Bogleheads

[Rick Ferri]

There is an inherent trade-off between managing a taxable portfolio to control risk and managing the portfolio to avoid taxes.

Managing for risk means rebalancing asset classes when they go out of parameter with the investment policy. Assume a portfolio is targeted to 50 percent in stocks and 50 percent in bonds. When a rising market pushes the stock allocation over a certain percentage, a rebalancing should occur. Those sales will generate a taxable event.

Rebalancing in a taxable account will create taxable gains over time. There is no way around this. However, there are some steps that an investor can take to mitigate the effect of taxes from risk control.

First, if the rise in stock value happens quickly, that rebalancing can be delayed temporarily until the sale of stock produce long-term capital gains. The tax rate on 35 percent maximum for short-term gains verses a maximum long-term capital gain rate of 15 percent. Of course, there is risk that the market will fall while an investor is waiting for short-term gains to become long-term.

Second, an investor should tax-loss harvest when some asset classes are at a loss in an account. In 2007, REITS and value stocks were negative for the year. Investors who had losses in those investments would do no harm to their portfolio by swapping from one value/REIT index fund to another for 30 days, and then swapping back. The tax loss can be used to offset future gains.

Third, it can be argued that “tax-location” is a good strategy. This is when an investor places stocks in their taxable account and bonds in their IRA or other non-taxable accounts. While this idea seems plausible in theory, it is debatable if this is an ideal strategy for everyone. Tax location could actually create higher realized capital gains and more taxes during rebalancing if the only account stocks are held in is taxable.

I am a big advocate of using low-cost index funds and exchange-traded funds (ETFs) because those investments are inherently tax-efficient. However, portfolio asset class allocation strategies typically consist of many different types of funds (U.S. stock, international stock, REITs, bonds, etc.). Rebalancing risk is the essence of Modern Portfolio Theory (MPT). Controlling portfolio risk through MPT requires that these different asset classes be rebalanced periodically.

Unfortunately, rebalancing is not compatible with tax-avoidance. MPT strategies were designed for non-taxable retirement accounts. When these strategies are applied to taxable money, the strategy can get very messy. Investors ultimately must choose between risk management and tax avoidance. They cannot have it both. Investors who use a widely diversified asset class strategy must understand and accept that fact that they will pay long-term capital gain taxes if the strategy is successful.

IMO, risk management takes precedence over tax avoidance. Portfolios should first be managed to control risk though rebalancing, and second to be managed to control taxes by using tax-efficient index funds and ETFs, take only long-term capital gain during rebalancing, and tax-loss harvesting when applicable.

Risk control and tax avoidance are not compatible. However, acting when you can to mitigate the tax effect of MPT is always a good idea. I am writing because we are in a period when investors should be harvesting tax losses in their taxable accounts. Those losses will help offset gains that will inevitable have to be taken if a taxable account is to follow a rebalancing strategy.

Rick Ferri

[snowman9000]

I think that is an excellent post. It goes to the heart of my own situation. Thanks.

[Taylor Larimore]

Hi Rick:

Quote:

I am writing because we are in a period when investors should be harvesting tax losses in their taxable accounts. Those losses will help offset gains that will inevitable have to be taken if a taxable account is to follow a rebalancing strategy.

Thank you for this important and timely advice.

Quote:

There are some steps that an investor can take to mitigate the effect of taxes on risk control.

I think we can add two more "steps":

1. Add contributions to underperforming funds, and take withdrawals from overperforming funds.

2. If possible, rebalance the portfolio by exchanging funds in tax-deferred accounts.

Again, thank you and best wishes.
Taylor

[AThiker]

Rick,

Thanks for this post. It's very helpful to me, as an investor with everything in a taxable account. I'd never thought about tax/risk in this way.

AThiker

[Derek Tinnin]

I would add that holding more TSM or Core Equity and less component funds further reduces the need to rebalance, at least within the equity portion of the portfolio. Too much slice and dice usually results in buying and selling from yourself, creating unncessary trade costs and taxes.

