(The Absence of) ERE in Academic Literature

Ask your investment, budget, and other money related questions here
7Wannabe5
Posts: 4856
Joined: Fri Oct 18, 2013 9:03 am

Re: (The Absence of) ERE in Academic Literature

Post by 7Wannabe5 » Mon Jan 09, 2017 8:52 am

Interesting.
Fish said: This lacks the investment part of modern FIRE in assuming a real return of 0%, meaning FI is achieved through one's own eventual death.
Depending on your age, this might or might not be more reflective of rational utility. Obviously, if your rational expectation of remaining life expectancy is less than 33 years, then unless you are desiring to leave some part of capital as estate after your death, applying the 3% rule rather than D-R/R will tend towards over-valuation of present time devoted to work towards savings. So, for example, since I am 51 and my optimistic expectation of life expectancy might be another 40 years:

$250,000 in savings/40 years= $6250/year
$250,000 X (.03) =$7500/year
$250,000 X (.04) =$10,000/year
$250,000 X (.025) =$6250/year

So, not too much difference between calculating one way or the other. However, if my current health and/or family history led me to believe that 20 years was a better estimate for life expectancy then:

$250,000 in savings/20 years=$12,500/year
$250,000 X (.03) = $7500/year
$416,667 X (.03) =$12,500/year

So, if my yearly expenses were $12,500 and my current income was $25,000, I might needlessly work an extra (416,667-250,000)/12,500= 13 years out of the slim remnant of years left to me, before choosing to retire if I was applying the 3% rule rather than death clock calculation with assumption of overall 0% return. Since my expectation is that I will likely be able to collect at least $500/month in early withdrawal SS benefits starting at age 62, and my current expenses are not much above that level, my clock is set at only 11 years, so it makes very little rational sense for current me to favor future me by saving money towards time, except to the extent that I want to set a chunk aside for a specific project that requires it. IOW, my current rate of semi-retirement leisure is directly determined by my rate of semi-employment pay with little benefit to be derived from compound interest. Given a 40/hour work week , 50 week/work year, and $8000 yearly expenditure, and $20/hr pay rate, each hour of work will purchase me 4 hours of leisure, but nary a penny more (sigh.) In fact, it is likely that my most rational behavior would be to greatly favor current time spent towards maintenance/improvement of health/fitness over current time spent making/saving money, since it is unlikely that 36 hours/week would ever be optimal and more than 4 hours/week would likely be optimal.

Thanks for posting that equation!

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Mon Jan 09, 2017 12:03 pm

@7w5: Even though the "divide by 40 years" and 3% SWR results are similar, the 3% assumption assumes preservation of capital while dividing by 40 depletes it. Mortality really gives you the option to either invest more conservatively at 0% real, OR retire earlier at a higher SWR which is not sustainable over longer durations.
Fish wrote:(Proportion of life spent working) = (1 - savings rate)
Actually, it turns out this relationship can be recovered more readily from equation 1 of the paper:

(Savings rate)*(working years) = (Post-retirement spending)*(Life expectancy - working years)

by setting post-retirement spending equal to pre-retirement spending which is (1 - savings rate). As Jacob noted above, it's apparently so obvious to the academics that they don't bother to discuss the implications. Meanwhile this was still news to me a few years ago when I learned about FIRE and ERE... :? :oops:

7Wannabe5
Posts: 4856
Joined: Fri Oct 18, 2013 9:03 am

Re: (The Absence of) ERE in Academic Literature

Post by 7Wannabe5 » Mon Jan 09, 2017 1:04 pm

the 3% assumption assumes preservation of capital while dividing by 40 depletes it.
Right, and the reason why you want to preserve capital is to provide income until you are dead in some distant future. If the future in which you will be dead looms nearer, and you do not care about leaving an estate, then it seems to me that the assumption of 0% return is enough of a security blanket. No?

Anyways, it doesn't really matter because nothing is likely to change my high level of optimism and low level of conservatism relative to the population of this forum. I think it might be kind of fun to work 10 hours/week as a greeter at Wal-Mart when I am 80, but I doubt that will happen. I wonder what will happen to me in the world of the future?

