saved you 13 percentage points of the $3000, or $390 in real tax savings. You can then add that to the $96 for a total of $486 in gross savings as a result of using Betterment. After subtracting the fee of $255, you have a net savings of $198. Looks like Betterment wins in your example.

In fact, despite your claim that it is merely tax deferral, tax loss harvesting usually results in real tax savings so long as you use the losses to offset ordinary income or short-term gains and then don't close the position until it is a long-term holding. Even if your earned income marginal tax rate is the same on both sides, you have real tax savings.

Good point.  But if you're holding for 5 years, and keep getting 11.5% gains (so you are unable to tax loss harvest past the first year), you will end up paying a cumulative $1,419 in fees at Betterment, compared to $310 at Vanguard.

You only need to hold for 1 year to get the tax benefit, not 5 years. I only said 5 years in the original version of my post because of how you wrote "5 years" in your post. However, despite you writing "5 years", your example was written for 1 year, and in the 1 year case, Betterment wins.

Betterment may win in the 5 year case as well, because the analysis becomes a lot more complicated over the 5 year horizon, and is not anywhere near as simple as what you just said.

+1

It's almost as if Dodge is hell-bent on coming up with unrealistic scenarios to prove a point that he wishes to be true.

Here is a simple point. Based on the amount of tax loss harvesting that betterment has already performed for me in the last six months,and my tax bracket, my expenses at betterment will be cheaper than the same amount of money invested at Vanguard for the next three years.

Simple fact.

Of your current portfolio size, yes.  If you get the 11.5% growth for the next 3 years, your portfolio will be around $92,375, making your yearly Betterment fee a much higher $380.  You will have paid over $1,300 in fees compared to my $217, and the fees will never be recovered with tax loss harvesting.  As shown in my earlier examples, the gap only widens from there.

[milesdividendmd]

I've paid 95 dollars in fees in over a year. I have saved over $400 in taxes.  In other words I have spent much less in fees than you. This is an indisputable fact that requires no assumptions.

Your math is once again flawed. If I have 92,375 in 3 years this will not effect what I have already paid in the past year (in fact my savings are already compounding Unlike your fees.)

I Anticipate that I will harvest more losses next year too.

Good luck with your 11.5 % assumed returns!

Remember:  garbage in, garbage out.

[Dodge]

I've paid 95 dollars in fees in over a year. I have saved over $400 in taxes.  In other words I have spent much less in fees than you. This is an indisputable fact that requires no assumptions.

Your math is once again flawed. If I have 92,375 in 3 years this will not effect what I have already paid in the past year (in fact my savings are already compounding Unlike your fees.)

I Anticipate that I will harvest more losses next year too.

Good luck with your 11.5 % assumed returns!

Remember:  garbage in, garbage out.




Good l

The fees are calculated by looking at the current year only.  Instead of looking forward 4 years, if we look back 4 years, we still see it.  This is a chart of all of Betterment's stocks:



They all hit their bottom in the 6th month.  A theoretical Betterment portfolio which started 4 years ago, would have had 0 tax loss harvesting after the 6th month.  While the portfolio, and the fee, would have nearly doubled.  Thanks to reinvesting dividends, it is unlikely this money will ever have the opportunity to tax loss harvest again.  It will end up looking like the chart I posted earlier (red line):



Maybe 2010 was a bad year to start, lets try some others:

2011: 1.5 years



2012: 1 year



2009: 3 months



2004-2007 all hit their bottom at the predictable 2008-2009 bottom, which was easily captured with manual tax loss harvesting.  So between 1-3 years was the maximum tax-loss-harvesting time.

2003: 3 months



2002: 9 months:



...etc

As you can see, after the first year or so, tax loss harvesting ceases.  Your portfolio, however, rises significantly, along with the percentage-based fee.

[brooklynguy]

Yes, to exploit harvesting opportunities, it's better to hold the individual funds directly rather than a "fund of funds."

Depending on when your investments were made, a harvest of 1.6% of your principal this year would have been pretty easy to achieve without much work or monitoring.  You paid Betterment's computers to do it for you, which is fine but less consistent with the mustachian ethos, but admittedly it is also only true looking back with 20/20 hindsight -- in periods of low volatility, Betterment's computers will be able to tax loss harvest when we mere humans cannot.

I'll take 3.2%/year in annualized tax loss harvesting any year I can.

Can you honestly claim that you have done as well with your taxable accounts this year, Brooklyn?

If so you are a much more meticulous invester than I am. Which is entirely possible, I guess. (I have my doubts)

No, I harvested exactly zero dollars of losses this year.  My contributions have been spread out equally across the year and I didn't consider the dollar amounts of my taxable investing to be high enough to jump on any harvesting opportunities.  That's why I said "depending on when your investments were made".  If I had invested $60K as a lump sum at a point during the year right before the markets took a 7% dip (as they did in the third quarter), it would have been easy to harvest a significant loss.

I was just trying to point out that it's not really an apples-to-apples to comparison if you include the benefit of algorithmic TLH for Betterment but entirely discount the possibility of manual tax loss harvesting in traditional investing.  Obviously, during periods of extreme volatility or extended downward market movement, manual tax loss harvesting will be easy, while during periods of low volatility and upward market movement, manual tax loss harvesting will be difficult or impossible.  Paying for Betterment's algorithmic TLH is like buying insurance against the absence of opportunities for manual TLH, which is a rational thing to do.  (But for me personally, as we discussed before, the maximum amount I stand to gain by paying for TLH is very low, so I'd prefer to go without that insurance.)

[milesdividendmd]

Dodge,

You seem to assume that you can only tax loss harvest at the lowest level of a trough which is quite simply wrong.

Assuming ongoing contributions (a pretty reasonable assumption based on how people actually invest)  you can tax loss harvest indefinitely as long as the value of any one asset drops below its basis.

If you want to limit the discussion to an emergency fund, then fine, but most people don't invest in an emergency fund and then stop investing in taxable accounts.

And in that case it would still be smarter to invest in betterment until your fees caught up with your money saved in tax losses, and then switch your ETFs over to another provider. In this way you would compound  your savings for a few years.

But again for most investors, investing does not stop once they save 50K.

[milesdividendmd]

Yes, to exploit harvesting opportunities, it's better to hold the individual funds directly rather than a "fund of funds."

Depending on when your investments were made, a harvest of 1.6% of your principal this year would have been pretty easy to achieve without much work or monitoring.  You paid Betterment's computers to do it for you, which is fine but less consistent with the mustachian ethos, but admittedly it is also only true looking back with 20/20 hindsight -- in periods of low volatility, Betterment's computers will be able to tax loss harvest when we mere humans cannot.

