The Ivy Portfolio

102 replies [Last post]
chief123's picture
Offline
Joined: 2008-10-28

Has anyone looked into this at all?

snaggletooth's picture
Offline
Joined: 2007-07-13

What portfolio?  Like all the funds, or any one inparticular?

noggin's picture
Offline
Joined: 2004-11-30

I like English Ivy and I dislike poison ivy with the exception of Drew Barrymore or Alyssa Milano.

B24's picture
B24
Offline
Joined: 2008-07-08

Yes, I've read a lot about it.  The only concern I have is that you really have to commit to doing it, without fail.  That may require some serious intestinal fortitude at times.  Taxes are another consideration for NQ accounts.  I can't imagine doing this for a large number of clients on a non-discretionary basis.  I am following EMA's for my larger client's portfolios, but do not use it simply for those 5 asset classes.  So I use sort fo a blended buy-and-hold and EMA strategy.  For example, for conservative clients, I may be buy-and-hold for fixed income, though I may shift between nominal Treasuries and TIPS, as well as moving in adn out of high-yield at times.  For equities, I also incorporate emerging market equities.  But I am still partial to active management (with some assets), so I ause this concept in conjunction with funds like First Eagle Global, IVY Asset Strategy, Mutual Discovery, etc. funds that don't necessarily represent one asset class.

Moraen's picture
Offline
Joined: 2009-01-22

noggin wrote: I like English Ivy and I dislike poison ivy with the exception of Drew Barrymore or Alyssa Milano.

Only Alyssa Milano. Drew Barrymore is disgusting.

noggin's picture
Offline
Joined: 2004-11-30

She has definitely taken a turn for the worse.

B24's picture
B24
Offline
Joined: 2008-07-08

I never liked her that much to begin with. 
 
Alyssa Milano...oh yeah.

Mike Damone's picture
Offline
Joined: 2004-12-01

Moraen wrote: noggin wrote: I like English Ivy and I dislike poison ivy with the exception of Drew Barrymore or Alyssa Milano.   Drew Barrymore is disgusting.
 

This is the most ridiculous thing ever posted on this board.  <?: prefix = o ns = "urn:schemas-microsoft-com:office:office" />

Moraen's picture
Offline
Joined: 2009-01-22

Mike Damone wrote: Moraen wrote: noggin wrote: I like English Ivy and I dislike poison ivy with the exception of Drew Barrymore or Alyssa Milano.   Drew Barrymore is disgusting.
 

This is the most ridiculous thing ever posted on this board.  <?: prefix = o ns = "urn:schemas-microsoft-com:office:office" />

Truth hurts sometimes. She is downright NASTY looking.

You remind me of a friend of mine. When told that he had no standards, he replied, "I have standards, I just don't try to meet them".

Mike Damone's picture
Offline
Joined: 2004-12-01

Moraen wrote: Mike Damone wrote: Moraen wrote: noggin wrote: I like English Ivy and I dislike poison ivy with the exception of Drew Barrymore or Alyssa Milano.   Drew Barrymore is disgusting.
 
This is the most ridiculous thing ever posted on this board.  <?: prefix = o ns = "urn:schemas-microsoft-com:office:office" />
Truth hurts sometimes. She is downright NASTY looking. You remind me of a friend of mine. When told that he had no standards, he replied, "I have standards, I just don't try to meet them".
 

Spaceman Spiff's picture
Offline
Joined: 2006-08-08

Drew has her moments, but I vote for Alyssa too.  I grew up watching her grow up.  I don't know that I've seen those movies.  I know the titles, but I can't say I remember watching the movies. 

SometimesNowhere's picture
Joined: 2008-12-22

Drew Barrymore is a heifer (though I think I could drink her cute) .