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Mon Jan 09, 2017 4:28 pm

Another reason that I do not consider the Simonovits paper to have priority on the concept of extreme early retirement (not that my opinion matters...) is that a 0% real return, while mathematically convenient, is very limiting in practice.

Consider a standard case where a person works for X years starting at age 20, then retires until death at age 90 (all at constant income and expenses). With a 0% return, this means that the work from X years must cover spending for the entire 70-year period. For example, if X=10 years, a savings rate of 1-(R/D) = 1-(10/70) = 86% is needed. On a median income of 50k this works out to a spending level of 7k/year.

While many people on this forum and in the FIRE community have demonstrated the feasibility of living well on a "one Jacob" budget, consider going down to an accumulation period of X=5 years with the same median income. This would mean a 3.5k/year budget ($10/day) which might be still possible but for most lifestyles would require some form of individual advantage in addition to an expert "skill of living."

What I'm getting at is if 5 years is the accumulation target, extreme early retirement is no longer robust (works for everyone) starting at an early age with 0% real returns. I realize the 5 years was arbitrary based on Jacob's personal experience, but this clearly demonstrates there is a limit to how much can be accomplished with frugality alone. The investment piece is a very important, perhaps essential part of the FIRE canon. Jacob still seems to have been the first person to tie it all together with one string.

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Tue Jan 10, 2017 4:27 pm

@7w5: If you found the Simonovits equation insightful, it could be interesting to reread the "Financial independence and investing" section of the ERE book with the new context on savings rate/duration of working career. Jacob covers it quite well and his equations do account for investment returns.

You raise a good point that a large fund may be overly conservative, especially for someone who is already close to traditional retirement age. Accounting for dying broke or fully depleting assets by the time one can receive a pension (assuming pension income alone is sufficient) may significantly reduce the required fund size depending on age and assumed investment returns. However, this is very much in "retire at your own risk" territory because removing the conservatism also means removing the safety nets.

I think the finite-horizon equations tend to get overlooked and retiring on a smaller fund is viable, especially if there is another proven source of income heading into (semi-)retirement so that one is not fully living off of capital. (Related: Deliberately coasting to FI?)

7Wannabe5
Posts: 4856
Joined: Fri Oct 18, 2013 9:03 am

Re: (The Absence of) ERE in Academic Literature

Post by 7Wannabe5 » Wed Jan 11, 2017 11:33 am

@Fish:

I did do a quick skim of that section of "ERE" after reading your post. I'm likely just being a little bit silly in my attempt to mathematically rationalize my unique combination of frugality, sloth and lack of risk-aversion (AKA anti-anxiety disorder.) My problem is that my variety of frugality is such that I think starting from scratch under adverse conditions might be fun, whereas if I had a million dollars in the bank it might be worrisome in the same way that carrying around the designer handbag my BF gave me for Xmas is rather worrisome.

What if we lived under some sort of meritocratic dictatorship which required that all citizens hand over all assets every 7 years and then resume play from new deal after being rewarded with trophies or badges? Like when you slide all the hotels off of the monopoly board and re-sort all the colorful paper slips back into the bank. Would this offer greater or lesser likelihood of development of internal capital than continuing play? Maybe all other contracts such as marriages and professional certificates etc. could also be automatically annulled at this 7-year-skin exfoliation festival?

ducknalddon
Posts: 249
Joined: Fri May 20, 2016 5:55 am

Re: (The Absence of) ERE in Academic Literature

Post by ducknalddon » Wed Jan 11, 2017 2:03 pm

7Wannabe5 wrote: What if we lived under some sort of meritocratic dictatorship which required that all citizens hand over all assets every 7 years and then resume play from new deal after being rewarded with trophies or badges? Like when you slide all the hotels off of the monopoly board and re-sort all the colorful paper slips back into the bank. Would this offer greater or lesser likelihood of development of internal capital than continuing play? Maybe all other contracts such as marriages and professional certificates etc. could also be automatically annulled at this 7-year-skin exfoliation festival?
Wars are a good way of doing that, particularly for the loosing side.

