Author Topic: Is Value Averaging superior  (Read 6131 times)

DrF

  • Bristles
  • ***
  • Posts: 464
Is Value Averaging superior
« on: October 17, 2014, 10:13:52 AM »
I've read a lot about Value Averaging (VA), and there is still a pretty big debate out there as to it's superiority. It makes sense to me that it would be superior over all other investing methods under the right circumstances. For background on VA just google it or follow these links.

http://en.wikipedia.org/wiki/Value_averaging
http://www.efficientfrontier.com/va.htm
http://www.amazon.com/Value-Averaging-Strategy-Investment-Classics/dp/0470049774/sr=1-1/qid=1162309979/ref=pd_bbs_sr_1/102-4650990-0102506?ie=UTF8&s=booksl
http://jasonkelly.com/resources/strategies/

Here are the circumstances that would make this strategy the most amazing out there.

1) 100% stock allocation immediately with 100% of your funds
2) The stock market MUST go up greater than your target % (probably 3% per quarter) for an undetermined # of quarters (3, 5, 10?) to build up the side account before the 1st correction
3) Your side account must be large enough to be able to buy up to the target % EVERY time the formula indicates
4) Additional steps are necessary if you are adding to your position on a regular basis (paycheck contributions)

With Value Averaging, proponents suggests that a realistic return would be ~12.55% annually (about 20% better than the historic return of the S&P500, ~10%).

What are your thoughts? Bullshit, or easy to do?

To start the debate, here is a blog post discrediting VA.

http://www.michaeljamesonmoney.com/2012/10/value-averaging-doesnt-work.html
« Last Edit: October 17, 2014, 11:24:26 AM by DrFunk »

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Is Value Averaging superior
« Reply #1 on: October 17, 2014, 10:30:53 AM »
Yet another foolproof method to buy low sell high? Oooh, but it has math!

yeah, I'm going to go out on a limb and say no it will not work (i.e. return would be less than buy and hold).

Just one thing I noted in the wiki. You supposedly put in the expected return of stocks (10%?) and then the formula will tell you whether to buy or sell based on current return? 10% is average based on decades, but using that based on 1 year returns? Does not make sense. The market can be above or below that for years. So you'd sell out at the start of a bull market (~2010), then invest nothing until it dips again (not yet..)?

DrF

  • Bristles
  • ***
  • Posts: 464
Re: Is Value Averaging superior
« Reply #2 on: October 17, 2014, 10:53:44 AM »
Just one thing I noted in the wiki. You supposedly put in the expected return of stocks (10%?) and then the formula will tell you whether to buy or sell based on current return? 10% is average based on decades, but using that based on 1 year returns? Does not make sense. The market can be above or below that for years. So you'd sell out at the start of a bull market (~2010), then invest nothing until it dips again (not yet..)?

If I am reading your post correctly you have the strategy nearly correct. Instead of merely putting your desired return equal to the historic return (~10%) you would put it slightly higher (3% per quarter, if you want to re-balance every quarter) which would return ~12.55% annually.

You are correct that in a bull market you would be selling the majority of quarters into a cash/bond side account for an undetermined amount of time. So, a buy and hold strategy would have been particularly amazing over the last 5 years (since the bottom on March 6 2009). If you push your investment timeline out 30-40 years though, would you return ~12.55% annually with the buy and hold strategy?

Scandium

  • Handlebar Stache
  • *****
  • Posts: 2317
  • Location: EastCoast
Re: Is Value Averaging superior
« Reply #3 on: October 17, 2014, 11:24:36 AM »
I read the link at the bottom of your post now. It gives a much better argument for why this is a bad idea then I could.

DrF

  • Bristles
  • ***
  • Posts: 464
Re: Is Value Averaging superior
« Reply #4 on: October 17, 2014, 11:26:22 AM »
Yes, but is he correct? Do the modifications I suggest make VA the optimal strategy?

TLV

  • Bristles
  • ***
  • Posts: 492
  • Age: 32
  • Location: Bellevue, WA
Re: Is Value Averaging superior
« Reply #5 on: October 17, 2014, 11:27:30 AM »
I tried it for about 6 months, a few years ago. I can believe that it has a decent chance of outperforming the alternative of dollar cost averaging, since it tends to amplify the good effects (with DCA you buy the same dollar amount each pay period, so you buy more shares when its cheap and fewer when its expensive. With VA you lower the dollar amount when stocks have gone up more than the amount you predicted and raise it when they've gone up less (or gone down) so you buy many more shares when its cheap and many fewer when its expensive). The biggest caveat mathematically is that you have to start with a reasonably good prediction of growth rates - if you guess too high you'll never hold anything back for the dips, and if you guess too low you'll never be fully invested.

For me it was far too costly on an emotional level to look at stock valuations and do that analysis every pay period. I have a much easier time staying the course if I just dump everything extra from my checking account without looking at the markets.

Sid Hoffman

  • Pencil Stache
  • ****
  • Posts: 751
  • Location: Southwest USA
Re: Is Value Averaging superior
« Reply #6 on: October 17, 2014, 11:31:40 AM »
So you'd sell out at the start of a bull market (~2010), then invest nothing until it dips again (not yet..)?

