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

[almostthere]

This has been a productive discussion. These are the questions that I have added to the monitoring and evaluation section of an investment policy document that I am developing for new strategies that I am exploring:

Questions to ask myself periodically:
Describe some situations where you made wise purchases.
Describe some situations where you made unwise purchases.
Describe some situations where you made wise sales.
Describe some situations where you made unwise sales.
How well did you understand _______ (investment name)?
What investments made you uncomfortable? What were the reasons for they made you feel uncomfortable?
What types of investments did you feel confident making? Was the confidence justified?

The point is to know reality as it is, not as I would like it to be.

[Tyler9000]

The point is to know reality as it is, not as I would like it to be.

I like that mindset. Too often in investing discussions people idolize math and logic while demonizing all emotion as something to be ignored. The downside is that it treats reality as they'd like it to be and exposes people to blind spots. Sustainability and maximization are often mutually exclusive, but many people have to first learn that the hard way.

[almostthere]

@Tyler - Perfectly stated.

[IlliniDave]

jacob, okay, I took ignorant more like uncertain than agnostic.

The Jack Bogle foundation for indexing (which is more similar to my attitude--the simple arithmetic approach) does not depend on EMH at all. We've beaten that topic to death in the past so I'll let it alone here. The Boglehead community members are all over the spectrum.

Any one who invests (versus speculates) in a market (or it's subsets or even elements) should be of the opinion that the returns on the investment will be positive over some time frame that is meaningful to them. There's a lot of uncertainty in that--at minimum there will generally be ups/downs along the way and it may never pan out as anticipated in the time window of interest. That's the sort of uncertainty I read into "ignorant" (e.g., I don't know what the SP500 will be 30 days from today, 90 days from today, 1 year from today, 20 years from today, or even at the close this afternoon; same is true for any individual stock or other index). That's the perspective from which I concluded I) is the predominant matrix element on a day-to-day basis, even year-to-year basis for most participants.

Although I'm admittedly uncertain of what the future holds, for those things I invest in, I am of the opinion that I'm increasingly likely to be up as time goes on. If I did not have an opinion on whether some investment would be up or down 20 years from now, I would not put money at risk in it.

[7Wannabe5]

How long does it take an index fund to change its basket to match the abrupt,deep,narrow action of a single player?

[jacob]

@7wb5 - IIRC The index committees discuss revisions about once a year. The Dow Jones Industrial is (IIRC again) decided by the editors at the WSJ. The S&P500 is decided by some committee at the Standard & Poors rating agency. So in a way, if you're buying the DJIA, your portfolio is in some sense being managed by a bunch of newspaper people

These revisions are announced in advance and you'll often see traders bidding up components that are about to enter the index and selling out components that are about to leave. Being in an index comes with inherent demand which translates into a premium.

Note that these are just two of the most well-known indexes. Anyone is free to invent and name their very own index. Index just refers to a weighted sum of prices of a collection of securities.

Or did you mean how long it takes for the index market price to reflect a sudden move in one of it's components? In that case, the answer is less than 50 milliseconds. More accurately, the individual entries act as a back stop to the index itself (if something moves fast, they simply pull their quotes and the spread goes wide --- usually in a directional sense only) ... The actual indexes are mainly driven by the action on the futures market on the intraday scale. On the monthly scale, it's very much driven by supply and demand from institutional investors (e.g. most of you guys via Vanguard, etc.) which is especially evident during the last five days of the month and the first couple (when people get and deposit their paychecks).

[BRUTE]

The point is to know reality as it is, not as I would like it to be.