7Wannabe5
Posts: 4856
Joined: Fri Oct 18, 2013 9:03 am

Re: (The Absence of) ERE in Academic Literature

Post by 7Wannabe5 » Wed Jan 11, 2017 2:42 pm

ducknalddon said: Wars are a good way of doing that, particularly for the loosing side.
Good point. I guess I was thinking about it more in terms of behavioral psychology. As in, how would you change your behavior if you only had 7 years to live/work/save/spend/invest/profit/etc. vs. 70, but instead of dying, you just commit to going blank slate?

I wonder why there aren't more 80 year old people taking out student loans to study art history? I wonder how loans/debts were handled differently in warrior cultures with short life expectancy for young men? I wonder how much faster Jacob could get to $100,000 or $1,000,000 net worth again if he gave all his assets to charity?

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Thu Jan 31, 2019 7:50 am

Even earlier than Simonovits (2001), a surprisingly modern take on FIRE is contained in this 30-page pamphlet that was written in 2000. (Paywall, costs $5)

“Drive Your Financial Advisor’s Porsche and Retire Before 40” by John P. Greaney, PE
http://retireearlyhomepage.com/reports1.htm
I retired in 1994 at age 38. I didn’t win the lottery, receive a big inheritance, or cash in on an IPO -- heck, when I was working I didn’t even get any stock options.
[...]
There are three keys to retiring before age 40: 1) managing your living expenses, 2) maximizing your disposable income available for investment, and, 3) minimizing investment fees and commissions.
[...]
It’s actually quite easy to estimate when you can retire early. The four variables that have the most effect on the calculation are 1) percent of gross salary saved, 2) annual rate of salary increases, 3) annual inflation rate, and 4) annual investment return. A fifth variable, your life expectancy has a somewhat lesser effect.
A 2-D table of “years to retirement” (i.e. time to a “100% safe” withdrawal rate) is presented with savings rate (10-40%) and investment return (5-20%) as the independent variables. An assumption of 10% CAGR for wage increases is made. It shows that lifestyle inflation (expenses increasing proportional to income) results in a lifetime of wageslavery UNLESS savings rate AND investment returns are both high (but note, SR tops out at 40%). By limiting spending increases to the rate of inflation, early retirement becomes feasible with high savings rate OR investment returns.

Then the calculation is repeated with a different set of income assumptions, with identical results, to illustrate the point that the savings rate is what matters.

It looks like most elements of the FIRE standard formula (as collected by Jacob on this page: https://earlyretirementextreme.com/about-the-blog ) are discussed here. Savings rate, lowering expenses more effective than increasing income, and early retirement being possible at age <55 with less than $1M. There is a tiny bit of “Millionaire Next Door” frugality advice about inexpensive housing and used cars, but it’s not systems-theoretic like ERE is.

This publication anticipated the main ideas behind FIRE but appears to have had little impact on today’s movement. Back then FIRE was discussed on the Motley Fool “Retire Early” board, followed by early-retirement.org starting in 2002. There were no blogs or social media to disseminate the ideas, so it stayed within the community. I found this while researching the etymology of the term FIRE for this other pet project: viewtopic.php?t=10210

jacob
Site Admin
Posts: 11062
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: (The Absence of) ERE in Academic Literature

Post by jacob » Thu Jan 31, 2019 8:46 am

Great find!

I wonder whether the paywall was also in effect back then. It makes a big difference in dissemination to the (ultra)leanFIRE crowd whether something is free or not---charge a dollar and most will never see it---whereas the money is quite loose with the FatFIRE crowd.

Absolutely love that he kept the 1995-style design until today (still updating). If wordpress ever crashes on ERE and I redo the website, I'm going to revive it in hand coded [html] format.