That seems to be the greatest risk with any sort of buy/sell timing.  I am OK with timing when you enter the market (although articles have shown that generally speaking, the best thing is to simply put the money in the market as soon as you save it) but it seems like where people get bit is by selling.  Since the long term average is for prices to go up, on average, it is the wrong time to sell because on average, prices are going up.  I like the MMM style of buy and hold and rely on things of tangible value, like buy and hold a rental property while it provides you with rental income, or buy & hold value stocks while they provide you with dividend income.  You can't make any money from dividends if you don't first own the stocks and taxes are less on dividends (in America) than they are on capital gains.  Similarly, there's tax advantages available with rental properties.

DrF

  • Bristles
  • ***
  • Posts: 464
Re: Is Value Averaging superior
« Reply #7 on: October 17, 2014, 12:00:04 PM »
The biggest caveat mathematically is that you have to start with a reasonably good prediction of growth rates - if you guess too high you'll never hold anything back for the dips, and if you guess too low you'll never be fully invested.


Wouldn't anything above the historical average of ~10% be worth it? Even if you leave a little on the table because you were never again "fully invested", you would still be ahead of the game.


VirginiaBob

  • Bristles
  • ***
  • Posts: 429
    • LRJ Discounters
Re: Is Value Averaging superior
« Reply #8 on: October 17, 2014, 12:09:47 PM »
The problem with Value averaging is that you need to keep a lot of cash on the side for it to work correctly.  When in reality, you would have just been better investing that idle cash.

Best strategy - invest as much as possible all the time.

DrF

  • Bristles
  • ***
  • Posts: 464
Re: Is Value Averaging superior
« Reply #9 on: October 17, 2014, 12:34:11 PM »
The problem with Value averaging is that you need to keep a lot of cash on the side for it to work correctly.  When in reality, you would have just been better investing that idle cash.

I agree with that statement, but I am wondering if anyone thought that the modifications I suggest would correct for some of the inherent flaws.

waltworks

  • Magnum Stache
  • ******
  • Posts: 4052
Re: Is Value Averaging superior
« Reply #10 on: October 17, 2014, 12:47:10 PM »
No amount of mathematical trickery is likely to beat buy/hold, invest as quick as you can. Period. You *might*, of course, beat buy and hold, but you need to win by a lot offset your transaction and tax drag. That's even more unlikely.

-W

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1700
  • One Does Not Simply Work Into Mordor
Re: Is Value Averaging superior
« Reply #11 on: October 17, 2014, 01:34:46 PM »
It seems like it would have results intermediate between lump-sum-buy-and-hold and dollar-cost-averaging, while requiring more work than either. It doesn't seem worth the time to me, if you have a pile of cash just dump it in.

Personally I have considered value averaging a continuous stream of income, with repaying a mortgage as the alternative. The idea is to calculate a 1% increase for every month. If you have exceeded, don't sell just keep paying the mortgage until you meet the 1% growth target in a few months. If your stocks are not meeting the 1% keep buying while paying the minimum on the mortgage, or whatever proportion meets the 1% goal. It could be a good way to end up with a paid-off house and 7 years of 12% compound growth in investments!

gecko10x

  • Bristles
  • ***
  • Posts: 420
    • SawyerPF
Re: Is Value Averaging superior
« Reply #12 on: October 17, 2014, 02:40:49 PM »
A couple thoughts, based solely on my recollection of researching the strategy a few years ago:

- It requires a fair amount of work
- It wasn't designed as a way to maximize returns, but as a way to get a consistent return from a volatile asset (more like what raven15 suggested above)
- In theory, it could beat a buy/hold strategy if the conditions are right:
   - It exploits the volatility of the investment, so you have to use something highly volatile (like small value)
   - The investment must have a positive future return; the higher the better
   - You must be able to predict the future return nearly perfectly
   - Your trading costs are near zero

I decided against implementing it, as I did not fully understand all the nuances, and feared I was missing something. There are no free lunches.

zinnie

  • Pencil Stache
  • ****
  • Posts: 676
  • Location: Boston
Re: Is Value Averaging superior
« Reply #13 on: October 17, 2014, 03:07:08 PM »
I tried it for about 6 months, a few years ago. I can believe that it has a decent chance of outperforming the alternative of dollar cost averaging, since it tends to amplify the good effects (with DCA you buy the same dollar amount each pay period, so you buy more shares when its cheap and fewer when its expensive. With VA you lower the dollar amount when stocks have gone up more than the amount you predicted and raise it when they've gone up less (or gone down) so you buy many more shares when its cheap and many fewer when its expensive). The biggest caveat mathematically is that you have to start with a reasonably good prediction of growth rates - if you guess too high you'll never hold anything back for the dips, and if you guess too low you'll never be fully invested.

For me it was far too costly on an emotional level to look at stock valuations and do that analysis every pay period. I have a much easier time staying the course if I just dump everything extra from my checking account without looking at the markets.

We tried it for a while and this was about our experience. I had just read a ton of Bernstein books and was convinced it was the best.

Now we do a twist on DCA in that while we invest the same amount every month, we contribute to each fund in a way that keeps our asset allocation where we want it (instead of rebalancing at certain times). It requires making all of the trades manually, but it is actually kind of fun to work out the numbers. And the few times that one fund has been too low to get the AA right, we have pulled out of cash savings to add the extra amount.

I like the concept of value averaging, but it seems like unless you can somehow earn extra money in the months you need more, you are just holding a lot of cash that could have been in investments all along.