How to reconcile Value Averaging with Portfolio Rebalancing?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

How to reconcile Value Averaging with Portfolio Rebalancing?

Post by InvestmentJunkie »

Hi
New member here. I just finished reading Value Averaging by Michael Edleson, would definitely recommend the book. 5 stars.

My problem is this. In the book, the author mostly focused on implementing value averaging with the assumption that you're only going to own one mutual fund.

However, I own a portfolio of funds (ETFs). I plan to VA quarterly (every 3 months) and rebalance my portfolio yearly.

Now when I VA quarterly, how do I calculate the amount to VA into each fund? Do I calculate it individually for each fund, or as a whole portfolio?

Would appreciate some help. Great to have found this board!
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

I'm using a spreadsheet to allocate my assets along my value path. So if my large-cap portion goes up and small cap down, I will allocate more in new investments in small cap, or even possibly sell some large cap, to stay on my value path. That means that I'm really doing a bit of rebalancing every quarter as I allocate the funds that have accumulated into my side account.

I'm very new at this but I don't see it as much of a difficulty. The downside is that back-tested studies on rebalancing show the optimum rebalancing frequency to be one or two years. I would be interested to know what impact using a VA approach might have on returns and if the quarterly "best" VA results hold up.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

WiseNLucky wrote:I'm using a spreadsheet to allocate my assets along my value path. So if my large-cap portion goes up and small cap down, I will allocate more in new investments in small cap, or even possibly sell some large cap, to stay on my value path. That means that I'm really doing a bit of rebalancing every quarter as I allocate the funds that have accumulated into my side account.

I'm very new at this but I don't see it as much of a difficulty. The downside is that back-tested studies on rebalancing show the optimum rebalancing frequency to be one or two years. I would be interested to know what impact using a VA approach might have on returns and if the quarterly "best" VA results hold up.
WiseNLucky
Thanks for the reply. That's something I'd like to know too.

Do you know of a way one could contact Michael Edleson and post this question to him?
User avatar
bilperk
Posts: 548
Joined: Tue Feb 20, 2007 7:00 am
Location: Florida

Post by bilperk »

Haven't read the book, but I thought the idea was NOT to sell those funds that were above your value path, just don't add any new money until the value path caught up. Those funds below the value path would be given the new money. During a bull, nothing new might be invested at all, during a bear, you might have to look for other sources of money to stay on the path.

Selling the shares above the value path every quarter would also run into problems with VG's frequent trade provisions if it happens a fund (REITs say) was going up every quarter for years.

I would think a value averager would not want to rebalance more than once every year or two unless a fund got seriously out of whack. Part of value averaging, I would think, is letting hot funds run and adding to lagging funds with new money.

But then as I said, I haven't read the book :lol:

best,
Bill
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

InvestmentJunkie wrote:Do you know of a way one could contact Michael Edleson and post this question to him?
I don't. William Bernstein obviously knows Mr. Edleson and sometimes responds to requests on the other forum. Also, the publisher's website has a lot of information and may have a way to submit a question to the author.

I'm not too worried about it. A lot of people rebalance by putting new funds to under-performing asset classes and that's essentially all I'm doing. I think we can make this investing thing too much of a science and not enough of an art. As long as I'm getting where I'm going, I'm not going to obsess too much about whether I'm squeezing out the exact amount of return I could be getting. I stick with low-cost funds, allocate my assets in a manner that doesn't lose me sleep, and focus on trying to buy low and sell high. It's worked well so far for me.
Last edited by WiseNLucky on Tue Mar 27, 2007 8:53 am, edited 1 time in total.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

bilperk wrote:Haven't read the book, but I thought the idea was NOT to sell those funds that were above your value path, just don't add any new money until the value path caught up. Those funds below the value path would be given the new money. During a bull, nothing new might be invested at all, during a bear, you might have to look for other sources of money to stay on the path.
No, in the ideal plan, you sell funds in excess of your "Value Path" when the market makes extraordinary returns. Author also advises you to wait a month before you sell, so that the outperformance is not a short term spike.