AdvisorControl.com's picture
Joined: 2009-05-29

The Ivy Portfolio is a decent read and a simple yet effective method for beating the market with lower risk (at least historically).  I've talked with Mebane Faber a few times and the dude is actually a lot smarter than the book reads.  He's capable of doing much, much more from a quant perspective; but the book is supposed to be for anyone to benefit from - including DIYers.I would NOT use the method for client assets in any major way.  There will be prolonged periods where the market pounds you and that more often than not will result in clients jumping ship no matter how sound the strategy is.For fun - DB is below average and always has been in the looks department.

B24's picture
B24
Offline
Joined: 2008-07-08

AdvisorControl.com wrote:The Ivy Portfolio is a decent read and a simple yet effective method for beating the market with lower risk (at least historically).  I've talked with Mebane Faber a few times and the dude is actually a lot smarter than the book reads.  He's capable of doing much, much more from a quant perspective; but the book is supposed to be for anyone to benefit from - including DIYers.I would NOT use the method for client assets in any major way.  There will be prolonged periods where the market pounds you and that more often than not will result in clients jumping ship no matter how sound the strategy is.For fun - DB is below average and always has been in the looks department.
 
Can you elaborate on what you mean by this?  I have followed the numbers in his white paper, and the strategy makes great sense, especially in down markets.  How would the markets pound you?  I guess I view the purpose of the strategy is to AVOID getting pounded.  Yes, you get get pounded while you are waiting for the 200SMA sell signal, but at least there is downside protection.  It beats basic buy and hold.  Are you referring to upside splippage?

chief123's picture
Offline
Joined: 2008-10-28

I would think there would be a great deal of upside you miss during a market bounce back, but I think by missing the great collapse you would still be ahead while mitigating risk... For example if had done this with Ishares, it looks like you would dump out around Jan 4-7 2008... Buy back in briefly in may, but then be out again by mid june all the way til late june early july of 09.. which means you missed the upswing from march-june, but also means you missed a lot of the collapsse from july 2008-dec 2008.. The only issue would be tax considerations for non-qualified accounts.

B24's picture
B24
Offline
Joined: 2008-07-08

Agreed.  However, there can be "manual overrides" on this as well.  I know some people do not buy back in immediately after crossing the SMA line to avoid a "whipsaw" (as mentioned in May 2008).  And you could have conceivably bought back in earlier than June 2009 (or DCA'd back in) if you felt that a bottom was in.  Although that sort of "speculation" would run counter to the purpose of the technique.  But if followed to the "T", and depending on what happens the rest of this year, you could have a losing year if you got back in around June and the rest of the year floated down a bit, while the market is up considerably.  I guess you have to give some to gain a lot.

AdvisorControl.com's picture
Joined: 2009-05-29

B24 wrote:AdvisorControl.com wrote:The Ivy Portfolio is a decent read and a simple yet effective method for beating the market with lower risk (at least historically).  I've talked with Mebane Faber a few times and the dude is actually a lot smarter than the book reads.  He's capable of doing much, much more from a quant perspective; but the book is supposed to be for anyone to benefit from - including DIYers.I would NOT use the method for client assets in any major way.  There will be prolonged periods where the market pounds you and that more often than not will result in clients jumping ship no matter how sound the strategy is.For fun - DB is below average and always has been in the looks department.
 
