"How to Retire With $2 Million On a $50,000 Salary"

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
prioritarian
Posts: 381
Joined: Tue Jul 16, 2019 6:00 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by prioritarian »

BrandonBogle wrote: Tue Nov 16, 2021 4:55 pm When I made $20k/year in the early 2000s...
Image
phxjcc
Posts: 1093
Joined: Thu Aug 23, 2018 3:47 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by phxjcc »

Kenkat wrote: Tue Nov 16, 2021 3:52 pm
phxjcc wrote: Tue Nov 16, 2021 3:06 pm Yah remember that math teacher in high school that said: "show all work"?

I am that guy.
I am the guy that says you’ve got to read all of the posts on the thread:

viewtopic.php?p=6329751#p6329751

The above post shows that you can get there @15%.
phxjcc wrote: Tue Nov 16, 2021 3:06 pm Now....when you add all that up, tell me where the "YOU SHOULD SAVE 15-20% of your income" is going to come from.

Becaus, unless you are an orthodontist living in Topeka (e.g.) it ain't happening.
Where I work, new employees can put 6% in the 401(k), get a 3% match plus the company kicks in another 5% since they closed the pension plan to new hires (I am still in it fortunately). So that’s 14% right there.

Now, maybe not everyone has those kinds of opportunities, but they probably have something or can save something.

It can be done. I basically did it (I started in 1987) as I suspect many on this board did.
Please post details requested.
User avatar
Kenkat
Posts: 7746
Joined: Thu Mar 01, 2007 11:18 am
Location: Cincinnati, OH

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Kenkat »

phxjcc wrote: Tue Nov 16, 2021 7:06 pm
Kenkat wrote: Tue Nov 16, 2021 3:52 pm
phxjcc wrote: Tue Nov 16, 2021 3:06 pm Yah remember that math teacher in high school that said: "show all work"?

I am that guy.
I am the guy that says you’ve got to read all of the posts on the thread:

viewtopic.php?p=6329751#p6329751

The above post shows that you can get there @15%.
phxjcc wrote: Tue Nov 16, 2021 3:06 pm Now....when you add all that up, tell me where the "YOU SHOULD SAVE 15-20% of your income" is going to come from.

Becaus, unless you are an orthodontist living in Topeka (e.g.) it ain't happening.
Where I work, new employees can put 6% in the 401(k), get a 3% match plus the company kicks in another 5% since they closed the pension plan to new hires (I am still in it fortunately). So that’s 14% right there.

Now, maybe not everyone has those kinds of opportunities, but they probably have something or can save something.

It can be done. I basically did it (I started in 1987) as I suspect many on this board did.
Please post details requested.
You’re the one contending that it can’t be done. I posted details above showing it can. Now it’s your turn to post some work showing how it can’t be done. Remember, find a decent company with some benefits and you don’t need to come up with the whole 15% yourself.
User avatar
BrandonBogle
Posts: 4204
Joined: Mon Jan 28, 2013 11:19 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by BrandonBogle »

prioritarian wrote: Tue Nov 16, 2021 5:53 pm
BrandonBogle wrote: Tue Nov 16, 2021 4:55 pm When I made $20k/year in the early 2000s...
Image
I’m not sure what point you are trying to make. My point was that one can still save for retirement even if you don’t come from a family with means. I don’t see anything counter to that in that chart.

I didn’t say I made $50k/year in the early 2000s nor did I say anything about the cost of a college education. I took out 6 figures in student loans while making $20k/year and contributed to my IRA and 401k throughout. That hasn’t changed to today. Student loans are still available and the case presented in the article has 250% higher pay than the example I gave.
oldfort
Posts: 2781
Joined: Mon Mar 02, 2020 8:45 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by oldfort »

Kenkat wrote: Sun Nov 14, 2021 3:28 pm Yes but the person making $50,000 / yr. is probably not going to be making that same salary for the next 40 years either.
Median income for non-family households is $40,464. Having a career ceiling of $50,000 a year in today's dollars is the most realistc part of the scenario.
calwatch
Posts: 690
Joined: Wed Oct 02, 2013 1:48 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by calwatch »

I do kind of love looking back at these kind of reports back 40 years ago. Los Angeles County had created a non-contributory pension option and management was selling employees on switching to the non-contributory plan, Plan E. They even allowed folks to cash out their previous pension contributions. In order to convince people to join the non-contributory plan they printed this chart:

Image

(http://apps.hr.lacounty.gov/digest/pdf/ ... 001981.pdf)

Now, the non-contributory plan is a 2% at 65 plan but is steeply progressive, such that it's reduced to 1.46% at 62 and 1% at 59. The contributory plan is a 2% at 61 and drops off more gradually such that it pays 1.5% at 55 and 2.43% at 65. (The post-pension reform plan grows linearly, such that the formula is 1.5% at 57, 2% at 62, and 2.5% at 67.)

For those who stayed until 65 and invested what they would have paid into the contributory plan to a 457 or IRA, they would have made out like bandits. Of course, many people just spent their contributions, and old county employees told me a lot of folks bought boats and cars with that cashout money. And if they retired before 65 and had to claim on leaving the workforce (instead of deferring until 65) they would have lost out. There are quite a few old timers on Plan E ("empty") hanging on until 65 to max out their benefit. And nothing is stopping the super saver from doing the contributory retirement plan, investing in the 457, and in the 401(k) (which is only offered to management) for $39,000 + 8% of salary of tax deferral.
prioritarian
Posts: 381
Joined: Tue Jul 16, 2019 6:00 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by prioritarian »

BrandonBogle wrote: Tue Nov 16, 2021 8:10 pm
BrandonBogle wrote: Tue Nov 16, 2021 4:55 pm When I made $20k/year in the early 2000s...

I’m not sure what point you are trying to make. My point was that one can still save for retirement even if you don’t come from a family with means. I don’t see anything counter to that in that chart.

I didn’t say I made $50k/year in the early 2000s nor did I say anything about the cost of a college education. I took out 6 figures in student loans while making $20k/year and contributed to my IRA and 401k throughout. That hasn’t changed to today. Student loans are still available and the case presented in the article has 250% higher pay than the example I gave.
My point is that lower/lower-middle income people require far higher loan levels to pay for tuition (due to stratospheric inflation) even though student loans/grants are less available (on a per student basis) than they were 20 years ago.

Individual anecdotes will not change my mind.
Last edited by prioritarian on Wed Nov 17, 2021 10:49 am, edited 1 time in total.
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Tue Nov 16, 2021 4:01 pm
Kenkat wrote: Tue Nov 16, 2021 3:16 pm
JackoC wrote: Tue Nov 16, 2021 2:37 pm
Kenkat wrote: Tue Nov 16, 2021 12:03 pm Again, no where in the article was an 8% real return mentioned. It just says 8%.
OK so maybe it's a meaningless number, a nominal rate of return quoted over multi-decades, rather than a ridiculously high number for expected real return. :happy But I'd say 8% nominal is still unrealistically high as midpoint expectation, myself, unless one projects higher than market expected inflation.

Though I don't 100% disagree with your later post. The general idea of saving robustly all along is one I agree with. I just still think there's a tendency in pop-finance toward the shiny object of a big eventual NW, which isn't a realistic expectation for median-ish income households IMO starting from where assets are valued now (though two $50k people would be a $100k household, well above median). But if they don't save much, they'll have to accept a diminution in living standard if SS isn't made more generous, let alone if it's eventually cut.
Yes, maybe 8% is too high from this point - if I had to pick, I’d say it probably is. That said, when you are investing regularly, the 2000-2002 and 2008-2009 type events might change all of that - i.e., not every dollar will go in at current valuations. It is highly variable though - the $2 million is definitely an attention getter; you’ll end up with a lot of money but I can’t say how much would probably be better but that doesn’t sell papers…er, clicks.
1. If you are just getting going now then low expected rates don't matter. The sequence in the decade before and after retirement are what matter. The rational expectation would be something like historical returns for someone starting out now. Late returns will dominate the portfolio performance and new additions will almost completely make up for any low initial returns.

2. Reaching high net worth relative to your income is relatively simple if you can live below your means. People that have trouble with this at 200k will have trouble at 100k and at 50k, but there are those that can adapt and always live below their means.
1. Re: the bolded the assumption is some tendency for returns to 'eventually' return to whatever they were in the last several decades. But actually there's no reason to think this. In fact 'reversion to the mean' if applied to stock valuations would mean poorer returns than implied by today's high valuations. A simple theoretical estimate of real expected return is E/P. It does not have a time dimension because a stock is a perpetual. If we take 'E' to be the inflation adjusted average of E over a recent business cycle-type period, 1/CAPE, for the world stock market that would be under 4% real expected return (looks worse just considering the US market but let's stick to the sunny side). That implicitly assumes the CAPE is constant. If the valuation 'reverts' to where expected return becomes 6% or 8% at constant CAPE, there'd be less or much less than 4% expected return for the 'reversion period'. But if the market actually expected stock return much less than 4% real for years, arguably it would just sell off now. 3-4% real is neither a very wide nor very narrow margin to 10 yr expected real bond return (-1.1%). It looks pretty much the ballpark by that measure and also in terms of return=dividend yield+EPS growth, 1.6% global div yield assuming real EPS growth somewhat lags world real GDP growth trend taken as around 3% (EPS growth lagged GDP growth in the US in whole 20th century by >1% point). A 'bank shot' planning estimate of lower return sooner to set up higher return later is not rational IMO. And past return, from a different valuation, yield and economic fundamentals, is basically irrelevant.

And this is as usual more obvious with bonds. Say horizon 30 yrs. The assumption that the *expected* return in the period 25 yrs from now to 30 yrs from now is much higher than the 25X30 forward rate is either assuming a) a big negative term premium* or b) the market is grossly inefficient. If a younger person making somewhere in the ballpark of $50k can see that their bond investment return starting 25 years from now will be much higher than the 25X30 forward rate now, why can't the market see that (and bid up that rate right now)? The rational assumption is that expected bond return is an implied by today's curve, could be less or more (if you don't just lock it in for 30 yrs) but the midpoint expectation can't be much better than implied by the curve in an efficient market.

If the market is efficient valuation now gives the best estimate of expected return now and past returns are basically irrelevant to the expected return now (to the extent they matter, they should already be factored into today's pricing along with all other relevant info).

2. If you modify it to, 'reaching the highest net worth relative to your income that market returns will allow is relatively simple if you live below your means' then I agree. But the return assumption in the article heavily flatters the absolute terminal number vs expected reality.

*term premium: the difference between the forward rate between two dates on todays' curve and the market's actual expectation of the spot rate when the first of those dates becomes spot. A positive term premium is where the forward rate is higher than that expectation, the usual case historically, but NY Fed's ACM model now says the term premium out to 10 yrs is ~zero.
Last edited by JackoC on Wed Nov 17, 2021 10:51 am, edited 1 time in total.
User avatar
BrandonBogle
Posts: 4204
Joined: Mon Jan 28, 2013 11:19 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by BrandonBogle »

prioritarian wrote: Wed Nov 17, 2021 10:36 am
BrandonBogle wrote: Tue Nov 16, 2021 8:10 pm
BrandonBogle wrote: Tue Nov 16, 2021 4:55 pm When I made $20k/year in the early 2000s...

