re-balancing granularity

[NYEXPAT]

Quote:

Originally Posted by bingybear I was not implying re-balancing more often. I was posing the question if more pieces (investments) such as breaking up VTI into pieces, say LCG, LCB, LCV, and Mid/small cap. Much of recent market gains losses from what I hear on the news (maybe a bad source) is due to tech. If true, I would think re-balancing would be a positive thing.

FIFY, The SOX is currently in a bear market!

[OldShooter]

Quote:

Originally Posted by pb4uski #2 above is a bit of an issue because sector weights wobble around some... easily solved by using a long-term moving average or something like that... ...

Isn't this pretty much saying that the choice of sector weights is arbitrary and that the act of rebalancing to these arbitrary weights is the secret sauce? That doesn't make any sense to me. What if you picked an allocation during the tech bubble or during the housing bubble?

Quote:

Originally Posted by SecondCor521 ... I can quit any time I want to. Really.

OK. Whatever turns your crank. Life would be pretty boring if our activities were limited to doing only things that are logical. The key, IMO, is self-understanding.

[pb4uski]

Quote:

Originally Posted by OldShooter Isn't this pretty much saying that the choice of sector weights is arbitrary and that the act of rebalancing to these arbitrary weights is the secret sauce? That doesn't make any sense to me. What if you picked an allocation during the tech bubble or during the housing bubble? ...

Not at all arbitrary. I presume that Vanguard has some sensible measure that they use in determining sector weightings for the purpose of the Portfolio Watch analysis.... I use those because it is convenient. If that wasn't available then would likely use a 3 or 5 year moving average of quarterly sector weights or something like that so the targets would move slowly and more in line with big picture changes than arbitrary market movements.

Rebalancing prods your to sell sectors that have had a good run and buy sectors that have not... just like it prods you to sell equities when they have had a good run or buy them when they have not.

[samclem]

Quote:

Originally Posted by pb4uski Rebalancing does have advantages. You can play around with the same portfolio with and without rebalancing in Portfolio Visualizer for different periods. For example, 60% Total Stock Adm and 40% Total Bond Adm Jan 2002 to Mar 2018: no rebalancing 6.69%, annual rebalancing 6.94% Jan 2008 to Dec 2017: no rebalancing 7.03%, annual rebalancing 7.38% Jan 2002 to Dec 2012: no rebalancing 5.14%, annual rebalancing 5.71%

Over time, rebalancing tends to >reduce< returns (slightly). That is because, in general, the assets with the highest volatility also have the highest returns. If we never rebalanced, our portfolios, over time, would tend to accumulate more and more of the highest return assets, and the portfolio as a whole would have higher returns. And higher volatility.



At one time I considered dividing my taxable US holdings into a lot more parts--using sector ETFs in place of the Total Market fund. I don't think I'd get higher returns, but there would be increased opportunity for tax loss harvesting as sectors gained and lost favor over the short term. To stay fully invested you'd need to find a few similar low-cost sector ETFs that you could buy into/sell out of. I don't know if the IRS would smile on this, or call it a violation of the wash sale rules.

[pb4uski]

Quote:

Originally Posted by samclem Over time, rebalancing tends to >reduce< returns (slightly). ....

The data says otherwise and I think it makes sense.... to the extent that a single asset class or sector ebbs and flows, rebalancing takes advantage of those ebbs and flows in a disciplined manner.

[samclem]

Quote:

Originally Posted by pb4uski The data says otherwise and I think it makes sense.... to the extent that a single asset class or sector ebbs and flows, rebalancing takes advantage of those ebbs and flows in a disciplined manner.

That's what many people think. Automatically buy low (the assets that haven't appreciated as much) and sell high (the assets that have appreciated more). And sometimes rebalancing does improve returns over time. Other times, more frequently, it does not. Regardless, the primary benefit of rebalancing is to maintain the desired AA and volatility ("risk") characteristics of the portfolio. Over many cases and long time periods, the impact of rebalancing on overall return is generally small, and slightly negative.
A primary reason investors do less well than the market overall is the impact of emotion. Rebalancing can help reduce the impact of emotions, especially a mechanical rebalancing scheme, so that is a good thing. I rebalance once per year, even though my AA might get quite out of whack over that time. History shows this is unlikely to have much impact on my long-term returns, and the impact is quite likely to be positive. The power of laziness!.
Source for graph below: Kitces

[bingybear]

Quote:

Originally Posted by walkinwood With more distinct asset investment and rebalancing between them, you may very well lower your risk. However, unless these investments are in tax sheltered accounts, you will end up paying more in taxes from more distributions and your rebalancing. Also, if you wish to simplify in the future, you'll end up with taxes on the gains as you change your portfolio composition. I face that problem now. I want to consolidate to fewer funds, but the tax cost of doing so in one shot is prohibitive to me, so I have to do it slowly to keep the cap gains taxes to a minimum.

