Can't Prune a Target Fund

[Rich_by_the_Bay]

For those of you who rely heavily on target funds (Target 2015, Strategic funds, etc.), how do you account for yearly rebalancing?

That is, say bonds are up and stocks are down. At year's end under the typical diversified fund holdings, you would sell off bonds for income, then rebalance the remainder. Next year maybe just the opposite. Or international is up, domestic is down, etc.

With a balanced or target fund, aren't you always selling off in a fixed proportion as set by the fund? If a fund is 60:40, you are selling 60:40 even if you'd otherwise sell mostly bonds in a given year. Isn't "pruning the winners" an important nvestment strategy?

Does automatic rebalancing within such funds correct for this? I can't convince myself that this is the case - in fact, it feels like autobalancing does almost the opposite: great in the accumulation phase, but counter productive in the draw-down phase.

Can anyone help me out with this, or is my concern valid?

 

[wabmester]

I'm not sure I get the rebalancing bonus.   Maybe somebody can explain it to me.

It seems obvious to me that selling your winners and buying your losers on some random periodic basis should be just as likely to hurt you as it is to benefit you.

If that's true, what's the bonus everybody is talking about?

According to Gummy, it's just the difference between the annualized returns of the rebalanced portfolio vs the weighted average returns of the component classes.    I'm not sure why that should excite me.

article

 

[Rich_by_the_Bay]

I'm not sure I get the rebalancing bonus. Maybe somebody can explain it to me.

Not to drift from the original post about strategic or target funds and rebalancing, my understanding is that the rebalance lets you sell when an asset class is at a relatively high price, and defer selling another uncorrelated asset class when it is relatively low. Sooner or later, that theoretically is superior to selling everything proportionately on a periodic basis. If that premise is not true, then the OP is invalid.

 

[wabmester]

If that premise is not true, then the OP is invalid.

Yours is the popular interpretation, which I don't believe is true. If you sell "high," that asset is just as likely to continue running and you're giving up your upside.

Anyway, take a look at the Gummy article I linked to above (and several other articles on rebalancing on his site), and tell me if you still believe in rebalancing when you're done.

 

[Rich_by_the_Bay]

Anyway, take a look at the Gummy article I linked to above (and several other articles on rebalancing on his site), and tell me if you still believe in rebalancing when you're done.

I read Gummy and understood uh... some of it, but I trust his analyses somehow. The problem is that he is addressing growth over time. I am not sure that this works during draw-down -- maybe I'll email him and see what he says.

Like dollar-cost-averaging: just cause it works going in doesn't mean it's best getting out. At certain unavoidable periods, you are forced to sell holdings to meet expenses. You are looking at a sector that is high relative to other sectors. Which do you sell?

I know that this stuff isn't always as it appears...

P.S. I emailed Gummy and will post his response if and when.

 

[wabmester]

People often argue about the best frequency and algorithm for rebalancing.   I think people always argue about this because there is no right answer.   I just can't imagine how periodic rebalancing should be likely to pay a "bonus" most of the time.

I think the main benefit to keeping your allocations fixed is to match your risk tolerance.

So, to answer your original question, the "target" funds will adjust to your new (lower) risk tolerance, but there's no guarantee of a "bonus" using any automated market timing rebalancing strategy.

 

[saluki9]

I think you're talking about two different things

Rebalancing as a return generating mechanism has been pretty much shown to be a failure. The fact is that momentum exists and asset classes that have done well in the past have a fairly good probability of doing well in the next quarter, year etc.

That being said, what Rich is talking about it using rebalancing as a risk reducing mechanism. As we know, you reduce risk, you reduce return. So from a strategic asset allocation perspective, it does make sense.

Also, most target funds do rebalance at some point. Read your fund company literature for details

 

[Rich_by_the_Bay]

I think you're talking about two different things...

Well said.

To be honest, I think that I had assumed, perhaps wrongly, that rebalancing improved actual rates of return by opportunistically and selectively selling the asset classes in favor at the time of the withdrawal.

I see your point about volatility reduction, etc. If that's the only advantage of diversification and periodic rebalancing,, I'd just as soon stick with a sensible strategic fund or two and forget about much more slicing and dicing.

 

[cute fuzzy bunny]

Yeah, understanding the question at hand never stops wab from weighing in.

One of the things I've wondered is how the tax management is done on the 'lifecycle' funds...they're rebalancing their domestic and international stock and bond holdings to maintain their correct proportions, but also shifting from equities to bonds and reducing their foreign holdings over time. I havent looked at vanguards because they havent been around long enough to show how anything other than the yearly rebalancing works. I'm interested to see what happens at one of those 5 year changes. Not even sure if they do it in one year or a little at a time over 5, i'll have to read the prospectus again.