I’m not sure what point you are trying to make. My point was that one can still save for retirement even if you don’t come from a family with means. I don’t see anything counter to that in that chart.

I didn’t say I made $50k/year in the early 2000s nor did I say anything about the cost of a college education. I took out 6 figures in student loans while making $20k/year and contributed to my IRA and 401k throughout. That hasn’t changed to today. Student loans are still available and the case presented in the article has 250% higher pay than the example I gave.
My point is that low income people require far higher loan levels to pay for tuition (due to stratospheric inflation) even though student loans/grants are less available (per student) than they were 20 years ago.

Individual anecdotes will not change my mind.
Gotcha. A valid point. Please note, my point was not trying to change your mind and your point goes along with mine with issue. Low income folks require more in loans (which seem to be harder to find), but one should still be able to save something if they are willing to make the tough choices others have pointed out in this thread. Definitely easier if you have the means, but still doable for low income folks.
dknightd
Posts: 2770
Joined: Wed Mar 07, 2018 11:57 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by dknightd »

I'm still trying to figure out why you would need $2 million if you were used to living on $50k
If you value a bird in the hand, pay off the loan. If you are willing to risk getting two birds from the market, invest the funds. Retired 9/19. Mortgage payed off 5/21. I have some 0% loans to pay off.
User avatar
willthrill81
Posts: 27088
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by willthrill81 »

dknightd wrote: Wed Nov 17, 2021 11:08 am I'm still trying to figure out why you would need $2 million if you were used to living on $50k
Because the $2 million would be in 2062 dollars. If the investor earned 5% inflation-adjusted returns and adjusted the contributions for inflation, the inflation-adjusted ending balance would be $959k, providing about $38k in withdrawals, assuming a 4% withdrawal rate.
Last edited by willthrill81 on Wed Nov 17, 2021 1:55 pm, edited 1 time in total.
“Good and ill have not changed since yesteryear; nor are they one thing among Elves and Dwarves and another among Men.” J.R.R. Tolkien, The Lord of the Rings
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

dknightd wrote: Wed Nov 17, 2021 11:08 am I'm still trying to figure out why you would need $2 million if you were used to living on $50k
'Fortunately' a person who never earns more than $50k in 2021 dollars is quite unlikely to ever amass $2million in 2021 dollars (Ok the article wasn't clear on nominal v 2021 dollars but only the latter makes any sense to talk about, 'I saved avidly and amassed 100 trillion Zimbabwe Dollars', who cares?). Though anything is possible.
Maverick3320
Posts: 816
Joined: Tue May 12, 2015 2:59 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Maverick3320 »

dukeblue219 wrote: Mon Nov 15, 2021 7:48 pm
stoptothink wrote: Mon Nov 15, 2021 3:32 pm My younger sister rents an apartment (with two roommates) literally a block away from my home, in an area that is certainly not San Francisco but significantly above the median for the country in COL. She saves a significant part of her ~$40k/yr income (she's a 2nd year, 2nd grade teacher)
Crazy part of this is that school teachers in some parts of the US are still paid $40k/year, and don't give me the "it's only 10 months of work" thing.
My mom was a teacher. And you're right: it was only 9 months of work a year. The other three months she taught at a local community college for an extra 5-6k or so of income.

I think she topped out at around 60k/year in an extremely LCOL area (houses in town are around 150k). She now has a lake house in the area she retired from and a condo down in Fort Myers that she lives in half the year.
prioritarian
Posts: 381
Joined: Tue Jul 16, 2019 6:00 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by prioritarian »

Maverick3320 wrote: Wed Nov 17, 2021 11:30 am
dukeblue219 wrote: Mon Nov 15, 2021 7:48 pm
stoptothink wrote: Mon Nov 15, 2021 3:32 pm My younger sister rents an apartment (with two roommates) literally a block away from my home, in an area that is certainly not San Francisco but significantly above the median for the country in COL. She saves a significant part of her ~$40k/yr income (she's a 2nd year, 2nd grade teacher)
Crazy part of this is that school teachers in some parts of the US are still paid $40k/year, and don't give me the "it's only 10 months of work" thing.
My mom was a teacher. And you're right: it was only 9 months of work a year. The other three months she taught at a local community college for an extra 5-6k or so of income.

I think she topped out at around 60k/year in an extremely LCOL area (houses in town are around 150k). She now has a lake house in the area she retired from and a condo down in Fort Myers that she lives in half the year.
Educators typically have benefits that total a significant fraction of their total wages but this is exceptionally rare for most lower-middle/lower-income people.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Wed Nov 17, 2021 10:49 am
abc132 wrote: Tue Nov 16, 2021 4:01 pm 1. If you are just getting going now then low expected rates don't matter. The sequence in the decade before and after retirement are what matter. The rational expectation would be something like historical returns for someone starting out now. Late returns will dominate the portfolio performance and new additions will almost completely make up for any low initial returns.

2. Reaching high net worth relative to your income is relatively simple if you can live below your means. People that have trouble with this at 200k will have trouble at 100k and at 50k, but there are those that can adapt and always live below their means.
1. Re: the bolded the assumption is some tendency for returns to 'eventually' return to whatever they were in the last several decades. But actually there's no reason to think this. In fact 'reversion to the mean' if applied to stock valuations would mean poorer returns than implied by today's high valuations. A simple theoretical estimate of real expected return is E/P. It does not have a time dimension because a stock is a perpetual. If we take 'E' to be the inflation adjusted average of E over a recent business cycle-type period, 1/CAPE, for the world stock market that would be under 4% real expected return (looks worse just considering the US market but let's stick to the sunny side). That implicitly assumes the CAPE is constant. If the valuation 'reverts' to where expected return becomes 6% or 8% at constant CAPE, there'd be less or much less than 4% expected return for the 'reversion period'. But if the market actually expected stock return much less than 4% real for years, arguably it would just sell off now. 3-4% real is neither a very wide nor very narrow margin to 10 yr expected real bond return (-1.1%). It looks pretty much the ballpark by that measure and also in terms of return=dividend yield+EPS growth, 1.6% global div yield assuming real EPS growth somewhat lags world real GDP growth trend taken as around 3% (EPS growth lagged GDP growth in the US in whole 20th century by >1% point). A 'bank shot' planning estimate of lower return sooner to set up higher return later is not rational IMO. And past return, from a different valuation, yield and economic fundamentals, is basically irrelevant.
Assuming the historical average doesn't need a mean reversion, it simply says
1) the first 10 years of returns hardly matter at all to someone working and accumulating for 40 years and investing for another 20-30 after that.
2) anyone that tries to give you a better 40-year value does not recognize the deviation on that value, nor what time added money does to portfolio expectations

The person no longer accumulating may receive and care about your prediction because their portfolio is large and without future contributions, but the person starting out and investing 15% per year should ignore your predictions completely.

Please plot your CAPE prediction on the date accumulation begins vs historical portfolio outcome 40 years later with 15% annual additions and tell me what you get. I'm on record that I think your year one metric has very little bearing on the portfolio size after 40 years of investing.
Last edited by abc132 on Wed Nov 17, 2021 3:32 pm, edited 1 time in total.
Cubs Fan
Posts: 33
Joined: Wed Nov 11, 2020 12:22 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Cubs Fan »

I started investing at age 50. I am now 78. I never made more than $100.000.00 per year. Currently our home is worth $450,000.00, and we do not have any debt. Our portfolio balance is $1,700,000.00. I practice the concept of being a self-directed investor. If I withdraw 4% per year for living expenses why pay an advisor 2% per year? I am not going to give someone half of my retirement income when all I need to do is invest in low cost ETFs or mutual funds based on my IPS of 60/40. Use a formula, turn the crank, and pull out the investment answer. This is so easy I don't know why people think that an advisor is working in their best interest. "How big is my advisor's boat?" Thank you Jack Bogle and Taylor.
sandan
Posts: 288
Joined: Wed Apr 03, 2013 12:48 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by sandan »

abc132 wrote: Tue Nov 16, 2021 4:01 pm
Kenkat wrote: Tue Nov 16, 2021 3:16 pm
JackoC wrote: Tue Nov 16, 2021 2:37 pm
Kenkat wrote: Tue Nov 16, 2021 12:03 pm Again, no where in the article was an 8% real return mentioned. It just says 8%.
OK so maybe it's a meaningless number, a nominal rate of return quoted over multi-decades, rather than a ridiculously high number for expected real return. :happy But I'd say 8% nominal is still unrealistically high as midpoint expectation, myself, unless one projects higher than market expected inflation.

Though I don't 100% disagree with your later post. The general idea of saving robustly all along is one I agree with. I just still think there's a tendency in pop-finance toward the shiny object of a big eventual NW, which isn't a realistic expectation for median-ish income households IMO starting from where assets are valued now (though two $50k people would be a $100k household, well above median). But if they don't save much, they'll have to accept a diminution in living standard if SS isn't made more generous, let alone if it's eventually cut.
Yes, maybe 8% is too high from this point - if I had to pick, I’d say it probably is. That said, when you are investing regularly, the 2000-2002 and 2008-2009 type events might change all of that - i.e., not every dollar will go in at current valuations. It is highly variable though - the $2 million is definitely an attention getter; you’ll end up with a lot of money but I can’t say how much would probably be better but that doesn’t sell papers…er, clicks.
If you are just getting going now then low expected rates don't matter. The sequence in the decade before and after retirement are what matter. The rational expectation would be something like historical returns for someone starting out now. Late returns will dominate the portfolio performance and new additions will almost completely make up for any low initial returns.

Reaching high net worth relative to your income is relatively simple if you can live below your means. People that have trouble with this at 200k will have trouble at 100k and at 50k, but there are those that can adapt and always live below their means.
Timing does matter. It looks bleak even with nominal dollars.

The 30 year risk free interest rate is ~2% (this is not an expected rate). In order to get 8%, the risk premium for stock is going to be 6%. That value is outrageously high.

For a simple example, I picked the year 2000 (VTI's inception). The nominal returns are 8.6%. At the same time treasury yields were 5%. Using this naïve measure, the suggested risk premium over treasuries is around 4%.

Overall, I think many can agree that there has been a hunt for finding investments providing risk premiums for at least 30 years (before the internet). Its really unlikely risk premiums are higher now than before.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

sandan wrote: Wed Nov 17, 2021 3:40 pm
abc132 wrote: Tue Nov 16, 2021 4:01 pm
Kenkat wrote: Tue Nov 16, 2021 3:16 pm
JackoC wrote: Tue Nov 16, 2021 2:37 pm
Kenkat wrote: Tue Nov 16, 2021 12:03 pm Again, no where in the article was an 8% real return mentioned. It just says 8%.
OK so maybe it's a meaningless number, a nominal rate of return quoted over multi-decades, rather than a ridiculously high number for expected real return. :happy But I'd say 8% nominal is still unrealistically high as midpoint expectation, myself, unless one projects higher than market expected inflation.

