The continued viability of the 4% rule in the US in the 21st century

[BRUTE]

brute prefers relatively mellow portfolios. he doesn't like checking stock prices every day, or being afraid that if he misses something and doesn't act upon it immediately, he'll lose a lot of money. he also likes simplicity.

thus a very passive, stable, simple approach seems best.

what appeals to brute about the GB is the very short (3-4 year) drawdown periods, and still reasonable returns (both historically, of course). brute can always work another year or be more frugal for a year or two. but ten years (as in some 100% stock periods) doesn't sound like brute could soldier through it.

brute keeps thinking that past results don't guarantee future results, meaning both the high-stock and the GB are kind of dangerous in that regard - they both rely heavily on past results, predicting the future will be similar.

the GMP is more of a max-diversification approach, but at the same time it's more difficult to build and likely to change a lot over time, requiring more effort and maintenance.

huh.

[jennypenny]

It would be great if Tyler's site also had a behavioral test that would match the score with the allocation best suited to the individual, almost like a dating site. Is PortfolioMatch.com available?

[OTCW]

I read a short article recently that suggests replacing the 4 percent rule of thumb with a withdrawal rate equal to your age divided by 20 as a safe withdrawal rate. Thus if you are 40 years old, you should aim for a 2 percent withdrawal rate. If you are 60 it would be 3 percent. It gets less conservative the older you get and vice versa. Probably too conservative at the ends - does a 95 year old really need to worry about out living his money? What does one spend on at that age anyway?

Anyway, just another rule of thumb based on some sort of data analysis. Has the benefit of changing with age, but still has a bit much of the one size fits all (other than differentiating for said age) about it imo.

[IlliniDave]

BRUTE, I would agree with Tyler9000. Optimal is almost impossible to define ahead of time, plus it is a function of what your goals are and what your temperament is. I could easily be wrong, but the first two you mention appear to me to be the simplest to implement and maintain (correlations are a moving target), and sometimes there's an elegance to simplicity that make it a little easier to ride the emotional roller coaster. I've heard it said that there are many strategies that will get most people to their goals as long as they stick with the one they pick. Continually switching strategies is one of the bigger cockpit errors investors make.

My approach is a hybrid between the 70/30 US and the Global portfolio. I am overweight in US equity and I don't buy non-US bonds at all. I thought both those approaches had merit so I split the difference in a way that best allows me to sleep well at night.

ETA: I see you said later you value stability very highly, so in your case maybe the Golden Butterfly (is that similar to the PP family?) might be the SWAN (sleep well at night) choice for you.

Also, if you do a GMP by cap weighting, it should largely self-adjust as markets rise and fall in their proportion of the global market. Getting the proportions right to start with and how to direct contributions could get tricky. I don't know how often the data one would base contribution spreads is refreshed.

[almostthere]

Just to add a little more 'color' as they say to how I define my behavioral issues. I have kept track of all the trades I believe I made mistakes on or where I could have done much (much) better. Two things that show up repeatedly for me are 1) panic selling where I did not completely understand the investment 2) Anchor bias to my 'basis' price. Luckily, I have never lost money, but in the case of income CEFs or even very specific ETFs that I may make a six figure trade on, I am running a risk that I could sell in a panic and face a major impairment of capital. I never lost money b/c of the anchor bias to my basis, but at least in one case, had I not panicked, I'd be up in the very high five figures. That same anchor bias allowed me to think my panic sale was okay b/c 'I wasn't losing any money'.
On the good side, I have shown repeatedly an ability to average down on things I did believe in. For me, I can easily average down into beta indexes like VWO or VGK. Any portfolio that contained something I may not completely understand is bad idea. I could never average down into gold or any income index b/c I just don't have any faith in my macro knowledge. I see now that even averaging down into 'value tilt' indexes could be trying when value is in a long under performance stretch.
In other words, behavioral issues are not esoteric intellectual ideas for me. They are the mistakes I see myself repeatedly making.

