Portfolio Charts

Ask your investment, budget, and other money related questions here
User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Tue Jan 17, 2017 6:56 pm

Tyler9000 wrote: BTW, I've been working a lot with the guy who maintains the Simba spreadsheet and there's a big update coming. One of the likely changes is that EM may be removed prior to 1985. The more we look into data before that date the less we trust it. Just something to think about.
Tyler9000 wrote: That doesn't mean I'll be removing EM from the calculators. I'm working on something too complicated to explain just yet, but it should be pretty cool once it's done.
If you haven't already noticed, the calculator update is now live and it includes the improved data and also the new system I alluded to above.

https://portfoliocharts.com/2017/01/09/ ... lculators/

Lucky C
Posts: 153
Joined: Sat Apr 16, 2016 6:09 am

Re: Portfolio Charts

Post by Lucky C » Thu Feb 02, 2017 7:45 pm

I'm less than two years from my expected FI date and I want to dramatically reduce the risk in my portfolio. The way it is now, if my returns significantly beat expectations maybe I'll retire half a year earlier, but if my returns match some of the worst scenarios historically, that could push me back years! Way more downside than the potential upside for my situation.

I also think it's especially risky to own US stocks and long term bonds right now if you care about short term returns. I want to stick to low-priced international equities and bonds, and only Short Term for US bonds. Throw in some gold and I think I've got some decent diversification.

And so I present this low drawdown FI home stretch portfolio:
25% international bonds - in reality I would do a smaller but riskier combo of EM bonds + LendingClub
25% short term treasury
25% cash
15% international value
10% gold

Returns and Drawdowns

Using the Long Term Returns calculator, over 2 years the upside CAGR is 15% but the downside CAGR is only -3%! Compare to the Permanent Portfolio where the upside CAGR over 2 years is 15% but the downside is -7%. I'm a lot more comfortable with a historical drawdown of about 5% with this portfolio vs. the 13% of the PP to get me to my FI date within a window of a few months - this is more important to me than a percent or two difference in returns over the next couple of years.

As always, past correlations do not guarantee future correlations and all that jazz, but this looks to me like a great way to almost-guarantee meeting my financial goals.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Thu Feb 02, 2017 9:45 pm

I understand where you're coming from. Having security in your FI date is certainly valuable, and that's extremely impressive from a drawdown perspective. However, keep in mind that your true financial goal is not to reach a specific date with a certain amount of money, but to build a sustainable system you can stick with for the long run. When I was on your position I preferred to start building towards the long-term goal rather than focusing too closely on the next guidepost. Just something to think about.

Lucky C
Posts: 153
Joined: Sat Apr 16, 2016 6:09 am

Re: Portfolio Charts

Post by Lucky C » Sat Feb 04, 2017 9:34 am

My situation is atypical in that I'm looking to get a house that's far away from work this year rather than waiting until FI, for various reasons. This will lock me into a commute that I can deal with for about 1.5 years, or 2 years if the conservative portfolio I describe above turns out to not do so well. If I stick to my current risk level and the markets go south, I'll then be stuck with a hellish commute and not much free time for maybe 4 years instead of <2. To me that would be a nightmare! I'm willingly adding on a couple months of work (based on expected returns) to completely eliminate the risk of a couple extra years of work. Besides, last year my investments and income exceeded expectations, so my predicted end date going the conservative route still makes me very happy - time to lock it down as much as possible!

Also, without getting into an active investing / predicting the future debate, I believe the odds are pretty good that there will be more bargains to buy in 1.5 - 2 years compared to today. I certainly won't regret holding a lot of cash or short term bonds if some prices start to drop, and I'll be comfortable going higher risk again when the time is right. I know that I can successfully manage a riskier portfolio, it's just time for me to take it easy for the short term.

userqname
Posts: 27
Joined: Thu Dec 29, 2016 9:19 am

Re: Portfolio Charts

Post by userqname » Mon Feb 06, 2017 9:16 pm

Tyler9000, what Asset Allocation performs best on the Financial Independence calculator? (lowest Maximum time to FI)
I would assume it is the one with the highest perpetual withdrawal rate, and the lowest volatility. Probably resembling the Golden Butterfly.