Though I don't 100% disagree with your later post. The general idea of saving robustly all along is one I agree with. I just still think there's a tendency in pop-finance toward the shiny object of a big eventual NW, which isn't a realistic expectation for median-ish income households IMO starting from where assets are valued now (though two $50k people would be a $100k household, well above median). But if they don't save much, they'll have to accept a diminution in living standard if SS isn't made more generous, let alone if it's eventually cut.
Yes, maybe 8% is too high from this point - if I had to pick, I’d say it probably is. That said, when you are investing regularly, the 2000-2002 and 2008-2009 type events might change all of that - i.e., not every dollar will go in at current valuations. It is highly variable though - the $2 million is definitely an attention getter; you’ll end up with a lot of money but I can’t say how much would probably be better but that doesn’t sell papers…er, clicks.
If you are just getting going now then low expected rates don't matter. The sequence in the decade before and after retirement are what matter. The rational expectation would be something like historical returns for someone starting out now. Late returns will dominate the portfolio performance and new additions will almost completely make up for any low initial returns.

Reaching high net worth relative to your income is relatively simple if you can live below your means. People that have trouble with this at 200k will have trouble at 100k and at 50k, but there are those that can adapt and always live below their means.
Timing does matter. It looks bleak even with nominal dollars.

The 30 year risk free interest rate is ~2% (this is not an expected rate). In order to get 8%, the risk premium for stock is going to be 6%. That value is outrageously high.

For a simple example, I picked the year 2000 (VTI's inception). The nominal returns are 8.6%. At the same time treasury yields were 5%. Using this naïve measure, the suggested risk premium over treasuries is around 4%.

Overall, I think many can agree that there has been a hunt for finding investments providing risk premiums for at least 30 years (before the internet). Its really unlikely risk premiums are higher now than before.
Pretend someone adds 10,000 on (Jan 1, 2000) and continues to do so annually through 2021, investing dividends.

The CAGR of each 10,000 investment from Jan 1 2000 to Jan 1 2021 is:
7.27%
8.11%
9.22%
11.17%
10.30%
10.28%
10.64%
10.33%
10.70%
15.55%
14.68%
14.67%
16.04%
16.10%
14.28%
14.41%
16.77%
17.82%
16.93%
25.53%
21.92%
26.13%

Every new investment after the initial Jan 2000 investment gets higher than 8%.

If we weight the above returns by the number of years we got those returns, the time average of the portfolio is 11.96%. (Multiply the the Jan 2000 rate by 22 years, 2001 by 21, .... Add them up and divide by (22+21+20+...+1).

(Note I counted 2021 as a full year with 1.26% dividends)

I am shocked that so many people do not consider additional investments when discussing valuations.

If my time averaging is *reasonable, their risk premium was 7% over the initial bond rate, which with bonds at 2% gets us to 9% nominal rather than the 8% suggested in the article. Low initial returns just means higher expected returns for future additions, so in a poor expected sequence one should expect better than the current predictions when making additions over a long period of time.

Conclusion: It is not necessarily a bad thing to start in a poor sequence. Ignore the noise and make a "good enough" plan.

* The rule of 72 is an approximation that says it will take X years to double, where X=72/interest rate. Getting 4% for 18 years should double your money, and getting 9% for 8 years should double your money. Thus it seams reasonable that time averaging the rates by number of years is reasonable, as 4*18=9*8.
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Wed Nov 17, 2021 3:04 pm
JackoC wrote: Wed Nov 17, 2021 10:49 am
abc132 wrote: Tue Nov 16, 2021 4:01 pm 1. If you are just getting going now then low expected rates don't matter. The sequence in the decade before and after retirement are what matter. The rational expectation would be something like historical returns for someone starting out now. Late returns will dominate the portfolio performance and new additions will almost completely make up for any low initial returns.

2. Reaching high net worth relative to your income is relatively simple if you can live below your means. People that have trouble with this at 200k will have trouble at 100k and at 50k, but there are those that can adapt and always live below their means.
1. Re: the bolded the assumption is some tendency for returns to 'eventually' return to whatever they were in the last several decades. But actually there's no reason to think this. In fact 'reversion to the mean' if applied to stock valuations would mean poorer returns than implied by today's high valuations. A simple theoretical estimate of real expected return is E/P. It does not have a time dimension because a stock is a perpetual. If we take 'E' to be the inflation adjusted average of E over a recent business cycle-type period, 1/CAPE, for the world stock market that would be under 4% real expected return (looks worse just considering the US market but let's stick to the sunny side). That implicitly assumes the CAPE is constant. If the valuation 'reverts' to where expected return becomes 6% or 8% at constant CAPE, there'd be less or much less than 4% expected return for the 'reversion period'. But if the market actually expected stock return much less than 4% real for years, arguably it would just sell off now. 3-4% real is neither a very wide nor very narrow margin to 10 yr expected real bond return (-1.1%). It looks pretty much the ballpark by that measure and also in terms of return=dividend yield+EPS growth, 1.6% global div yield assuming real EPS growth somewhat lags world real GDP growth trend taken as around 3% (EPS growth lagged GDP growth in the US in whole 20th century by >1% point). A 'bank shot' planning estimate of lower return sooner to set up higher return later is not rational IMO. And past return, from a different valuation, yield and economic fundamentals, is basically irrelevant.
1. Assuming the historical average doesn't need a mean reversion, it simply says
1) the first 10 years of returns hardly matter at all to someone working and accumulating for 40 years and investing for another 20-30 after that.
2) anyone that tries to give you a better 40-year value does not recognize the deviation on that value, nor what time added money does to portfolio expectations

The person no longer accumulating may receive and care about your prediction because their portfolio is large and without future contributions, but the person starting out and investing 15% per year should ignore your predictions completely.

2. Please plot your CAPE prediction on the date accumulation begins vs historical portfolio outcome 40 years later with 15% annual additions and tell me what you get. I'm on record that I think your year one metric has very little bearing on the portfolio size after 40 years of investing.
1. the first thing is true, but any relevance depends on the assumption you have a better idea what the return will be in the last 20-30 yrs than the first 10, as by the assumption 'in the long run, return will be whatever it was in the last X decades'. But there is again no reasonable basis for that assumption. Hence while it's elementary that if you mainly save the money in years 20-30 that the return in years 20-30 will matter more, there is no basis to think the best estimate for expected return of stocks in that period is any different than the best estimate from now.

2. Stocks are *risk assets*. If the realized return in each period had always been equal or close to the ex-ante expected return there would not have been any risk! :happy So we know that's not going to be the result of plotting any past estimate of ex-ante return against ex-post realized return. However if you go back and look at ex-ante expected return based on 1/CAPE in the past, it was generally significantly higher than now. Plus, valuation generally increased, a significant tailwind up to now. The *theoretical* estimate (look up the derivation, we shouldn't have to go through it each time) that real expected return=P/E implicitly assumes P/E is as likely to go up in the future as down. *If* the estimate was assuming the expected future value of CAPE in 10 yrs was lower *then* it might be reasonable to say the expected return thereafter was higher. But if estimating it as 1/CAPE there's no reason to think the 20x30 expected return is any different than the 0x10 expected return. There is no additional piece of information to split the expected return of the *perpetual* into a near term and far term expected return.

And as already mentioned you can do at least two reality checks on 1/CAPE as estimate of expected return. An alternative fundamental estimate is return=div yield+real EPS growth. World div yield is 1.6%, a reasonable estimate of world real GDP growth trend is 3%, real EPS growth in the US 20th century lagged real EPS growth by 1+% so that would also come out (1.6+3-1+) in the 3's, 1/CAPE gives high 3's for expected real return of the world stock market. And if we prefer historical analysis but realize it would not make sense for risky expected return to be a constant regardless of riskless expected return, we can compare the risky expected return estimate to the riskless curve and realize 3.some% risky expected return is 4+% premium over the riskless return which is not that narrow.

I don't know what actual factual analysis can lead to the conclusion expected return now for stocks is 8% real, or even 6%, whether years 0-10 or 20-30 except the arbitrary assumption that 'normal stock return' is at that level regardless of valuation and riskless rates, a very hard to defend assumption IMO.
User avatar
Garco
Posts: 886
Joined: Wed Jan 23, 2013 2:04 am
Location: U.S.A.

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Garco »

$50,000?? My SALARY??? No way.

After 10 years of college (paid for by my parents and scholarships while in grad school), I got out at age 27 with a PhD but without any debt. THE LACK OF DEBT at the start is pretty crucial to that simplistic formula.

And another thing: My first job after grad school paid the princely sum of $14,000 per year.

On the way to $2 million at age 70? Not on an annual salary of $14K.

But I got raises every year, and I stayed employed, and I invested in a 403b plan -- with combined employee and employer contributions of 15% every year to the plan.

And guess what? After 43 years of employment (after I'd finished grad school) I retired at age 70 with $2 million in my 403b plan.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Thu Nov 18, 2021 10:12 am
I don't know what actual factual analysis can lead to the conclusion expected return now for stocks is 8% real, or even 6%, whether years 0-10 or 20-30 except the arbitrary assumption that 'normal stock return' is at that level regardless of valuation and riskless rates, a very hard to defend assumption IMO.
If true, there is no reason to plan for a historically low value either.

You can't have it both ways, and you are misunderstanding me as declaring a future value as "the value". I said a historical value is as good as any other value, which is much different than what you seem to be representing. The rational was the uncertainty of the value - which is in direct opposition to me predicting a value.

I made the point that the current expectations are almost meaningless to someone accumulating and provided a historical example of why that was true. The burden is on the complainers about the 8% value to give a better value. Using current expectations is fundamentally flawed for an accumulator, as their money will enter the market further in the future when expectations will be much different. I showed this through your very own case - getting 12% nominal returns starting in 2000 when future expected returns were very low. It wasn't a case of the luck of the draw, it was a case of the forward estimate when portfolio=$0 being irrelevant to the accumulator. Feel free to repeat the experiment for every time period.

I guess I will end with the concept that present expectations really don't matter when your future portfolio is dominated by future contributions. It is a mathematical concept that has nothing to do with whatever we can or can not predict, and anyone arguing exclusively about what we can predict has not grasped the fundamental concept that we are not affected by returns when our accumulated portfolio has little exposure to those initial returns.

Please ignore the low performance predictors if you are just starting out. They have not made the transition that what happens to their accumulated portfolio will be much different than what happens to your accumulation portfolio.
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Thu Nov 18, 2021 1:32 pm
JackoC wrote: Thu Nov 18, 2021 10:12 am
I don't know what actual factual analysis can lead to the conclusion expected return now for stocks is 8% real, or even 6%, whether years 0-10 or 20-30 except the arbitrary assumption that 'normal stock return' is at that level regardless of valuation and riskless rates, a very hard to defend assumption IMO.
1. If true, there is no reason to plan for a historically low value either.

2. You can't have it both ways, and you are misunderstanding me as declaring a future value as "the value". I said a historical value is as good as any other value, which is much different than what you seem to be representing. The rational was the uncertainty of the value - which is in direct opposition to me predicting a value.

