SP500 Sigma Events

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

SP500 Sigma Events

Post by tadamsmar »

I thought this was an interesting article on the daily behavior of the stock market.

http://blogs.cfainstitute.org/investor/ ... ma-events/

Fat tails imply more predictability in than there is a much higher percentage of 1 sigma events than for a normal distribution.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

So, you are getting into the RBD strategy and the probability theory behind it? I like it. Now you can make some real money.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:So, you are getting into the RBD strategy and the probability theory behind it? I like it. Now you can make some real money.
Actually, I was looking at the accuracy of my current estimates of win probabilities the 2013 Boglehead Contest, those being based on normal distributions with only 2 trading days.

But the data indicates that you'll average about 1 RBD per month if you use -1% as the definition of a RBD, one per quarter for -2%, one per year for -3%.

I guess you can only do this for ETFs? Or you would have to rebalance late in the day and hope it stays bad.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I think one should use a floating definition of RBD based on relatively recent past price changes. This has been explained a number of times previously here in the forum(s). That is, sometimes 2.5% will be an RBD, but sometimes it will take a bigger drop. A 1% drop will not be an RBD because it is typical and not really bad. The article you linked has lots of insights in this regard, but also leaves out some others.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
nisiprius
Advisory Board
Posts: 42552
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: SP500 Sigma Events

Post by nisiprius »

tadamsmar wrote:I thought this was an interesting article on the daily behavior of the stock market.

http://blogs.cfainstitute.org/investor/ ... ma-events/

Fat tails imply more predictability in than there is a much higher percentage of 1 sigma events than for a normal distribution.
Surely you mean LESS predictability?

The article is a little bogus IMHO, as it is still stuck in a very traditional, statistically tractable, unrealistic view of financial data. He loosens one of the assumptions--a straw man, in my opinion: "What if you feel that mean and standard deviation are not the only way to describe a probability distribution? Then these data do not tell the entire story."

Well, who thinks they do tell the entire story? Mean and standard deviation only tell the entire story if you know, a priori, the shape of the distribution. There are some situations in the real world where you are justified in thinking you might know that. For example, you might reasonably assume the shape is the normal distribution if you think what you are seeing can be modeled as a sum of independent random variables, each with a "reasonable" distribution. If it is an industrial process, in which small changes in temperature or time or composition produce small changes in some characteristic of the end product, that might be reasonable. But there is no reason in the world to think financial data follows that model, and all sorts of reasons to think it doesn't.

The questions that need to be asked, the assumptions that need to be challenged, go far deeper. To begin with, why do we think financial data has a stationary, unchanging distribution? Why would we think that the differences between the stock market of 1920-1940 and those of 1990-2010 are just the luck of the draw, overlying two things that quantitatively have the same underlying model?

Another question, extreme but must be asked: why would we think that stock market data HAS a mean or a standard deviation? It might follow the Cauchy distribution, in which the tails are so fat that no number of observations, however great, can ever determine the mean or the standard deviation, because no matter how many thousands or millions of observations you take, it is capable of throwing out a sample that is so incredible extreme that one single example can outweigh everything that has gone before it.

The relevant book here is The Misbehavior of Markets: A Fractal View of Financial Turbulence, by Mandelbrot and Hudson. According to Mandelbrot, the truth of financial data series is not as bad as the Cauchy distribution, but it is still very bad.

As nearly as I can tell, to the investment community this book is obscene in the literal sense of the word--it is so awful that everybody averts their eyes and will not look at it. They pretend it does not exist. For example, my recently acquired used copy of the textbook, Investments, by Bodie, Kane, and Marcus does not even include Mandelbrot in the "names" index. He is an unperson. Why? Has his work been discredited? I don't think so. I think it is simply ignored, because if he is right, then you simply cannot do the kinds of financial statistics that everybody does.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:I think one should use a floating definition of RBD based on relatively recent past price changes. This has been explained a number of times previously here in the forum(s). That is, sometimes 2.5% will be an RBD, but sometimes it will take a bigger drop. A 1% drop will not be an RBD because it is typical and not really bad. The article you linked has lots of insights in this regard, but also leaves out some others.
Perhaps I will pay more attention to it in an attempt to debunk it. I did a quick search of the forum and it tends to confirm my suspicion that it's one of those tar-baby ideas. It's too ill-defined to make heads or tails of, a waste of time. It's one saving grace is that it is too meaningless to be refuted.