[BRUTE]

SWAN (sleep well at night) choice

brute likes that term

and yes, GB is similar to PP, but it's divided into fifths instead of quarters. here's more: https://portfoliocharts.com/portfolio/golden-butterfly/

[FBeyer]

My reasoning goes like this: from the minute I decided to index my investments rather than choose a more active approach I forfeited a lot of the boons that active investing would give me.
Hence, my investments are now bound by the approximate same sentiment that also governs exercise: the best program is the one that you actually follow.

[BRUTE]

so how does FBeyer tell which program he will follow?

brute supposes that a relatively non-volatile portfolio will prevent panic sells, and one that has reasonable if not crazy returns will prevent humans from jumping towards the high gain train.

[Tyler9000]

It would be great if Tyler's site also had a behavioral test that would match the score with the allocation best suited to the individual, almost like a dating site. Is PortfolioMatch.com available?

[Frantically searching for that domain]
...looks like it's taken. Shucks. Bonus points if it belongs to you and you've been playing the long game for this reference.

I've actually been working on a framework that does just this. I guess I'm not yet satisfied with the selection criteria, but I really like the idea. Pairing an investment style to an individual is a lot more psychological than people realize.

[FBeyer]

so how does FBeyer tell which program he will follow?

brute supposes that a relatively non-volatile portfolio will prevent panic sells, and one that has reasonable if not crazy returns will prevent humans from jumping towards the high gain train.

Low volatility and low drawdowns at the potential risk of lower returns, yes.
I think that is actually the most common type of portfolio IIRC.

I've broken down the portfolio in my journal, so I'll just quickly recap here: portfolio is split evenly between the only major assets available to me: bonds, stocks and real estate.

bonds will be as globally diversified as I can and diversified over junk, corp and gov/mortgage in a 10, 25, 65 percent split.
stocks are split evenly between the MSCI minimum volatility index, dev. world large cap value stocks, and dev. world small cap stocks.
Real estate will most likely be in the form of two smallish rentals only if I can find a 'Dhando offer'.

That way I'll have three slightly less correlated asset classes and the two assets that I have outsourced, I have outsourced to as global an allocation as possible.

That's going by the global portfolio mantra: if the world as a whole is making money, so are you. I know that my portfolio wont be 'correct' in the global portfolio sense, but it's a very easy 1/3 split that I can remember and the overall diversification seems quite good. Even if stocks and bonds take a dive, I'll still potentially have a rental or two for those who lost their jobs or, now, cant get a loan because the economy is in the shits. Backtesting on portfoliocharts gave me about 2-3 years drawdowns with
11 dev. world
11 intl. small
11 intl value
34 global bonds
33 REIT
and I can work as a carpenter or analyst (probably) if needs be. Or mailman, or delivery boy, or christmas gift wrapper, or personal math tutor, or pallet stacker or...


TOO LONG DIDNT READ: three major asset classes that I can somewhat understand. Trust in the global economy to carry my portfolio for me. Trust in my education in construction and my mathematical education to wring value out of real estate.
It's simple, conceptually, but plotting the actual contents of the portfolio makes it look rather convoluted.

[Tyler9000]

That's going by the global portfolio mantra: if the world as a whole is making money, so are you. ... Backtesting on portfoliocharts gave me about 2-3 years drawdowns with
11 dev. world
11 intl. small
11 intl value
34 global bonds
33 REIT

Just so there's no confusion, the international numbers on PC are all EX-US and the bonds are all US-centric. I don't want anyone to misinterpret the results.

Carry on.

[FBeyer]

Yes, I found out that the performance of the bond fund I found almost exactly matched the US total bonds fund, so I deemed it a good proxy!
Switching the numbers on 11 dev world to 4 us total and 7 dev world, and doing the same thing for the value and small cap as well, changes very little.
Annual returns go from 7.1 and sd 11.4 to 7.2 and 11.9 sd.
Bernstein's mantra about getting the percentages in the right ball park still holds. It's the behaviour of the chosen assets that matter, not necessarily their strict ratios in the portfolio, which also should calm some people down if they are tempted to rebalance too often.