Also http://www.retireearlyhomepage.com/extreme.html

jennypenny
Posts: 6191
Joined: Sun Jul 03, 2011 2:20 pm
Location: Stepford USA

Re: (The Absence of) ERE in Academic Literature

Post by jennypenny » Thu Jan 31, 2019 10:13 am


7Wannabe5
Posts: 4856
Joined: Fri Oct 18, 2013 9:03 am

Re: (The Absence of) ERE in Academic Literature

Post by 7Wannabe5 » Thu Jan 31, 2019 11:00 am

His top book shelf recommendation is "Cashing in on the American Dream: How to Retire at 35" by Paul Terhorst 1988, which I clicked along to see had already been given a 5 star recommendation on Goodreads by Jacob :lol: Short rabbit hole plunged.

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Thu Jan 31, 2019 4:29 pm

@jacob - The paywall dates to the original release. Thinking about it some more, the downside of Greaney’s model was its complexity (4 independent parameters). Scientists and engineers could appreciate it, but I don’t see it catching on with the general public if a spreadsheet has to be opened to understand the effect of a parameter variation. The intuition gets lost in the mechanics of performing the calculation.

The ERE model has 1 parameter (savings rate) and its simplicity allows a person to quickly estimate time to retirement without getting bogged down in the details of salary growth, inflation and investment returns. No spreadsheet is required and it can be read off of a table or graph. True, it is not accurate for projecting traditional retirement, but it doesn’t need to be. Loss of accuracy wrt ER is not a serious concern because the other parameters cannot be known in advance with much precision anyway. So even if you may have been “scooped” in identifying the importance of savings rate, the decision to singularly focus on SR was a contribution that facilitated mass adoption of these ideas.

jacob
Site Admin
Posts: 11062
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: (The Absence of) ERE in Academic Literature

Post by jacob » Thu Jan 31, 2019 4:40 pm

Fish wrote:
Thu Jan 31, 2019 4:29 pm
... the decision to singularly focus on SR was a contribution that facilitated mass adoption of these ideas.
I think the bigger innovation [with ERE] was in considering rates way above 50%. After all, the equations were well known to academia and the CFP crowd. It just never occurred to "people" to investigate all the parameter space. Thus that 80% is not merely twice as fast/good as 40% but six (SIX!) times faster/better was not really appreciated and thus not really tested systematically/as part of a plan.

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Mon Feb 18, 2019 10:26 pm

jacob wrote:
Thu Jan 31, 2019 4:40 pm
I think the bigger innovation [with ERE] was in considering rates way above 50%.
I think you’re right. There is a book called “Engineering Your Retirement” by Mike Golio published in December 2006. In it, he presents an equation which can be used to calculate time to retirement (similar to ERE equation 7.15), but only considers savings rates up to about 43%. See page 26 of this PDF: https://web.archive.org/web/20150701111 ... deShow.pdf

On that page, there is a plot of years to retirement vs savings ratio (denoted as “f” in Golio’s formula). The savings rate is SR=f/(1+f)... but the way it’s presented by Golio suggests a limit of f=1 when in fact it can be infinite. His graph only goes up to f=0.75 (43% SR).

Golio retired at 49, from what appears to be a distinguished career as a PhD electrical engineer. He was a fellow of the IEEE and edited some of its journals. (The IEEE was even involved in publishing his retirement book.)

Just another example of how extreme early retirement was hiding in plain sight, but remained unnoticed by the academics. However, this book demonstrates that the FIRE math in ERE chapter 7 was known before ERE. (Golio’s book came out a full year before Jacob started blogging.) I had been previously crediting Jacob with the first derivation of eq. 7.15. Now I stand corrected.

jacob
Site Admin
Posts: 11062
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: (The Absence of) ERE in Academic Literature

Post by jacob » Tue Feb 19, 2019 9:02 am

I'm too lazy to convert the notation (it's beyond what I can juggle in my head) but does Golio have (7.10), (7.15), or/and (7.16)? It's (7.16) that's the "hard" one as it can only be solved numerically. If it's (7.15) only, one way would be to assume that 25x (SWR=4%) is enough (in the book, I use 30x) to last "forever". What (7.16) does is to connect (7.15) to (7.10) to spend the money back down.

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Tue Feb 19, 2019 11:04 am

Golio’s equation is an improved version of (7.15) that allows for having nonzero savings as an initial condition. Setting the initial savings (“p” in his formula) to zero recovers (7.15). I haven’t read Golio’s book aside from the free preview (it is really fortunate that equation was included, otherwise I never would have looked further), which promises a few more equations in the non-free part of the book. I’ll report back once I get it from the library.