3. I made the point that the current expectations are almost meaningless to someone accumulating and provided a historical example of why that was true. The burden is on the complainers about the 8% value to give a better value. Using current expectations is fundamentally flawed for an accumulator, as their money will enter the market further in the future when expectations will be much different. I showed this through your very own case - getting 12% nominal returns starting in 2000 when future expected returns were very low. It wasn't a case of the luck of the draw, it was a case of the forward estimate when portfolio=$0 being irrelevant to the accumulator. Feel free to repeat the experiment for every time period.

4. I guess I will end with the concept that present expectations really don't matter when your future portfolio is dominated by future contributions. It is a mathematical concept that has nothing to do with whatever we can or can not predict, and anyone arguing exclusively about what we can predict has not grasped the fundamental concept that we are not affected by returns when our accumulated portfolio has little exposure to those initial returns.

5. Please ignore the low performance predictors if you are just starting out. They have not made the transition that what happens to their accumulated portfolio will be much different than what happens to your accumulation portfolio.
1. Again the estimate real expected return=1/CAPE assumes the future expected value of CAPE is *today's* high CAPE. It does not assume CAPE goes down. If you assume the expected future value of CAPE is lower than now's the expected return is <1/CAPE.

2. There's nothing 'both ways' in my argument. And if 'a historical value' is as good any other value why isn't *any* value as good as any other value? Again consider bonds. The long term historical real pre tax return of long term bonds was 2-2.5% roughly depending how far back and what bond (for example 30 yr bond data doesn't go back that far, very long term data is the 10 yr). Now the 30 yr TIPS yield is -0.50%. So, 2-2.5% real pre tax return is just as good an estimate of my expected 30 yr return now if I buy the 30 yr TIPS now at -0.50%? :shock: I don't think so. And by slight extension 2.5% is also irrelevant if I choose to buy 5 yr bonds now and roll them over out to 30 yrs. If the midpoint expectation of that strategy was a real return of 2.5%, who are the idiots locking it for 30 yrs at -0.50% by buying the 30 yr? For bonds 'the historical return is as good an estimate of the expected return as as any' is a clearly ridiculous statement. You'd have to find a good argument why this concept would be valid for stocks when it clearly isn't for bonds (hint: you won't find such an argument, the concept is invalid in either case).

3. The expected return from 0 to 30 is only 'meaningless' if you have a particular reason to think the forward expected return from say 10 to 30 is different. You don't, is my entire point. When 'future money enters the market' the expected return may be different, but you've no reason to think the expected return now is a biased estimator of it. Without assuming valuation will decline to some intermediate point, which the estimate expected return=1/CAPE *does not assume*, there is no reason to assume a lower expected return near term and a higher one later. And 'the complainers' have given an alternative estimate to the obviously too high 8% real, which comes out similarly 3 different ways of looking at it: 1) real expected return=1/CAPE=<4% for world stocks (again, please look up and understand the theoretical derivation of E[r]=E/P), 2) real expected return=div yield+real EPS growth rate, 1.6% div, 3% world real GDP growth trend, 1%+ shortfall in EPS growth v GDP growth in the US in 20th century, also comes out in the 3's 3) reality check: how wide a premium is a 3-4% real expected return for stocks over the riskless rate now? a pretty healthy one because the riskless real return is now negative all the way to 30 yrs.

4. A baseless concept unless you have some better reason to say the expected return starting at a point in the future is higher than now's as by 'convergence' to whatever the long term historical return was. But there's no reason to think that's true.

5. You've therefore given no good reason to ignore valuation and fundamental measure of expected return, for savers at any stage of life. Uncertainty increases with time but stock investors with various horizons (or starting points) have no reason to assume different expected returns, according to simple estimates like 1/PE or div yield+EPS growth which assume zero expected change in future valuation.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Thu Nov 18, 2021 4:38 pm
abc132 wrote: Thu Nov 18, 2021 1:32 pm 1. Again the estimate real expected return=1/CAPE assumes the future expected value of CAPE is *today's* high CAPE. It does not assume CAPE goes down. If you assume the expected future value of CAPE is lower than now's the expected return is <1/CAPE.
We don't really need get into this since you are applying 1/CAPE into perpetuity, which is severely flawed logic.

JackoC wrote: Thu Nov 18, 2021 4:38 pm 2. There's nothing 'both ways' in my argument. And if 'a historical value' is as good any other value why isn't *any* value as good as any other value?
Any value is as good as any other value when you are talking 40-60 years of returns with 40 years of additions. 1/CAPE was never intended for such long periods. Have you tried using 1/CAPE over 40-60 year periods to see how well it predicts? It is terrible and 1/CAPE is usually used for 10 year returns.

Assuming the future returns for 10 years is equal to the future returns for 60 years is completely flawed logic. High initial valuations wash out over very long periods of time. If P/E hit 100 it would not mean the future return of the stock market for the next 60 years are expected to be 1%. It would mean there is a shorter term preference for stocks, and that 10 year returns might be muted because of this.


JackoC wrote: Thu Nov 18, 2021 4:38 pm Again consider bonds. The long term historical real pre tax return of long term bonds was 2-2.5% roughly depending how far back and what bond (for example 30 yr bond data doesn't go back that far, very long term data is the 10 yr). Now the 30 yr TIPS yield is -0.50%. So, 2-2.5% real pre tax return is just as good an estimate of my expected 30 yr return now if I buy the 30 yr TIPS now at -0.50%? :shock: I don't think so. And by slight extension 2.5% is also irrelevant if I choose to buy 5 yr bonds now and roll them over out to 30 yrs. If the midpoint expectation of that strategy was a real return of 2.5%, who are the idiots locking it for 30 yrs at -0.50% by buying the 30 yr? As a general proposition across assets classes, 'the historical return is as good an estimate of the expected return as as any' is a clearly ridiculous statement. You'd have to find a good argument why this concept would be valid for stocks when it clearly isn't for bonds (hint: you won't find such an argument, the concept is invalid in either case).
I would expect stock returns to be juiced by low bond returns, as we all choose between these two assets. You would have predicted very low stock returns for the last 10 years by adding the historical premium to low bond returns which is contrary to what anyone else would expect. I prefer bonds when they pay me highly and that actually reduces my preference for stocks. I expect a lower premium for stocks when bonds pay well (why take the risk?) and a higher one when they don't (more people need that risk). That is why your prediction for the last 10 years would have been terrible - low rates increased stock returns.

JackoC wrote: Thu Nov 18, 2021 4:38 pm 3. The expected return from 0 to 30 is only 'meaningless' if you have a particular reason to think the forward expected return from say 10 to 30 is different. You don't, is my entire point. When 'future money enters the market' the expected return may be different, but you've no reason to think the expected return now is a biased estimator of it. Without assuming valuation will decline to some intermediate point, which the estimate expected return=1/CAPE *does not assume*, there is no reason to assume a lower expected return near term and a higher one later. And 'the complainers' have given an alternative estimate to the obviously too high 8% real, which comes out similarly 3 different ways of looking at it: 1) real expected return=1/CAPE=<4% for world stocks (again, please look up and understand the theoretical derivation of E[r]=E/P), 2) real expected return=div yield+real EPS growth rate, 1.6% div, 3% world real GDP growth trend, 1%+ shortfall in EPS growth v GDP growth in the US in 20th century, also comes out in the 3's 3) reality check: how wide a premium is a 3-4% real expected return for stocks over the riskless rate now? a pretty healthy one because the riskless real return is now negative all the way to 30 yrs.
I have a great reason to believe the current prediction is biased because it is nowhere near a constant value and it was only meant for 10 years. There is no way a 50% drop in the market means our 60 year returns will go from 4% to 8%. A 50% market drop would create 5x as much future value 60 years later, which is ridiculous. If we apply this as intended to something like a 10 year period, a 50% drop giving us 8% returns for 10 years recovers only some of the value lost from the drop.
JackoC wrote: Thu Nov 18, 2021 4:38 pm 4. A baseless concept unless you have some better reason to say the 'very long term' expected return will converge to whatever the long term historical return was.
It doesn't need to be historical, it just can't create 5x future value from a 50% drop. A rational constraint on predictions is that a drop may help future returns, but is not a net benefit to the economy.
JackoC wrote: Thu Nov 18, 2021 4:38 pm 5. You've given no good reason to ignore valuation and fundamental measure of expected return, for savers at any stage of life. Uncertainty increases with time, but there just isn't any valid concept of expected return, over any horizon, being equal to long term past return without considering the valuations/yields, which generated those past returns, vs now's.
I gave excellent reasons to ignore your expected returns. I hope you will take the time to digest them.
sandan
Posts: 288
Joined: Wed Apr 03, 2013 12:48 pm

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by sandan »

abc132 wrote: Wed Nov 17, 2021 10:55 pm I am shocked that so many people do not consider additional investments when discussing valuations.
Valuations don't even need to be discussed.

Its not difficult to subtract the 20 year yield from the 30 year yield to figure out the market interest rate between year 20 and 30.

The rates are still bleak.

People shouldn't need a false promise of rosy returns to save consistently. I'm not sure why some bogleheads feel adamantly about endorsing a website marketing piece.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

sandan wrote: Fri Nov 19, 2021 10:53 am
abc132 wrote: Wed Nov 17, 2021 10:55 pm I am shocked that so many people do not consider additional investments when discussing valuations.
Valuations don't even need to be discussed.

Its not difficult to subtract the 20 year yield from the 30 year yield to figure out the market interest rate between year 20 and 30.

The rates are still bleak.

People shouldn't need a false promise of rosy returns to save consistently. I'm not sure why some bogleheads feel adamantly about endorsing a website marketing piece.
Nobody knows the correct value, which is what makes all the angst about 8% nominal so silly.

The new investor that would benefit from reading this article doesn't fully understand the implications of the numbers being used, and would make the same conclusion if you drop the rate and extend the years to get the same 2 million dollar figure. What they retain is that money compounds and they can accumulate a large sum even on a modest salary - which is 100% true.

Adding additional details "real vs nominal", "1/CAPE" "20 vs 30 year interest spreads" will only muddy the conclusion and is not in the interest of the reader. I certainly never concerned myself with these things starting out - they only became relevant as my portfolio size became meaningful towards my long term retirement goals.
User avatar
CyclingDuo
Posts: 4599
Joined: Fri Jan 06, 2017 9:07 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by CyclingDuo »

sandan wrote: Fri Nov 19, 2021 10:53 am People shouldn't need a false promise of rosy returns to save consistently. I'm not sure why some bogleheads feel adamantly about endorsing a website marketing piece.
Now that everyone has “article shamed her”, we can move on - or better yet, invest $625 a month for 41 years and see what really happens.😎

CyclingDuo
"Save like a pessimist, invest like an optimist." - Morgan Housel
User avatar
Portfolio7
Posts: 1054
Joined: Tue Aug 02, 2016 3:53 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Portfolio7 »

Most people move upwards in income over their lifetime in real terms.

I started after college in 1990 making $24.5K for a couple years. Lost my job, spent my savings down to zero (jobless for 9 months) before getting a new job making perhaps $15-20K annually. Two years later was back up to $30K annual pay, paid down college debt, then started investing in 1996. I didn't even save 15% every year, but I did so for perhaps the first 8 years of investing, and with employer matching always managed to save 8-10% minimum. I hit 6 figures in income a few years back, and 7 figures in retirement investments shortly after. The point being:

- People shift between income deciles over time, typically moving up as they age.
- Modest saving efforts generally result in significant net worth results 25-40 years later.
"An investment in knowledge pays the best interest" - Benjamin Franklin
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Thu Nov 18, 2021 5:29 pm 1. We don't really need get into this since you are applying 1/CAPE into perpetuity, which is severely flawed logic.