I like that mindset. Too often in investing discussions people idolize math and logic while demonizing all emotion as something to be ignored.

brute finds it easy to be pragmatic in the abstract, but how this pragmatism is applied can be difficult. for example, the 2 camps about indexing - is it the holy grail (Malkiel) because the future of this complex system is unknowable, or is it foolish (jacob) because only amateurs don't know how to actively invest well?

unfortunately, reality is so messy that it seems almost impermeable by rigorous science, the kind that uses experiments to validate hypotheses. humanity and its economic activities are just too complex to run experiments on - even to tell if they're knowable or unknowable. (how would a scientist validate that something about reality is knowable or unknowable?) for every <x years of great investment success, indexers can say it was just a lucky guess. yet there are people who guess luckily enough to make enough money for the rest of their lifetime.

and even most indexers display (in brute's opinion) strong "picking" behavior, which tells brute that they actually think some things ARE knowable about the future - be it that US stocks will dominate vs. global, that stocks will go up in the long run, and so on. these are all predictions, yet they usually don't seem to be made consciously. when asked by brute, most humans have just brushed the topic aside or answered in a way that demonstrated they a)hadn't even considered this, or b)were actually making a huge bet based on their assumptions, but assumed certainty.

[IlliniDave]

How long does it take an index fund to change its basket to match the abrupt,deep,narrow action of a single player?

If the fund tracks a cap-weighted index and is a cap-weighted traditional mutual fund, it's basically instantaneous. The basket is the basket, the cap weighting automatically self-adjusts, the only question is exactly what is it worth which is calculated at the end of each trading day using the closing price of the portfolio components.

Jacob I think explained about an ETF, which might get a temporary deviation in price from its NAV.

For index mutual funds that are not cap-weighted (I don't know of any indexes that aren't cap weighted, so these funds are likely quasi index funds), it will take until the manager adjusts the holdings for the fund to be valued as intended. I don't know how managers handle those. They are always chasing a moving target and trading too much hurts fund performance. I'd guess though that for a stock that has an extreme movement they'd adjust sooner rather than later.

[FBeyer]

@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).

The bond index I found tracks bonds globally, but focuses mainly on Europe. The index's behavior mirrors the total US bond market so close it's almost weird, hence the total bond market seemed like a good proxy.

I don't have good access to an ETF that tracks, say, Barclay's Global aggregate bond index, and Danish mutual funds are easily above the 1% expense ratio, so we're not exactly spoiled for choice.

I'm winging my way through the portfolio construction here and so far I know I don't know enough to make money on picking stocks, so passive investing is the only way forward for me currently. So in regards to your question about what strategy to choose: the one I expect will induce the least panic while I learn some actual investing skills.

[jacob]

@iDave - The DJIA is price-weighted. That was somewhat an issue during the last revision when they added in GS, NKE, and AAPL because these are highly priced stocks. This means that GS which trades at $160 has about five times the impact on the DJIA as PFE which trades as $35 even though the market cap of the latter is 3+ times that of the former. As such the DJIA is mostly dominated by a few almost idiosyncratically selected stocks. Before the change, IBM was a big influencer on the DJIA. So I think adding in a shoe company, an iThing company, and a vampire squid did balance out the big blue to make the DJIA more representative of the US market than before.

The S&P is cap-weighed. This means that Apple which employes about 50k people is considered twice as representative of the US economy than Walmart which employs 2+ million and does twice the revenue. So ... there's that.

There are non-trivial weightings as well. They're usually called "smart beta" funds. I suppose it's a way of calling them index funds without admitting that they're pure quant-funds. Index schmindex ...

[7Wannabe5]

Jacob saidL So in a way, if you're buying the DJIA, your portfolio is in some sense being managed by a bunch of newspaper people

Yeah, that doesn't sound so good. Some of them probably drink too much and trade secrets for sex with their polyamours.

On the monthly scale, it's very much driven by supply and demand from institutional investors (e.g. most of you guys via Vanguard, etc.) which is especially evident during the last five days of the month and the first couple (when people get and deposit their paychecks).

So, the index fund is like a whorehouse located nearby a large truck-stop?