Anyway, you seem enthusiastic, but are you enthusiastic enough to cough up a real definition of what you are talking about?
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

nisiprius wrote:
tadamsmar wrote:I thought this was an interesting article on the daily behavior of the stock market.

http://blogs.cfainstitute.org/investor/ ... ma-events/

Fat tails imply more predictability in than there is a much higher percentage of 1 sigma events than for a normal distribution.
Surely you mean LESS predictability?
Nope, I meant more. Should have said a higher percentage of one sigma or less events. That's what the article means by 1 sigma events.
Last edited by tadamsmar on Fri Dec 27, 2013 1:36 pm, edited 1 time in total.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

tadamsmar wrote:
livesoft wrote:I think one should use a floating definition of RBD based on relatively recent past price changes. This has been explained a number of times previously here in the forum(s). That is, sometimes 2.5% will be an RBD, but sometimes it will take a bigger drop. A 1% drop will not be an RBD because it is typical and not really bad. The article you linked has lots of insights in this regard, but also leaves out some others.
Perhaps I will pay more attention to it in an attempt to debunk it. I did a quick search of the forum and it tends to confirm my suspicion that it's one of those tar-baby ideas. It's too ill-defined to make heads or tails of, a waste of time. It's one saving grace is that it is too meaningless to be refuted.

Anyway, you seem enthusiastic, but are you enthusiastic enough to cough up a real definition of what you are talking about?
And perhaps cough up an actual intellectual justification and perhaps even some evaluation of past performance of a real actionable definition? How often has this definition provided some benefit and if it has provided an on average benefit, what is its magnitude?
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

It can't be Cauchy for the obvious reason that Cauchy has two infinite tails. Even the Bolsheviks could not get a stock market to go below zero.

Maybe you mean the log?
robertalpert
Posts: 457
Joined: Wed Aug 22, 2007 10:09 pm

Re: SP500 Sigma Events

Post by robertalpert »

nisiprius wrote:
The relevant book here is The Misbehavior of Markets: A Fractal View of Financial Turbulence, by Mandelbrot and Hudson. According to Mandelbrot, the truth of financial data series is not as bad as the Cauchy distribution, but it is still very bad.

.... ...... .......

I think it is simply ignored, because if he is right, then you simply cannot do the kinds of financial statistics that everybody does.

Is this a reasonable rationale for the need for investing in CCFs?
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

tadamsmar wrote:Anyway, you seem enthusiastic, but are you enthusiastic enough to cough up a real definition of what you are talking about?
Was this definition not real and precise enough? -> http://www.bogleheads.org/forum/viewtop ... 7#p1028077

Also note that tadamsmar participated in that thread (see post just immediately previous to the linked one).

Unfortunately, one cannot just "look up" what is a RBD. One needs to write a program that implements an algorithm. Nevertheless, the algorithm is precise.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:
tadamsmar wrote:Anyway, you seem enthusiastic, but are you enthusiastic enough to cough up a real definition of what you are talking about?
Was this definition not real and precise enough? -> http://www.bogleheads.org/forum/viewtop ... 7#p1028077

Also note that tadamsmar participated in that thread (see post just immediately previous to the linked one).

Unfortunately, one cannot just "look up" what is a RBD. One needs to write a program that implements an algorithm. Nevertheless, the algorithm is precise.
An algorithm that is designed to look good in a back-test using historical data that is not actionable on Vanguard or TSP because of trading curbs!

Why don't you twerk it some more so it will back-test even better? What's stop to stop you? You obviously have no qualms about fitting it to the historical data.
Last edited by tadamsmar on Fri Dec 27, 2013 3:19 pm, edited 2 times in total.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I use ETFs and since I have IRAs and 401(k)s I have no problem making all this actionable.

I have announced all my RBD trades for the past few years as well. One day I will write a book about it and sell millions of copies and give all the profits to charity.
Wiki This signature message sponsored by sscritic: Learn to fish.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

livesoft wrote:
tadamsmar wrote:Anyway, you seem enthusiastic, but are you enthusiastic enough to cough up a real definition of what you are talking about?
Was this definition not real and precise enough? -> http://www.bogleheads.org/forum/viewtop ... 7#p1028077

Also note that tadamsmar participated in that thread (see post just immediately previous to the linked one).