2. Any value is as good as any other value when you are talking 40-60 years of returns with 40 years of additions.

3. I would expect stock returns to be juiced by low bond returns, as we all choose between these two assets.
1. We actually do need to get into it due to your insistent misunderstanding and refusal to look up the derivation of the measures I'm using. I've given two, Real Expected Return Estimate 1: 1/CAPE. Real Expected Return Estimate 2: div yield+real EPS growth rate. Both are derived from the math of perpetuities, since stocks are perpetuities. Neither has any particular attachment to 10 yrs. You seem to be confusing the first measure with regression analysis of past starting CAPE and subsequent 10 yr return in the US stock market. Different topic, as I've pointed out twice already. Then, you are ignoring the second estimate and its basic congruence now with the first, both put world stock real expected return in the ballpark of 3-4% and no way 8%.

Consider first a perpetual fixed rate bond. Price %=coupon/yield, the most basic bond math there is. Yes if yield halves, price doubles, if yield doubles, price halves. UK consols existed for around 200 yrs and this happened along the way in both directions.

A stock is a perpetual paying Div, but Div is not fixed. Assuming for simplicity Div grows at rate g, the price of a stock P=Div/(r-g) P the stock price, r the return, and g the dividend growth rate. Or IOW r=P/Div+g, the return the perpetual payout of the dividend plus dividend growth. Real Expected Return Estimate 2 is directly adapted from that form of the equation, r=Div yield+EPS growth rate assuming Div and EPS growth rate is equal in the long run, then we further associate long run EPS growth with long run GDP growth trend, but historically EPS growth has notably lagged GDP growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as P= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/P.

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real expected return. This is a standard derivation you can easily look up. I'm just using CAPE as a smoothed version of now's PE, *NOT* doing a regression analysis of past starting CAPE and subsequent return in 10 yr periods, also *NOT* assuming the future expected value of CAPE is lower (or higher) than now's. If you have a future expected value for CAPE in 10 yr's that's lower than now's, state it. In that case the expected return 0-10 would be <1/CAPE (we can calculate when you state your future expected value of CAPE) and thereafter >1/CAPE. But assuming the expected value of CAPE is a constant, 1/CAPE has no particular relation to 10 yrs or any particular maturity.

And note if you halved P today and nothing else changed, Estimates 1 and 2 would not necessarily line up as well as they do now. You act as if I've just given 1/CAPE and other measures give a much different answer. Actually other measures tend to give a similar answer, and it's you who have give no actual basis for 8% (8% real is higher actually than the very long term past average, it isn't even that).

2. OK I'll plan using a real expected return of -2%. No wait I changed my mind, I'll use 18%, since 'any value is as good as any other value'. It seems you really mean 'the very long term return can be expected to converge to the return of US stocks in last X decades' which is a weak hypothesis IMO and wouldn't necessarily be quite as high as 8% real depending how far back, but 'any value is as good as any other' really makes no sense at all for expected return.

3. Wrong way around. Low bond yields have juiced stock *values*, implying lower future returns. This relates to Sanity Check on Expected Return Estimates 1 and 2, which you've also ignored. The two estimates both come out in ballpark of 3-4%. Isn't that very thin compensation for investors to take the risk of stocks? No, because the expected real return of riskless bonds is solidly below 0%. In the past the real return of bonds averaged something like 2.5%, the ex-ante expected values were presumably similar (though we don't have long term TIPS history to see it ex ante) so the market had to price stocks at a much higher expected return than now to give a similar margin of stock over bond expected return. And while there's no reason to think that margin, the expected Equity Risk Premium, is an absolute constant either the risk of stocks is something it's at least plausible to say might be relatively constant. There's no solid reason at least to say the risk of stocks over bonds from now is higher than it's been, making investors demand a higher expected ERP now. For the market to price stocks now at an expected return of 8% real, investors would have to be demanding an extraordinarily wide expected ERP. I don't see why that would be and it surely hasn't been explained by you.
User avatar
JupiterJones
Posts: 3196
Joined: Tue Aug 24, 2010 3:25 pm
Location: Nashville, TN

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JupiterJones »

Okay, we can pick apart the assumptions of the article all day long, just as we can (justifiably) criticize "The Millionaire Next Door" and other things like that.

But I think the basic point is useful and helpful: A level of wealth that the reader might have never considered possible, actually might be possible. Sacrifices and trade-offs now can lead to amazing financial improvement later for anyone... it's not just for "lucky" people.

I personally know plenty of people who just assume they'll always have credit card debt and car payments and bupkis in their retirement accounts. They figure they're work until they die. They see a comfortable retirement as a club they'll never be allowed into--something reserved for people raking in 6-figure+ incomes with big houses and fancy cars.

If they read this article and it inspires them, well I think that's pretty good.
"Stay on target!" -- Davish Krail
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Sat Nov 20, 2021 11:29 am
1. We actually do need to get into it due to your insistent misunderstanding and refusal to look up the derivation of the measures I'm using. I've given two, Real Expected Return Estimate 1: 1/CAPE. Real Expected Return Estimate 2: div yield+real EPS growth rate. Both are derived from the math of perpetuities, since stocks are perpetuities. Neither has any particular attachment to 10 yrs. You seem to be confusing the first measure with regression analysis of past starting CAPE and subsequent 10 yr return in the US stock market. Different topic, as I've pointed out twice already. Then, you are ignoring the second estimate and its basic congruence now with the first, both put world stock real expected return in the ballpark of 3-4% and no way 8%.

Consider first a perpetual fixed rate bond. Price %=coupon/yield, the most basic bond math there is. Yes if yield halves, price doubles, if yield doubles, price halves. UK consols existed for around 200 yrs and this happened along the way in both directions.

A stock is a perpetual paying Div, but Div is not fixed. Assuming for simplicity Div grows at rate g, the price of a stock P=Div/(r-g) P the stock price, r the return, and g the dividend growth rate. Or IOW r=P/Div+g, the return the perpetual payout of the dividend plus dividend growth. Real Expected Return Estimate 2 is directly adapted from that form of the equation, r=Div yield+EPS growth rate assuming Div and EPS growth rate is equal in the long run, then we further associate long run EPS growth with long run GDP growth trend, but historically EPS growth has notably lagged GDP growth.

But dividend growth comes from earnings growth comes from return. Assume in equilibrium the real expected return on the stock is the return on capital of the company.
So restate the first equation as P= p*E/(r-(1-p)*r), where p is the payout ratio, ie Div=p*E, div growth rate is (1-p)*r
Simplify and you get r=E/P.

Thus if you assume the real return is the reinvestment rate, 1/PE ratio is equal to the real expected return. This is a standard derivation you can easily look up. I'm just using CAPE as a smoothed version of now's PE, *NOT* doing a regression analysis of past starting CAPE and subsequent return in 10 yr periods, also *NOT* assuming the future expected value of CAPE is lower (or higher) than now's. If you have a future expected value for CAPE in 10 yr's that's lower than now's, state it. In that case the expected return 0-10 would be <1/CAPE (we can calculate when you state your future expected value of CAPE) and thereafter >1/CAPE. But assuming the expected value of CAPE is a constant, 1/CAPE has no particular relation to 10 yrs or any particular maturity.

And note if you halved P today and nothing else changed, Estimates 1 and 2 would not necessarily line up as well as they do now. You act as if I've just given 1/CAPE and other measures give a much different answer. Actually other measures tend to give a similar answer, and it's you who have give no actual basis for 8% (8% real is higher actually than the very long term past average, it isn't even that).
Plot a regression of how well your assumptions do over a 60 year period with 40 years of additions at 15% income, and retirement expenses equal to 0.85* expenses for 20 years. Post it here for us.

JackoC wrote: Sat Nov 20, 2021 11:29 am
2. OK I'll plan using a real expected return of -2%. No wait I changed my mind, I'll use 18%, since 'any value is as good as any other value'. It seems you really mean 'the very long term return can be expected to converge to the return of US stocks in last X decades' which is a weak hypothesis IMO and wouldn't necessarily be quite as high as 8% real depending how far back, but 'any value is as good as any other' really makes no sense at all for expected return.
Personally, I would look at the range of historical returns as shown in #1, noting that the year 2000 8% returns for a fixed amount of money was actually 12% for someone making additions. I would consider anything within one deviation likely, two standard deviations unlikely but possible and 3 or more standard deviations not worth insuring against. I consider your methods deeply flawed, as you will see if you ever decide to measure how they actually do historically for an accumulator. It doesn't matter that you think now is worse than ever, the same method needs to be highly predictive over past 60 year time period if it is self evident that your choice of prediction method predicts we must get lower than 8% returns. Greater than 95% confidence interval would be appropriate for such a statement.

JackoC wrote: Sat Nov 20, 2021 11:29 am
3. Wrong way around. Low bond yields have juiced stock *values*, implying lower future returns. This relates to Sanity Check on Expected Return Estimates 1 and 2, which you've also ignored. The two estimates both come out in ballpark of 3-4%. Isn't that very thin compensation for investors to take the risk of stocks? No, because the expected real return of riskless bonds is solidly below 0%. In the past the real return of bonds averaged something like 2.5%, the ex-ante expected values were presumably similar (though we don't have long term TIPS history to see it ex ante) so the market had to price stocks at a much higher expected return than now to give a similar margin of stock over bond expected return. And while there's no reason to think that margin, the expected Equity Risk Premium, is an absolute constant either the risk of stocks is something it's at least plausible to say might be relatively constant. There's no solid reason at least to say the risk of stocks over bonds from now is higher than it's been, making investors demand a higher expected ERP now. For the market to price stocks now at an expected return of 8% real, investors would have to be demanding an extraordinarily wide expected ERP. I don't see why that would be and it surely hasn't been explained by you.
They won't get the expected value, they will get an accumulation over time value. This can easily be 4% over the expected value, as my year 2000 example showed. A bad ending sequence for someone no longer accumulating does not necessarily mean a bad starting sequence for someone accumulating, as they have little money in the game. Again, go graph your assumptions historically and see how they do over the timeframe we are discussing. I would just pick a country with the most desirable CAPE if I believed what you were saying was true. It is not, so nobody does this for the obvious reason that it is not predictive enough to ensure better returns. There simply isn't one metric that will tell us whether future returns actual experienced over a 60 year period will be under or over 8% nominal with any reasonable degree of certainty. It seems silly to have to argue against someone doing so.

I'll wish you the best on your predictions, but I will encourage you to at least backtest them. I'm on record that backtesting should be used to eliminate the worst 50% of ideas, and I think you have found one of them as applied to 60 years of accumulation. CAPE has never been constant and should be expected to fluctuate in a way that makes extending the present value into perpetuity ridiculous. An accumulator simply doesn't experience the present CAPE for 60 years, nor will this idea backtest well.




WARNING: MATH
It is not rational for two investors, investor A and investor B to both get your predicted returns if CAPE is not a constant.

