Portfolio Charts

Ask your investment, budget, and other money related questions here
ThisDinosaur
Posts: 492
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Wed Feb 22, 2017 9:43 am

banker22, are you suggesting the calculators should illustrate flash crashes and intraday volatility? That seems unnecessarily distracting to me.

If I know that a particular portfolio can have a huge drop in principle that takes several years to recover from, but still sustain a 3 or 4% initial withdrawal rate, that may give me confidence to stay the course instead of switch strategies at the worst possible time.

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Wed Feb 22, 2017 11:18 am

banker22 wrote:Leveraged or not, the max drawdown figured you have stated are wrong and misleading. A portfolio with a better worst year (December to December) performance may have a much worse max DD. Please clarify this on your site.
Just so we're speaking the same language (and for the benefit of others following along), the Max Drawdown on the site is the maximum compound drawdown over multiple years. I use year-end annual returns for my source data in all site calculations, as that is what is freely and readily available. The drawdown numbers are correctly calculated and are as accurate as the data allows.

It is absolutely correct that if one were to look at finer daily data starting and ending on different days, then the Max DD numbers would almost assuredly be worse than what is reported. Readers who click the "More Info" link on the Drawdowns calculator will note that I specifically mentioned this point in the original post, but this is not the first time I've been asked about it so clearly it requires a more explicit callout. I'm on it.

However, I think it's important to keep a bit of perspective. When you increase the resolution on a computer monitor, it improves the detail but does not change the picture. Even though the absolute numbers will be different using daily data, I would not expect something like the Golden Butterfly to suddenly have worse drawdown characteristics than the total stock market. Both portfolios will have more severe drawdowns using that method! That's what I mean about the site being appropriate for relative portfolio comparisons. But if you need absolute certainty in every number, I cannot offer that with the available data.

In any case, I appreciate the feedback and will do what I can to make it more clear. Thanks!

User avatar
C40
Posts: 1737
Joined: Thu Feb 17, 2011 4:30 am
Location: Western U.S.
Contact:

Re: Portfolio Charts

Post by C40 » Wed Feb 22, 2017 11:50 am

Reminder: you don't need to cater to all complaints. There's always going to be someone complaining... It always seemed obvious to me (IDK if from insight or from reading the notes) that the numbers are year-end and that the charts are meant to be used for long-term projections. It seems this was obvious to Banker22 also, but he/she/they** just personally want your charts to be something they are not meant to be.

** ;) ;)

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Wed Feb 22, 2017 12:14 pm

C40 wrote:Reminder: you don't need to cater to all complaints.
I think I've identified my first Portfolio Charts business expense. :D

banker22
Posts: 94
Joined: Mon Aug 03, 2015 1:17 pm

Re: Portfolio Charts

Post by banker22 » Tue Feb 28, 2017 9:50 am

Sorry, not intraday.

The current portfolio charts setup is inadequate and misleading. It shows historic returns for ONLY Dec 31 - Dec 31 periods, essentially n=1. The returns and losses (and comparative returns an losses) would be totally different looking at Jan 10 - Jan 10 for example, and likely even more so looking from June 30 - June 30. You should look at returns across every possible 12m time period, not just one.

Also, max drawdown should technically be from peak to trough, rather than just the lowest arbitrarily defined 12 month period (again, n=1).

User avatar
stayhigh
Posts: 78
Joined: Sun Dec 06, 2015 4:20 pm

Re: Portfolio Charts

Post by stayhigh » Wed Mar 01, 2017 2:44 pm

banker22 wrote:Sorry, not intraday.

The current portfolio charts setup is inadequate and misleading. It shows historic returns for ONLY Dec 31 - Dec 31 periods, essentially n=1. The returns and losses (and comparative returns an losses) would be totally different looking at Jan 10 - Jan 10 for example, and likely even more so looking from June 30 - June 30. You should look at returns across every possible 12m time period, not just one.