I'll take 3.2%/year in annualized tax loss harvesting any year I can.

Can you honestly claim that you have done as well with your taxable accounts this year, Brooklyn?

If so you are a much more meticulous invester than I am. Which is entirely possible, I guess. (I have my doubts)

No, I harvested exactly zero dollars of losses this year.  My contributions have been spread out equally across the year and I didn't consider the dollar amounts of my taxable investing to be high enough to jump on any harvesting opportunities.  That's why I said "depending on when your investments were made".  If I had invested $60K as a lump sum at a point during the year right before the markets took a 7% dip (as they did in the third quarter), it would have been easy to harvest a significant loss.

I was just trying to point out that it's not really an apples-to-apples to comparison if you include the benefit of algorithmic TLH for Betterment but entirely discount the possibility of manual tax loss harvesting in traditional investing.  Obviously, during periods of extreme volatility or extended downward market movement, manual tax loss harvesting will be easy, while during periods of low volatility and upward market movement, manual tax loss harvesting will be difficult or impossible.  Paying for Betterment's algorithmic TLH is like buying insurance against the absence of opportunities for manual TLH, which is a rational thing to do.  (But for me personally, as we discussed before, the maximum amount I stand to gain by paying for TLH is very low, so I'd prefer to go without that insurance.)

Brooklyn,

I think that's kind of the point.

If my taxable accounts were in say Schwab, or Vanguard, I doubt I would've tax loss harvested a single dollar this year.

Tax loss harvesting is one of those things that's conceptually easy to understand, but difficult in practice.  to do it well requires constant vigilance, and focus, and algorithmic decision-making, which most humans are quite simply not capable of.

It is exceptionally easy with betterment to take full advantage of tax loss harvesting opportunities because it requires zero effort on the investor's part.

buying and holding, rebalancing once a year, maintaining the discipline to stick to your plan, these are all conceptually easy but difficult in practice. And the evidence suggests that most investors are simply not up to the task.

When you add in tax loss harvesting that more than pays for itself,as well as all of the rest of it, betterment becomes a very attractive option.

And behaviorally I can't tell you how heartwarming it is on a down market day to get a message that you've just saved another couple of hundred dollars on next year's taxes. Talk about positive reinforcement for good behavior!

[Dodge]

Dodge,

You seem to assume that you can only tax loss harvest at the lowest level of a trough which is quite simply wrong.

Assuming ongoing contributions (a pretty reasonable assumption based on how people actually invest)  you can tax loss harvest indefinitely as long as the value of any one asset drops below its basis.

If you want to limit the discussion to an emergency fund, then fine, but most people don't invest in an emergency fund and then stop investing in taxable accounts.

And in that case it would still be smarter to invest in betterment until your fees caught up with your money saved in tax losses, and then switch your ETFs over to another provider. In this way you would compound  your savings for a few years.

But again for most investors, investing does not stop once they save 50K.

Don't you see?  That's exactly how the government looks at it when determining capital gains/losses.  It treats each individual deposit as its own separate container, a tax lot.  If you invested $100,000 at the beginning of each year I charted, that's exactly how it would have played out.

Your $100,000 deposit from 2010 would have received no more tax loss harvesting after the 6th month, and would have cost you about $1,250 more in fees over going directly with Vanguard.  With your fee in year 5 is poised to hit $532, increasing the total fee paid by about 33%.  A similar outcome would have occurred for all your other yearly deposits:



This is the ladder I tried to visualize earlier in the thread:



Is it possible to transfer just the tax lot from deposit #1 to another provider for the lower fee, while leaving the other tax lots alone?  If so, then yes that would be a much smarter move.  But then you would be stuck manually managing possibly all 12 of their ETFs, plus the 12 alternates it uses for tax loss harvesting.  Manually managing a 24 fund portfolio, for the rest of my investment horizon (50 years), for a benefit that only lasts 2 years doesn't seem very "set it and forget it".  I do not believe this is a good recommendation for someone seeking a "set it and forget it" portfolio.

[Dodge]

Dodge, you seem to be acting like the entire Betterment portfolio is a single security. That's not the case. The fact that it consists of multiple securities actually offers more tax harvesting opportunities and is not equivalent to it being an ETF.

In fact, there would be even more tax harvesting opportunities if instead of owning ETFs, you bought the individual securities directly. If you have over $500,000, Wealthfront will actually buy the individual companies inside the S&P 500 to do that for you.

Your entire fancy spreadsheet seems to be based on treating the entire Betterment portfolio as if it were a single security with harvesting only being possible if the value of that security drops below the cost basis. That is not how it works.

Check this post to see the graphs, showing each individual stock ETF in Betterment's portfolio.  Had you made deposits on Jan 1 of each of those years, all possible tax loss harvesting would have stopped for that tax lot after the indicated time (3 months after the 2009 deposit, 3 months after the 2002 deposit...etc).

[Dodge]

You are actually correct that you eventually stop getting a benefit for certain lots. However, the claim is that it may be worthwhile if you keep contributing because the benefits on the future lots may outweigh the costs of the lots no longer harvestable.

The other thing that all of your analyses ignore is that Betterment is actually basically free because it's very easy to rack up a few free years with referrals. The total amount I've actually paid to Betterment so far is $0 and I won't have to pay anything for another couple years, assuming I get no more free time in in the interim.

If you re-run the models with Betterment being free, I think you'll find it comes out comfortably ahead.

I calculated this too, in this post:

"In short, even if you make six $100,000 deposits, one a year, and have market conditions which allow you to tax loss harvest a +$600 gain each year for the first 5 years (which won't happen), the amount of fees accumulated by the year after your 6th deposit, will be larger than the total amount of tax loss harvesting you will ever accumulate."

While Betterment might not charge you the extra 0.35-0.15 ER if you have some credits saved up, you will still be charged the 0.16 ER from the underlying funds of course, so it's not completely free.  Since your portfolio is comparatively low during the first few years, it's not as big an impact as you might think.  Also, everyone won't have this opportunity, so it seems weird calculating it in like it's a given...but I'll do it for comparison's sake:



If you have no extra Betterment fee for the first three years, the new "Catch-up Year" is now year 4, when you consider the tax-adjusted portfolio value (Betterment has an additional $450 tax burden in this example).  Every year thereafter that this tax lot is with Betterment, you will be losing more and more money.

[milesdividendmd]

I've already shared my concerns with your assumptions above, which are far less realistic than Betterment's in their TLH white  paper. (And they're selling something.)

Your's is an unrealistic model with unrealistic assumptions.  Do you know anybody who invests 100k in a taxable account for 5 Straight years?  I certainly don't.

If the costs got too high, one could simply transfer their entire lot to a new provider, not specifically identified shares.