Monte Carlo testing showed that selling reaps a higher return than not selling. However, no-selling VA is an alternative if one were not inclined to do it, the returns are about 1 - 1.5% lower a year but still good.
Selling the shares above the value path every quarter would also run into problems with VG's frequent trade provisions if it happens a fund (REITs say) was going up every quarter for years.
It's important to read the book if one were to implement VA because previously I also thought it was simply just making sure your portfolio grows by a fixed amount each month. However, there's much more to it than that.

VA accounts for not only market growth, but contribution growth. Hence, your Value Path has to grow by an estimated percentage that mimics market return each year (say 9 - 10%) and so does your contribution amount. He provides a spreadsheet for you to tweak the figures so that the plan is realistic for you.
I would think a value averager would not want to rebalance more than once every year or two unless a fund got seriously out of whack. Part of value averaging, I would think, is letting hot funds run and adding to lagging funds with new money.
Actually this is my problem. Value averaging is done quarterly for best results. Portfolio balancing should be done no more than once a year for best restuls. As the author did not go into how to value average for a portfolio of funds, I'm basically just confused and not sure how to proceed.

VA does allow funds to run, but not at an unrealistic pace. If it's running at an unrealistic rate, you sell off a bit at a time, then buy it back when it tanks. There is where the excess returns comes from, and also results in a lower average buying price than DCA.
User avatar
Petrocelli
Posts: 2871
Joined: Mon Feb 19, 2007 6:29 pm
Location: Fenway Park, between 2nd and 3rd base

Investment Junkie

Post by Petrocelli »

My advice: Don't over-think it. The difference between quarterly and yearly rebalancing is not all that great. Do whatever is easiest for you in terms of effort and minimizing transaction costs.

I also read the book and recently started value averaging. I intend to add cash to the positions which are down each quarter. In effect, I am rebalancing quarterly with new cash. This seems like the easiest way to approach the situation.
Petrocelli (not the real Rico, but just a fan)
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

bilperk wrote:Selling the shares above the value path every quarter would also run into problems with VG's frequent trade provisions if it happens a fund (REITs say) was going up every quarter for years.
Your comments/questions are all quite valid and most of them were well addressed by InvestmentJunkie. I would add, for the comment above, that I do all my VA adjustments in my 401(k) account through my employer where I have no tax issues and no frequent trader penalties.

Selling would not happen very often since I do not add funds to any asset class until the end of the quarter. Since my additions to my portfolio are still pretty substantial, the value path including new funds is fairly steep. The funds accumulate in a money market fund within the 401(k) until my quarterly reconciliation to my value path. In order to have to sell a fund, it would have had to increase on its own with no added funds pretty substantially during a quarter. Normally, it would not have exceeded its value path on its own and I would just allocate fewer dollars to it than the other slower growth or negative growth asset classes.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Re: Investment Junkie

Post by InvestmentJunkie »

Petrocelli wrote:My advice: Don't over-think it. The difference between quarterly and yearly rebalancing is not all that great. Do whatever is easiest for you in terms of effort and minimizing transaction costs.

I also read the book and recently started value averaging. I intend to add cash to the positions which are down each quarter. In effect, I am rebalancing quarterly with new cash. This seems like the easiest way to approach the situation.
Petrocelli,
Thanks for the reply.
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

I don't. William Bernstein obviously knows Mr. Edleson and sometimes responds to requests on the other forum. Also, the publisher's website has a lot of information and may have a way to submit a question to the author.
May I know which forum is that?

I also went to the Wiley & Son's website, but could not find a way to contact the author.
User avatar
alvinsch
Posts: 1613
Joined: Mon Feb 19, 2007 10:16 pm
Location: Northwest

Post by alvinsch »

I haven't read the book either but previous comments by Petrocelli have me curious enough that I've ordered it.

What I have/would consider doing as a poor man's value averaging is to linearly increase my target allocation percentage over the specific time and then just do normal rebalancing based on the ever increasing target. Seems like it would have a similar affect of buying more when the asset is down and less when the asset is up while keeping it much simpler than what I've gleaned about the authors method.