Also, max drawdown should technically be from peak to trough, rather than just the lowest arbitrarily defined 12 month period (again, n=1).
Personal portfolio is just a tool, that make it easier to create, preserve and grow ones wealth. This process usually take decades. Short, few months long trends doesn't really matter, when you consider long term process of building or living off your portfolio. Changing dates of past long term returns won't make any difference, as future will be different for sure. Past performance is only a indicator of future performance, giving idea of what you can expect in long run, based on simple assumption what worked well in the past. Well diversified portfolio is the key to success for passive investor (and probably the only free lunch one can get). Every fraction of percent based on past returns is not.

User avatar
jennypenny
Posts: 5195
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: Portfolio Charts

Post by jennypenny » Wed Mar 15, 2017 4:29 pm

@T9000- have you done an article about rebalancing? How, when, etc, and theories about that wrt allocations?

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Wed Mar 15, 2017 5:42 pm

jennypenny wrote:@T9000- have you done an article about rebalancing? How, when, etc, and theories about that wrt allocations?
Not yet, but that's a good idea. I looked at it briefly a while back before realizing I'd need to update my calculation method to study other rebalancing strategies. I ended up punting and moving on to other topics, but maybe it's about time to circle back.

User avatar
jennypenny
Posts: 5195
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: Portfolio Charts

Post by jennypenny » Wed Mar 15, 2017 7:43 pm

Sorry Tyler, I wasn't trying to give you an assignment. I was looking around for the recommendations of Swedroe and others and then wondered if you'd already gathered that data somewhere.

It seems like an important component to me. It's strange that they don't make that info easier to find.

User avatar
theanimal
Posts: 952
Joined: Fri Jan 25, 2013 10:05 pm
Location: Gates of the Arctic
Contact:

Re: Portfolio Charts

Post by theanimal » Wed Mar 15, 2017 8:37 pm

@JP- Can't offer much help but re-balancing info for the PP is in Craig Rowland's book. I imagine there's an article or info on his site as well. Not sure if you're looking for that one or not.

ThisDinosaur
Posts: 492
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Thu Mar 16, 2017 8:15 am

http://www.aaii.com/journal/article/the ... wealth.pdf

Craig Rowland says PP should be rebalanced when one of your 25% allocations goes up to 35% or down to 15%. As far as I can tell, those numbers are arbitrary. However, if you rebalance too often, you have to consider the Tax consequences. Vs. too seldom and you lose the "buy low, sell high" benefit of rebalancing.

I imagine any answer Tyler has for this question would have to include more tax consideration than other types of questions PortfolioCharts answers.

User avatar
jennypenny
Posts: 5195
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: Portfolio Charts

Post by jennypenny » Thu Mar 16, 2017 10:30 am

@TD--Thanks for the link.

I'm sure there's no one right answer for this. What I've been digging around for is what rebalancing parameters each portfolio designer used/recommended for their own portfolio. Tax implications aside, each much have used certain rebalancing assumptions when constructing their portfolio. Some may have used percentages like with the PP, but others may recommend a specific time frame, like yearly or at the end of every quarter. Some suggest adding each paycheck's contribution to the lagging fund (seems like a lot of work unless you dump it into a holding account first).

Anyway, banker22's question got me thinking about this, and now that I'm helping someone else with their allocation I realize I should have a firm recommendation for when to rebalance based on the intent of the allocation's developer.

Sorry, I'm not sure that's any clearer.

User avatar
Dragline
Posts: 4281
Joined: Wed Aug 24, 2011 1:50 am

Re: Portfolio Charts

Post by Dragline » Thu Mar 16, 2017 11:03 am

I am curious about this as well -- the only tests I have seen on rebalancing have been kind of ad hoc affairs. I also have to believe that the correct re-balancing parameters for a particular portfolio would vary based on the components of the portfolio.

OTOH, sometimes I wonder if somewhat randomized re-balancing would work just as well as anything (aside from the transaction costs) in the same way that monte carlo simulations yield varying but consistent-in-direction results.