Can you elaborate on what you mean by this?  I have followed the numbers in his white paper, and the strategy makes great sense, especially in down markets.  How would the markets pound you?  I guess I view the purpose of the strategy is to AVOID getting pounded.  Yes, you get get pounded while you are waiting for the 200SMA sell signal, but at least there is downside protection.  It beats basic buy and hold.  Are you referring to upside splippage?You nailed it.  Client are funny like this.  You could save them from a 40% decline, but as soon a their neighbor is up 30% and you're still on the sidelines (remember, Mebane only used monthly data to remove noisy trades) the clients begin to question the strategy - they think the model failed.Case in point, one of my models was up nearly 50% through June 30 this year.  It's since given back 10% while the market has risen - client phone calls come in swarms and one person even pulled their account.  Investor behavior is a fickle thing; the great inefficiency in supposedly efficient markets.So this could be a portion of ones strategy, but I would not use it in a major way.In the quant geek world (which I guess I'd be a card carrying member) we'd call this strategy a single factor model.  It's not adaptive (which is bad), it's very simple (which is good), and it's based on the past 100 years using monthly data (which is very bad).  I won't waste everyone's pixels with crazy long explanations; but think honestly about the past 100 years: will there really be a high correlation the next 100 years?This is really a trick question.  We're talking about 1200 units of data for each market analyzed.  The strategy seemed to work on each market.  The main problem is the high degree of correlation among total returns of the units of data.  In non geek-speak, this simply means there is not a real measurable probability of the next 100 years of the markets analyzed replicating the next 100 years.As a somewhat goofy example; the US went through one of the greatest industrial booms of any nation in history during the last 100 years.  In reality, nearly all the markets analyzed in Mebane's book had equally historic booms for their respective markets.  So the questions are, will this continue?  Will a new market leader emerge?  Will new asset classes emerge?  Will investor behavior change?Honestly, I have no clue.My point: I don't really have one - but I would not put a major chunk of my clients dough in a single factor model that has some lack of large numbers type data to it.  Anyone remember reading Beating the Dow?  Hasn't worked so well since publication - markets changed, accounting rules changed, etc.The book, fwiw, is very good.  The strategy, is better than what I see 90% of advisors doing (unscientifically, of course).  A little tweaking to the parameters and maybe an additional factor or two - it could be way better.I'd be willing to prove this and post a little software app for free if you all are interested.  Something that takes Mebane's strategy into Excel with Macros in one sheet and then some mods and tweaks for making it better in another.  If requested - I'll deliver it and any feature requests can be PM'd to me.Good night,J

Squash1's picture
Offline
Joined: 2008-11-19

Hey buddy, then how come your returns on your site say this:
                           ytd(7/30/09)      1yr     avg(11/04)
RWA Aggressive                    -1.73%     -20.18%     12.62%
RWA Moderate Aggressive             2.46%     -15.42%     12.37%
RWA Moderate                    6.76%     -10.62%     11.98%
RWA Moderate Conservative     11.16%     -5.81%     11.47%
RWA Conservative             15.67%     -1.02%     10.84%
RWA High Income                 20.29%      3.75%      10.09%
RWA Income                      13.30%      3.16%     7.05%
S&P 500                             9.52%     -22.89% -3.33%

Not bad returns but not what you say either..

AdvisorControl.com's picture
Joined: 2009-05-29

Squash -
 
Those are allocations made up of various models.  We have 14 models all together.  I presume you're referring to the post I made about a model being up 50% and now only being up 40% (or so).  That model is part of the "high income" allocation - essentially a treasury arbitrage model that takes positions in 20 year treasuries when it's probable rates will fall and rising rate (inverse) funds when it's probable rates will rise.  The model makes up roughly 1/2 of the high income and 1/3 of the regular income.
 
IM(not so)HO - nobody should put all their $$ in any one model.  So allocations of models is a better solution for our clients.
 
Hope this helps - I hate unscrupulous liars as much as anyone.
 
J

B24's picture
B24
Offline
Joined: 2008-07-08

Jason, in your previous post above, are you suggesting that a lack of major moves in the market in the future will make moving-average difficult to execute?  I'm having a hard time understanding why it matters what type of market changes occur in the future - the market will always go up, down or sideways, regardless of the cause.  I'd be inetrested in the app you are referring to.
Thanks.

Moraen's picture
Offline
Joined: 2009-01-22

I'd like to see the app as well.

Back to Alyssa Milano though...

Jason don't you know that you can't beat the market with your fancy models? As an advisor, you are supposed to put your money in the hands of EXPERIENCED money managers, or create great ALLOCATION models out of index funds and ETFs.