Investor A starts investing in the beginning of year 0.
CAPE=25 in year 0 and they expect 1/CAPE = 4% real returns.

Investor B starts investing a year later at the beginning of year 1
By the beginning of year 1, CAPE had dropped to to 20 due to a 20% price drop and 1/CAPE=5% real returns.

Investor A and B then share the same returns for the next 59 years.

It is mathematically improbable for investor A to get 4% real and investor B to get 5% real because 59 of their 60 year returns are shared.



CONCLUSION: Two people that share 59 of 60 years of returns should be expected to get very similar results. It is impossible for 1/CAPE to be a good predictor for two people starting one year apart but with different CAPE values. The reason is that 1/CAPE predicts a much different outcome and their shared timeline and accumulation should give them largely the same outcome.

Note that 1/CAPE predictions can give accurate predictions for both investor A and B over ~10 year time frame. This has nothing to do with curve fitting historical performance. It comes from mathematically solving what time frame investor A can expect 4% returns, experience a 20% drop in the first year, and have investor B start that next year and get 5% returns. Over time periods less than a decade the 1/CAPE predictions can be good for both investors.

1/CAPE is simply not meant for lifetime predictions.
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Tue Nov 23, 2021 8:12 pm
JackoC wrote: Sat Nov 20, 2021 11:29 am
1. We actually do need to get into it due to your insistent misunderstanding and refusal to look up the derivation of the measures I'm using. I've given two, Real Expected Return Estimate 1: 1/CAPE. Real Expected Return Estimate 2: div yield+real EPS growth rate. Both are derived from the math of perpetuities, since stocks are perpetuities. Neither has any particular attachment to 10 yrs. You seem to be confusing the first measure with regression analysis of past starting CAPE and subsequent 10 yr return in the US stock market.

And note if you halved P today and nothing else changed, Estimates 1 and 2 would not necessarily line up as well as they do now. You act as if I've just given 1/CAPE and other measures give a much different answer. Actually other measures tend to give a similar answer, and it's you who have give no actual basis for 8% (8% real is higher actually than the very long term past average, it isn't even that).
Plot a regression of how well your assumptions do over a 60 year period with 40 years of additions at 15% income, and retirement expenses equal to 0.85* expenses for 20 years. Post it here for us.

JackoC wrote: Sat Nov 20, 2021 11:29 am
2. OK I'll plan using a real expected return of -2%. No wait I changed my mind, I'll use 18%, since 'any value is as good as any other value'. It seems you really mean 'the very long term return can be expected to converge to the return of US stocks in last X decades' which is a weak hypothesis IMO and wouldn't necessarily be quite as high as 8% real depending how far back, but 'any value is as good as any other' really makes no sense at all for expected return.
Personally, I would look at the range of historical returns as shown in #1,
JackoC wrote: Sat Nov 20, 2021 11:29 am
3. Wrong way around. Low bond yields have juiced stock *values*, implying lower future returns.
They won't get the expected value, they will get an accumulation over time value. This can easily be 4% over the expected value, as my year 2000 example showed. A bad ending sequence for someone no longer accumulating does not necessarily mean a bad starting sequence for someone accumulating, as they have little money in the game. Again, go graph your assumptions historically and see how they do over the timeframe we are discussing.
a) I would just pick a country with the most desirable CAPE if I believed what you were saying was true.
b) There simply isn't one metric that will tell us whether future returns actual experienced over a 60 year period will be under or over 8% nominal with any reasonable degree of certainty. It seems silly to have to argue against someone doing so.

I'll wish you the best on your predictions, but I will encourage you to at least backtest them. I'm on record that backtesting should be used to eliminate the worst 50% of ideas, and I think you have found one of them as applied to 60 years of accumulation. CAPE has never been constant and should be expected to fluctuate in a way that makes extending the present value into perpetuity ridiculous. An accumulator simply doesn't experience the present CAPE for 60 years, nor will this idea backtest well.
1. You're insistently ignoring the derivation of expected return=1/PE and returning to 'regression analysis' though I've explained three times how the derivation has nothing directly to do with regression of past starting CAPE to subsequent return over 10 yrs (applying past regression constants of CAPE v subsequent return would give an expected return *less than* 1/CAPE since the past pattern was from CAPES this high to be followed by falling valuations, but the theoretical estimate 1/PE doesn't assume that). To answer the point you must IMO
a) show what is wrong with that derivation, based on the math of perpetuities, a stock being a perpetuity; '10 years' does not figure in anywhere
b) why it basically agrees with the other fundamental estimate expected return=div yield+real EPS growth, and how both those answers, world real expected stock return=3-4% imply a healthy Equity Risk Premium over a now solidly negative riskless rate decades out the yield curve. As already stated your argument in bold against reality now by way of 'what if the stock market dropped by 1/2 and nothing else changed' ignores the congruence of those three measures now. If in an imaginary scenario they came to wildly different answers, that imaginary scenario would be different than reality now. :happy

2. Again no basis to assume the expected return is 'the range of past values' of realized return without considering past start/end points of yield (riskless and stock earnings yield) and valuation of the past realized returns. In general yields were higher, valuations lower in past than now. Taking the unconditional (on valuation) realized return of the past as the expected return now is implicitly assuming ever lower yield/higher valuation. You've seemed at times to propose instead a 'bank shot' assumption of falling valuation (and world stock return less/much less than 3-4% real for some years) followed by higher expected return. But the latter seems wishful thinking tailored to the investor's schedule. One person says 'I'll do most of my saving far in the future so I'll assume terrible returns for awhile (falling valuations) to set up better returns for most of my savings'. Another says 'I've finished accumulating so I'll assume higher returns in the near term (continued rise in valuations) and eventually poorer ones when I don't care anymore'. Everyone can have 'their own truth'! What could possibly go wrong with that kind of thinking? :happy

3. As is illustrated clearly in the case of bonds, but again you didn't answer that point. If the 'range of historical values' is the estimate of expected return, why don't you estimate bond expected return now as around 2.5% real pre tax (roughly the past average) when the real yield curve now is negative all the way out? Because obviously that would be wrong. It's also wrong with stocks.

a) I've stated every time: the estimate is for world stock return, the CAPE is world CAPE. This is another idea you are importing from some other debate, that of comparing the attractiveness of different country stock markets by their CAPE.

b) and this seems a basic misunderstanding the meaning of expected return after all this discussion of it. 'What they will get' is the the *realized* return. The realized return cannot be 'predicted' or else there would be no risk (similarly as previously explained: any comparison of ex ante expected return to ex post realized return of stocks can't possibly have a very high correlation coefficient if stocks remain a risk asset). But your formulation 'whether it will be above or below 8%' assumes your conclusion, that 8% is a reasonable estimate for the *expected* return (roughly speaking the centroid of the distribution of possible future, non-predictable, realized returns). It simply isn't by any reasoning except that implicit in all your arguments but which you won't explicitly state (or you don't understand you're implying? I'm not sure): that the expected return equals the past realized return over some arbitrary period in the past, 'eventually', 'after everything washes out' etc. There's no reason to think this.
Grt2bOutdoors
Posts: 24272
Joined: Thu Apr 05, 2007 8:20 pm
Location: New York

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by Grt2bOutdoors »

Brianmcg321 wrote: Sun Nov 14, 2021 5:45 pm
manatee2005 wrote: Sun Nov 14, 2021 2:12 pm "by age 67, you'll have a cool $2.1 million in savings"

How about we teach how to get $2.1 million by age 37 or even 27. Nobody teaches that.

Learn to code.
The suggestion is like saying "learn how to perform surgery". Someone reading your post might want to understand where to begin on learning how to code. Code what? What languages? Point the reader in the right direction much like this forum does in investing.
"One should invest based on their need, ability and willingness to take risk - Larry Swedroe" Asking Portfolio Questions
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Wed Nov 24, 2021 11:14 am
1. You're insistently ignoring the derivation of expected return=1/PE and returning to 'regression analysis' though I've explained three times how the derivation has nothing directly to do with regression of past starting CAPE to subsequent return over 10 yrs (applying past regression constants of CAPE v subsequent return would give an expected return *less than* 1/CAPE since the past pattern was from CAPES this high to be followed by falling valuations, but the theoretical estimate 1/PE doesn't assume that). To answer the point you must IMO
a) show what is wrong with that derivation, based on the math of perpetuities, a stock being a perpetuity; '10 years' does not figure in anywhere
b) why it basically agrees with the other fundamental estimate expected return=div yield+real EPS growth, and how both those answers, world real expected stock return=3-4% imply a healthy Equity Risk Premium over a now solidly negative riskless rate decades out the yield curve. As already stated your argument in bold against reality now by way of 'what if the stock market dropped by 1/2 and nothing else changed' ignores the congruence of those three measures now. If in an imaginary scenario they came to wildly different answers, that imaginary scenario would be different than reality now. :happy
I disputed 1/CAPE as a forward estimate as you were using it- without context for a time period. Saying I am using historical estimates when I am using your own estimate of 1/CAPE doesn't make your statement true. I ignored history completely and just looked at if it is possible for 1/CAPE to be a good predictor for two people with similar time series. It is not possible over long periods of say 60 years of investing.

Person A: CAPE = 25 predicting 4% returns.
Person B, one year later after at 20% price drop CAPE=20, predicting 5% returns

0.8*(1+???)^59 = 1.0*(1.05)^59 (can both people get their expected return when they share the same returns for 59 years?? - NO!)

If person B's estimate is accurate, the ??? has to also be 5%, and person A can not experience 4% returns over a long time period. Person A will get very close to 5% returns if person B's estimate is accurate. Both estimates can not be good.


0.8*(1+0.05)^??? = 1.0*(1.05)^??? (??? is the number of years in which they both could get their expected return)

Now solve for the number of years in which the prediction can be correct.


Repeat with a 50% market drop in one year from CAPE=25 to CAPE=12.5 an person B expecting 8% returns.

0.5*(1+0.08)^??? = 1.0*(1.08)^??? (??? is the number of years in which they both could get their expected return)

Taking the two solutions for a 50% drop and 20% drop, we can determine what type of time periods in which 1/CAPE can be an accurate forward estimate for both investors.
JackoC wrote: Wed Nov 24, 2021 11:14 am 2. Again no basis to assume the expected return is 'the range of past values' of realized return without considering past start/end points of yield (riskless and stock earnings yield) and valuation of the past realized returns. In general yields were higher, valuations lower in past than now. Taking the unconditional (on valuation) realized return of the past as the expected return now is implicitly assuming ever lower yield/higher valuation. You've seemed at times to propose instead a 'bank shot' assumption of falling valuation (and world stock return less/much less than 3-4% real for some years) followed by higher expected return. But the latter seems wishful thinking tailored to the investor's schedule. One person says 'I'll do most of my saving far in the future so I'll assume terrible returns for awhile (falling valuations) to set up better returns for most of my savings'. Another says 'I've finished accumulating so I'll assume higher returns in the near term (continued rise in valuations) and eventually poorer ones when I don't care anymore'. Everyone can have 'their own truth'! What could possibly go wrong with that kind of thinking? :happy
My mathematical argument works whether CAPE goes up or down. CAPE can go up for investor A due to stock price increase, and the two returns can not both be true. You are confusing an example that proves my point with some sort of wishful thinking that this example has to happen or is likely to happen. The math shows that for ANY two CAPE's with meaningfully different predictions your method can not be correct over very long time frames (~60 years). Try as many CAPE values as you want, and feel free to use gains or losses in the first year.