[waitforit]

I'm trying to digest how this applies to my situation. I'm following a S&D style that has me holding only VEU in taxable. Because my taxable account is so small relative to my tax-exempt holdings, I only have to sell in my taxable account to harvest losses... all the rebalancing happens on the tax-exempt side.

I can see how in 10-15 years this subject could start to become a problem for me. If my holdings in VEU in taxable become sufficient to account for all of my international developed / EM holdings, I'll need to add a second holding to the taxable account. If I'm not careful, this may set me up for the problems described here.

Good stuff Rick - very thought-provoking.

[Rick Ferri]

Taylor wrote:

[drjdpowell]

[Rick Ferri]

[larryswedroe]

few brief comments
It is a fact as demonstrated in several papers that locating tax efficient equities in taxable accounts is the right strategy--and the studies did not even include the benefits of tax loss harvesting nor the ability to donate appreciated shares

Second, I completely agree that one should avoid realizing ST gains (unless small) to rebalance. Other than that risk control is the more important issue. Remember you can basically only defer taxes not avoid them.

Third, as noted to extent possible one should use new cash flows and divs and distributions to rebalance (not reinvest them) to minimize capital gains that results from rebalancing

Fourth, this issue is why the new DFA core funds are such an important evolution of fund strategies as the reduce the need for rebalancing (among other benefits)

Last, the issue is not tax avoidance or tax minimization anyway. It is what gives you the best chance of achieving your goals without taking more risk than needed (etc).

Forgot to add that rebalancing is NOT a RTM strategy. That is simply incorrect as I have previously stated and it is easy to show why that is not true.

Say you want to be 50% stocks and 50% bonds. And stock returns are expected to be say 8% and bonds 4%. And next year stocks return 7 and bonds 5. If you were betting on RTM you would buy stocks and sell bonds. But a rebalancer would sell stocks to buy bonds--to keep the portfolio at 50/50. Thus rebalancing is a means to maintain portfolio allocations, not an RTM strategy.

[drjdpowell]

[Rick Ferri]

[drjdpowell]

[larryswedroe]

Rick
I am not incorrect at all. You just added that additional factor. What I said was the core strategies reduce the costs of rebalancing--that is a fact and certainly not incorrect. They reduce the costs of rebalancing between the equity accounts.

And as I have demonstrated to you it is clearly not an RTM strategy. Sometimes it happens that is what occurs, but that is coincidence, nothing more, as the simple example I gave demonstrated. Saying it is RTM is obviously incorrect as the example demonstrates.

[Rick Ferri]

[Wagnerjb]

Rick said:

[larryswedroe]

Rick
This conversation has NOTHING to do with hiring an advisor or not.
It is simply about whether a core fund provides advantages--and the obvious answer is yes on two fronts.
First, it cuts down on the internal turnover of the funds.
But the second reason is it cuts down on the need to rebalance. It amazes me that you either cannot see it or refuse to acknowledge it.

Here is example, say you own Vanguard TSM and Vanguard SV to achieve a certain loading on risk factors. When a small cap stock leaves the SV index the Vanguard small fund must sell it, and when it enters it must buy it. A core fund that owns the same loading will have less turnover as it can avoid some of those trades.

The same is true for rebalancing. If SV outperforms the individual will have to sell some SV. But the core fund will be able to use cash flows to rebalance the portfolio. Clearly a significant advantage

To repeat, this has NOTHING to do with DFA but whether core funds have advantages. The same benefits hold true for any broader fund like say for Vanguard's total international that includes EM. That avoids the need to rebalance between them. At least to extent that the fund holds the allocations you want.

[ScottW]

Both Rick and Larry make valid points. Although I do not have access to DFA (and assume that 95% of the people on this forum are in the same boat--if you do, you likely have an advisor, and have little need for a do-it-yourself investor site), it's my understanding that the Core funds are essentialy a Total Stock Market fund titled towards small value. Since most investors who S&D must do so using several funds, it's clear that a Core fund eliminates the need for rebalancing between 2 or more US equity funds. Plus, as Larry mentioned, the internal turnover is likely to be lower.

However, as Rick pointed out, the biggest discrepancy between asset class returns is stocks versus bonds. A Core fund will cut down on your equity rebalancing, but it won't do a thing for the asset classes you'll need to rebalance the most.