And before you dismiss managing a 12 asset slice and dice portfolio, you should ask yourself, why bother tilting small?  Why bother tilting to value?  Why bother diversifying internationally?

Since betterment gets a flat 0.15-0.35% AUM fee, it would certainly be cheaper for them to have a simpler portfolio with fewer transactions.

And the answer is: expected returns. They have constructed the portfolio to maximize expected returns.

These are not guaranteed returns, they are just the best guess on future returns based on past evidence. For more on this concept, I would suggest reading this book:

http://www.amazon.com/Expected-Returns-Investors-Harvesting-Rewards/dp/1119990726

I invested in betterment before they offered tax loss harvesting, and I still thought it was worth it. Now is a complete no-brainer.

[Dodge]

Dude, what you're missing is that Betterment is free forever under their current rules, assuming you're willing to put in a minimum amount of effort. If that stops being the case, I'll re-evaluate.

I'm not sure why you mention the expenses of the underlying funds because that will apply no matter where you buy them. I am obviously aware of them. The impact would be the same whether you buy those funds through Vanguard, Bank of America, or Interactive Brokers.

Also another misunderstanding you may or may not have is that the fee is based on the total of your contributions minus your withdrawals. The fee is not affected by the market value of the securities.

I wouldn't count on Betterment giving away their product for free for very long.  They got some big venture capital investments, and as such are likely under pressure to maximize profits.  In any case, I calculated the numbers assuming Betterment were free forever as you stated:



The new "Catch-up" year is 7.  If your investment horizon is longer than 7 years, you are better off going with with Vanguard.

[Dodge]

I've already shared my concerns with your assumptions above, which are far less realistic than Betterment's in their TLH white  paper. (And they're selling something.)

Your's is an unrealistic model with unrealistic assumptions.  Do you know anybody who invests 100k in a taxable account for 5 Straight years?  I certainly don't.

If the costs got too high, one could simply transfer their entire lot to a new provider, not specifically identified shares.

And before you dismiss managing a 12 asset slice and dice portfolio, you should ask yourself, why bother tilting small?  Why bother tilting to value?  Why bother diversifying internationally?

Since betterment gets a flat 0.15-0.35% AUM fee, it would certainly be cheaper for them to have a simpler portfolio with fewer transactions.

And the answer is: expected returns. They have constructed the portfolio to maximize expected returns.

These are not guaranteed returns, they are just the best guess on future returns based on past evidence. For more on this concept, I would suggest reading this book:

http://www.amazon.com/Expected-Returns-Investors-Harvesting-Rewards/dp/1119990726

I invested in betterment before they offered tax loss harvesting, and I still thought it was worth it. Now is a complete no-brainer.

The math is clear, the deposits don't matter, they all have the same result.  Here is the same math, following the same investing schedule from their chart, for the same number of years:



There are some typos in the screenshot above, but the story stays the same.  Vanguard wins due to lower fees over time.  The best part is, you don't really see the difference until we look at the long term, since again the fee is percentage based.  If you stopped contributing at this point, the gap grows to $90,000 in Vanguard's favor after another 10 years.  Add another 10 years on top of that, and the gap grows to over $400,000.  Add another 10 years, and it's $2,000,000 in Vanguard's favor...etc.

I dismiss managing a 12-24 fund for the rest of your investment horizon (30-60 years for Mustachians) as a viable option, in threads where the OP is specifically looking for a "set it and forget it" portfolio.

[Dodge]

Dodge, there are four things to address with your latest post.

1. Your annotation of the website does not mean what you think it does. The way Betterment defines "balance" is as the sum of your contributions minus the sum of your withdrawals, with some complicated special cases if you withdrew too much. You will have to read the full terms and conditions to understand this: https://wwws.betterment.com/document/BettermentCustomerAgreements.pdf?documentType=3

In particular I draw your attention to pages 39-40, where it is stated that "No asset-based Fee adjustment will be made at any time for appreciation or depreciation in Account asset value during a given period."

Strange, I know, but that's how they do it.


2. I'm already locked in free for 3 years including the time I already had. Even if I don't get anymore, that's three years of a great product. At this time, I might be tempted to go beyond three years and simply average the fee over the years I got free for a more fair comparison.


3. Your latest suggestion that Vanguard beats Betterment even if Betterment is assumed to have no fees is completely silly. At that point, you're just comparing hypothetical future returns of the underlying securities -- i.e. your crystal ball is somehow predicting that Betterment's portfolio will do worse than whatever you are defining as "Vanguard" (is it just an S&P 500 ETF?). I have no idea what you are trying to show there, but if you assume that Betterment charges no advisory fee, it purely comes down to which portfolio you are using, which is something you cannot model into the future.


4. Betterment charges no fee for an ACAT transfer out, unlike virtually every other broker-dealer including Vanguard, so you can do that whenever the fees become too large for you to stomach.

1.  I'll read up on that, but as noted in my "completely free forever" example, it doesn't matter.

3.  I detailed this in prior posts.  Vanguard wins because the fees are lower.  Yes, even when Betterment isn't charging you anything.

4.  I dismiss managing a 12-24 fund for the rest of your investment horizon (30-60 years for Mustachians) as a viable option, in threads where the OP is specifically looking for a "set it and forget it" portfolio.

[milesdividendmd]

Dodge,

Your graphic is unreadable. Literally.

I have no idea what you are attempting to show here. Literally.

The one constant in all of your models is that your assumptions are bizarre.

You assume that you can get identical returns from two completely different portfolios.

You assume that you can only tax loss harvest for one year.

You assume that the value anomaly and the size anomaly are fiction.

You assume that the average investor contributes $100,000 a year to their taxableaccount on January 1.

You assume that international diversification makes no difference to future portfolio returns.

All of these faulty assumptions lead to worthless output.

But my assumption is no assumption at all. It is simply the observation that my taxable account has been much cheaper than yours for the past year. And all things considered it is likely to continue to be cheaper than yours for at least the next 2 to 3 years.

It is also easier to use, A better user experience overall, and for me, at least, an approach that is less likely to lead to my own behavioral errors.

[Dodge]

Dodge,

Your graphic is unreadable. Literally.

I have no idea what you are attempting to show here. Literally.

Haha, I don't blame you. Trying to show both portfolios with yearly contributions matching the contributions in Betterment's example.  Each column is a new deposit, made in the next year. I'm keeping each deposit (tax lot) separate from the main portfolio, so you can see how tax loss harvesting provides an early boost for each deposit (tax lot), but after a few years is beaten by Vanguard's lower fee.

I'll try again tomorrow.

[Dodge]

Dodge,

Your graphic is unreadable. Literally.