Edit: by the way, I decided on the portfolio because I trusted the design before I even checked the idea out on portfoliocharts. I've not decided on this portfolio because of past returns or something like that. The three spokes in the portfolio and even distribution between them is simply something I can relate to.

[BRUTE]

I've actually been working on a framework that does just this. I guess I'm not yet satisfied with the selection criteria, but I really like the idea. Pairing an investment style to an individual is a lot more psychological than people realize.

does Tyler9000 care to elaborate on this framework? sounds interesting.

@FBeyer: what class does FBeyer use for global bonds on portfoliocharts? brute can't find that, and using 34 Total Bond Market instead, seems to get different results (5y longest drawdown).

[jennypenny]

It would be great if Tyler's site also had a behavioral test that would match the score with the allocation best suited to the individual, almost like a dating site. Is PortfolioMatch.com available?

[Frantically searching for that domain]
...looks like it's taken. Shucks. Bonus points if it belongs to you and you've been playing the long game for this reference.

Haha ... I own a lot of domain names, but that's not one of them. It didn't occur to me to add any with 'portfolio' after you started the site.

Too bad. I would have given you a good deal.

[Tyler9000]

does Tyler9000 care to elaborate on this framework? sounds interesting.

I put together a rough decision tree a while back that asks a series of behavioral questions to guide people to the appropriate portfolio for their personality. It's completely unscientific, and mainly relies on common complaints I read from people on various forums for specific portfolios. Stuff like:

What worries you more?
A) Missing out on a big market rally
B) Experiencing an unexpected market drop
C) Neither -- I really don't pay attention to the markets

I temporarily put it on ice as I'm not comfortable making specific recommendations to people I've never met, but I'm still playing with the idea of adding more behavioral features to the site.

[BRUTE]

brute would say it's both A and B about equally. maybe weighting is better than having to make a choice?

[Tyler9000]

Yeah, the nuance involved is why I'm not yet satisfied with the results. But it's an interesting exercise to try to break things down in qualitative terms.

[jacob]

Here's my model: (3x3 matrix)

A) Being right when most others are right
B) Being right when most others are wrong
C) Being right when most others are ignorant
D) Being wrong when most others are right
E) Being wrong when most others are wrong
F) Being wrong when most others are ignorant
G) Being ignorant when most others are right
H) Being ignorant when most others are wrong
I) Being ignorant when most others are ignorant

And my question is which pair of choices do you have for the letter you'd most like to be and the letter you'd least like to be.

[IlliniDave]

Since most behavioral "errors" come from emotion/personality overriding dry logic, it would be interesting to create matrix that more directly speaks to an investing temperament focused on the investor rather than the investor relative to others. The kinds of things I see through interacting with a lot of investors that could be relevant are: overall optimism/pessimism, intensity of aversion to short-term loss, degree of self confidence, susceptibility to doom/exuberance porn (financial porn in general), degree of patience, degree/desire to be "in control", degree of connectivity of self-image to investment strategies and results, opinions/abilities re delayed gratification, strength/lack of interest in the topic of investing, etc. It's possible those could be mapped to something like a BM scheme where readily available evaluation systems exist. How to map all that to suitable investment strategies is not clear to me.

Regarding jacob's question, I) is where most of us exist (if you assume that the net of being right part of the time (through luck or skill/knowledge), wrong part of the time (through bad luck or poor skill/knowledge), and ignorant the rest of the time, averages out to ignorance). Omniscience, being "right" and knowing it, ( A)-C) ) means the omniscient one would be able to choose the best strategy/portfolio as a function of conditions and would need no recommendation. It would be interesting to think through mapping the other 6 to various passive/defensive strategies, but I haven't the energy this late in the week. It seems rather complicated as it's most likely that the specifics are important.

[jacob]

@iDave - The EMH foundation for index investing is (G). The self-referential herding that I've been warning about if majority became indexers is (I)---at that point the market would have turned into a giant savings account where the market value would strictly depend on the demand/supply of cash needs.

In my matrix, "right" means being correct about the direction. "wrong" means incorrect. "ignorant" means not having an opinion.