I don't care what you've done the past few years... you can't REPLICATE that. Your clients will leave in the future.

The markets are soooooooooooo EFFICIENT.

Sorry guys, had to bring this argument back. I am getting absolutely NO work done today.

AdvisorControl.com's picture
Joined: 2009-05-29

Moraen wrote:I'd like to see the app as well. Back to Alyssa Milano though... Jason don't you know that you can't beat the market with your fancy models? As an advisor, you are supposed to put your money in the hands of EXPERIENCED money managers, or create great ALLOCATION models out of index funds and ETFs. I don't care what you've done the past few years... you can't REPLICATE that. Your clients will leave in the future. The markets are soooooooooooo EFFICIENT. Sorry guys, had to bring this argument back. I am getting absolutely NO work done today.
 
Good stuff.  I'm guessing that you're joking.

AdvisorControl.com's picture
Joined: 2009-05-29

I'm not getting jack done today either - so off to the golf course.  Have a nice weekend everyone.

Moraen's picture
Offline
Joined: 2009-01-22

AdvisorControl.com wrote: Moraen wrote:I'd like to see the app as well. Back to Alyssa Milano though... Jason don't you know that you can't beat the market with your fancy models? As an advisor, you are supposed to put your money in the hands of EXPERIENCED money managers, or create great ALLOCATION models out of index funds and ETFs. I don't care what you've done the past few years... you can't REPLICATE that. Your clients will leave in the future. The markets are soooooooooooo EFFICIENT. Sorry guys, had to bring this argument back. I am getting absolutely NO work done today.
 
Good stuff.  I'm guessing that you're joking.

Yes, joking.

Mike Damone's picture
Offline
Joined: 2004-12-01

XBOX Live?

B24's picture
B24
Offline
Joined: 2008-07-08

iceco1d wrote:Moraen,If and when you decide to man-up and get on XBox Live, we are going to have an in-depth chat about this. 
 
ICE, you need a wife and kids....

chief123's picture
Offline
Joined: 2008-10-28

Or a friend...

chief123's picture
Offline
Joined: 2008-10-28

You can't be much younger than me.

chief123's picture
Offline
Joined: 2008-10-28

U said I was too cool.. I figure we are similar ages, we can be friends..

Moraen's picture
Offline
Joined: 2009-01-22

I'm over thirty. Ice and I are friends. I like XBOX. I have a wife and kid. That's why I play XBOX. Problem is, I have a wife and kid. Hard to play XBOX.

Ice's wife still let's him do stuff.

But that's all right. Soon, I will begin "studying" late at night, while she is asleep. I'm all about it Ice. Next week starts school, then I have an excuse to be up late.

howie's picture
Offline
Joined: 2008-04-14

Jason - I'd like to see the app that you mention. Please post when you get a chance.

B24's picture
B24
Offline
Joined: 2008-07-08

Moraen wrote: I'm over thirty. Ice and I are friends. I like XBOX. I have a wife and kid. That's why I play XBOX. Problem is, I have a wife and kid. Hard to play XBOX.

Ice's wife still let's him do stuff.

But that's all right. Soon, I will begin "studying" late at night, while she is asleep. I'm all about it Ice. Next week starts school, then I have an excuse to be up late.

Hopefully ICE understood that my comment was just a joke. It was more self-depricating humor than anything....as in "I'm married with kids so can't imagine having time to play XBox". I should have used one of those goofy little emoticons...like this , or maybe this

It's funny, one of my good friends at Jones used to talk about XBox all the time. I went to his wedding a few years back, and since then the XBox talk slowed a bit, he had a little girl about 6 months ago, and now I haven't heard him talk about it once.   

B24's picture
B24
Offline
Joined: 2008-07-08

...and by the way ICE, you keep playing XBox, you'll have no shot at that kid.

Moraen's picture
Offline
Joined: 2009-01-22

B24 wrote: ...and by the way ICE, you keep playing XBox, you'll have no shot at that kid.