I have no idea what you are attempting to show here. Literally.

The one constant in all of your models is that your assumptions are bizarre.

You assume that you can get identical returns from two completely different portfolios.

You assume that you can only tax loss harvest for one year.

You assume that the value anomaly and the size anomaly are fiction.

You assume that the average investor contributes $100,000 a year to their taxableaccount on January 1.

You assume that international diversification makes no difference to future portfolio returns.

All of these faulty assumptions lead to worthless output.

But my assumption is no assumption at all. It is simply the observation that my taxable account has been much cheaper than yours for the past year. And all things considered it is likely to continue to be cheaper than yours for at least the next 2 to 3 years.

It is also easier to use, A better user experience overall, and for me, at least, an approach that is less likely to lead to my own behavioral errors.

It's clear there's a misunderstanding here. The math is quite simple, making those assumptions irrelevant. Tomorrow I will model the same time period as Betterment's data, with the same contributions, so you can see more clearly.

[Dodge]

If the only thing a person is looking for, is an “automatic fund” that handles everything for them, because they are a complete newbie and don’t understand any of this stuff, the correct answer to their question should be one of Vanguard’s 18 or so fully automatic funds.  They will be mathematically guaranteed to beat or match 50% of the dollars invested in the market, and will never have to worry about underperforming.  They will also have the most diversified, in terms of unique assets held, portfolio available in a single fund, for the cheapest cost. You need to have a good reason to deviate beyond these choices.

For some people, like Milesdividend, that reason is a tilt.  If these people don’t want to worry about managing the 12+ funds needed to construct that tilt, and this pain is worth the additional yearly percentage fee to them, then going with Betterment, or Wealthfront, or some other fund manager who will manage this portfolio can make sense.  I personally don’t agree with tilting, but it’s a valid argument that can be made.

Is Tax Loss Harvesting alone a good enough reason to justify paying an additional yearly percentage fee?  The most important thing they need to understand for this question, is that the benefits from TLH have a limited shelf-life.  Paying an extra percentage fee for TLH mathematically stops making sense when one of these four things happen (whichever comes first):

1.   You have more than a certain amount of money invested.  This number can be anywhere from $0 to $700,000 depending on your income, your tax bracket, and when you make the withdrawal.

2.   You stop contributing new money for too long.

3.   There aren’t enough tax loss harvesting opportunities.

4.   You stop having taxable income (retired/lost your job) or are living off capital gains income of less than $72,500 (assuming you’re married).

Eventually, one of these four things will happen.  It is inevitable.  I do not believe paying an additional yearly percentage fee solely for Tax Loss Harvesting is a good recommendation for a complete newbie, as the fee will be higher than the gain over the long term, or they will have to move their assets of around 12-24 funds to another company so they can manage it themselves without the extra fee, which is the opposite of what they’re looking for (an automatic fund).

[milesdividendmd]

Dodge,

My thoughts.

Where we disagree:

1.  I believe Vanguards all in one funds Have poor International diversification, and no exposure to the value or size factors. It is virtually impossible to manually tax loss harvest with them unless there's a major bear market.  And they are expensive for what they are.

2.  If past is prologue (as in the last 100 years), and the value factor and the size factor anomies persist, then the betterment portfolio is a good bet to beat the vanguard all in one portfolios, even after expenses, and before tax loss harvesting.

With ongoing contributions to the betterment portfolio, I feel it is likely the betterment with tax loss harvesting will actually be cheaper for taxable accounts in the Vanguard portfolio for a long period of time.  (betterments white paper supports this thesis.)

In the end we have a difference of opinion. And it's unlikely that we will bridge this gap.  Which is to be expected.

And it should be said that I have zero problem with someone owning a slice and dice Vanguard portfolio and rebalancing once a year. For me behaviorally, I feel such an approach would likely underperform betterment. 
But each investor has to figure this out for themselves.

The exchange was enjoyable!

AZ

[Fallenour]

I think...I think I'm in love.

The math. The beautiful math. The excel sheets. The data graphs. Structured information! *tears up* So beautiful Q_______Q

[Fallenour]

In fact, as much as I'd like to call it a "safety net" or "rainy day" funds, I'd probably just leave it open and flexible to even become a home down-payment or car fund.

To answer your original question, it really doesn't make sense to use Tax Loss Harvesting in this situation.  Tax Loss Harvesting isn't free money, whatever you deduct is owed back when you sell.  If you earn $100,000 a year (28% tax bracket), and get the full $3000 deductible, you'll get $840 back in taxes this year.  However, if you plan on cashing out your investments in 5 years to buy a house/car, while you're still in the 28% tax bracket, you will owe the government an additional $840 in taxes that year.  In other words, while tax loss harvesting doesn’t necessarily produce any net tax savings, it is the equivalent of getting an interest-free loan from the Federal government.



If you invest that $840 back into the Betterment, and get an 11.5% gain for the year, your $840 will have grown to $936, for a profit of $96 from your interest free loan.  Not bad!  And how much will Betterment charge you for this service?  $255.  How much would Vanguard have charged you for a fully automatic "set it and forget it" 60/40 stock/bond fund that handles rebalancing across 10,752 unique assets?  $56

Don't pay $199 to earn $96.  Keep it simple.

"..is the equivalent of getting an interest-free loan from the Federal government."

That you get to earn 11.5% for 5 years, year over year, compounding interest on.

The most powerful force on the planet in the world of finance is compounding interest.

[Fallenour]

Soo...*cough* while I think I found my dinosaurs of the investment world (experienced power players).

Hypothetical question and questions.

Say you make, roughly 250k (before investments) as a household.

You max out both 401k, both roth, and you have roughly 80k left over after bills and taxes take out.

Currently playing your current tax game is getting old fast (international corps buying state side corps under state side parent corps under international holding corp, assets structured as raw material costs, preplanned corporate buyouts on projected losses, amortizing capital costs over time, really complicated.)

Say you want to get away from this, to a "set and leave it" concept.

I noticed with this "tax harvest", new term for me, that you guys project people dont have the money, what if they do? How does this tax harvesting help them? How does it hurt them?

I'm seeing lots of good debate on both sides, both of which are accounting for the concept of taxes, but one side argues no taxes paid, higher fees, while the other side argues no taxes paid, lower fees. (at least for loss periods, and if Im not mistaken, amortized losses over time for periods of gains?)

General gist of everything from what I'm seeing. Am I grasping this concept correctly? This is a very unique, very complex concept. I like the idea of it, but I never bite on things I don't understand.

[Dodge]

In fact, as much as I'd like to call it a "safety net" or "rainy day" funds, I'd probably just leave it open and flexible to even become a home down-payment or car fund.