If they're throwing off capital gains as well as the dividends, you might be fine just taking both of those as your distribution rather than reinvesting and later selling shares. Heck, you already paid taxes on them. If they're using some loss selling to offset the gains to neutralize the tax implications of the rebalance, then that'd be a different story.

I'm not sure that one can accurately predict the effects of rebalancing going forward. Theoretically you're taking away an appreciated asset to buy a depreciated one, which should improve both volatility AND return as the cycles eventually reverse. I'm sure you could lay out a study that shows that this helps, hurts and does nothing at all, depending on the funds chosen and the length of your study.

 

[TromboneAl]

using rebalancing as a risk reducing mechanism.

I'd say not reducing your risk, but maintaining the same level of risk/expected return that you chose when you decided on your asset allocation.

For example, if bonds do better for a few years and you don't rebalance, your risk will be reduced, but so will your expected return. In this case, rebalancing increases your risk, but lets you maintain the asset allocation you originally chose.

 

[unclemick]

okie dokie - unclemick's left handed version:

Da "free lunch" from 'slice and dice' goes like this - if you have 'non/lowly correlated' asset classes - by rebalancing to predetermined slices periodically - in theory you are 'selling high/buying low while capturing the expected growth for each asset class over the long term. The lunch part is you still get expected growth(or a lot of it) while damping risk by matching zigs and zags. Lots of debate here - and nuts and bolts to master in execution.

The lifecycle type funds are as mentioned - age dated so to speak - damping market volitility as you age.

 

[wabmester]

Yeah, understanding the question at hand never stops wab from weighing in.

Thanks for that, but I *thought* I was saying the same thing as saluki.   Allow me to quote myself:

So, to answer your original question, the "target" funds will adjust to your new (lower) risk tolerance, but there's no guarantee of a "bonus" using any automated market timing rebalancing strategy.

Translation: these funds adjust to your chosen risk tolerance by adjusting their allocation mix, but the only "bonus" is reduced volatility.   The sell "winner" and buy 'loser" rebalancing return bonus is a myth.

 

[unclemick]

I know 'we' pretty much got it - but how good are we at explaining it?

heh heh heh heh heh heh heh heh heh heh heh heh

 

[ats5g]

Hi Rich,

Perhaps a between-the-lines question in your first post was something to the effect of do the balanced/lifecycle funds run the risk of reverse dollar cost averaging? See also bob’s explanation. For example, since liquidating shares of balanced funds liquidates both stocks and bonds at the same time, does this sell the worse performing asset class as well as the better performing asset class?

I don’t think selling the shares of a balanced/lifecycle fund would be reverse dca-ing. This is because in a sense, the money is just being transferred an extra step or through another fund before the withdrawal.  The balanced fund rebalances from the better performing asset class to the worse.  When you sell shares of the balanced fund, you can think of it like selling the better performing asset as well as selling the gains of the better performing asset that have been transferred into the worse performing asset.

If you didn’t have a balanced fund, but say a stock and bond fund, if stocks were doing worse than bonds, in addition to the fund distributions, you’d just (1) sell the enough of the bonds for your withdrawal, and then (2) sell some more of the bonds to rebalance fully if necessary.

With the balanced fund, 1 and 2 are just reversed, with the rebalancing coming first.  If you want to mimic the balanced fund, you would (1) sell enough of the bond fund to rebalance, and then (2) immediately sell enough shares of both the stock and bond fund for your withdrawal.

The gains that are sold for withdrawal are still gains from the bond fund, they’re just transferred to the stocks/stock fund first.  I don’t think this changes the fact that the money you’re withdrawing are the gains from the better performing asset class.

Did that make sense?

- Alec

 

[mickeyd]

I have always considered Target Retiremant type funds as an oasis for the many folks that are invested in a MF but have no clue why they are there.  At least with a fund date of say 2045, they should be able to come pretty close to when they will retire or get to age 65.  As long as the ER's stay low, these funds seem to have a good future IMHO.

I have instructions for my DW (bored with anything investment-related) to exchange all of my tax-advantaged funds to either the VG TR 2015 or 2025  fund if I kick off before she does. She will have to decide which one to select, but can't really make a serious mistake by selecting any one of them.

 

[unclemick]

Ah yes - toss in a few tax considerations, maybe some RMD and:

yardwork doesn't look so bad - even at 95 and climbing.

105 heat index predicted.

heh heh heh heh heh heh - Target Retirement, iced tea and a straw hat.