Whenever my wife "lets" me play. She watches for about twenty minutes and then goes and puts on lingerie. Next thing I know, I could care less that an eight your old just sniped me in Halo.

He'll have a kid before he knows it.

noggin's picture
Offline
Joined: 2004-11-30

iceco1d wrote:Sorry guys, but I don't think kids are my "thing."
I don't know man, I was babysitting my 5 yr old and her best friend this weekend while Mom was out. The daughter's friend started calling me poopyhead when I wouldn't let them do something.......you can't really put a value on that.......

AdvisorControl.com's picture
Joined: 2009-05-29

howie wrote:
Jason - I'd like to see the app that you mention. Please post when you get a chance.

 
Just an update:
 
I coded out my enhanced version of The Ivy Portfolio in VBA this weekend.  I'll convert it to excel and disable macros so you all can have a fully functional, easy to use Ivy on Streroids to use as you wish by the end of this week.
 
The early details:
 
No leverage
No shorting
Light turnover
Better than 15% annualized returns
Lower than 13% standard deviation
Can all be done with NL NTF funds
Went back to 1996 for the backtest since not all the funds had longer history
 
If there's any special requests for ways to make the strategy better in anyones eyes let me know and I'll throw some macros in for optimization.  Keep in mind this is all in sample as we obviously know in retrospect what would've worked and what wouldn't have.
 
JW

chief123's picture
Offline
Joined: 2008-10-28

Any update?

AdvisorControl.com's picture
Joined: 2009-05-29

chief123 wrote:Any update?
Will be done by Friday.

AdvisorControl.com's picture
Joined: 2009-05-29

Alright - so here's another update:Switching to excel with no
macros was a little more difficult than I originally thought.  I did
this so advisors could see the minor changes I made to Mebane's
algorithm.  I have it done and also did the tweaking of using actual
funds advisors could use.  This shortened the backtest and changed the
#'s from using raw index data.Here's the yearly numbers for the past 10 years:Strategy1998 - 15.17%1999 - 13.52%2000 - 0.92%2001 - 1.79%2002 - (1.73%)2003 - 22.97%2004 - 6.37%2005 - 38.51%2006 - 10.26%2007 - 14.44%2008 - 3.09%2009 ytd as of 7/31 - 3.02%This
is using on 6 asset classes, which lowers the return potential but
makes curve fitting less an issue.  For fun - I plugged in the NAS 100
(which is not otherwise used) and the 1999, 2003 and 2009 returns all
nearly doubled.  That, is fitting (somewhat), and the excercise was
only to improve Mebane's returns - which was done.The mechanics
used were a simple momentum ranking system to dynamically change the
position size to favor the single asset class with the best
intermediate term momentum.  The end result is a super easy tactical
allocation program that would beat 95%+ of supposedly smart,
professional fund/sma managers the past 10 years with only 10 minutes
trading per month.I'll publish over at my site by Friday in the
'free stuff' section.  Any questions or tweaks people would like to see
just let me know.

B24's picture
B24
Offline
Joined: 2008-07-08

Jason, how would something like this perform during an extended bull-market run, like the 80's/90's?  It is obviously a great program during extremely turbulent/secular bear markets, but would it underperform over long positive periods?
 
Also, how do you measure momentum?

AdvisorControl.com's picture
Joined: 2009-05-29

B24 wrote:Jason, how would something like this perform during an extended bull-market run, like the 80's/90's?  It is obviously a great program during extremely turbulent/secular bear markets, but would it underperform over long positive periods?
 