For example, let's say I wanted to add a 6% allocation to a new asset class over a two year period. I would just increase my target allocation by 0.75% per quarter (or 0.25% / month), or whatever your rebalancing period is and just do your normal rebalancing.

WARNING: I reserve the right to change my mind after I read the book. ;)

- Al
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

After giving it some thought, I could only come up with 2 possible scenarios.

1) Split my final target goal into the various
percentages represented by these funds (e.g. 10% US
large cap, 10% US large cap value, 10% US microcap, 10% US small cap value, 10% US REITs, 10% MSCI EAFE, 10% MSCI EAFE Value, 10% Int'l Small Cap, 10% Emerging Mkts), and treat
each fund seperately.

Hence I just use value averaging
every quarter and treat each fund as a seperate entity
with a seperate Value Path, and no longer do
rebalancing at the end of each year, as I think it's no longer necessary.

2) Treat the entire portfolio as one entity. So I
value average into the entire portfolio depending on
the value of the entire portfolio each quarter, then
rebalance once a year.

However, intuitively, this also
means that I'll be buying more of certain funds in the
portfolio even if it has moved up, if another fund has
gone down significantly, and instead of buying more
into the fund that has dropped significantly, I'm
buying more of all funds, of which some may have
risen.

This goes against the Value Averaging advantage (in my mind).

To me (1) seems more intuitive. What does everyone else think?
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

InvestmentJunkie wrote:To me (1) seems more intuitive. What does everyone else think?
I'm not sure I understand. I don't think either of your choices are necessarily intuitive since neither choice takes into consideration your desired asset allocation.

I use a spreadsheet to do my value averaging. I set my expected return at the rate I expect my blended portfolio to hit. Quarterly, I compare my total portfolio value to the value on my path and put any extra into the side account (MMA) or take from it if necessary.

Once I have the total portfolio balance figured out, I apply it in my desired allocation across the various assets.

A simple example:

I have $200,000 in a 50/50 TBM/TSM portfolio to start and plan to maintain that 50/50 split going forward. My target is $1,000,000 after 55 quarters and I assume an overall portfolio return rate of 7% with a 2% per year increase in my contribution level. Put all that in the model and it spits out ~$4,300 per quarter as my starting contribution rate to hit my goal.

Fast forward a quarter. I have put the $4,300 in my side account as I get my paycheck each month. At the end of the quarter, I look in my portfolio and see that my stock balance is sitting at $103,000 and my bond balance is sitting at $101,700. My total portfolio is now $209,000 including the side account.

My value path says I should be at $208,510 split evenly between stocks and bonds since I want to hold a 50/50 allocation. Since I want $104,255 in stocks, I use $1,255 of my side account to buy stocks. Since I want $104,255 in bonds, I use $2,555 out of my side account to buy bonds. That leaves me with $104,255 in each of my stock and bond accounts and $490 in my side account after I make my VA adjustments.

End example.

That is a fairly simple example. In my real model, I have 6 asset classes and varying percentages of each that I have set as my desired asset allocation. Following the plan is not any harder because I just set up a small spreadsheet section right in the model itself where all the verbiage and that graph used to be before I took them out.

As I said in a previous post, I use my 401(k) plan assets to make my VA adjustments so I have no tax or frequent trading penalties. In the event one of my asset classes really blows up in a quarter, I would sell some and put the proceeds in my side account or buy lower performing assets.

If my side account begins to build and keeps building, I would start a new model with either an increased return assumption or decreased contribution assumption. I would do the reverse if my side account didn't seem to keep up.

This is hard to describe with words. I hope what I've said makes sense.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

WiseNLucky,
Thanks for the example. Perhaps I will ask my questions using your example, hope that will be clearer.

In your example, when you go forward one quarter and you topped up the stock and bond funds so that both are $104,255 in value, you're also applying the market growth factor (r in the book) to the bond fund. Won't this result in you putting significantly more funds at the end of the day into bond funds than you should? Actually I'm not too clear about this myself. :D

What you described is quite close to Scenario #1 I described in my first post. What we're doing if we do follow this strategy, is we're in effect rebalancing the portfolio every quarter. I've read research by Larry Swedroe (if I remember correctly) that doing rebalancing too often will hurt rather than help returns. This is mainly due to the fact that trends take a long time to develop and by rebalancing too often, you're not buying at the "bottom" or selling at the "top", but are buying on the way down, and selling on the way up. But I won't go off on a tangent...hehe!