Unfortunately, one cannot just "look up" what is a RBD. One needs to write a program that implements an algorithm. Nevertheless, the algorithm is precise.
That is a fine definition of RBD. Now what is the trading strategy?

A possible answer could be that you desire a 30/30/40 portfolio (US TSM, X-US TSM, TBM) and you rebalance every time there is a RBD in either stock market and otherwise you let things ride. Or, you hold your monthly new investment in a money market fund and invest on the first RBD that comes along, or you invest new money either on the first RBD or end of the month whichever occurs first and start again the next month.

Or maybe you have something else in mind.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

One doesn't need to use RBDs in any "trading" strategy. They appear to be good indicators of when to look to rebalance and when to tax-loss harvest.

Let me ask a few questions:

1. What day or days were the best days to rebalance into VWO in all of 2013 so far?
2. What day or days were the best days to tax-loss harvest VWO in all of 2013 so far?
3. Did VWO have any RBDs in 2013? What day(s) did they occur?
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:One doesn't need to use RBDs in any "trading" strategy. They appear to be good indicators of when to look to rebalance and when to tax-loss harvest.
If you use only RBDs to rebalance the TSM, then you will typically end up selling, not buying, on low days, because the TSM will have increased more between the RBDs more than the decline on the RBD. The average gain between RBDs will probably exceed 6% while the decline will on a RBD will typically be less than 3%.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

Very interesting. I never would have thought to rebalance out of something that has dropped. Maybe you are on to something?
Wiki This signature message sponsored by sscritic: Learn to fish.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

1. What day or days were the best days to rebalance into VWO in all of 2013 so far?
Very likely not RBD and would take some work to prove one way or the other. It depends a great deal on what happened before and after the RBD. It and of itself it is unlikely RBD matters much at all.

If I had $1,000,000 and had 60% in stocks and 40% in bonds and stocks dropped 2% I would have $588k in stocks. Let's say bonds are flat at $400K. To rebalance I move $5000 from bonds to stocks, or a move of less than 1% of my stock money (about half of one percent in total assets).

If the market bounces right back I have a win of some small percent of less than 1%. If the market heads down (I was induced to rebalance too soon, say compared to rebalance bands) then I have a loss that is some small percent of less than 1%.

Not at all clear rebalancing on a RBD is the right thing to do, but it is clear that the effect is far too small to care.
Last edited by Rodc on Fri Dec 27, 2013 3:51 pm, edited 1 time in total.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
gvsucavie03
Posts: 1460
Joined: Sat Feb 16, 2013 8:30 am

Re: SP500 Sigma Events

Post by gvsucavie03 »

RBD: June 24?
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I must say I enjoy all the handwaving that folks are doing here and lots of unsubstantiated assertions opinions by Rodc and tadamsmar who clearly do not want to "take some work". But no worries.

The lowest price VWO occurred around June 25th. From yahoo finance the 52-week low was 36.02 for VWO. The historical prices around that time:

Code: Select all

Date,Open,High,Low,Close,Volume,Adj Close
2013-06-27,38.27,38.64,38.26,38.57,43154000,38.05
2013-06-26,37.60,38.05,37.56,37.99,37348400,37.47
2013-06-25,37.11,37.26,36.68,37.20,45888900,36.69

2013-06-24,36.50,36.92,36.02,36.50,51498400,36.00

2013-06-21,37.71,38.01,37.10,37.85,45082800,36.84
2013-06-20,38.00,38.02,36.75,37.19,79385400,36.19
2013-06-19,39.90,39.99,38.85,38.88,36388300,37.84
2013-06-18,39.92,40.19,39.81,40.16,19146900,39.09
2013-06-17,40.22,40.37,39.87,40.07,20658000,39.00
But June 24, 2013 was NOT an RBD. However, it was the day a dividend was paid. RBDs occurred on June 20th and June 19th. Since the dividend announcement was known on June 20th, do you think someone would "buy the dividend" or would wait until after the dividend to work with VWO?