To answer your original question, it really doesn't make sense to use Tax Loss Harvesting in this situation.  Tax Loss Harvesting isn't free money, whatever you deduct is owed back when you sell.  If you earn $100,000 a year (28% tax bracket), and get the full $3000 deductible, you'll get $840 back in taxes this year.  However, if you plan on cashing out your investments in 5 years to buy a house/car, while you're still in the 28% tax bracket, you will owe the government an additional $840 in taxes that year.  In other words, while tax loss harvesting doesn’t necessarily produce any net tax savings, it is the equivalent of getting an interest-free loan from the Federal government.



If you invest that $840 back into the Betterment, and get an 11.5% gain for the year, your $840 will have grown to $936, for a profit of $96 from your interest free loan.  Not bad!  And how much will Betterment charge you for this service?  $255.  How much would Vanguard have charged you for a fully automatic "set it and forget it" 60/40 stock/bond fund that handles rebalancing across 10,752 unique assets?  $56

Don't pay $199 to earn $96.  Keep it simple.

"..is the equivalent of getting an interest-free loan from the Federal government."

That you get to earn 11.5% for 5 years, year over year, compounding interest on.

The most powerful force on the planet in the world of finance is compounding interest.

In the example above, you don't get to earn compound interest, because your overall portfolio value ends up going down.  As Cathy and Milesdividendmd pointed out, this example only applies if you sell before it becomes long term gains.

[Dodge]

I noticed with this "tax harvest", new term for me, that you guys project people dont have the money, what if they do?

I'm not sure what you mean.  Have the money for what?

[Dodge]

Dodge, there are four things to address with your latest post.

1. Your annotation of the website does not mean what you think it does. The way Betterment defines "balance" is as the sum of your contributions minus the sum of your withdrawals, with some complicated special cases if you withdrew too much. You will have to read the full terms and conditions to understand this: https://wwws.betterment.com/document/BettermentCustomerAgreements.pdf?documentType=3

In particular I draw your attention to pages 39-40, where it is stated that "No asset-based Fee adjustment will be made at any time for appreciation or depreciation in Account asset value during a given period."

Strange, I know, but that's how they do it.


2. I'm already locked in free for 3 years including the time I already had. Even if I don't get anymore, that's three years of a great product. At this time, I might be tempted to go beyond three years and simply average the fee over the years I got free for a more fair comparison.


3. Your latest suggestion that Vanguard beats Betterment even if Betterment is assumed to have no fees is completely silly. At that point, you're just comparing hypothetical future returns of the underlying securities -- i.e. your crystal ball is somehow predicting that Betterment's portfolio will do worse than whatever you are defining as "Vanguard" (is it just an S&P 500 ETF?). I have no idea what you are trying to show there, but if you assume that Betterment charges no advisory fee, it purely comes down to which portfolio you are using, which is something you cannot model into the future.


4. Betterment charges no fee for an ACAT transfer out, unlike virtually every other broker-dealer including Vanguard, so you can do that whenever the fees become too large for you to stomach.

"No asset-based Fee adjustment will be made at any time for appreciation or depreciation in Account asset value during a given period."

Based on the screenshot I took of their page, I'd guess the "given period" is a year.  After the period is over, expect to pay a fee on your appreciation.

[Dodge]

Dodge, there are four things to address with your latest post.

1. Your annotation of the website does not mean what you think it does. The way Betterment defines "balance" is as the sum of your contributions minus the sum of your withdrawals, with some complicated special cases if you withdrew too much. You will have to read the full terms and conditions to understand this: https://wwws.betterment.com/document/BettermentCustomerAgreements.pdf?documentType=3

In particular I draw your attention to pages 39-40, where it is stated that "No asset-based Fee adjustment will be made at any time for appreciation or depreciation in Account asset value during a given period."

Strange, I know, but that's how they do it.


2. I'm already locked in free for 3 years including the time I already had. Even if I don't get anymore, that's three years of a great product. At this time, I might be tempted to go beyond three years and simply average the fee over the years I got free for a more fair comparison.


3. Your latest suggestion that Vanguard beats Betterment even if Betterment is assumed to have no fees is completely silly. At that point, you're just comparing hypothetical future returns of the underlying securities -- i.e. your crystal ball is somehow predicting that Betterment's portfolio will do worse than whatever you are defining as "Vanguard" (is it just an S&P 500 ETF?). I have no idea what you are trying to show there, but if you assume that Betterment charges no advisory fee, it purely comes down to which portfolio you are using, which is something you cannot model into the future.


4. Betterment charges no fee for an ACAT transfer out, unlike virtually every other broker-dealer including Vanguard, so you can do that whenever the fees become too large for you to stomach.

"No asset-based Fee adjustment will be made at any time for appreciation or depreciation in Account asset value during a given period."

Based on the screenshot I took of their page, I'd guess the "given period" is a year.  After the period is over, expect to pay a fee on your appreciation.

On the other hand, the same page you took a picture of has some language that could suggest otherwise. See attached picture.

That's not talking about their fee.  Considering it's a yearly fee, they specifically mentioned "during a given period", and "of your average annual balance" it's pretty clear in my opinion.

[tj]

Also another misunderstanding you may or may not have is that the fee is based on the total of your contributions minus your withdrawals. The fee is not affected by the market value of the securities.

This is news to me. It's industry standard to charge the AUM based on assets invested, I'd be shocked if Betterment worked differently.

[tj]

Also another misunderstanding you may or may not have is that the fee is based on the total of your contributions minus your withdrawals. The fee is not affected by the market value of the securities.

This is news to me. It's industry standard to charge the AUM based on assets invested, I'd be shocked if Betterment worked differently.

They confirmed via e-mail that it is indeed applied as a normal AUM.

The phrase “It is not impacted by market changes” applies to the pricing tier, rather than how the fee is applied.

The fee is applied to the balance, no matter the market change. The pricing tier, only changes if there is growth, never when there is a loss.

[Wolf359]

Hey guys,

I started reading up a little more on how some are using Betterment for their safety nets. It seems this is kind of a controversial topic where the naysayers wouldn't advise *investing* your savings money whereas proponents advise that it's only as risky as you want (or can tolerate) and in the end shouldn't leave you worrying a whole lot. So what's the consensus? I'm assuming it comes down to how comfortable you are with the risk of "investing emergency money" so maybe I'm just answering my own question.

