In Investing, It’s When You Start And When You Finish

[obstinate]

http://www.nytimes.com/interactive/2011/01/02/business/20110102-metrics-graphic.html?_r=0

There's nothing wrong with this chart, per se, but it leads people to think that early retirement is a pipe dream. One important thing to note: it does not count dividend reinvestment! Another important thing to note: it assumes it matters quite a lot what the stock market was at exactly twenty years after you retired. It does matter, but it doesn't matter quite as much as the chart might lead you to believe.

Take the worst twenty year period. '61-'81. With appreciation only, you're -1.3% real annualized (the chart says -2%, but this does not match Schiller's numbers). With dividend reinvestment, you're actually up 2% per year. I made a spreadsheet to model the 4% rule retiree in '61. Spoiler alert: successful retirement.

https://docs.google.com/spreadsheets/d/1VXYx12gBECG537mswqRMFz655HD6tGEQ2TF0XiwoG9Y/edit?usp=sharing

[Mezzie]

Isn't that the very risk we've all accepted by investing, though? That our target retirement date will be the next 1929 or 2008? That's why many investors become more conservative as they near retirement.

Sometimes I feel like some people have deluded themselves that there is no risk in investing since over the long haul there are consistent returns, but our individual long hauls, however long, have an end date, and depending on our age and ability, we may not be able to extend that date long enough to recover financially from a crash.

I think knowing and accepting that risk is important. I didn't feel discouraged reading about it, but maybe that's just me because I tend towards imagining worst case scenarios. :)

[SwordGuy]

There was an interesting article a year or two back which told the story of an imaginary person who only ever invested the day before the market tanked.   They tracked his investments over a multi-decade time period and he still came out ahead.

http://awealthofcommonsense.com/2014/02/worlds-worst-market-timer/

[maizefolk]

The color scheme they chose is making things look worse than they actually are. They use two colors (red and green) and the convention when you do this for a group of positive and negative values is to set zero as the boundary between the two colors, and make sure that equal levels of saturation/brightness of each color represent represent equal absolutely values (so if bright red is -10% then bright green should be 10% and so on). By setting their "zero point" at 4% the emotional take away people will get is that the only good times to invest in the stock market are the green bits which are actually the areas returned more than 7% per year.

Also they include "average taxes and fees" in their calculated. Who wants to guess that "average" is to a new york times reporter?

[obstinate]

Also they include "average taxes and fees" in their calculated. Who wants to guess that "average" is to a new york times reporter?

I didn't notice that! Good spotting.

[JAYSLOL]

There was an interesting article a year or two back which told the story of an imaginary person who only ever invested the day before the market tanked.   They tracked his investments over a multi-decade time period and he still came out ahead.

http://awealthofcommonsense.com/2014/02/worlds-worst-market-timer/

I absolutely love this story, I've probably read it 10 times (lol) and shared it with a few "but! you can loose everything in the market!" type friends.

[gerardc]

One important thing to note: it does not count dividend reinvestment!

How do you know? They claim their numbers "include dividends". Standard practice is that total returns include dividend reinvestment.

[obstinate]

One important thing to note: it does not count dividend reinvestment!

How do you know? They claim their numbers "include dividends". Standard practice is that total returns include dividend reinvestment.

Well, hrm. It differs so much from schiller's numbers w/ dividend reinvestment that I just assumed it was not included. I don't know how they came up with -2%, then. Maybe they included 4% annual fees?

[NykkiC]

Sometimes I feel like some people have deluded themselves that there is no risk in investing since over the long haul there are consistent returns, but our individual long hauls, however long, have an end date, and depending on our age and ability, we may not be able to extend that date long enough to recover financially from a crash.

Which is yet another good reason for early retirement. If your target date leaves you with several productive years left, then you're in a much better position than someone who waited and, if something goes wrong, needs to work when they're less physically able. This is problem is compounded if, like many older workers, the potential retiree has trouble getting hired in a job appropriate for their skills and experience.

[maizefolk]

One important thing to note: it does not count dividend reinvestment!

How do you know? They claim their numbers "include dividends". Standard practice is that total returns include dividend reinvestment.

Well, hrm. It differs so much from schiller's numbers w/ dividend reinvestment that I just assumed it was not included. I don't know how they came up with -2%, then. Maybe they included 4% annual fees?

Depending on how they calculate those average taxes and fees, those could easily be the explanation. But they don't provide nearly detailed enough methods for us to say for sure. 0.6 or 0.7% annual fee would not be outside the realm of what could be justified as average. If the nytimes considers "average" income to be above $200k/year* capital gains and dividends would be taxed at 18.8% (and remember these taxes apply to the nominal gains, not inflation adjusted ones, so in periods of high inflation can take out a much bigger than expected bite out of "real" returns).

*Which is ridiculous, but not entirely out of character for the paper.

[gerardc]

Probably their biggest flaw is assuming you withdraw a lump sum 20 years later. In reality, you'll withdraw gradually for 30-50 years after that point. That'd greatly contribute to reduce the variance further, i.e. push most periods towards the mean (4.1% according to them).

ETA: Actually, you can see that from their chart by picking a diagonal further up and to the right. Most squares are grey there (median).