In Scenario #2 which I tried to describe in my previous post, what you will do is this.

After the first quarter, the value path states that you should be at $208,510 but your actual portfolio is $204,700. So you need to top up $3,810 to match the Value Path. Without regard to the individual values of the stock and bond funds at this point, you just divide $3,810 according to your original asset allocation percentage, for your case, 50/50. So you will put $1,905 into the stock fund, and $1,905 into the bond fund.

Note that at this point, the stock/bond percentages will no longer be 50/50. You do the same thing for 2 more quarters, then at the end of the year (quarter 4) you rebalance the portfolio back to the 50/50 percentage.

If we follow Scenario #1 which I tried to describe, we would have two seperate value paths of $500,000, one for stock and one for bonds. We will then pursue these two value paths seperately and independantly, and hope to arrive at our target of $1,000,000. However, if we do follow this path, we will need to use a seperate "r" (market growth) for the stock portion and the bond portion.

I hope this is clearer. Great discussion so far!
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

OK, I've got you. Examples really do help! :wink:
InvestmentJunkie wrote:In your example [WNL renames this Scenario #3], when you go forward one quarter and you topped up the stock and bond funds so that both are $104,255 in value, you're also applying the market growth factor (r in the book) to the bond fund. Won't this result in you putting significantly more funds at the end of the day into bond funds than you should? Actually I'm not too clear about this myself. :D
You are correct. This process results in constant rebalancing and may not provide the optimum rebalancing bonus. However, it does take advantage of swings in asset prices, which is what VA is all about, since it would have you adjust your allocation to various asset classes based on their recent trend.

If I remember correctly, the loss in overall portfolio return (rebalance bonus) by rebalancing quarterly was not that far off from the optimum timing of 1 to 2 years. The real benefit of this model is that it keeps the asset allocation in line with ones willingness to take risk.
In Scenario #2 which I tried to describe in my previous post, what you will do is this.

After the first quarter, the value path states that you should be at $208,510 but your actual portfolio is $204,700. So you need to top up $3,810 to match the Value Path. Without regard to the individual values of the stock and bond funds at this point, you just divide $3,810 according to your original asset allocation percentage, for your case, 50/50. So you will put $1,905 into the stock fund, and $1,905 into the bond fund.

Note that at this point, the stock/bond percentages will no longer be 50/50. You do the same thing for 2 more quarters, then at the end of the year (quarter 4) you rebalance the portfolio back to the 50/50 percentage.
I don't see why this couldn't work from a rebalancing perspective. The only problem I see is that the blended return combined with allocation of new funds based on AA percentages would result in your losing the VA bonus. If one asset class really jumped, instead of selling some of that class you would buy some more. You would, in effect, be trading the VA bonus for an attempt to gain rebalancing bonus.
If we follow Scenario #1 which I tried to describe, we would have two seperate value paths of $500,000, one for stock and one for bonds. We will then pursue these two value paths seperately and independantly, and hope to arrive at our target of $1,000,000. However, if we do follow this path, we will need to use a seperate "r" (market growth) for the stock portion and the bond portion.
This idea is really interesting and is, perhaps, the best way to optimize both the VA bonus and rebalancing bonus. It is much more complex (you would have to keep track of the side account against each of your asset classes) if you have a number of asset classes but is certainly do-able with a (relatively) simple spreadsheet.

For myself, I think I like Example 3 the best for my current personal circumstance. My portfolio balance is pretty high at this point and I like the comfort of managing my risk level. I'm willing to give up a little off the optimum rebalancing premium to mitigate my overall portfolio risk.

If I were younger, I would probably gravitate toward your Scenario #1. The calculations wouldn't be that hard and the overall return would probably be the highest. My least favorite would be your Scenario #2. That said, I think any of them would work to reach the stated goal.