IlliniDave said: If the fund tracks a cap-weighted index and is a cap-weighted traditional mutual fund, it's basically instantaneous. The basket is the basket, the cap weighting automatically self-adjusts, the only question is exactly what is it worth which is calculated at the end of each trading day using the closing price of the portfolio components.

So, in theory, a large move in a large component could possibly cause some people who suffer from an irrational fear of loss to unload their shares in the index fund, thereby causing predictable secondary effect on the other smaller components of the index? IOW, are the movements of the smaller components of a popular index relatively predictably turbulent, yet low-risk within margin of protected realm of index member status? Of course, this seems pretty obvious, so likely already accounted for, and also the accounting that takes into account the ramifications of this accounting, etc. etc. That's why I stick to investing my energies and moneys in the more inefficient markets.

[jacob]

@7wb5 - Well, in defence of the WSJ deciding on the DJIA, they probably do read more financial news than the average cookie. And whatever goes on at those analyst meetings at the S&P, they probably do have a better understanding of US companies than the average money-magazine-reading over-confident high-income JD or MD from the Midwest.

And yes, these days large index moves drag everyone along even if they don't really deserve it from a valuation standpoint. IOW, if Tim Cook says something like "we too are looking into Pokemon Go stuff", then, that will affect the price of a company that sells fasteners or paint. And, also, yes, this works in theory, but in practice people are smart enough in theory to simply pull their quotes. You know a pimp's love is very different from that of a square. Same with Wall Street vs index funds

[IlliniDave]

IlliniDave said: If the fund tracks a cap-weighted index and is a cap-weighted traditional mutual fund, it's basically instantaneous. The basket is the basket, the cap weighting automatically self-adjusts, the only question is exactly what is it worth which is calculated at the end of each trading day using the closing price of the portfolio components.

So, in theory, a large move in a large component could possibly cause some people who suffer from an irrational fear of loss to unload their shares in the index fund, thereby causing predictable secondary effect on the other smaller components of the index? IOW, are the movements of the smaller components of a popular index relatively predictably turbulent, yet low-risk within margin of protected realm of index member status? Of course, this seems pretty obvious, so likely already accounted for, and also the accounting that takes into account the ramifications of this accounting, etc. etc. That's why I stick to investing my energies and moneys in the more inefficient markets.

I suppose that is possible. From the math perspective, Apple I think makes up about 3% of the SP500 as the largest component. So if Apple were to vanish completely in one day and everything else in the SP500 was a wash for the day, someone who owns a SP500 fund would experience a 3% loss. That's a pretty bad day, but I don't know if it rises to blood in the streets/panic levels. And that's the most extreme loss in the largest single component of the index. Something more tame like a 40% loss in a company in the top 20 percent range would be lost in the noise.

That math is not specific to index funds--it's the same for any portfolio that holds the stock in question, and is proportional to the weight of the stock in the portfolio.

Every US stock of reasonable liquidity is included in some index/index fund.

I think the emotional impact of the news of an Apple or Amazon collapsing would have a much larger impact on things in the market than the math. And index funds in the general sense are far more towards the reactive back end than the causal front end of market fluctuations. In other words, the price of an index/index fund share reacts to the market prices of the individual stocks in the basket, not the other way around (indexes are a thermometer, not a thermostat). Extreme inflows or outflows will contribute to overall price movements, but they are usually a reaction to movements that are already underway; and that is not a unique characteristic of index funds. Stock markets displayed a lot of correlation long before index funds existed.

Speculators that play around with ETFs a la SPY and the like on a daily basis of course contribute to the noise, but they are a separate universe from the "typical" index investor.

My take is that the reason to invest in stocks is because you are interested in getting a share of the ongoing profits and growth in the business(es). That means you have to endure the speculative noise. If you want to game the speculative noise, well, that's an endeavor I bowed out of more than 20 years ago, so can't comment on that.