Note the date and time of this post: http://www.bogleheads.org/forum/viewtop ... 3#p1732833 At that date and time there was no way to know what would happen afterwards. But we know now what happened afterwards.
Wiki This signature message sponsored by sscritic: Learn to fish.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

handwaving
Snort.

Pot meet kettle. :)

I at least offered up a general example of how the math works.

You have offered up what? Nothing as far as I can tell.

It is your idea. Prove it has value, run a real example.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

So that link with its date/time stamp was not real enough?

If I posted trade confirmations I think you would believe they were photoshopped.
Wiki This signature message sponsored by sscritic: Learn to fish.
gvsucavie03
Posts: 1460
Joined: Sat Feb 16, 2013 8:30 am

Re: SP500 Sigma Events

Post by gvsucavie03 »

One day off.... crud. I'll admit, I just scrolled through the Google Finance chart to find the low. And yes, the symbol for the dividend payout was there, but I honestly didn't know what that was either.

Let me know when you finish that book of yours, Livesoft.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

livesoft wrote:So that link with its date/time stamp was not real enough?

If I posted trade confirmations I think you would believe they were photoshopped.
That is only a list of RBDs. Means nothing. RBDs happen. No one denies that. The issue is can you exploit them for meaningful benefit. That is the issue you have dodged repeatedly in thread after thread.

Assume a strategy of your liking, say a $1,000,000 lump sum 60/40 TSM/TBM. Apply a standard yearly rebalancing and a RBD rebalancing for the last 10 years, a great time for RBD to shine. That is about the most simple basic plausibility test of your ideas.

What happens?
Last edited by Rodc on Fri Dec 27, 2013 7:25 pm, edited 1 time in total.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I don't see a list of RBDs in the link I am writing about.
Wiki This signature message sponsored by sscritic: Learn to fish.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: SP500 Sigma Events

Post by Rodc »

livesoft wrote:I don't see a list of RBDs in the link I am writing about.
One cherry-picked purchase means nothing. You and everyone else knows that.

How does it work through the ups and downs over time? How often does it make a good purchase and how often not? If it wins or loses, by how much?

I see your list above and the one post in the lastest linked thread. Is that all you have? Or is there more?
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:Very interesting. I never would have thought to rebalance out of something that has dropped. Maybe you are on to something?
Historically, stock indexes grow at 10% per year and bonds at 6%. RBDs as defined are going to happen about once per year for stocks and weigh in at >-3% on average. So, of course, most of the rebalancing will be out of stocks on an RBD for stocks.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:I must say I enjoy all the handwaving that folks are doing here and lots of unsubstantiated assertions opinions by Rodc and tadamsmar who clearly do not want to "take some work". But no worries.

The lowest price VWO occurred around June 25th. From yahoo finance the 52-week low was 36.02 for VWO. The historical prices around that time:

Code: Select all

Date,Open,High,Low,Close,Volume,Adj Close
2013-06-27,38.27,38.64,38.26,38.57,43154000,38.05
2013-06-26,37.60,38.05,37.56,37.99,37348400,37.47
2013-06-25,37.11,37.26,36.68,37.20,45888900,36.69

2013-06-24,36.50,36.92,36.02,36.50,51498400,36.00

2013-06-21,37.71,38.01,37.10,37.85,45082800,36.84
2013-06-20,38.00,38.02,36.75,37.19,79385400,36.19
2013-06-19,39.90,39.99,38.85,38.88,36388300,37.84
2013-06-18,39.92,40.19,39.81,40.16,19146900,39.09
2013-06-17,40.22,40.37,39.87,40.07,20658000,39.00
But June 24, 2013 was NOT an RBD. However, it was the day a dividend was paid. RBDs occurred on June 20th and June 19th. Since the dividend announcement was known on June 20th, do you think someone would "buy the dividend" or would wait until after the dividend to work with VWO?