ThisDinosaur
Posts: 492
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Thu Mar 16, 2017 11:19 am

Some suggest adding each paycheck's contribution to the lagging fund
Definitely, during accumulation, that's the way to do it. Say you have a 50/50 target allocation, but there's a $500 difference between the two assets. You get 2K to invest this month, you put $1250 in the lesser and $750 in the greater. Otherwise you buy both and immediately have to sell one of them.

Swedroe's rebalancing advice:
http://awealthofcommonsense.com/2014/03 ... cing-rule/

Rebalancing in a taxable account(which involves trading costs and capital gains taxes) should only be done once you're living off of your portfolio OR the difference between two of your assets is way bigger than the amount you have to invest this month.

So the question for Tyler would be, in a tax advantage account, assuming no trading costs or taxes, is rebalancing continuously/high frequency better or worse than rebalancing rarely, say immediately after huge crashes. My guess is that depends a lot on the relative volatility of the assets involved.

Aside; I would rather rebalance on my birthday or some other time of year than in January when everyone else does it.
http://www.investopedia.com/terms/j/januaryeffect.asp

User avatar
bryan
Posts: 644
Joined: Sat Nov 29, 2014 2:01 am
Location: mostly Bay Area

Re: Portfolio Charts

Post by bryan » Thu Mar 16, 2017 1:21 pm

ThisDinosaur wrote: Aside; I would rather rebalance on my birthday or some other time of year than in January when everyone else does it.
http://www.investopedia.com/terms/j/januaryeffect.asp
Not a bad idea.. though it might be better to just do:

Code: Select all

python -c 'import random; from datetime import date; days=random.randint(1, 365); print(date.fromordinal(date(date.today().year, 1, 1).toordinal() + days - 1))'
on Jan 1?

User avatar
bryan
Posts: 644
Joined: Sat Nov 29, 2014 2:01 am
Location: mostly Bay Area

Re: Portfolio Charts

Post by bryan » Thu Mar 16, 2017 1:22 pm

ThisDinosaur wrote: Aside; I would rather rebalance on my birthday or some other time of year than in January when everyone else does it.
http://www.investopedia.com/terms/j/januaryeffect.asp
Not a bad idea.. though it might be better to just set a reminder for:
> python -c 'import random; from datetime import date; days=random.randint(1, 365); print(date.fromordinal(date(date.today().year, 1, 1).toordinal() + days - 1))'
on Jan 1?

JohnnyFactor
Posts: 10
Joined: Sun Jun 05, 2016 3:44 pm

Re: Portfolio Charts

Post by JohnnyFactor » Thu Mar 16, 2017 3:39 pm

I was just re-listening yesterday to the Harry Browne Money Show episode where he discussed rebalancing bands. The general idea was that anything below 15% didn't provide enough effectiveness and anything over 35% exposed you to too much risk.

classical_Liberal
Posts: 97
Joined: Sun Mar 20, 2016 6:05 am

Re: Portfolio Charts

Post by classical_Liberal » Sun Mar 19, 2017 4:50 pm

Dragline wrote:OTOH, sometimes I wonder if somewhat randomized re-balancing would work just as well as anything (aside from the transaction costs) in the same way that monte carlo simulations yield varying but consistent-in-direction results.
Maybe I'm just restating @draglines thoughts in a more unscientific way. I would think theoretically, assuming null transactional and tax costs, the more often one rebalances between uncorrelated assets, the more one would see narrowing of the bands between max and min CAGR... Correct? So if the goal is to minimize volatility, one could rebalance fairly regularly in a deferred account with free transactions (ie a 60/40 between VTSAX & VBTLX in a Vanguard IRA).
Some suggest adding each paycheck's contribution to the lagging fund
Do the opposite in taxable accounts during withdrawal. If one is going to maintain a set AA and one plans to incur the tax liability/transaction cost in a calendar year anyway, it seems the only logical way to drawdown (unless dealing with an oddly taxed asset like gold). Then use any deferred accounts to rebalance the difference. I would assume anyone near ERE levels of cash outflow wouldn't be too concerned with Cap Gains in a drawdown situation.