I have most of my equity in taxable accounts. I let the distributions (mostly qualified) be distributed so that I can use it to rebalance (buy what I need to increase an asset). I've taken some of my orphaned ETFs and made proxies for other funds (take IVV and add mid cap and small cap ETFs to act more like VTI. Selling IVV would be a huge tax hit.

I feel your pain. Unless I have to sell a large chunk of appreciated ETFs, the taxes don't seem too bad. Note I don't have any MF in taxable accounts.

[pb4uski]

Quote:

Originally Posted by samclem That's what many people think.....

Then go to Portfolio Visualizer in two tabs and set up the same portfolio of stocks and bonds... run one without and one with annual rebalancing for a number of different periods of time of your chosing and compare the results. I ran a number of different periods and different length of periods and annual rebalancing won in all my trials.... frequently by a substantial margin... so I'm skeptical of that claim.

Maybe this time it is different.

[flintnational]

I don't split by sector but I do use a couple of extra funds. Stocks are about 65% - 70% of our AA. Below is the stock split . I rebalance back to this once per year. Works for us but, YMMV.

US Total Market Index* 40%
US Value Index 15%
US Small Cap Value Index 15%
International Total Mkt Index 20%
Emerging Market Index 10%

Each shown as % of total stock allocation not % of total AA.
*There is a small smattering of individual stocks included in Total Mkt. Index.

[MichaelB]

Here’s an article by Michael Kitces on this https://www.kitces.com/blog/how-reba...gement-anyway/
From the executive summary

Quote:

As a result, rebalancing may be helpful as a risk management strategy – otherwise higher-returning stocks would compound to the point that they are significantly overweighted relative to lower-returning bonds – but it’s only when rebalancing amongst investments with similar returns in the first place that rebalancing provides a return-enhancement potential.

[samclem]

Quote:

Originally Posted by pb4uski Then go to Portfolio Visualizer in two tabs and set up the same portfolio of stocks and bonds... run one without and one with annual rebalancing for a number of different periods of time of your chosing and compare the results. I ran a number of different periods and different length of periods and annual rebalancing won in all my trials.... frequently by a substantial margin... so I'm skeptical of that claim.

Skeptical is fine. Kitces already did the runs for me, they are in that chart. 30 year rolling returns for time periods starting in 1926 through 1984. He used large cap US stocks (probably due to data availability) and intermediate bonds.
I'm not arguing that there are no time periods and no asset class combinations where rebalancing won't improve returns. Sometimes it does. But rebalancing generally has a negligible impact on returns (and it can be quite harmful to returns, see 30 year series started in 1933 - 1944). Mechanical rebalancing is primarily a tool to keep a portfolio's volatility in check and to help reduce the impact of emotion. Impact on long-term returns is hit-and-miss.

[flintnational]

It is my understanding if you rebalance 2 assets with the same return, but they are negatively correlated, your return does not change but your SD is reduced. The greater the negative correlation, the greater the reduction in SD. IMO, finding assets that have adequate returns that are negatively correlated is the problem.

Edit: If the returns are not the same, the overall return should be approximately an average of the 2, but the SD should still be reduced based on the negative correlation value.

[OldShooter]

This thread has drifted off the OP's original question. He was asking about sector rebalancing within his equity tranche, not about the classical equity/fixed income AA rebalancing.

I am still scratching my head over @pb4uski's arguments. For one thing, @sengsational's point applies: Rebalancing among sectors is rebalancing among highly correlated assets. It is not at all like rebalancing between equities and fixed income.