Note the date and time of this post: http://www.bogleheads.org/forum/viewtop ... 3#p1732833 At that date and time there was no way to know what would happen afterwards. But we know now what happened afterwards.
VWO declined less than 2% on the day you bought, the 25th. If that was the right day to buy, then this is evidence against your RBD theory as you defined it earlier. Perhaps you are using the 2-consecutive-day intraday high to low?
User avatar
JoMoney
Posts: 10522
Joined: Tue Jul 23, 2013 5:31 am

Re: SP500 Sigma Events

Post by JoMoney »

livesoft wrote:Very interesting. I never would have thought to rebalance out of something that has dropped. Maybe you are on to something?
That would be momentum / trend trading, the opposite side of contrarian trading that would be similar to most "rebalancing" transactions.
Either way, if you're expecting that it's a long term way to beat the market, it's most probably going to disappoint. Lots of people out there more aggressively using strategies similar (and opposite) to yours. It's a slippery slope if you think that's the path to making money in the market.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

tadamsmar wrote:VWO declined less than 2% on the day you bought, the 25th. If that was the right day to buy, then this is evidence against your RBD theory as you defined it earlier. Perhaps you are using the 2-consecutive-day intraday high to low?
Can someone else please explain to tadamsmar what a dividend is and how it affects the price of a stock or ETF?
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:
tadamsmar wrote:VWO declined less than 2% on the day you bought, the 25th. If that was the right day to buy, then this is evidence against your RBD theory as you defined it earlier. Perhaps you are using the 2-consecutive-day intraday high to low?
Can someone else please explain to tadamsmar what a dividend is and how it affects the price of a stock or ETF?
Nevermind, I did the math wrong.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:
But June 24, 2013 was NOT an RBD. However, it was the day a dividend was paid. RBDs occurred on June 20th and June 19th. Since the dividend announcement was known on June 20th, do you think someone would "buy the dividend" or would wait until after the dividend to work with VWO?

Note the date and time of this post: http://www.bogleheads.org/forum/viewtop ... 3#p1732833 At that date and time there was no way to know what would happen afterwards. But we know now what happened afterwards.
Yeah, we know what happened. The Buy and Holders of the SP500 made 16 percent while you made 9ish percent.

We all feel like geniuses in a bull market.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I suppose if one was a trader, then buying the S&P500 or a US small-cap value fund was the way to make the most money in ETFs in 2013.

However, if one had an asset allocation plan that they followed with allocations to bonds, US large-caps, US small-caps, int'l large, int'l small, REITs, …, and emerging markets, then one would probably at some point be rebalancing into emerging markets in 2013. When was a good time to rebalance into emerging markets in 2013?
Wiki This signature message sponsored by sscritic: Learn to fish.
MindBogler
Posts: 1045
Joined: Wed Apr 17, 2013 12:05 pm

Re: SP500 Sigma Events

Post by MindBogler »

tadamsmar wrote:
livesoft wrote:Very interesting. I never would have thought to rebalance out of something that has dropped. Maybe you are on to something?
Historically, stock indexes grow at 10% per year and bonds at 6%. RBDs as defined are going to happen about once per year for stocks and weigh in at >-3% on average. So, of course, most of the rebalancing will be out of stocks on an RBD for stocks.
You have the flow of money totally backwards. Why would you rebalance out of an asset class that just had a RBD? You wouldn't, you would rebalance into the asset class. I can't figure out if you're making a straw man or you are really this confused.

Disclaimer: I'm totally agnostic wrt the idea of a RBD and whether it means anything should or shouldn't be done.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

MindBogler wrote:
tadamsmar wrote:
livesoft wrote:Very interesting. I never would have thought to rebalance out of something that has dropped. Maybe you are on to something?
Historically, stock indexes grow at 10% per year and bonds at 6%. RBDs as defined are going to happen about once per year for stocks and weigh in at >-3% on average. So, of course, most of the rebalancing will be out of stocks on an RBD for stocks.
You have the flow of money totally backwards. Why would you rebalance out of an asset class that just had a RBD? You wouldn't, you would rebalance into the asset class. I can't figure out if you're making a straw man or you are really this confused.

Disclaimer: I'm totally agnostic wrt the idea of a RBD and whether it means anything should or shouldn't be done.
You are missing a lot of this thread. Livesoft defined process for RBD rebalancing and all I was doing was just describing what happens if you use that process on a typical stock index/bond index portfolio. Nothing more, I was not advocating anything.

Perhaps it is backwards, not sure what the why would be. I have no interest in expressing an opinion on those matters.