ThisDinosaur
Posts: 492
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Mon Mar 27, 2017 8:58 am

Tyler, I noticed you use a bond fund for your LTT allocation, as opposed to individual treasuries. How do you think this would compare to using individual 20 or 30 year Treasuries?

Also, would tax free muni bonds/funds be a substitute? Are they a closer match to investment grade corporate? Just looking at the performance of VUSTX to VWLTX, they are pretty similar since inception, but they had opposite trajectories to 2008.

classical_Liberal
Posts: 97
Joined: Sun Mar 20, 2016 6:05 am

Re: Portfolio Charts

Post by classical_Liberal » Tue Mar 28, 2017 12:23 pm

@Tyler9000 I enjoy the changes to your withdrawal rate visuals. Whenever portfolio charts mixes things up it's like a free expansion pack to a video game...Maybe I'm getting old :cry:

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Tue Mar 28, 2017 4:17 pm

@ThisDinosaur -- A bond fund like TLT should be really similar to a ladder of individual treasuries between 20-30 years maturity.

@Classical_Liberal -- Thanks! From my perspective, the beauty of choosing a passive asset allocation is that it helps free up time for more video games. ;) But if playing with the data is also fun, all the better!

herp
Posts: 37
Joined: Sat Apr 08, 2017 1:11 pm

Re: Portfolio Charts

Post by herp » Wed Apr 19, 2017 9:58 am

ThisDinosaur wrote:
Thu Mar 16, 2017 11:19 am
Definitely, during accumulation, that's the way to do it. Say you have a 50/50 target allocation, but there's a $500 difference between the two assets. You get 2K to invest this month, you put $1250 in the lesser and $750 in the greater. Otherwise you buy both and immediately have to sell one of them.

Swedroe's rebalancing advice:
http://awealthofcommonsense.com/2014/03 ... cing-rule/

Rebalancing in a taxable account(which involves trading costs and capital gains taxes) should only be done once you're living off of your portfolio OR the difference between two of your assets is way bigger than the amount you have to invest this month.
Slightly OT: I think this is really solid advice and it's also the practice that I follow. I like how it's completely mechanical and takes away any decisions about which asset to buy more of.

Aditionally, any proceeds from dividends can be lumped together with the contribution.

I like Swedroe's 5/25 bands as well. There's no need to fuss about a shift of a few percentage points.

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Thu Apr 20, 2017 6:12 pm

So I wrote a long post, thought better of it, and decided to go the PM route for now. Sometimes my excitement outpaces my rational side. :D I'm sure I'll post it again sometime soon.

User avatar
BRUTE
Posts: 2260
Joined: Sat Dec 26, 2015 5:20 pm

Re: Portfolio Charts

Post by BRUTE » Wed Apr 26, 2017 8:20 pm

Tyler9000 wrote:
Thu Apr 20, 2017 6:12 pm
oh hai
how hard would it be to create a calculator that'll judge portfolios not based on lump-sum investing, but on investing $X every year? this sort of forced DCAing would likely lead to a very different outcome for most portfolios over different periods, and more realistically models how humans would approach FI (unless they inherit large lump sums).

input variables could be as follows (imagining heatmap calculator):
portfolio (like on all the existing calculators)
working years

then instead of displaying the development of the portfolio from lump sum X to Y, each rectangle could show development of DCAing for that period.

it would probably be more of a modality in existing calculators than its own calculator? does this make sense to display?

User avatar
Tyler9000
Posts: 1256
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Wed Apr 26, 2017 8:40 pm

It depends on the calculator. I think it makes sense for some more than others.

BTW, two already have this functionality. The Portfolio Growth calculator has the option to set a fixed amount of annual contributions, and the Financial Independence calculator assumes that you save a set percentage of your income every year.

Post Reply