My additional argument here is that the deviation around expected value is big enough that we can't make the kind of conclusion you are making with any confidence interval. The market doesn't just give higher returns with low cape and lower returns with high CAPE. There is large deviation in actual outcomes from any prediction.
JackoC wrote: Wed Nov 24, 2021 11:14 am 3. As is illustrated clearly in the case of bonds, but again you didn't answer that point. If the 'range of historical values' is the estimate of expected return, why don't you estimate bond expected return now as around 2.5% real pre tax (roughly the past average) when the real yield curve now is negative all the way out? Because obviously that would be wrong. It's also wrong with stocks.
I don't add a historical premium to bond returns to predict stock returns. I see there is a historical premium and I would hope to capture this premium. The specific future events of my timeline will determine this outcome. I think you are guilty of extending the present for the next 60 years, which is a form of recency bias.

The range of historic values is not my estimated return. Within one standard deviation of historical returns seems likely, two standard deviation unlikely, and three standard deviations not worth insuring against. Your prediction would fit within my criteria as possible, but so would the author of the article. Unlike you, I do not believe picking one value is meaningful. The reason will be obvious if you backtest anything you have said in this thread - realized value is unlikely to be the predicted value and may not even be close to it.
JackoC wrote: Wed Nov 24, 2021 11:14 am a) I've stated every time: the estimate is for world stock return, the CAPE is world CAPE. This is another idea you are importing from some other debate, that of comparing the attractiveness of different country stock markets by their CAPE.
If you are using the world CAPE, that is meaningless to the 95% of investors that don't have a 100% world stock portfolio. My point was that 1/CAPE can not be accurate for all individual countries, and has to be very bad for some of them. As such it can not be a good solo indicator of the realized stock returns that the investor cares about.
JackoC wrote: Wed Nov 24, 2021 11:14 am b) and this seems a basic misunderstanding the meaning of expected return after all this discussion of it. 'What they will get' is the the *realized* return. The realized return cannot be 'predicted' or else there would be no risk (similarly as previously explained: any comparison of ex ante expected return to ex post realized return of stocks can't possibly have a very high correlation coefficient if stocks remain a risk asset). But your formulation 'whether it will be above or below 8%' assumes your conclusion, that 8% is a reasonable estimate for the *expected* return (roughly speaking the centroid of the distribution of possible future, non-predictable, realized returns). It simply isn't by any reasoning except that implicit in all your arguments but which you won't explicitly state (or you don't understand you're implying? I'm not sure): that the expected return equals the past realized return over some arbitrary period in the past, 'eventually', 'after everything washes out' etc. There's no reason to think this.
If the realized return can't be predicted, 8% nominal is a fine estimate to show compounding.

The person reading the article cares about their realized return when making investing decisions.

The problem still seems to be that you think I expect 8% nominal, but this has never been true, and should be obvious from my statements. Please read this last statement as many times as you need to.
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

AerialWombat wrote: Sun Nov 14, 2021 11:11 pm
sandan wrote: Sun Nov 14, 2021 10:43 pm This motley fool post is untimely and out of touch. Rent is sky high for many people making $50k (urban) in 2021. This might be more appropriate back in 2010. Alternatively, someone today with a financially responsible partner could also pull it off with $100k combined.
I disagree. Seems perfectly reasonable to me. Because of intentional lifestyle creep, it now costs me a whopping $60k a year to live (single, no kids). That’s 2x what it was 3 years ago. If I sold the mini-yacht, started cooking at home, fired the lawn services, etc. then I could live very well on $40k, including the payment on a 15 year mortgage.
Except that the median income in the United States in 2019 was a little over $31,000.

That doesn't change the message, which I believe to be correct. Delayed gratification pays off. Long term investing pays off. Those are the key messages.

I just wanted to point out that saving 15% of the median income is not as easily attainable to everyone as many bogleheads believe it to be.
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

JupiterJones wrote: Tue Nov 23, 2021 12:00 pm Okay, we can pick apart the assumptions of the article all day long, just as we can (justifiably) criticize "The Millionaire Next Door" and other things like that.

But I think the basic point is useful and helpful: A level of wealth that the reader might have never considered possible, actually might be possible. Sacrifices and trade-offs now can lead to amazing financial improvement later for anyone... it's not just for "lucky" people.

I personally know plenty of people who just assume they'll always have credit card debt and car payments and bupkis in their retirement accounts. They figure they're work until they die. They see a comfortable retirement as a club they'll never be allowed into--something reserved for people raking in 6-figure+ incomes with big houses and fancy cars.

If they read this article and it inspires them, well I think that's pretty good.
I agree 100%. The % of Americans who believe such things are unattainable is quite high. Sadly, many Americans would view earning $50,000 a year as just as unlikely an event as amassing a $2 million portfolio.
jharkin
Posts: 3030
Joined: Mon Mar 28, 2016 7:14 am
Location: Boston suburbs

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by jharkin »

I don’t have time to read 4 pages so apologies if this is redundant… but did anyone ask the author why someone who earns 50k even wants 2 million?

Our household earns over 4x that and 2 million is enough even for us to take a slightly early, fairly lavish retirement maintaining a 6 figure lifestyle forever.I suspect somebody making 50k and getting reasonable SS would never even come close to spending down 2MM. It would probably just keep growing into a massive estate. They might get more overall life satisfaction by getting off the hamster wheel early than by working 40years just to reach some artificially chosen dollar target.
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

prioritarian wrote: Wed Nov 17, 2021 1:54 pm
Maverick3320 wrote: Wed Nov 17, 2021 11:30 am
dukeblue219 wrote: Mon Nov 15, 2021 7:48 pm
stoptothink wrote: Mon Nov 15, 2021 3:32 pm My younger sister rents an apartment (with two roommates) literally a block away from my home, in an area that is certainly not San Francisco but significantly above the median for the country in COL. She saves a significant part of her ~$40k/yr income (she's a 2nd year, 2nd grade teacher)
Crazy part of this is that school teachers in some parts of the US are still paid $40k/year, and don't give me the "it's only 10 months of work" thing.
My mom was a teacher. And you're right: it was only 9 months of work a year. The other three months she taught at a local community college for an extra 5-6k or so of income.

I think she topped out at around 60k/year in an extremely LCOL area (houses in town are around 150k). She now has a lake house in the area she retired from and a condo down in Fort Myers that she lives in half the year.
Educators typically have benefits that total a significant fraction of their total wages but this is exceptionally rare for most lower-middle/lower-income people.
These benefits are slowly being eroded nationwide, which is leading to a workforce shortage that could be significant in the coming years.

This does not change the central message of this thread, which is that routine savings over a long period of time can result in larger sums of money than many would believe
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

jharkin wrote: Thu Nov 25, 2021 7:40 am I don’t have time to read 4 pages so apologies if this is redundant… but did anyone ask the author why someone who earns 50k even wants 2 million?

Our household earns over 4x that and 2 million is enough even for us to take a slightly early, fairly lavish retirement maintaining a 6 figure lifestyle forever.I suspect somebody making 50k and getting reasonable SS would never even come close to spending down 2MM. It would probably just keep growing into a massive estate. They might get more overall life satisfaction by getting off the hamster wheel early than by working 40years just to reach some artificially chosen dollar target.
That is an interesting question. I would say uncertainty is definitely one factor. The opportunity to be more extravagant may be another??

Does anyone ever ask why someone wants a new Mercedes, when a Toyota sits in the parking lot just the same? Does anyone ask why someone wants a designer handbag, when a knock off works the same?

Money is a tool. It is used to buy goods and services. Some people value some goods and services more than others. Some on this site seem to have lost that perspective, and believe money is the goal.
jharkin
Posts: 3030
Joined: Mon Mar 28, 2016 7:14 am
Location: Boston suburbs

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by jharkin »

coachd50 wrote: Thu Nov 25, 2021 9:52 am
jharkin wrote: Thu Nov 25, 2021 7:40 am I don’t have time to read 4 pages so apologies if this is redundant… but did anyone ask the author why someone who earns 50k even wants 2 million?

Our household earns over 4x that and 2 million is enough even for us to take a slightly early, fairly lavish retirement maintaining a 6 figure lifestyle forever.I suspect somebody making 50k and getting reasonable SS would never even come close to spending down 2MM. It would probably just keep growing into a massive estate. They might get more overall life satisfaction by getting off the hamster wheel early than by working 40years just to reach some artificially chosen dollar target.
That is an interesting question. I would say uncertainty is definitely one factor. The opportunity to be more extravagant may be another??

Does anyone ever ask why someone wants a new Mercedes, when a Toyota sits in the parking lot just the same? Does anyone ask why someone wants a designer handbag, when a knock off works the same?

Money is a tool. It is used to buy goods and services. Some people value some goods and services more than others. Some on this site seem to have lost that perspective, and believe money is the goal.
But you miss the point that to save 2MM on a 50k salary they have to be ultra frugal super savers for decades. Do you really think such a person is going to spend 40 years driving beat up used Corolla’s and then suddenly wake up one morning wanting to buy a brand new S class at age 67?
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

jharkin wrote: Thu Nov 25, 2021 11:49 am
coachd50 wrote: Thu Nov 25, 2021 9:52 am
jharkin wrote: Thu Nov 25, 2021 7:40 am I don’t have time to read 4 pages so apologies if this is redundant… but did anyone ask the author why someone who earns 50k even wants 2 million?

Our household earns over 4x that and 2 million is enough even for us to take a slightly early, fairly lavish retirement maintaining a 6 figure lifestyle forever.I suspect somebody making 50k and getting reasonable SS would never even come close to spending down 2MM. It would probably just keep growing into a massive estate. They might get more overall life satisfaction by getting off the hamster wheel early than by working 40years just to reach some artificially chosen dollar target.
That is an interesting question. I would say uncertainty is definitely one factor. The opportunity to be more extravagant may be another??

Does anyone ever ask why someone wants a new Mercedes, when a Toyota sits in the parking lot just the same? Does anyone ask why someone wants a designer handbag, when a knock off works the same?

Money is a tool. It is used to buy goods and services. Some people value some goods and services more than others. Some on this site seem to have lost that perspective, and believe money is the goal.
But you miss the point that to save 2MM on a 50k salary they have to be ultra frugal super savers for decades. Do you really think such a person is going to spend 40 years driving beat up used Corolla’s and then suddenly wake up one morning wanting to buy a brand new S class at age 67?
Yes. Some people save because they are worried about the uncertainty of the future, then spend to treat themselves.
It may not be the majority, but there is a segment of the population that would do exactly that.