[stayhigh]

I think people trust too much in past performance of US in last century. But they forget why last century was so great. Market long term performance is based on economy. And economy is made by people, for people. Take a look at demography, if you want to have an idea about the future. Now try to build your portfolio again keeping this in mind. I believe 4% rule can still work for future ERs, but entry point (current valuations) is more important than ever. I'd never pull the trigger today at 4%.

[Noided]

I think people trust too much in past performance of US in last century. But they forget why last century was so great. Market long term performance is based on economy. And economy is made by people, for people. Take a look at demography, if you want to have an idea about the future. Now try to build your portfolio again keeping this in mind. I believe 4% rule can still work for future ERs, but entry point (current valuations) is more important than ever. I'd never pull the trigger today at 4%.

If you mean demographics in terms of economic growth. From what I've studied, with population growth you get nominal increase in GDP, not an increase in per capita GDP. I don't know if the first type of growth is that positive.

If you mean demographics as in the baby boomers having a huge stake in the US stock market, there is a Vanguard study that does not support that idea. This makes me doubt the doomsday predictions with regards to demography.

I agree that one must be conservative in the future expected returns. I personally like to count the years before ER assuming a future real rate of return of 3%. Because I am still at the start of my journey I will probably adjust my target accumulation amount as I go.

[WFJ]

I'm a little late to this discussion, but the 4% will experience many cracks this year. I've been FI for a while and only recently learned of the prevalence and almost cult like adherence to the 4% rule. It's dangerous as very few of the followers even know what goes into building the model. 4% is better than nothing, but I could never recommend anyone retire early (before 60) that did not have a <2% SWR that is subject to equity and bond risk (pension fund would increase this #).

Model assumptions that turned out to be false caused the GFC (models estimated the probability of a negative home price for any T+ period was 0.00). 4% is better than nothing but wish there was more of a focus on the duration, call it the "30-year retirement rule" or focus on spending relative to portfolio income "Initial AUM 1% per year drawdown + portfolio income" which would frame the number in a different light and be more robust for ER.

[CactusSurfer]

4% is better than nothing, but I could never recommend anyone retire early (before 60) that did not have a <2% SWR that is subject to equity and bond risk (pension fund would increase this #).

Keep in mind that pension funds generally invest their money into various assets like stocks, bonds, and real estate. If we see historically low returns you'll see many pension funds fail to meet their obligations, and in that scenario the overall shortfall will be too large for a politically feasible bailout.

[conwy]

With retirement spending, I think flexibility is the ultimate winning strategy.

The 4% rule was a good source of motivation for me during my early accumulation phase. But over time, and after learning more, I've come to believe less in any % rules. Not because I don't think shares will pay off, but because I don't think one can predict when and how much they will pay off. Thus it doesn't make sense to use one % at all times, but makes more sense to adjust the % depending on one's overall financial situation.

Nowadays I adopt a more pragmatic "investments as tools" mindset, seeing each component of my portfolio as a tool, which I might or might not exercise, depending on the situation.

For example, in the present situation, with fairly low prices, falling stock market, low cash interest rates and strong job market, I would prefer to work a part-time job and rely on that for income without spending the portfolio or cash.

If the job market goes south but interest rates go up, maybe I will switch to part-time work, supplemented by spending some of the cash interest.

Or if the job market goes south, interest rates go down and the stock market booms again, perhaps I will live from some of my stocks.

So I'm trying to see my investments as tools, each of which might be useful at a given time in a given situation, but not one of them as a golden ticket.

I want to avoid selling my stocks at all during downturns. But during stock market booms I'm happy to spend 2-4% or even more.

That said, based on my current living expenses and my current stock holdings (even after the recent downturn) I can survive on 2% per year drawdown. And I think that would be fairly reliable over the long-term (50+ years), especially taking into account that I would probably do some paid work from time to time and supplement my stocks.

But I think realistically I wouldn't stick to 2% but would try to adjust my living expenses based on stock returns, e.g. probably find a way to spend less than 2% during downturns and 2%+ during upturns.