Also, is there any reason the Portfolio Finder has to cap at 11 assets? It would obviously be preferable to select All Assets and subtract from there.

User avatar
Fish
Posts: 148
Joined: Sun Jun 12, 2016 9:09 am

Re: Portfolio Charts

Post by Fish » Tue Feb 07, 2017 11:01 am

@Tyler9000 Great site! Thanks for making these tools freely available. I've recommended it to my friends and they love it.

I'd like to suggest an all-cash portfolio, to be added to the portfolios page. It could be 100% TMM or a new asset that always returns 0% nominal. The reason for this is that it's a very common "portfolio", especially among those with lower NW. Currently one has to go through each of the calculators and generate the fancy graphs and charts one at a time. Viewing them all on a single page would provide motivation to get off the sidelines and start investing!

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Tue Feb 07, 2017 12:35 pm

userqname wrote:Tyler9000, what Asset Allocation performs best on the Financial Independence calculator? (lowest Maximum time to FI)
I would assume it is the one with the highest perpetual withdrawal rate, and the lowest volatility. Probably resembling the Golden Butterfly.

Also, is there any reason the Portfolio Finder has to cap at 11 assets? It would obviously be preferable to select All Assets and subtract from there.
Good questions.

I'm actually not sure what the single (mathematically) best portfolio is on the FI chart, but I suspect your instinct is correct. High return/low volatility portfolios not only get to the target faster but also have lower targets (thanks to higher withdrawal rates).

Selecting all assets and subtracting from there is how I originally designed the Portfolio Finder, but it was extremely data intensive and I ultimately decided to change my algorithm to allow for more flexibility. Bounding the problem to solving for only 10 asset options at a time limits the calculations to a manageable level, which allows me to build a more flexible tool that can handle more asset choices and calculation options than I would otherwise be able to pull off. But I'm looking at new methods and tools for ramping up the horsepower.
Last edited by Tyler9000 on Tue Feb 07, 2017 5:58 pm, edited 4 times in total.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Tue Feb 07, 2017 12:59 pm

@Fish -- Thanks for spreading the word!
Fish wrote: I'd like to suggest an all-cash portfolio, to be added to the portfolios page. It could be 100% TMM or a new asset that always returns 0% nominal. The reason for this is that it's a very common "portfolio", especially among those with lower NW. Currently one has to go through each of the calculators and generate the fancy graphs and charts one at a time. Viewing them all on a single page would provide motivation to get off the sidelines and start investing!
That's a nice insight! I tend to think of just putting all of your money in a total stock market fund as the investing baseline, but you're absolutely correct that the most common money management start is likely a simple savings account. I also sorta like it from the perspective of explaining how cash works, as people today tend to have zero sense of history with regards to interest rates (remember when saving accounts made 6% a year?). I'll have to think about how to message it, but I like the idea.

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Sun Feb 12, 2017 4:51 pm

Tyler, your Thinking Beyond Stocks article has a figure that overlays TSM, Coward's Portfolio, and Golden Butterfly.

https://portfoliocharts.files.wordpress ... .jpg?w=840
https://portfoliocharts.com/2016/07/25/ ... tion-plan/

They all have the same lower bound, but TSM has a wicked large upside. Doesn't this argue in favor of an All Stock Portfolio? Compared with the other two, TSM seems antifragile.

Same with Risk in the Real World. TSM vs. PP.

https://portfoliocharts.com/2015/08/12/ ... eal-world/

I know the Sustainable Withdrawal Rates for these allocations will differ, but the article argues against volatility in *accumulation*.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Sun Feb 12, 2017 6:29 pm

A few things:

1) The larger point of both articles is that there's more to accumulation planning than simply maximizing desired returns without regard to tradeoffs. Putting all of your money in a TSM fund is not particularly dependable, efficient, or sustainable for most investors, and that makes planning for important life goals a challenge.