Any pointers on why I should or shouldn't consider this though? I read somewhere that one person stated (or maybe it was based off of advice from Betterment) he would only go this route if he had $12k in SAVINGS, and anything in excess would be the "safety net" - does that make sense? I'm in a position where I think I could maintain a true savings and put down probably around $50k for a 'safety net' investment. I have much more stashed away in other taxable investments with Fidelity in addition to my non-taxable/retirement funds, so it's not like this is the last leg or anything. The reason Betterment is appealing is because there are fewer restrictions on withdrawals and thus easier access to that money in case it becomes "need it now" or "need it soon" money. As far as the fees go, it seems reasonable to pay the .25% when it seems like they take care of A LOT of things that you'd otherwise be worrying about managing yourself. At this point, I think I'd rather prefer to have a "set and forget" it type of investment in Betterment with this shorter-to-mid term money. In fact, as much as I'd like to call it a "safety net" or "rainy day" funds, I'd probably just leave it open and flexible to even become a home down-payment or car fund. Digression: besides Betterment are there other/better of these "robo" management tools out there? Wealthfront? Learnvest? etc? There's so many of these guys that it's hard keeping up with what it is they are or are even offering!

All that said, if it sounds like stashing this $50k in Betterment might be a good idea, what would you suggest my target allocation be?

Thoughts?

Going back to your original question...

I started out by putting $10,000 using Betterment Safety Net.  My original plan was to scale that amount up to use it to cover my emergency fund.  The Betterment recommendation is to have 6-12 months of living expenses + 30% in the Safety Net option (to allow for market fluctuation.)

My experience so far is that the Safety Net option is pretty stable.  Day-to-day fluctuations are minor, but the balance is below my original deposit (not by a lot).  Since the original deposit, however, I have rethought my overall investing strategy.  (I will probably either be changing the risk category of that money or removing it from Betterment altogether.)

If you plan to use the 3- or 4- passive index fund strategy for your primary investments, then you can park the bond fund component in your taxable investment account.  Use a tax-free (municipal or treasury) bond fund to minimize taxes. Put all your non-immediate emergency funds into that bond fund, and let it do double-duty.  All new money goes into the stock index funds only until your overall account is balanced (since this will start you out with an over-sized bond component.)

If an emergency comes up, you can tap the bond fund.  This strategy has the following advantages:
- It lets you start your primary investing plan right away.  It reduces the lost opportunity cost of having an emergency fund.
- It lets you have an emergency fund "invested" right away.
- If your overall portfolio is large enough, your bond component may already be bigger than your emergency fund.  In that case, there is no opportunity cost, and you can fully invest the entire amount per your desired mix.
- The emergency fund is liquid (takes 1-2 days to transfer electronically to your bank account if you need it.)
- The emergency fund is in a taxable account, so there are no penalties for withdrawing it from IRA or Roth early.
- If you tap the emergency fund, you can either re-fund it using new money, or it will be taken care of when you rebalance.  Basically, it gives you options.
- The concept of treating all your accounts (taxable and non-taxable) as a single big portfolio will let you concentrate all your bond fund requirements in a single location.  This means the funds available to you in an emergency will actually grow over time. 
- With your bond component in a taxable account, your limited tax-deferred or tax-free space may be used for the assets with the greatest appreciation potential (stocks.)

By incorporating the emergency fund into the bond component of your overall investments, you're technically investing it right away, but you're investing it into something that is still liquid and less volatile.

Wealthfront, Betterment, and the DIY Vanguard schemes all discussed here are really the same basic strategy.  That is passive investment into low-expense index funds, managed in a way to increase returns over long timeframes while minimizing taxes and reduce risk (volatility). 

In fact, all the schemes use mostly the same funds to achieve this goal (Vanguard ETFs).  The primary difference is in the number of funds used, the balance of how it is done, and extra features (TLH) and fees (.25% or .15% above fund expenses) resulting from using a managed account (WealthFront or Betterment).  It's impossible to know which variant provides the most return in the long run.  They have the same concept using the same underlying funds, so they'll probably all be in the same ballpark of returns. Betterment has a slight value tilt, a higher international percentage, and a lower bond allocation.  Wealthfront has a real estate component for greater diversification.  The 3- and 4- fund DIY Vanguard strategies are easier to understand and maintain.

In the end, getting started is the most important part. 

[Wolf359]

Soo...*cough* while I think I found my dinosaurs of the investment world (experienced power players).

Hypothetical question and questions.

Say you make, roughly 250k (before investments) as a household.

You max out both 401k, both roth, and you have roughly 80k left over after bills and taxes take out.

Currently playing your current tax game is getting old fast (international corps buying state side corps under state side parent corps under international holding corp, assets structured as raw material costs, preplanned corporate buyouts on projected losses, amortizing capital costs over time, really complicated.)

Say you want to get away from this, to a "set and leave it" concept.

I noticed with this "tax harvest", new term for me, that you guys project people dont have the money, what if they do? How does this tax harvesting help them? How does it hurt them?

I'm seeing lots of good debate on both sides, both of which are accounting for the concept of taxes, but one side argues no taxes paid, higher fees, while the other side argues no taxes paid, lower fees. (at least for loss periods, and if Im not mistaken, amortized losses over time for periods of gains?)

General gist of everything from what I'm seeing. Am I grasping this concept correctly? This is a very unique, very complex concept. I like the idea of it, but I never bite on things I don't understand.

I never bite on things I don't understand, either.  I'm also trying to understand this, but I might be further along.

For simplicity, I'm focusing on the "set it and forget it" strategy.  On the Vanguard camp side, this means that you put your money into a Vanguard Life Strategy Fund or a Vanguard Targeted Date fund.  Depending upon the fund you select (based on your risk tolerance or date requirements), this fund will have an expense ratio (probably around .18%).  This expense ratio is how much you pay the fund managers -- that rate multiplied by your total assets.

On the robo-advisor side, that means you give your money to Betterment, who manages your money for you.  For this, they charge you .15%.  The catch is that they're managing your money by putting them into (mostly) Vanguard funds, which themselves have an expense ratio.  So your costs are slightly higher because you're paying Vanguard + Betterment management fees. 

So why do Betterment customers pay the extra money?  Tax loss harvesting.  When Betterment is buying ETFs for you, some of them will drop in value.  They will then sell those at a loss, then replace it with a proxy (similar asset but not the same.)  At tax time, they will provide a statement so you can claim those losses on your taxes.

There are limits.  You can only claim $3,000 in passive losses per year. This reduces your income by that amount, saving you from paying the income taxes. If you are conducting trades in Vanguard funds outside of Betterment, some of the Betterment trades may be disallowed (this is eliminated if you give Betterment everything to manage).  After 2-3 years, the stock market has probably gone up enough that your older assets are no longer generating losses.  Losses will typically come only from new money.  When you sell, your assets will have a lower cost basis.  That means the taxes will be higher.  Betterment will manage your money to minimize taxes when you withdraw money. 