It's great to have someone who has read the book to bounce ideas off of. :D
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

WNL,
Thanks for the insight, I agree it's great to be able to talk to others who've read the book too. Let's hope this thread gets a little revival after more people have read the book and come back with questions.

I've tried to contact Mike Edleson through Wiley's customer service email, the lady who answered was very nice and said she'll try to forward my question to him, but it depends if he want to answer or not.

I'll post back if I get a reply.
johnb
Posts: 282
Joined: Sat Mar 17, 2007 12:11 pm

Post by johnb »

When I backtested using the Scenario 1 method in a spreadsheet, I found that in the early years stocks went far higher than 50% of the portfolio. By the time stocks hit $500k, the portfolio was very unbalanced, and for the next several years I was buying only bonds.

So it seems best to value average on the entire portfolio rather than individual components.

John
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

johnb wrote:When I backtested using the Scenario 1 method in a spreadsheet, I found that in the early years stocks went far higher than 50% of the portfolio. By the time stocks hit $500k, the portfolio was very unbalanced, and for the next several years I was buying only bonds.

So it seems best to value average on the entire portfolio rather than individual components.

John
Thanks, John, that's very interesting.

Did you rebalance the backtested portfolio every year or two? I think that's what InvestmentJunkie intended. VA on the entire portfolio would effectively rebalance quarterly. IJ would let the portfolio subcategories run longer than a quarter but not for more than a year or two before rebalancing.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

WiseNLucky wrote:
johnb wrote:When I backtested using the Scenario 1 method in a spreadsheet, I found that in the early years stocks went far higher than 50% of the portfolio. By the time stocks hit $500k, the portfolio was very unbalanced, and for the next several years I was buying only bonds.

So it seems best to value average on the entire portfolio rather than individual components.

John
Thanks, John, that's very interesting.

Did you rebalance the backtested portfolio every year or two? I think that's what InvestmentJunkie intended. VA on the entire portfolio would effectively rebalance quarterly. IJ would let the portfolio subcategories run longer than a quarter but not for more than a year or two before rebalancing.
John,
As WNL said, great job.

To be honest, I did not think about rebalancing in Scenario one, as I thought of keeping the value paths seperate. I think Scenario 1 is not a good soluiton now on hindsight.

Scenario 2 is also not ideal as it totally takes away the advantages of VA. When we VA into the total portfolio, we are buying both funds that have gone up, and gone down. We are only VA-ing based on the total value of the entire portfolio. So in effect, if you look at each seperate fund, there is no VA effect at all. I hope I've not muddied the waters further with my explaination!

My brain hurts.

Why don't we lay out all the possible scenarios and walk through all of them looking for pros and cons, and perhaps we can come up with the best strategy?
User avatar
tetractys
Posts: 4836
Joined: Sat Mar 17, 2007 3:30 pm
Location: Along the Salish Sea

For VA, treat your portfolio like a fund of funds.

Post by tetractys »

Junkie,

There are a wealth of posts on all aspects of VAing on the M* forum; so for more info, including your question here, search there.

That being said, this is what I would do if I were VAing:

Decide on your target growth rate, and VA treating your whole portfolio as one fund, or like a fund of funds, always buying or selling towards your portfolio's target allocation. Naturally the growth rate, which determines the Value Path, would factor in your target allocation, and would be for your whole portfolio.

Example considering 2 funds, A and B:

When your portfolio has grown faster than your Value Path, if fund A is low and fund B is high, you would sell fund B to reduce your portfolio towards your target allocation, and then if more is needed, sell both funds A and B in proportion to your target allocation.

When you need to buy to reach your Value Path, if fund B is low buy that first to rebalance back to your target allocation, and then if more is needed, add to both A and B in proportion to your target allocation.

Everything else would apply just like in Edelson's book, including avoiding capital gains in taxable, only buying and not selling, using his different variations, using treasury yields to determine the value path, etc.

Tet
leahkim
Posts: 1
Joined: Sat Mar 31, 2007 4:35 pm

Do one VA curve for your entire portfolio

Post by leahkim »

I've read the book, and I am using VA for my entire portfolio (of 11 different Vanguard Index funds).