2) Note that the "wicked large upside" quickly corrects and does not sustain itself over other alternatives like the Coward's portfolio. The sharp peaks are the result of the greatest equity bubble in US history that popped in 2000 (every red peak on the chart is Dec 1999). One can argue the perceived upside leading up to the breaking point was an illusion.

3) I also think that putting all of your money in a TSM index is pretty much the opposite of antifragile. It's a concentrated bet on US prosperity, with all the black swan risk that implies. For reference, Taleb has said his personal portfolio includes gold, silver, commodities, Treasury bills, euros, income properties, and stocks.

But if you understand these points and prefer to invest more heavily in stocks, more power to ya! I include the 100% TSM portfolio on the site because I recognize that it works for certain people. We all have different needs and personalities.

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Sun Feb 12, 2017 7:58 pm

"every red peak is Dec 1999."

I see that now. But I also see that the maximum of those other portfolios corresponds closely with other TSM peaks. Doesn't that indicate that stocks are the primary driver of returns? It seems like other asset classes just serve as indicators for market overvaluation. You could run a "simulated" permanent portfolio along with your "real" stock portfolio and refer to the simulation to tell you when to get out of stocks, for example. Just buy what you would have reballanced into, if that makes sense.

Most importantly, what should I read into the fact that the lower bound in accumulation is nearly identical for all three?

Lucky C
Posts: 153
Joined: Sat Apr 16, 2016 6:09 am

Re: Portfolio Charts

Post by Lucky C » Sun Feb 12, 2017 8:05 pm

The Total Stock Market "financial independence" chart shows 11 to 22 working years.
The Coward's Portfolio shows 12 to 18 years.
The Golden Butterfly shows 12 to 15 years.
For accumulation, I would rather have 12 to 15 years of work than 11 to 22 years of work.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Sun Feb 12, 2017 8:11 pm

ThisDinosaur wrote: But I also see that the maximum of those other portfolios corresponds closely with other TSM peaks. Doesn't that indicate that stocks are the primary driver of returns?
Stocks are definitely an important part of any portfolio, but correlated peaks likely speak more to similar portfolio construction than to anything else.
ThisDinosaur wrote:Most importantly, what should I read into the fact that the lower bound in accumulation is nearly identical for all three?
It depends on your personal investing timeframe. Before you come to any conclusions, consider the same chart cropped at 15 years.
Image

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Sun Feb 12, 2017 9:06 pm

@Lucky, But the time to FI is based on the SWR, which is the inverse of the net Worth in years of expenses. So maybe you could get to 20 years expenses faster with stocks, then switch to golden Butterfly in retirement for the lower volatility, taxes notwithstanding.

The risk is stocks underperforming, which Tyler just illustrated is more common than I gave it credit for. BUT, isn't that stocks lower bound just another repeating worst year? The opposite of 1999?

ETA:Sorry for being argumentative. It's how I learn.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Sun Feb 12, 2017 11:08 pm

ThisDinosaur wrote: The risk is stocks underperforming, which Tyler just illustrated is more common than I gave it credit for. BUT, isn't that stocks lower bound just another repeating worst year? The opposite of 1999?
Yep -- that's true. Think of the highest and lowest points as the fat tails in the returns distribution. Some portfolios are more susceptible to those swings than others.

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Wed Feb 15, 2017 1:59 pm

Tyler, you define Small Cap as < 2Billion, and Mid Cap as 2-10 Billion. But the Vanguard Small Cap funds use CRISPR Small cap index, which includes assets with market caps up to almost 12B.

What sources of data do you use for small caps and mid caps? Any thoughts on the relevance of this discrepancy?