It's livesoft's specific defined process I did not make it, he made it.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

LOL! For the record somehow tadamsmar appears to have made my rebalancing process backwards as the other poster politely asked about. One should not sell out of the equity class that has had an RBD. One should be rebalancing INTO the equity class that had the RBD.

One may wish to tax-loss harvest, but that is selling the asset that had the RBD and buying a replacement asset in the same asset class. For example, sell large-cap index and buy total stock market index.
Last edited by livesoft on Sat Dec 28, 2013 12:04 pm, edited 1 time in total.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:I suppose if one was a trader, then buying the S&P500 or a US small-cap value fund was the way to make the most money in ETFs in 2013.

However, if one had an asset allocation plan that they followed with allocations to bonds, US large-caps, US small-caps, int'l large, int'l small, REITs, …, and emerging markets, then one would probably at some point be rebalancing into emerging markets in 2013. When was a good time to rebalance into emerging markets in 2013?
VWO has historically grown grown quite a bit faster than bonds. So, most of one's rebalancing events would have been moves out of VWO and into bonds, assuming a trigger that happens not too often like once a year or so. RBD as you have defined it is such a trigger.

You are obviously advocating selling on RBDs in most cases. But, in your defense, you don't seem to understand what you are advocating.
MindBogler
Posts: 1045
Joined: Wed Apr 17, 2013 12:05 pm

Re: SP500 Sigma Events

Post by MindBogler »

I read the whole thread and I didn't miss anything. You seem to be claiming that if there is a RBD in Emerging Markets that all other asset classes will also have one. Why wouldn't you take some out of bonds and place it into EM? How about if EAFE is up 1% and EM is down 3%? Or how about harvesting some money from any asset class that is flat or positive? Why on earth would you rebalance out of a down asset class? You've made the assertion twice now that Livesoft would be doing exactly that, so I think it is time for you to explain why that would ever be the case.

If I understand Livesoft correctly, this strategy can be used for TLH or rebalancing.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

tadamsmar wrote:Nevermind, I did the math wrong.
http://www.youtube.com/watch?v=V3FnpaWQJO0
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: SP500 Sigma Events

Post by Epsilon Delta »

I got to the second paragraph and decided that anybody who could write:
For example, from 3 January 1950 through 31 July 2012, the average daily return of the S&P 500 was 0.03%, and the standard deviation was 0.98% (source: Yahoo Finance, CFA Institute). These results are remarkably similar to the mean and standard deviation of the normal distribution of 0 and 1, respectively. This suggests that daily returns for the S&P 500 closely approximate the normal distribution, and that returns follow a random walk.
has no idea what they are talking about.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:LOL! For the record somehow tadamsmar appears to have made my rebalancing process backwards as the other poster politely asked about. One should not sell out of the equity class that has had an RBD. One should be rebalancing INTO the equity class that had the RBD.
Lets say you have 50/50 TBM amd VWO. VWO has a good year and doubles so now you have about 33/67 whereas you target AA is 50/50. At this point VWO has a 3% down day which is a RBD. So do you rebalance to 50/50? If not, when do you ever rebalance to your target?

Perhaps the term "rebalance" has some special meaning to you?

Not sure what the overall process you propose is.

I rebalance on my birthday to my target AA. The rebalancing has to be based relative performance over the year, not just what happened in the markets on my birthday. That's the only way I could possibly rebalance to my target AA.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

There is rebalancing into something when it drops and rebalancing out of something when it goes way up. For me the rebalancing triggers are different in each of those cases. They may not be for you, but they are for me.

You asked when I would rebalance to my target. If stocks have gone up, I rebalance out of stocks when they are 3% to 5% over my allocation. If stocks have gone down, I rebalance into stocks when they are 3% to 5% under my allocation. In addition to that, I always buy at the end of a RBD (or as soon as practical thereafter).

I do not use a calendar-based rebalancing trigger. I also strive for balance when making new additions to my portfolio. One can read about rebalancing in the Bogleheads wiki:
http://www.bogleheads.org/wiki/Rebalancing
Wiki This signature message sponsored by sscritic: Learn to fish.
gvsucavie03
Posts: 1460
Joined: Sat Feb 16, 2013 8:30 am

Re: SP500 Sigma Events

Post by gvsucavie03 »

One could also argue that your RBD philosophy can also act as a secondary rebalancing tool since you are buying an under-performing ETF....