But my big hangup is the arbitrariness of the target balance. Just any old average calculated or scraped from someone's web site. Sectors do wax and wane; sometimes they just wane. Look at the components of the original Dow Industrials (American Cotton Oil, American Sugar, American Tobacco, Chicago Gas, Distilling & Cattle Feeding, General Electric, Laclede Gas, National Lead, North American, Tennessee Coal and Iron, U.S. Leather pfd. and U.S. Rubber) as representing major sectors is dinosaur viewing and arguably too extreme a case. But sectors do die, and this rebalancing strategy, which would methodically increase investment in a waning sector certainly makes no sense.

I don't doubt that there are backtests that make this idea look good, but there are an infinite number of backtests -- too many to run. I first want to believe in the logic of a strategy and I just don't see it here.

[samclem]

Where's the "World Total Market and We Mean Everything" fund? Stocks, bonds, commodities, etc.
For true Efficient Market believers. After all, bonds compete against stocks for the favor of investors, who presumably choose the investments that offer the best risk-adjusted return. Investment money flows between stocks and bonds based on perceived value, so have an index fund based on the market values of the assets. Any investment priced on large and efficient exchanges would be in the fund.

[pb4uski]

Quote:

Originally Posted by samclem ....Source for graph below: Kitces

What I find interesting in this graph is that while there are no rolling 30 year periods where rebalancing outperforms prior to 1955 or so, but since 1955 it seems that rebalancing outperforms more often than not... IOW, the gray bar years from 1955 on exceed the non-gray bar years.... perhaps things really are different. I wonder if it has something to do with much broader ownership of equities over that period of time.

Also, the same article strongly states that rebalancing within subsets of an asset class, which was the central topic of this thread does indeed have significant benefits.

[samclem]

Quote:

Originally Posted by pb4uski . . .perhaps things really are different.

Tee-hee. Well, with regard to large caps and bonds we can see that since 1955 rebalancing made almost no difference in overall returns (see the yellow and blue lines). Prior to that, rebalancing often had a profoundly negative impact.

Yes, I agree that rebalancing between more similar asset classes more often has improved returns. But, if we use US industrial sectors as an example, or style boxes, etc. we still don't have an intellectually satisfying answer as to what the allocations should be rebalanced to and why. After all, one answer is most obvious--the sectors are automatically and continuously "rebalanced" every moment the market is open, to the allocation the market deems appropriate. The sectors (and style boxes) are entirely arbitrary anyway. Any artificially imposed "rebalancing" that is different from the judgement of the market will require some rigorous justification.

[bingybear]

Ideally one would like to find asset classes that have no correlation. However, that is likely becoming more difficult with globalization.
I looked around a bit and some advisors (edelman - I know, heresy) is using style boxes, sans bended as part of their allocations.

It sounds from this thread that there may be supporting data for extending the re-balancing beyond bonds, US equities and international equities.

[OldShooter]

Quote:

Originally Posted by samclem ... we still don't have an intellectually satisfying answer as to what the allocations should be rebalanced to and why. After all, one answer is most obvious--the sectors are automatically and continuously "rebalanced" every moment the market is open, to the allocation the market deems appropriate. The sectors (and style boxes) are entirely arbitrary anyway. Any artificially imposed "rebalancing" that is different from the judgement of the market will require some rigorous justification.

Yes. My feelings as well, but said in a different way.

Here's another way to look at this idea: There are something like 10,000 mutual funds availble, with the majority being stock pickers. So tens of thousands of individual portfolio managers executing tens of thousands of strategies. The S&P SPIVA and Manager Persistence statistics consistently show us that none of these guys consistently beats the market, right? Out of these tens of thousands, I have to believe that some are using this same strategy in an attempt to outperform. And it isn't working.

[racy]

I personally use Larry Swedroe's re-balancing rules: "Re-balancing should occur only if the change in an asset class’s allocation is greater than either an absolute 5 or 25 percent of the original target allocation, whichever is less".
More here: The Larry Swedroe 5/25 Rule

[sengsational]

The reason why rebalancing is good for me, flawed as I am, is that I get to do something. Of course this is off in that fuzzy, non-analytical area that spreadsheets don't cover. But what if there was no such thing as rebalancing? I can only imagine the really stupid things I'd do. But with rebalancing, I can take action, and even though it makes very little difference, at least I've done something, and something that's not really too costly.