Also, how do you measure momentum?B -Extended bull markets would've bee a problem for Mebane's version - which is why I tweaked it a little.  By making the *best* asset class (ie; the one with the most momentum) a bigger weight that problem is erased.I'm measuring momentum based on price, with two indicators blended and based on 6 month and 12 month combined measures.For kicks, I think I'll also blend a "hedged" version into the spreadsheet that actually shorts the least desirable asset class while holding long the other 5, and perhaps adding a little leverage to the top asset class.  This would essentially be like a 110/10 fund for net 100% investment but with a little hedging, of course.  I think this will drop the volatility and increase the return while also eliminating any negative years in entirety.I'll keep you all posted.J

B24's picture
B24
Offline
Joined: 2008-07-08

Thanks Jason.  What I meant to ask was, where do you gather data to measure (price) momentum.  Is there a service you subscribe to that provides the data, or do you simply write a program to extract daily prices from a particular service?

AdvisorControl.com's picture
Joined: 2009-05-29

I write the code myself and simply do auto EOD downloads from Yahoo Finance for something this simple.

LockEDJ's picture
Offline
Joined: 2009-07-06

AdvisorControl.com wrote: I write the code myself and simply do auto EOD downloads from Yahoo Finance for something this simple.

Looking forward to reading the code ... if you'll provide it. I love reading VB.

AdvisorControl.com's picture
Joined: 2009-05-29

LockEDJ wrote: AdvisorControl.com wrote: I write the code myself and simply do auto EOD downloads from Yahoo Finance for something this simple.

Looking forward to reading the code ... if you'll provide it. I love reading VB.Hate to disappoint - but I removed the macros and transferred the program piece meal to excel so non tech-junkies would get to see the algorithms (which were quite simple anyway).  I may turn this into a fully automated VB program with a desktop icon and .exe files...depends on how motivated and generous I'm feeling.  That way advisors could just click a couple buttons and it would run a full backtest on whatever funds they use and tell them exactly what the program says to do each month.We'll see...J

howie's picture
Offline
Joined: 2008-04-14

Jason, in measuring momentum are you adding in volume as a factor as well or just straight price movement? Have you considered adding in volume as a factor for momentum?

chief123's picture
Offline
Joined: 2008-10-28

i dont see it yet

AdvisorControl.com's picture
Joined: 2009-05-29

Sorry guys/gals - I forgot I was set to go hiking over the weekend and haven't had the time to finish post my work just yet.  There's a couple of things I wanted to do to get it ready for the average non-techie to use.It'll be done sometime this week and I'll let you all know when it's been posted.J

AdvisorControl.com's picture
Joined: 2009-05-29

Oh...and the volume thing - it's not nearly as important as the technical analysts make it out to be.  And since we're looking at trading global markets (some being non stock markets) - volume is non-existent.  In the world of data analytics, you need comparable and complete data sets with large numbers of non manipulable data.  So just prices and dates works just fine for this stuff.

AdvisorControl.com's picture
Joined: 2009-05-29

Forum Folks,If you go to our "free stuff" section you'll find the ultra-crude version of The Ivy Portfolio is excel.  Keep in mind it was a Visual Basic doc that I pretty much dismantled so the average excel user could decipher the excel formulas used to create the portfolio.With modest excel experience you all should be able to figure it out.  Pretty much within excel you just get the month end prices for the funds I lay out at the top of the doc and copy/paste the rest.  The system tells you what fund should have what percent and tabulates the rest.If there's enough interest (ie, more than just a couple people) I'll consider doing an API for the VB doc so the whole program works like any other desktop icon auto-run program.This is kind of like a quasi-social experiment, as this model is ultra simple, yet would beat 99% of all mutual funds over the past 5, 10, 20, etc year periods; yet, if I was a bettin' man - I would guess almost no advisors would use it because so many think they are somehow 'smarter' with their 'process' (er...lack thereof).So I'm curious what everyone thinks - so check it out and let me know.Cheers,JW

LockEDJ's picture
Offline
Joined: 2009-07-06

First pass through ... very simple code so anyone could understand what's going on. Intriguing.

Please or Register to post comments.

Industry Newsletters
Careers Category Sponsor Links

Sponsored Introduction Continue on to (or wait seconds) ×