So far, this is all good. You pay Betterment a little more for portfolio management, and they give you a tangible return through Tax Loss Harvesting.  The issue is that the "little more" you're paying is a % of assets.  Assuming you're paying .05% more than you'd pay by using the Life Strategy fund, when your assets hit $1Million, your fee is $500 more per year.  Meanwhile, that TLH loss can't grow bigger than $3,000. 

So if you truly want 100% hands off, Betterment isn't a bad choice.  But if you are willing to follow a simple indexing strategy (DIY), your fund costs are the same as Betterment, and you save the entire .15% management fee.  With assets at $1Million, your Betterment fees are $1,500 per year extra.  (Betterment estimates that if you contribute $80,000 per year, or $6,666/mo, you will hit $1Million in 9 years.)

When you finally stop adding new funds, there is no more Tax Loss Harvesting.  At that point, you are still paying extra Betterment fees as long as you stay with them.  Okay, some argue that that's when you leave Betterment.  But how?  If you cash out, you take the capital gains for a very large (hopefully multi-million dollar portfolio.) If you transfer the assets in kind to a brokerage, you get 1-2 dozen funds in investments you don't understand, with years of complex accounting records to boot.  We're no longer talking set it and forget it.  So you stay with Betterment forever, and consider the fee just the cost of having someone else managing your multi-million dollar portfolio.

[tj]

Soo...*cough* while I think I found my dinosaurs of the investment world (experienced power players).

Hypothetical question and questions.

Say you make, roughly 250k (before investments) as a household.

You max out both 401k, both roth, and you have roughly 80k left over after bills and taxes take out.

Currently playing your current tax game is getting old fast (international corps buying state side corps under state side parent corps under international holding corp, assets structured as raw material costs, preplanned corporate buyouts on projected losses, amortizing capital costs over time, really complicated.)

Say you want to get away from this, to a "set and leave it" concept.

I noticed with this "tax harvest", new term for me, that you guys project people dont have the money, what if they do? How does this tax harvesting help them? How does it hurt them?

I'm seeing lots of good debate on both sides, both of which are accounting for the concept of taxes, but one side argues no taxes paid, higher fees, while the other side argues no taxes paid, lower fees. (at least for loss periods, and if Im not mistaken, amortized losses over time for periods of gains?)

General gist of everything from what I'm seeing. Am I grasping this concept correctly? This is a very unique, very complex concept. I like the idea of it, but I never bite on things I don't understand.

I never bite on things I don't understand, either.  I'm also trying to understand this, but I might be further along.

For simplicity, I'm focusing on the "set it and forget it" strategy.  On the Vanguard camp side, this means that you put your money into a Vanguard Life Strategy Fund or a Vanguard Targeted Date fund.  Depending upon the fund you select (based on your risk tolerance or date requirements), this fund will have an expense ratio (probably around .18%).  This expense ratio is how much you pay the fund managers -- that rate multiplied by your total assets.

On the robo-advisor side, that means you give your money to Betterment, who manages your money for you.  For this, they charge you .15%.  The catch is that they're managing your money by putting them into (mostly) Vanguard funds, which themselves have an expense ratio.  So your costs are slightly higher because you're paying Vanguard + Betterment management fees. 

So why do Betterment customers pay the extra money?  Tax loss harvesting.  When Betterment is buying ETFs for you, some of them will drop in value.  They will then sell those at a loss, then replace it with a proxy (similar asset but not the same.)  At tax time, they will provide a statement so you can claim those losses on your taxes.

There are limits.  You can only claim $3,000 in passive losses per year. This reduces your income by that amount, saving you from paying the income taxes. If you are conducting trades in Vanguard funds outside of Betterment, some of the Betterment trades may be disallowed (this is eliminated if you give Betterment everything to manage).  After 2-3 years, the stock market has probably gone up enough that your older assets are no longer generating losses.  Losses will typically come only from new money.  When you sell, your assets will have a lower cost basis.  That means the taxes will be higher.  Betterment will manage your money to minimize taxes when you withdraw money. 

So far, this is all good. You pay Betterment a little more for portfolio management, and they give you a tangible return through Tax Loss Harvesting.  The issue is that the "little more" you're paying is a % of assets.  Assuming you're paying .05% more than you'd pay by using the Life Strategy fund, when your assets hit $1Million, your fee is $500 more per year.  Meanwhile, that TLH loss can't grow bigger than $3,000. 

So if you truly want 100% hands off, Betterment isn't a bad choice.  But if you are willing to follow a simple indexing strategy (DIY), your fund costs are the same as Betterment, and you save the entire .15% management fee.  With assets at $1Million, your Betterment fees are $1,500 per year extra.  (Betterment estimates that if you contribute $80,000 per year, or $6,666/mo, you will hit $1Million in 9 years.)

When you finally stop adding new funds, there is no more Tax Loss Harvesting.  At that point, you are still paying extra Betterment fees as long as you stay with them.  Okay, some argue that that's when you leave Betterment.  But how?  If you cash out, you take the capital gains for a very large (hopefully multi-million dollar portfolio.) If you transfer the assets in kind to a brokerage, you get 1-2 dozen funds in investments you don't understand, with years of complex accounting records to boot.  We're no longer talking set it and forget it.  So you stay with Betterment forever, and consider the fee just the cost of having someone else managing your multi-million dollar portfolio.

Wolf,

What about the risk of Betterment being acquired/disbanded? Then you're wtuck with 10-12 ETF's at Apex Clearing anyway. Personally, I'm waiting to see what Schwab's robo-advising product looks like. They claim that they will not charge fees for the service.

[GGNoob]

Personally, I'm waiting to see what Schwab's robo-advising product looks like. They claim that they will not charge fees for the service.

I really hope that launches on time (Q1 2015). I'll have some money to invest soon and don't really want to keep it in cash waiting for them to launch. Schwab would be a very safe place to invest in a "robo-advisor."

[tj]

Personally, I'm waiting to see what Schwab's robo-advising product looks like. They claim that they will not charge fees for the service.

I really hope that launches on time (Q1 2015). I'll have some money to invest soon and don't really want to keep it in cash waiting for them to launch. Schwab would be a very safe place to invest in a "robo-advisor."

Agree. I'm selling my rental property in late Q2 or early Q3, and  hope to use the Schwab product for my stock portion of my portfolio. (Bonds will go directly to a Vanguard mutual fund - I see no need to use a robo-advisor for bonds.

[gluskap]

If you plan to use the 3- or 4- passive index fund strategy for your primary investments, then you can park the bond fund component in your taxable investment account.  Use a tax-free (municipal or treasury) bond fund to minimize taxes. Put all your non-immediate emergency funds into that bond fund, and let it do double-duty.  All new money goes into the stock index funds only until your overall account is balanced (since this will start you out with an over-sized bond component.)