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Wed Feb 15, 2017 2:55 pm

Yeah, industry size definitions vary a lot more than you think and the numbers in the asset summaries are only approximate. For reference, Vanguard uses CRSP data which defines small caps as the 85th percentile to the 98th percentile of the total market when sorted by valuation. Obviously, the precise cutoff by company size will vary from year to year. Also, Vanguard (like many index funds) further complicates it by adding the equivalent of a rebalancing band on top of the index to minimize transaction fees within the fund. So when a company grows from a small cap to a mid cap by the index definition, Vanguard does not immediately sell it. That's why if you look at a Morningstar style box, a small cap fund like VB actually includes quite a few mid caps.

Every index fund is a little different in that regard, which is why returns numbers can vary more than you probably realize between funds. The guys who consolidate all the historical data used on the site do their best to minimize any discrepancies due to index methodologies as much as reasonably possible.

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Fri Feb 17, 2017 5:06 pm

For reference, Vanguard uses CRSP data which defines small caps as the 85th percentile to the 98th percentile of the total market when sorted by valuation.
The guys who consolidate all the historical data used on the site do their best to minimize any discrepancies due to index methodologies as much as reasonably possible.
So, would vanguards VSIAX perform more like Mid Cap Value or Small Cap Value on your calculators? Or a (50/50?) mix of the two? Or something else entirely? If so, how would Vanguard's Mid Cap funds compare to the Mid Cap performance on your calculators?

Somewhat related; I haven't found any allocation with a higher PWR or faster Time to FI than 70% MCV, 30% COM. Yes its data mining, but I thought I'd share. Any opinions on the mechanism and repeatability of that AA?

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Fri Feb 17, 2017 6:09 pm

ThisDinosaur wrote: So, would vanguards VSIAX perform more like Mid Cap Value or Small Cap Value on your calculators? Or a (50/50?) mix of the two? Or something else entirely? If so, how would Vanguard's Mid Cap funds compare to the Mid Cap performance on your calculators?
In general the calculators use Vanguard index data whenever available, and you can see the exact funds referenced on the Asset pages. Yes, VSIAX = SCV. The source data introduces progressively more error the farther back you go with different index sources but we do the best we can.

I don't really have an opinion on your proposed AA. The one thing I'd point out from a PWR perspective is that you'll note in the calculator that the results are only estimated for that portfolio. That means that there's a measurably high amount of error with some of the older data. So plan conservatively.

Lucky C
Posts: 153
Joined: Sat Apr 16, 2016 6:09 am

Re: Portfolio Charts

Post by Lucky C » Fri Feb 17, 2017 10:07 pm

ThisDinosaur wrote:@Lucky, But the time to FI is based on the SWR, which is the inverse of the net Worth in years of expenses. So maybe you could get to 20 years expenses faster with stocks, then switch to golden Butterfly in retirement for the lower volatility, taxes notwithstanding.

The risk is stocks underperforming, which Tyler just illustrated is more common than I gave it credit for. BUT, isn't that stocks lower bound just another repeating worst year? The opposite of 1999?

ETA:Sorry for being argumentative. It's how I learn.
The point I made in my previous post still stands. In my example, if you go with TSM you are risking up to 7 years of additional work in the hopes of getting one less year of work, compared to the golden butterfly. YMMV. I encourage you to see the time until retirement charts I was looking at, not just the accumulation charts that look pretty optimistic. I also suggest doing the math in your own spreadsheet projecting time until FI. How much longer will it take if real returns are 0% over the next decade or two instead of what you've been assuming?

There have been 20-year periods when the S&P total real return has been about 0%, and PortfolioCharts data sets do not contain these periods, only part of one of them. So the lower bound you see in Tyler's data are by no means true lower bounds. See the 20-year periods ending 1921, 1949, and 1982. Also, since the bubble in 2000 was so high, it's likely that the 20-year total return ending in 2020 will be pretty low (unless we end the decade in another big bubble).
https://www.advisorperspectives.com/ima ... 34a947.png

The good news though is that since the economy is cyclical, if you had the bad luck of a rough accumulation phase, you should end up with a more successful retirement phase (but the opposite is also true).