I find it fascinating... you take a well-researched and even scholarly approach with what is essentially "fun" money. Have you ever done a hypothetical comparison with the dollars you would have invested had you not done the RBD approach versus what is actually sitting there today? I know your rebalance triggers would have behaved differently in that scenario, but anything "rough and dirty"? I'm curious to know if this philosophy has significant gain.

Aside from what you have written about in your posts, is there any other research out there to justify this as a smart investing move? Why 150 days/4% drop? Is there any research to back this?
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

I have done my own research and backtesting*. I reported on that obliquely in this post: http://www.bogleheads.org/forum/viewtop ... 85#p877285 which I know is not exactly what you wanted. I am mostly curious if anyone else has done backtesting that they want to write about. So far, I only read of folks who don't want to understand the method and furthermore twisted words that I have written into their own interpretations, so I essentially gave up explaining it. Plus I don't feel like publishing my research without getting paid for it.

I have also "blogged" about all my RBD purchases in real-time elsewhere. I do not want to reveal that info on the internet because one day it will form a published track record for the book I will write. That is, there will be no "cherry-picked" results because they were all done in real-time without knowing the future (just like the VWO RBD I noted in this thread).

Nevertheless, I have fun with this and find it eye-opening how some folks cannot understand plain English and add things such as rebalancing out of an ETF that has had a RBD (when did I ever write that???). I understand now how some folks cannot get the idea of investing with a low-expense-ratio, passively-managed index fund approach with portfolio rebalancing and tax-loss harvesting. Their brains are wired differently from mine and they just don't understand some things that are plain and obvious to me.

*I do not use 150 days / 4%, see, e.g.: http://www.bogleheads.org/forum/viewtop ... 1#p1838451

Anyways, I think folks who are interested should do the work themselves. Those who are not interested are free to shoot me with arrows and throw rocks at me because that really doesn't bother me. You can imagine I like the attention whether good or bad. Their comments also give me ideas for things to put in the book. :)
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
nisiprius
Advisory Board
Posts: 42552
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: SP500 Sigma Events

Post by nisiprius »

livesoft wrote:...I don't feel like publishing my research without getting paid for it.

I have also "blogged" about all my RBD purchases in real-time elsewhere. I do not want to reveal that info on the internet because one day it will form a published track record for the book I will write...
Would it then, at this time, be unfair to describe this as a "secret formula?"
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

nisiprius wrote:
livesoft wrote:...I don't feel like publishing my research without getting paid for it.

I have also "blogged" about all my RBD purchases in real-time elsewhere. I do not want to reveal that info on the internet because one day it will form a published track record for the book I will write...
Would it then, at this time, be unfair to describe this as a "secret formula?"
It would be unfair since the "secret formula" is exactly as I have posted it on this forum, so it cannot be secret. Quite a number of people get it and do not try to make it more complicated than it is. Others try to make it more complicated than it is and I am not sure why.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
JoMoney
Posts: 10522
Joined: Tue Jul 23, 2013 5:31 am

Re: SP500 Sigma Events

Post by JoMoney »

livesoft's approach, and enthusiasm for a system, reminds me of books by Michael Covel... but probably trading in opposite directions, and probably less aggressive.
At least Covel acknowledges that trading (in the aggregate of all the market) is zero-sum, dog-eat-dog, and not everyone can be above average.

What I keep thinking when I see what's being suggested, is this:
http://www.dummies.com/how-to/content/d ... ategy.html
But trying to disguise it into a framework of portfolio "rebalancing", which should be about controlling the risk profile. The idea that everyone could follow that path and obtain a superior return is provably wrong.
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
chisey
Posts: 277
Joined: Mon Apr 16, 2007 10:47 am

Re: SP500 Sigma Events

Post by chisey »

livesoft wrote:There is rebalancing into something when it drops and rebalancing out of something when it goes way up. For me the rebalancing triggers are different in each of those cases. They may not be for you, but they are for me.

You asked when I would rebalance to my target. If stocks have gone up, I rebalance out of stocks when they are 3% to 5% over my allocation. If stocks have gone down, I rebalance into stocks when they are 3% to 5% under my allocation. In addition to that, I always buy at the end of a RBD (or as soon as practical thereafter).