If an emergency comes up, you can tap the bond fund.  This strategy has the following advantages:
- It lets you start your primary investing plan right away.  It reduces the lost opportunity cost of having an emergency fund.
- It lets you have an emergency fund "invested" right away.
- If your overall portfolio is large enough, your bond component may already be bigger than your emergency fund.  In that case, there is no opportunity cost, and you can fully invest the entire amount per your desired mix.
- The emergency fund is liquid (takes 1-2 days to transfer electronically to your bank account if you need it.)
- The emergency fund is in a taxable account, so there are no penalties for withdrawing it from IRA or Roth early.
- If you tap the emergency fund, you can either re-fund it using new money, or it will be taken care of when you rebalance.  Basically, it gives you options.
- The concept of treating all your accounts (taxable and non-taxable) as a single big portfolio will let you concentrate all your bond fund requirements in a single location.  This means the funds available to you in an emergency will actually grow over time. 
- With your bond component in a taxable account, your limited tax-deferred or tax-free space may be used for the assets with the greatest appreciation potential (stocks.)

By incorporating the emergency fund into the bond component of your overall investments, you're technically investing it right away, but you're investing it into something that is still liquid and less volatile.

We have about $60k that is our "emergency fund" that is just sitting in a savings account now.  We are thinking of doing exactly this.  Although we are more of the set it and forget it type so I'm thinking of keeping it really simple and just doing two funds in Vanguard...VTSAX and an intermediate municipal bond fund. Are there any downsides to this that I am missing?  This seems like a great place to put our emergency fund.  This is great because this also helps me have a more moderate asset allocation too as I have been 100% in stocks in my 401k/IRA since I thought before finding MMM that I wouldn't retire until a very long time frame so was okay with a riskier AA.  But now that the timeline to retirement is much shorter, I like the idea of having my EF in bonds as that will bring my AA down to a less risky 75% stocks/25% bonds.

[tj]

A poster on Bogleheads as highlighted a limitation of Betterment's TLH program:

Something interesting I've learned is to not expect a switchback after 30 days. In all of these TLH events the market has recovered by then, and Betterment won't switch back at a gain. So the TLH'ed funds tend to get "stuck" in the secondary ETFs until a bigger downturn occurs. At some point the unrealized gain of TLH'ed shares will be so big that there will be almost no chance of a downturn being big enough to TLH them again, and there's more or less a 50/50 chance of whether they'll be in the primary or secondary ETF at that time. Currently more than 50% of every asset class category is in the secondary ETF. So I may end up paying the higher expense ratio of the secondary ETFs "forever" due to the TLH. But I'm essentially reinvesting the tax savings, and expense ratios will change over time, so it's hard to say whether it's a net gain or loss in the long run in the case of, e.g., a mere 2% harvested loss in VOE (which occurred on 10/10) into IWS with 0.16% in extra expenses. My guess is that if TLH is considered "in aggregate" then it's a net benefit long-term but that some individual TLH events may not be.

As Dodge has repeatedly pointed out, most of the TLH happens in the beginning. After some time, all positions have large unrealized gains. There's nothing to harvest. The new money you are adding becomes a smaller and smaller percentage of the total invested. Meanwhile, you are stuck in the secondary ETFs paying higher expenses.

[uwp]

A poster on Bogleheads as highlighted a limitation of Betterment's TLH program:

Something interesting I've learned is to not expect a switchback after 30 days. In all of these TLH events the market has recovered by then, and Betterment won't switch back at a gain. So the TLH'ed funds tend to get "stuck" in the secondary ETFs until a bigger downturn occurs. At some point the unrealized gain of TLH'ed shares will be so big that there will be almost no chance of a downturn being big enough to TLH them again, and there's more or less a 50/50 chance of whether they'll be in the primary or secondary ETF at that time. Currently more than 50% of every asset class category is in the secondary ETF. So I may end up paying the higher expense ratio of the secondary ETFs "forever" due to the TLH. But I'm essentially reinvesting the tax savings, and expense ratios will change over time, so it's hard to say whether it's a net gain or loss in the long run in the case of, e.g., a mere 2% harvested loss in VOE (which occurred on 10/10) into IWS with 0.16% in extra expenses. My guess is that if TLH is considered "in aggregate" then it's a net benefit long-term but that some individual TLH events may not be.

As Dodge has repeatedly pointed out, most of the TLH happens in the beginning. After some time, all positions have large unrealized gains. There's nothing to harvest. The new money you are adding becomes a smaller and smaller percentage of the total invested. Meanwhile, you are stuck in the secondary ETFs paying higher expenses.

I mostly agree with Dodge that TLH on a lump sum loses it's benefit over time, but as the portfolio rebalances (they auto-rebalance, right?), won't that create new tax-lots to possibly harvest losses in?
I don't think this is something easily added to a spreadsheet,

[tj]

A poster on Bogleheads as highlighted a limitation of Betterment's TLH program:

Something interesting I've learned is to not expect a switchback after 30 days. In all of these TLH events the market has recovered by then, and Betterment won't switch back at a gain. So the TLH'ed funds tend to get "stuck" in the secondary ETFs until a bigger downturn occurs. At some point the unrealized gain of TLH'ed shares will be so big that there will be almost no chance of a downturn being big enough to TLH them again, and there's more or less a 50/50 chance of whether they'll be in the primary or secondary ETF at that time. Currently more than 50% of every asset class category is in the secondary ETF. So I may end up paying the higher expense ratio of the secondary ETFs "forever" due to the TLH. But I'm essentially reinvesting the tax savings, and expense ratios will change over time, so it's hard to say whether it's a net gain or loss in the long run in the case of, e.g., a mere 2% harvested loss in VOE (which occurred on 10/10) into IWS with 0.16% in extra expenses. My guess is that if TLH is considered "in aggregate" then it's a net benefit long-term but that some individual TLH events may not be.

As Dodge has repeatedly pointed out, most of the TLH happens in the beginning. After some time, all positions have large unrealized gains. There's nothing to harvest. The new money you are adding becomes a smaller and smaller percentage of the total invested. Meanwhile, you are stuck in the secondary ETFs paying higher expenses.

I mostly agree with Dodge that TLH on a lump sum loses it's benefit over time, but as the portfolio rebalances (they auto-rebalance, right?), won't that create new tax-lots to possibly harvest losses in?
I don't think this is something easily added to a spreadsheet,

Yes, but the additions are miniscule. If you start with a $200k portfolio and add $20k per year, as time goes on $20k is going from a 10% of your portfolio value to 1% or perhaps even less, and you could be stuck with the secondary ETF with high capital gain exposure.