The way I do it agrees with what was discussed above (of course a lot of different things were discussed above, so I'll be more specific!)

I created my VA path for the portfolio as a whole. My portfolio is 30% bond mutual funds and the rest stock mutual funds, so I went with an assumption of 10% return per year for the whole portfolio (I think this is a reasonable, if not slightly conservative number since I am fully diversified among domestic and foreign, growth and value, large and small cap, etc.) I used a growth rate of 4.5% per year. I set up my path monthly, so of course I used these numbers divided by 12 in the VA spreadsheet.

I invest quarterly based on the advice in the book, so what I do is look at the entire portfolio's value on a quarterly basis. Let's use an example since we can all understand better that way -- I'll also use easy numbers from a math perspective. Say my portfolio is worth 95,000 this quarter, and next quarter it should be worth 100,000 per my VA path (i.e. 3 months down the spreadsheet). All I do is put cash in each of my mutual funds (or put nothing in, if that particular asset class did really well that quarter) until it meets my target asset allocation. So if my portfolio allocation is 7.5% S&P 500 index, then I just put cash in that fund until its value is 7,500 (remember, my total portfolio should be worth 100,000 according to my hypothetical VA path). 10% in total bond fund means I add cash to that fund until it equals 10,000. Make sense?

From the way I understand the book, you DON'T REBALANCE when using Value Averaging; it rebalances itself every time you Value Average (whether it be monthly, quarterly, yearly, whatever). That's the whole point of VA. Using the system I described above, you are essentially doing VA individually for each individual mutual fund, but you don't need to set up separate VA paths for each one, just one for the entire portfolio. If one of my mutual funds is doing great this quarter, I will have to put in less cash at the end of the quarter to "rebalance" it to the target asset allocation. And if one of my funds is lagging, I will need to put in more cash to get it up to the target allocation. That's the whole concept of VA - putting more money into the lagging funds, and less into the ones that are doing well. I don't sell off anything; I put a regular amount of money monthly into my side money market fund, and if my portfolio is doing better than expected, it just means I have more cash in my money market fund (making 5% yearly) on hand for the following quarter in case the market declines.

Make sense? What do you all think? VA is definitely a REPLACEMENT for rebalancing. It takes more time to do it quarterly than yearly or every two years, but you are maximizing your return if you trust Edelson's reasoning behind his "matching" concept. There is no reason to rebalance since the whole point of rebalancing is to take advantage of highs and lows in your various asset classes, and that's what VA does.

But let's face it -- if you are fully diversified in index funds and are on this message board because you are attempting to maximize value averaging, you are already ahead of 95% of the people investing in the market, so what are you worried about?? :wink:
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Post by WiseNLucky »

leahkim,

You are doing exactly what I'm doing. I think any of the methods will work out fine, but like the risk adjustment that comes from doing it "our" way.
WiseNLucky
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Re: Do one VA curve for your entire portfolio

Post by WiseNLucky »

leahkim wrote:My portfolio is 30% bond mutual funds and the rest stock mutual funds, so I went with an assumption of 10% return per year for the whole portfolio (I think this is a reasonable, if not slightly conservative number since I am fully diversified among domestic and foreign, growth and value, large and small cap, etc.)
One difference that I wanted to go back and read a little about before commenting. I'm going with a 60/40 stock/bond allocation but I'm using a much lower assumed return rate of 8% on the entire portfolio. Bogle and others have substantially lowered their assumptions of expected stock returns over the next decade to something less than 10%; 8% for a blended portfolio may even be high by these standards. I think Edleson was even recommending not going higher than 8.8% on a pure stock portfolio.

If you're right, I will see my side account bloom as I under-invest until I change my assumptions. I will have lost out on the opportunity to invest those funds at cheaper prices along the way until I make the change. If I am right, you will find you have to add more savings to your portfolio to stay on your value path. However, you will have purchased more assets at lower prices than you anticipated. I'm interested to see what happens.
WiseNLucky
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Re: For VA, treat your portfolio like a fund of funds.