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

Re: Portfolio Charts

Post by banker22 » Mon Feb 20, 2017 8:21 am

I am little disappointed with the calculators, particularly with regard to what is erroneously called 'deepest drawdown'. This is actually just the greatest year on year loss, looking ONLY at calendar year and not all of the possible annual periods. The actual max drawdown since 2002 is around 30%. Could you please clear this up on your site, Tyler? It is potential fatal for someone running these portfolios with leverage.

ThisDinosaur
Posts: 573
Joined: Fri Jul 17, 2015 9:31 am

Re: Portfolio Charts

Post by ThisDinosaur » Mon Feb 20, 2017 11:19 am

I encourage you to see the time until retirement charts I was looking at, not just the accumulation charts that look pretty optimistic.
I am looking at those charts. My question to Tyler was prompted by my naïve reading of one of his articles where accumulation charts were compared.
There have been 20-year periods when the S&P total real return has been about 0%, and PortfolioCharts data sets do not contain these periods, only part of one of them.

That gets to my other concern about any fixed asset allocation, which is the soul of Portfolio Charts. It seems to me that equities are the primary driver of returns, with other assets as simply a placeholder for when the stock market underperforms. Look at the returns for the permanent portfolio in any country where the stock market has been flat for more than a decade (i.e., Japan.)

I ask these questions mainly to find justification for myself to hold gold and cash, which don't produce anything.

User avatar
Tyler9000
Posts: 1268
Joined: Fri Jun 01, 2012 11:45 pm
Location: Austin, TX

Re: Portfolio Charts

Post by Tyler9000 » Mon Feb 20, 2017 12:46 pm

@banker22 : The goal of the site is to provide relative portfolio comparisons for everyday buy and hold investors, and I've found that the daily data you're looking for is overkill for most asset allocation discussions. That said, I appreciate your point especially with regard to drawdown numbers and understand why the distinction is important for your personal goals. I'll look at adding a note to the calculator so that nobody is confused about the drawdown definition.

FWIW, I do not personally recommend investing with leverage and strongly discourage pushing it to the limits based on faith in historical data from any source. I've learned that reconstructing historical data is a dark art with way more hidden assumptions than you may realize, and even relying on PhDs and expensive professional firms is no guarantee of accuracy. So please be very careful about using historical data to justify investing in a manner where you can lose more than your original investment and otherwise reasonable tracking error can bankrupt you. Even if I had daily drawdown data, I would be uncomfortable using it for that purpose.

User avatar
bryan
Posts: 668
Joined: Sat Nov 29, 2014 2:01 am
Location: mostly Bay Area

Re: Portfolio Charts

Post by bryan » Mon Feb 20, 2017 3:59 pm

My own style of re-balancing is a few random (chosen at the beginning of the year) days in the year, so daily closing prices (and operations with that granularity) seem acceptable. Good for worst-case analysis at least? Though, it's certainly not a deal-breaker for me using these tools (and I suspect not that high on the prioritized TODO list relative to other possible features).

@ThisDinosaur, well I don't worry about it much. There are assets that absolutely killed equities and assets that absolutely got killed by cash out there, through history. I'm not sure how a tool can accommodate all these (and thus they try to stick to more common assets). I even have nits with some tools that use equity data from times when only “rich-folk” even had the knowledge, capability to invest in equities. Another nit would be the historical data using “official prices” instead of black market prices. One could come up with more nits..

One common solution is to design a portfolio in the abstract sense and change what those abstractions map to in the real world.

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

Re: Portfolio Charts

Post by banker22 » Wed Feb 22, 2017 4:29 am

@Tyler

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.

Post Reply