As for the RTM discussion, I agree that rebalancing often takes advantage of the reversion to the mean by accident. If a class becomes overvalued, rebalancing will essentially sell high and buy low in underperforming classes. This is a good thing, as people who rebalanced near the height of the large growth insanity in the late 90's can attest. The problem, and I assume this is where Larry has an issue is that:

1) You have no idea when RTM will occur.
2) RTM can only apply when rebalancing equities.

Since bonds have historically underperformed stocks, you can't argue that you're selling stocks high and buying bonds at a low. Rebalancing stocks and bonds is a risk-reduction exercise and nothing else. Only when rebalancing among equites does RTM come into play . . .

. . . except that you don't know when RTM is going to occur. Several authors (and I believe Bogle and Bernstein have discussed this), have found that since asset classes can "go on a run" for a while, it's sometimes better to rebalance less frequently. For example, investors would clearly have been better off rebalancing out of large growth at the end of 1999 (and letting there profits pile up) as opposed to rebalancing every year into classes that consistently underperformed. Rebalancing in early 2000 helped you, but rebalancing in the previous years did not.

The short version (from my perspective): rebalancing can take advantage of RTM, but it's hit or miss, and I wouldn't want to base a strategy off it. Consider rebalancing a risk-reduction exercise, and if you benefit in other ways, consider it a bonus.

[RiskyB]

Larry,

Your comments about the tax benefits of DFA Core funds leave me a bit confused. In a separate discussion I asked your opinion on using DFA TA US Core2 and DFA TA x-US Core as a simple two fund global stragety for my taxable account.

I claimed that I wanted to hold the two Core funds and ST bonds as my total portfolio because the funds were in my comfort zone regarding the SV tilt and they offer exposure to the total global stock market with minimal internal trading cost, therefore lower tax consequences.

However, you recommended adding several additional satellite funds around the Core funds to increase the expected return. I understand that adding additional funds MAY increase the expected return, however it WILL increase the tax burden and therefore diminish the benefit if holding Core funds in a taxable account. .

Your comments to Rick Ferri lead me to believe that you may have changed your position about adding funds to the Core funds. Your thoughts will be appreciated.

RiskyB

[Rick Ferri]

[larryswedroe]

Risky B

I don't believe I ever "recommended" any specific strategy. There is no right portfolio allocation, just one right for you. Having said that, the cores can be used either as stand alones or as CORES, and you add or subtract tilts from their, depending on your personal preferences. Also one can add other asset classes like RE and CCF.

You have to decide for yourself which is the best strategy. For example, you can tilt more to gain higher expected returns or tilt more and lower equity allocation to have same expected return but lower potential dispersion of returns.

The cores should be used for most people who have access because that is the most efficient way to gain exposure to the loading factors and minimize rebalancing. This is especially true for the new int'l core which avoids the need to rebalance EM. Now if you want more or less EM exposure than that core has you can add EM or add the core that doesn't include EM or you can add a fund like EM Value if you prefer.

I hope that helps

[ShinyTop]

This has been a very interesting thread; I have struggled with this tradeoff in my own portfolio and many of you have offered some good perspectives. One comment that I would add is that the decision about whether to take a taxable gain is further complicated when your AGI is at that border where many deductions and credits begin to phase out, which can substantially raise the marginal tax cost of taking any gain. This one has driven me crazy for years now. Annoying

[harry]

Larry

Here are several possible problems with the DFA core funds you were discussing above.

First, If DFA TA Core fund features continuous automatic rebalancing , how can it avoid reinvesting the dividends in a taxable account. Is it not true that reinvesting dividends and the frequent rebalancing should be avoided in taxable accounts.

Second, If the ex-US Core fund is a fund of funds it will not qualify for the foreign tax credit.