I do not use a calendar-based rebalancing trigger. I also strive for balance when making new additions to my portfolio. One can read about rebalancing in the Bogleheads wiki:
http://www.bogleheads.org/wiki/Rebalancing
Two questions about your approach then:

1) You always buy on or just after a RBD. Where does the money for the buy come from? Do you hold cash while waiting, sell some other asset class, or what? How long will you hold cash, if that's the approach, waiting for an RBD? This will answer a question on the mechanics that I'm unclear about.

2) The combo of rebalancing and RBD seem to work against one another. Example: suppose EM floats up near 5% above desired allocation, then a RBD drives it down to 4.7%. Your RBD buy could, theoretically, bump you above your rebalancing band and force you to sell. At the very least it puts you closer to it, and your RBD buy may actually force you to rebalance out of EM at a lower price than you would have with no RBD buy. I'm not sure how that washes out (and I know you've done the research on this), but it seems counterintuitive to me to buy an asset that's near a high side rebalancing band.
livesoft
Posts: 74489
Joined: Thu Mar 01, 2007 8:00 pm

Re: SP500 Sigma Events

Post by livesoft »

1) My asset allocation calls for having fixed income (bond) funds. I exchange from a bond fund into the equity fund. I do not hold cash waiting for an RBD.

2) Yes, I suppose something going up and being above a rebalancing band could have an RBD which would call for buying it. But I am going to say that when it has the RBD, that would put it below the rebalancing band. When it rises back above the rebalancing trigger, it should be sold. OTOH, it might be more likely that when something went above the rebalancing trigger that it would be sold. Then if it had an RBD the next day, that would cause re-buying. Then if it went back up, it would trigger rebalancing and it would be sold again.

So far it seems to me that an RBD is like the final whoosh coming out of a balloon that signals the end of outgassing and indicates a possible change in momemtum, so the 2nd scenario is probably unlikely. Nevertheless, one has to be careful and after a buy, one has to be ready to sell after a quick run-up. If one goes back to where the money comes from in (1) above, one has to be prepared to put the money back into fixed income after a pop.

To tie this in to the article linked in the OP, that article does NOT indicate that the high sigma (3, 4, …, 10 sigma) events usually have positive and negative events occurring close together in time. That's why it is ridiculous to talk about missing the 10 best days and the 10 worst days. Often (take a look at the history) a big drop is following by a big pop or if the big pop occurs first, then it is followed by a big drop. That is, volatility works both ways. A classic instance is what happened in August 2011.
Wiki This signature message sponsored by sscritic: Learn to fish.
User avatar
Topic Author
tadamsmar
Posts: 9164
Joined: Mon May 07, 2007 12:33 pm

Re: SP500 Sigma Events

Post by tadamsmar »

livesoft wrote:There is rebalancing into something when it drops and rebalancing out of something when it goes way up. For me the rebalancing triggers are different in each of those cases. They may not be for you, but they are for me.

You asked when I would rebalance to my target. If stocks have gone up, I rebalance out of stocks when they are 3% to 5% over my allocation. If stocks have gone down, I rebalance into stocks when they are 3% to 5% under my allocation. In addition to that, I always buy at the end of a RBD (or as soon as practical thereafter).

I do not use a calendar-based rebalancing trigger. I also strive for balance when making new additions to my portfolio. One can read about rebalancing in the Bogleheads wiki:
http://www.bogleheads.org/wiki/Rebalancing
When you buy or rebalance (you have now used both terms for the same thing in this thread) at the end of an RBD, do you rebalance to the ratios of the previous day, or do you rebalance to your target AA? I would deduce that it must be the ratios (in other words, the defacto AA) that held on the previous day, otherwise you still may find yourself rebalancing "backwards" to your target AA since that other type of rebalancing you use (3% to 5% bands) will not prevent it. But it could be you just have not thought through it. It will prevent it less if you use 5% rather than 3%, but even 3% will not prevent it given the definition that you gave earlier for a really bad day.
Last edited by tadamsmar on Sun Dec 29, 2013 12:59 am, edited 1 time in total.
Post Reply