Post by InvestmentJunkie »

Tet,
Thanks for the advice. Have some questions hope you can help me with.
There are a wealth of posts on all aspects of VAing on the M* forum; so for more info, including your question here, search there.
M* forum...is that where the Diehards forum started? Something to do with a constellation at dawn?
When your portfolio has grown faster than your Value Path, if fund A is low and fund B is high, you would sell fund B to reduce your portfolio towards your target allocation, and then if more is needed, sell both funds A and B in proportion to your target allocation.

When you need to buy to reach your Value Path, if fund B is low buy that first to rebalance back to your target allocation, and then if more is needed, add to both A and B in proportion to your target allocation.
How do you decide how much of each fund to buy or sell? Do you buy or sell until the funds return to their correct percentage of the entire portfolio?

Thanks!
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Re: Do one VA curve for your entire portfolio

Post by InvestmentJunkie »

leahkim wrote: Make sense? What do you all think? VA is definitely a REPLACEMENT for rebalancing. It takes more time to do it quarterly than yearly or every two years, but you are maximizing your return if you trust Edelson's reasoning behind his "matching" concept. There is no reason to rebalance since the whole point of rebalancing is to take advantage of highs and lows in your various asset classes, and that's what VA does.
Leahkim,
Thanks, your post definitely clears up a lot of things for me. I didn't know that we no longer need to rebalance if we're VA-ing. I was having problems trying to juggle VA and rebalancing.

Your post is very much appreciated, thank you!
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Re: For VA, treat your portfolio like a fund of funds.

Post by WiseNLucky »

InvestmentJunkie wrote:
WiseNLucky wrote:When your portfolio has grown faster than your Value Path, if fund A is low and fund B is high, you would sell fund B to reduce your portfolio towards your target allocation, and then if more is needed, sell both funds A and B in proportion to your target allocation.

When you need to buy to reach your Value Path, if fund B is low buy that first to rebalance back to your target allocation, and then if more is needed, add to both A and B in proportion to your target allocation.
How do you decide how much of each fund to buy or sell? Do you buy or sell until the funds return to their correct percentage of the entire portfolio?

Thanks!
To repeat from my example above:
A simple example:

I have $200,000 in a 50/50 TBM/TSM portfolio to start and plan to maintain that 50/50 split going forward. My target is $1,000,000 after 55 quarters and I assume an overall portfolio return rate of 7% with a 2% per year increase in my contribution level. Put all that in the model and it spits out ~$4,300 per quarter as my starting contribution rate to hit my goal.

Fast forward a quarter. I have put the $4,300 in my side account as I get my paycheck each month. At the end of the quarter, I look in my portfolio and see that my stock balance is sitting at $103,000 and my bond balance is sitting at $101,700. My total portfolio is now $209,000 including the side account.

My value path says I should be at $208,510 split evenly between stocks and bonds since I want to hold a 50/50 allocation. Since I want $104,255 in stocks, I use $1,255 of my side account to buy stocks. Since I want $104,255 in bonds, I use $2,555 out of my side account to buy bonds. That leaves me with $104,255 in each of my stock and bond accounts and $490 in my side account after I make my VA adjustments.

End example.
The same would hold true if my portfolio were below the value path at the end of the quarter. I would add investment money to bring the total portfolio up to my planned path value at the end of the quarter and would then rebalance, in effect, to get the allocation back to 50/50/.
WiseNLucky
User avatar
tetractys
Posts: 4836
Joined: Sat Mar 17, 2007 3:30 pm
Location: Along the Salish Sea

OK I'll try...

Post by tetractys »

Junkie,
M* forum...is that where the Diehards forum started? Something to do with a constellation at dawn?
Access to both Diehard forums is here: http://www.diehards.org

Morningstar is a big investment information firm.
How do you decide how much of each fund to buy or sell? Do you buy or sell until the funds return to their correct percentage of the entire portfolio?
You can use inflows to move closer to your target allocation, if not all the way there, by adding first to funds that are the farthest out. Same with outflows vis a vis. Even if you do this in a strict mathematical way, there's still going to be some subjective decisions, i.e., where to draw the lines.