[larryswedroe]

Harry, They don't continually rebalance by selling stocks

First, let me explain how DFA sets its asset class definitions--which provides a significant advantage over simple indices that reconstitute annually. They reconstitute daily their asset class definitions. Russell for example reconstitutes annually. And there are about 30% of the stocks in the R2KValue index for example that are there in June that are not there in July. You have same issue for the RAFI indices. That means that over the course of the year funds based on these indices DRIFT to lower loading factors, lowering returns (expected returns, though of course in years of negative premiums the index funds would have a tailwind)

As to rebalancing--the does two things. First it uses new cash flows and dividends to rebalance say between developed markets and emerging markets, to keep the allocations approximately right. Similarly as stocks migrate away from small and value they would have lower weights in the fund, but cash flows would be used to buy more of other stocks.

DFA does not care about random errors because they don't exactly match some index. That is one of the disadvantages of pure index funds. Exact duplication leads to higher trading costs and tax inefficiency.

The second thing is that they use a multiplier system (too tough to explain simply here) for each of 36 "style boxes" so that transactions costs are minimized as stocks migrate from small to large and value to growth. To make up a example it might be that a stock would to go from 1% of the portfolio to 0.8% as it migrated up--the entire stock would not be sold, and perhaps no sales would be needed if there are sufficient cash flows.

I hope that is helpful

[harry]

Thanks Larry

I am still learning so your answer was a bit over my head. What about the DFA ex-US core being a fund of funds. Dose this fund (with no ticker symbol) qualify for the foreign tax credit and dose it compare favorably with the Vanguard FTSE all World ex-Us fund?

[larryswedroe]

It is NOT a fund of funds.

What happened is to take advantage of the new law that allows a fund to start not with securities but with holdings of other funds they launched it with a small amount of holdings of the other funds. Now that the find is live the new cash flows will start to be used to buy individual securities. Eventually the amount of fund holdings should be tiny relative to the size of the fund and not an issue. And if they can harvest a loss at some point my assumption is that they will do so--and get rid of the fund holdings.

Thus it was only launched as a fund of funds to get it out--otherwise given the broad diversification needed it would have way too much tracking error. So this is likely to be the way new funds are now launched, allowing them to come out with them much faster and not have to get advisors to "commit" to a large amount on an exact specific day, which is difficult unless you have an institutional investor who wants a particular fund. Probably need $20mm or more to launch a fund.

[tadamsmar]

[harry]

Great info. I am beginning to see how this works.

Will someone help me to understand the following:

DFA TA Core funds feature automatic internal rebalancing.

Vanguard TSM funds requires no rebalancing.

How do you decide which would be a better single US fund for a taxable account?

[stan1]

[learning]

Hi Rick, or any other knowledgeable Boglehead,


There's one rebalancing issue I still can't get clear in my mind. Hoping for some guidance.

Right now, I'm re-balancing by adding new money to the portfolio to whatever asset class is underweighted.
It's tax-efficient.
But it seems like I'm "buying" low, but never "selling" high.
Am I giving up the second "potential" benefit of rebalancing -- potential higher performance -- by rebalancing in this way?

Thank you for a great post.

Learning

[Taylor Larimore]

Hi tadamsmar:

Quote:

I don't have any taxable accounts. But if I did, I would seriously consider not rebalancing. If you look at studies of the risk of running out of money at a given draw-down rate during retirement, you will find that that the risk is not really improved by rebalancing.

Nearly every recognized authority, recommends rebalancing.

SEC recommendation:

Quote:

Rebalancing is bringing your portfolio back to your original asset allocation mix. This is necessary because over time some of your investments may become out of alignment with your investment goals. You'll find that some of your investments will grow faster than others. By rebalancing, you'll ensure that your portfolio does not overemphasize one or more asset categories, and you'll return your portfolio to a comfortable level of risk. For example, let's say you determined that stock investments should represent 60% of your portfolio. But after a recent stock market increase, stock investments represent 80% of your portfolio. You'll need to either sell some of your stock investments or purchase investments from an under-weighted asset category in order to reestablish your original asset allocation mix. When you rebalance, you'll also need to review the investments within each asset allocation category. If any of these investments are out of alignment with your investment goals, you'll need to make changes to bring them back to their original allocation within the asset category.

http://www.sec.gov/investor/pu....cation.htm

In my opinion, rebalancing is very important.

Best wishes.
Taylor

[stan1]

[learning]

Hi Stan1:

Ah. Got it. Thank you. That's clear, and helpful.