Your question, which is a valid one, is based on the supposition that people with that level of income decide would not have any interest in expensive purchases once they felt comfortable that they had the means. I don't believe that supposition to be true.
User avatar
CyclingDuo
Posts: 4599
Joined: Fri Jan 06, 2017 9:07 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by CyclingDuo »

coachd50 wrote: Wed Nov 24, 2021 8:57 pm
AerialWombat wrote: Sun Nov 14, 2021 11:11 pm
sandan wrote: Sun Nov 14, 2021 10:43 pm This motley fool post is untimely and out of touch. Rent is sky high for many people making $50k (urban) in 2021. This might be more appropriate back in 2010. Alternatively, someone today with a financially responsible partner could also pull it off with $100k combined.
I disagree. Seems perfectly reasonable to me. Because of intentional lifestyle creep, it now costs me a whopping $60k a year to live (single, no kids). That’s 2x what it was 3 years ago. If I sold the mini-yacht, started cooking at home, fired the lawn services, etc. then I could live very well on $40k, including the payment on a 15 year mortgage.
Except that the median income in the United States in 2019 was a little over $31,000.

That doesn't change the message, which I believe to be correct. Delayed gratification pays off. Long term investing pays off. Those are the key messages.

I just wanted to point out that saving 15% of the median income is not as easily attainable to everyone as many bogleheads believe it to be.
We would agree. The following data shows how any particular Boglehead household might compare their income to that of a household where saving 15% is going to be a stretch for a certain percentage of the US population. A $50K household income in 2021 puts the household in the percentile of 38%. The following breaks down the percentage of the US population that would be hard pressed to save much - if anything - all the way up the percentile scale to where saving 15% or more would not be a stretch...

Latest look at household income for 2020 and 2021...

Image
https://dqydj.com/average-median-top-ho ... rcentiles/

Image
Image
Image
Image
Image
https://dqydj.com/average-median-top-ho ... rcentiles/

Distribution of household income in 2019...

Image
https://www.statista.com/statistics/203 ... in-the-us/
"Save like a pessimist, invest like an optimist." - Morgan Housel
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

jharkin wrote: Thu Nov 25, 2021 11:49 am But you miss the point that to save 2MM on a 50k salary they have to be ultra frugal super savers for decades. Do you really think such a person is going to spend 40 years driving beat up used Corolla’s and then suddenly wake up one morning wanting to buy a brand new S class at age 67?
This is similar to what I am doing, other than making more than 50k while living off less than 50k and accomplishing this in 20 years rather than 40 years. The benefit of being so frugal for 20 years is the option to retire and/or upscale lifestyle in the 40's and 50's. Having the choice is very powerful in life and employment decisions. Buying that S class before reaching financial independence seems like the crazy choice to me. The reality is that you just get used to living below your means and are protected from all the typical Boglehead portfolio fears because your retirement wants are more than 2x your retirements needs.

Inflation worries? LBYM and invest early
Bad boss? LBYM and invest early
SCV/International vs US? LBYM and invest early
Bad sequence worries? LBYM and invest early
Divorce? LBYM, invest early, and work into your 50/60's.

Investing early is so powerful that the objections to this article really do a disservice to those that might otherwise benefit from it.
coachd50
Posts: 733
Joined: Sun Oct 22, 2017 10:12 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by coachd50 »

abc132 wrote: Fri Nov 26, 2021 12:17 am

Inflation worries? LBYM and invest early
Bad boss? LBYM and invest early
SCV/International vs US? LBYM and invest early
Bad sequence worries? LBYM and invest early
Divorce? LBYM, invest early, and work into your 50/60's.

Investing early is so powerful that the objections to this article really do a disservice to those that might otherwise benefit from it.
This really sums it up nicely. The plan really is that simple. I would add that part of the success might come from being lucky enough to avoid any issues during that critical time of trying to get to solid financial footing, as for some the margin between living at your means and living below your means is fairly thin.

Not much one can do about that though
JackoC
Posts: 2840
Joined: Sun Aug 12, 2018 11:14 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by JackoC »

abc132 wrote: Wed Nov 24, 2021 7:30 pm
JackoC wrote: Wed Nov 24, 2021 11:14 am
1. You're insistently ignoring the derivation of expected return=1/PE and returning to 'regression analysis' though I've explained three times how the derivation has nothing directly to do with regression of past starting CAPE to subsequent return over 10 yrs (applying past regression constants of CAPE v subsequent return would give an expected return *less than* 1/CAPE since the past pattern was from CAPES this high to be followed by falling valuations, but the theoretical estimate 1/PE doesn't assume that). To answer the point you must IMO
a) show what is wrong with that derivation, based on the math of perpetuities, a stock being a perpetuity; '10 years' does not figure in anywhere
b) why it basically agrees with the other fundamental estimate expected return=div yield+real EPS growth, and how both those answers, world real expected stock return=3-4% imply a healthy Equity Risk Premium over a now solidly negative riskless rate decades out the yield curve. As already stated your argument in bold against reality now by way of 'what if the stock market dropped by 1/2 and nothing else changed' ignores the congruence of those three measures now. If in an imaginary scenario they came to wildly different answers, that imaginary scenario would be different than reality now. :happy
I disputed 1/CAPE as a forward estimate as you were using it- without context for a time period.
JackoC wrote: Wed Nov 24, 2021 11:14 am 2. Again no basis to assume the expected return is 'the range of past values' of realized return without considering past start/end points of yield (riskless and stock earnings yield) and valuation of the past realized returns. In general yields were higher, valuations lower in past than now.
My mathematical argument works whether CAPE goes up or down. CAPE can go up for investor A due to stock price increase, and the two returns can not both be true.
JackoC wrote: Wed Nov 24, 2021 11:14 am 3. As is illustrated clearly in the case of bonds, but again you didn't answer that point. If the 'range of historical values' is the estimate of expected return, why don't you estimate bond expected return now as around 2.5% real pre tax (roughly the past average) when the real yield curve now is negative all the way out? Because obviously that would be wrong. It's also wrong with stocks.
The range of historic values is not my estimated return. Within one standard deviation of historical returns seems likely, two standard deviation unlikely, and three standard deviations not worth insuring against. Your prediction would fit within my criteria as possible, but so would the author of the article. Unlike you, I do not believe picking one value is meaningful. The reason will be obvious if you backtest anything you have said in this thread - realized value is unlikely to be the predicted value and may not even be close to it.
JackoC wrote: Wed Nov 24, 2021 11:14 am a) I've stated every time: the estimate is for world stock return, the CAPE is world CAPE. This is another idea you are importing from some other debate, that of comparing the attractiveness of different country stock markets by their CAPE.
If you are using the world CAPE, that is meaningless to the 95% of investors that don't have a 100% world stock portfolio.
JackoC wrote: Wed Nov 24, 2021 11:14 am b) and this seems a basic misunderstanding the meaning of expected return after all this discussion of it. There's no reason to think this.
The person reading the article cares about their realized return when making investing decisions.
1. You're still ignoring the basic derivation of the two estimates Expected real return=earnings yield (1/PE, CAPE used for 'PE' to smooth out noise in spot PE) and Expected real return=div yield+real EPS growth (with latter taken as the historical discount to GDP growth). Both are based on valuation of a *perpetual* which a stock is. Neither estimate therefore has a time dimension. For some reason you also consistently ignore the second estimate and how it basically agrees with the first now.

2. This pseudo-paradox is based on your misunderstanding and mixing/matching of two different things. The expected return is basically the centroid of the future distribution of realized returns. It's not meaningful to mix the two over time as in 'what if the (realized) return is a giant negative/positive the next X years...' That whole line of 'mathematical' argument of yours is meaningless

3. This illustrates again your basic misunderstanding of the concept we've (supposedly) been 'debating'. The number from which the realized return will differ by X standard deviations, fewer std devs being more likely, loads and loads of std devs being really unlikely, is a reasonable description of the *expected return*. You are IOW just assuming the expected return is 8% and all your actual arguments are about future realized return...which isn't the topic. :happy But 8% is not a realistic estimate of the expected real return of stocks now. Again this should be 100% obvious in case of bonds, the nominal subject of point 3. It would be ridiculous to set estimate the expected return of long term bonds as the past realized return (~2.5%) when the current 30yr yield is <0% real. And this should be a clear clue that it would also be bogus to assume the expected return of stocks is just the past realized return.

3. a) using world stock expected return just avoids side debates about whether eg. the estimate 'real expected return=div yield+real EPS growth' should consider sources of EPS growth outside country X for companies included in country X's stock index. That washes out for the whole world. But there is no more plausible argument that the expected return of just US stocks (which are more than half the world index anyway) is 8% than for world stocks. As another poster simply noted long ago in the thread, assuming 8% real expected return, for world stocks real EPS growth would be ~6.4% and corporate profit as % of world GDP (growing at 3% real) would go from around 10% of world GDP now (an all time high) to 70% in 60 yrs. So looking at whole world also helps us realize it's an ultimately closed system where very high estimates of real expected return like 8% generate ridiculous future results in macroeconomic terms. If we looked at just the US the projection for 8% real stock return would be for corporate profits to be bigger than the whole US GDP in 60 yrs.

b) crystalizes your basic misunderstanding. Everybody wants to know the future realized return, there is no forward looking context where we wouldn't want to know it. But we can't know it for a risk asset, by definition. We can only estimate the centroid the distribution of future realized returns. That's called the expected return. It can be estimated by various measures of current yield and valuation. Now both an earnings yield (1/CAPE), div yield+expected EPS growth (at a historical discount to GDP growth) and consideration of bond return to give the expected Equity Risk Premium all give estimates in the range of 3-4% real for stocks. 8% real is out to lunch as an estimate of stock expected return now.
abc132
Posts: 704
Joined: Thu Oct 18, 2018 1:11 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by abc132 »

JackoC wrote: Fri Nov 26, 2021 11:37 am You are IOW just assuming the expected return is 8% and all your actual arguments are about future realized return...which isn't the topic. :happy But 8% is not a realistic estimate of the expected real return of stocks now.
abc132 wrote: Wed Nov 24, 2021 7:30 pm The problem still seems to be that you think I expect 8% nominal, but this has never been true, and should be obvious from my statements. Please read this last statement as many times as you need to.
I am arguing that any value that is a likely future 60 year return is a reasonable value the reader should care about, and you are arguing that one must use expected returns. I am not making the argument you have fabricated. I did respond to your expected returns methodology, but that was really to inform any readers of why I believe extending 1/CAPE predictions over 60 year timeframes was a poor idea. It is fine that you disagree. I look at the range of expected values from experts and conclude there is no one value that must be used for expected stock returns. It is okay that you disagree.

The way one figures out if advice is actionable is if changing the value meaningfully changes the conclusion of the reader.

It doesn't in this case.
User avatar
snackdog
Posts: 1255
Joined: Wed Nov 12, 2014 4:57 am

Re: "How to Retire With $2 Million On a $50,000 Salary"

Post by snackdog »

The article appears to ignore inflation, which is fine if you set it to 0% for everything: salary, savings, returns and withdrawals.

But why work until 67 to get $2 million?

If you are earning $50K, paying taxes, mortage and saving, you are living on max $35,000. At age 67, Bogleheads' VPW recommends 5.3% SWR which would require about $660,000 in savings. With $2 million you are suddenly spending $106,000 per year. You worked a couple decades too long or lived too frugally!
Post Reply