Good Luck Junkie!, Tet
johnb
Posts: 282
Joined: Sat Mar 17, 2007 12:11 pm

Post by johnb »

I've backtested value averaging on my portfolio and found it to be not all it's cracked up to be. I found it actually did worse than if I'd just allocated new funds to whatever part of my portfolio needed it the most.

From analyzing the data in my spreadsheet, I found 2 interesting phenomena:

1. In the years where most asset classes were skyrocketing higher than expected value path (like in the late 1990s), I lost out on a lot of appreciation by putting money into the money market account.

2. When your portfolio is going down (like in 2001 and 2002), your money market account will be depleted in a hurry. So, if your portfolio goes down 18% let's say, you'll have already emptied out your money market account when it was down 5%. So you won't be getting that many cheap shares.

I also found that if you ARE going to do value averaging, it's best to NOT be conservative in your expectations for performance. So if you expect your portfolio to go up 10%, then plan for 10%, not 9%. If you have a conservative estimate, then on the up years you'll end up with a massively bloated money market account, which will only become depleted in the earliest part of the bear market anyway.

In conclusion, if you're still considering diving into value averaging, I'd suggest backtesting it. You may find that for your portfolio, it won't work. Or you may find that it will -- but the key is to backtest.
Topic Author
InvestmentJunkie
Posts: 28
Joined: Tue Mar 27, 2007 4:56 am

Post by InvestmentJunkie »

All, thanks for the replies.

Johnb, could you share with us your backtesting methodology, and if it's possible for us to replicate it?

Was it done in excel or using some other program?
Gordon
Posts: 331
Joined: Tue Feb 20, 2007 4:56 pm
Location: Lake Havasu City, Arizona

Value averaging book

Post by Gordon »

The author originally wrote a paper for AAII explaining value averaging. The book written several years ago is essentially a reprint, however at the end of each chapter there is a current up date. that makes the book even more useful. I understand that AAII keeps a running conversation going on value averaging.
User avatar
oneleaf
Posts: 2512
Joined: Mon Feb 19, 2007 5:48 pm

Re: Do one VA curve for your entire portfolio

Post by oneleaf »

leahkim wrote:I've read the book, and I am using VA for my entire portfolio (of 11 different Vanguard Index funds).
Leakhim,
I do it exactly the way you do it, and think that is the easiest way to incorporate the VA strategy into a diehard portfolio.

So right now, I have my portfolio (which is 60/40), and I also have a large cash supply right now, since my Value Path is keeping up with very minimal contributions. When I do have to contribute money (quarterly), I put more into the low performing asset classes, and less into the high performing asset classes (which keeps my portfolio balanced).

I see this as a very important strategy that incorporates diehard/boglehead wisdom. If you are ahead of your goals, and your value path has exceeded your target, the VA approach will increase your cash supply, which is effectively decreasing your risk (lower ratio of risk/nonrisk assets). This means that you decrease your risk when your "need to take risk" goes down (i.e. when your value path is better than you expect). You increase your risk when your "need to take risk" goes up (i.e. when your value path is lagging), with your target allocation limiting the maximum risk you are taking (as determined by your risk tolerance).

Looking at it this way, I see VA is not necessarily there to provide you with the best returns, and thus, it doesn't bother me that backtesting may yield inferior results. It is not maximum returns I am looking for as much as I am looking for a method that keeps my need, ability, and willingness to take risk in balance, and is an approach that makes the most sense in the way i face risk.
User avatar
WiseNLucky
Posts: 634
Joined: Tue Mar 13, 2007 5:14 am
Location: South Florida

Re: Do one VA curve for your entire portfolio

Post by WiseNLucky »

oneleaf wrote:Looking at it this way, I see VA is not necessarily there to provide you with the best returns, and thus, it doesn't bother me that backtesting may yield inferior results. It is not maximum returns I am looking for as much as I am looking for a method that keeps my need, ability, and willingness to take risk in balance, and is an approach that makes the most sense in the way i face risk.
Excellent point. If the worst thing that happens to me is that I reach my goal after having incurred lower risk than I expected, my world is a happy place indeed. :D

I'm not in a big rush to retire.
WiseNLucky
Post Reply