 

[cute fuzzy bunny]

Allow me to quote myself:

What are you now, John Galt?

 

[Gone4Good]

I'm not sure I get the rebalancing bonus.   Maybe somebody can explain it to me.

It seems obvious to me that selling your winners and buying your losers on some random periodic basis should be just as likely to hurt you as it is to benefit you.

If that's true, what's the bonus everybody is talking about?

According to Gummy, it's just the difference between the annualized returns of the rebalanced portfolio vs the weighted average returns of the component classes.    I'm not sure why that should excite me.

article

The article you cite is perfectly accurate if you are comfortable letting your equity allocation, or even single asset concentration, grow to high percentages.  Most of us would not feel comfortable doing that, thus, rebalancing.

The results presented in the article might also come out different if the starting period happens to be just before a major bear market.

 

[Gone4Good]

With a balanced or target fund, aren't you always selling off in a fixed proportion as set by the fund? If a fund is 60:40, you are selling 60:40 even if you'd otherwise sell mostly bonds in a given year. Isn't "pruning the winners" an important nvestment strategy?

Does automatic rebalancing within such funds correct for this? I can't convince myself that this is the case - in fact, it feels like autobalancing does almost the opposite: great in the accumulation phase, but counter productive in the draw-down phase.

The fund rebalances for you. The results should be the same (except for the fact that the fund will actually rebalance whereas you might decide you don't really need more stocks after they've dropped 20%)

I think rebalancing works the same whether accumulating or withdrawing. Dollar Cost Averaging is the one that hurts you during liquidation.

 

[spideyrdpd]

I dont have these but ...
What they are: The Fidelity Freedom Funds are investment options that allow the investor to select the fund that best matches his or her expected retirement year. The Fidelity Freedom Funds invest in a diversified portfolio of other Fidelity® mutual funds to provide moderate asset allocation. They are designed for investors who want a simple, yet diversified approach to investing for their retirement. The allocation strategy for the underlying stock, bond, and money market mutual funds, is based on the number of years until the Freedom Funds' reach their target retirement dates. Each Freedom Fund with a target retirement date will gradually adopt a more conservative asset allocation as it approaches its target retirement date. Therefore, each fund's target asset allocation percentages will change over time to become more conservative, by gradually reducing allocations to stock funds and increasing allocations to bond and money market funds. The Fidelity Freedom Income Fund®, designed for those already in retirement, emphasizes bond and money market mutual funds and seeks to maintain a stable asset allocation from year to year.

 

[rmark]

Efficientfrontier.com from 1996

ABSTRACT
The actual return of a rebalanced portfolio usually exceeds the expected return calculated from the weighted sum of the component expected returns. A formula for estimating this excess return is derived and tested. It is demonstrated that assets with high volatility and low correlation with the remainder of the portfolio provide considerable excess return, or "rebalancing bonus."

 

[Rich_by_the_Bay]

Efficientfrontier.com from 1996

Nice find, rmark. Here's the full link.

Bernstein thinks the rebalancing bonus is real. I am not equipped to critique it technically.

 

[wabmester]

The actual return of a rebalanced portfolio usually exceeds the expected return calculated from the weighted sum of the component expected returns. A formula for estimating this excess return is derived and tested. It is demonstrated that assets with high volatility and low correlation with the remainder of the portfolio provide considerable excess return, or "rebalancing bonus."

Gummy's paper references Bernstein's paper and makes clear the distinction between getting an actual return bonus and getting a bonus vs the weighted sum of the component expected returns. It's a subtle difference, but one of them puts money in your pocket, and the other one (Bernstein's) doesn't necessarily.

 

[justin]

weighted sum of the component expected returns -

The fact that rebalancing allows one to beat the weighted sum of the component expected returns is significant. Let's say you have a portfolio consisting of a mix of US and intl stock index funds, and a short term bond fund. Further assume that these three asset classes are loosely correlated (r^2 of 0.6 or so). The weighted sum of the component expected returns might be 8%, but with annual rebalancing, you might expect your return to be 8.2% or so. That 0.2% is your rebalancing bonus (alpha). A little more return for the same amount of risk.

 

[wabmester]

The fact that rebalancing allows one to beat the weighted sum of the component expected returns is significant. 

I may need to reread Gummy's articles again, but I believe he is basically making two points:

1) The weighted sum of annualized returns is not the same as the annualized return of a mixed portfolio.

2) The expected bonus is dependent upon the covariance being fixed (and of course, it changes all the time).

I'd be thrilled to learn that I'm wrong about this.   But basically, Bernstein is not comparing rebalancing with buy-and-hold.