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Wed Feb 20, 2019 4:43 am

Read Golio’s EYR book. Sadly, the only part that is pertinent to FIRE math was included in the free book preview. To answer Jacob’s question: there is no equivalent to (7.10) or (7.16). He does suggest the 4% rule as a guideline for accumulation.

As mentioned earlier, he did have an improved version of (7.15), which is eq. (1.3) on page 15: https://books.google.com/books?id=9mrdw ... &q&f=false

The references to (B3) and (B4) are standard finance equations for compound interest and FV of an annuity. These are also part of the free preview (page 200). Last year, I derived a formula similar in intent to Golio’s (1.3) and used the same method, so maybe it can be considered somewhat obvious? Though if I didn’t have previous exposure to ERE/FIRE, it never would have occurred to me to combine the formulas in this way, and to make savings rate an independent variable and solve for years to retirement, instead of the other way around.

It should be noted that the EYR book does not include the plot of years to retirement vs SR in the earlier PDF. In the book Golio never suggests an upper limit to savings, nor does he attempt to explore what would happen. Overall, the EYR book is an excellent example of the state of the art for FIRE pre-ERE. Which is impressively modern; the book doesn’t feel dated with one exception: the lack of emphasis on savings rate (and the possibility of retirement before 30). It’s kind of sad that the book didn’t achieve more success because it’s well-written and presented (from my STEM perspective).

I liked this quote from section 3.1 on spending (p.46):
The control you exercise on spending is the most powerful weapon in your retirement planning arsenal. [...] Being frugal is not about sacrifice.
Golio suggests using equivalence scales to budget for expected changes in household size. :idea: The scale he uses is developed by the US Bureau of Labor Statistics. We might be able to use it here, I like this one better than modified OECD (sqrt) but maybe it’s because it makes me look better. :P This table comes from Golio’s blog.
Family type--BLS Normalized Family Budget
Single adult--0.360
Two adults--0.600
Two adults, One child--0.820
Two adults, Two children--1.000
Two adults, Three children--1.116
One adult, One child--0.570

For cases not shown in the table, the equivalence family budget can be roughly approximated using the expression:

E=[(A+pK)^F]/2.751

where A is the number of adults in the family, K is the number of children, p=0.92 and F=0.75.
Chapter 4 (investment instruments) was a concise introduction to various investment types. Vicki Robin attempted something similar in the 2018 YMOYL but I like Golio’s version better.

------

This post is becoming a wall of text, but I wanted to include one last quote from this pre-ERE era paper: “Saving and investing over the life cycle and the role of collective pension funds” (2007) by Bovenberg, Koijen, Nijman and Teulings (link). It expresses in words the same conclusion I reached after reading Simonovits (2001) (discussed here).
The savings rate is thus equal to the share of retired life in the overall adult life.
How can one possess such powerful insight, yet lack the imagination to carry it to its logical extreme?! :?

jacob
Site Admin
Posts: 11062
Joined: Fri Jun 28, 2013 8:38 pm
Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 73
Contact:

Re: (The Absence of) ERE in Academic Literature

Post by jacob » Wed Feb 20, 2019 1:06 pm

The savings rate is thus equal to the share of retired life in the overall adult life.
How can one possess such powerful insight, yet lack the imagination to carry it to its logical extreme?! :?
Generally speaking, I think a lot of insights are "vapor" in the sense that many can see them, yet can not quite grasp them. The first quote succinctly states what is usually expressed in table form but the quote takes a step back/up and looks at it from an "adult life" perspective. "If you save 50% of your income, you can talk half your adult life off", ... "If you save 10% of your income, you can take 10% of your adult life off". The clever part here (as I see it) is the "adult life" because it frames it differently than the Ferris-like mini-retirements that the table does. Of course the table does not take compound interest into account.

The hard part wrt vapor is that it depends on framing, because the framework is what allows us to grasp and fix it in ways we can understand it. Framing is often part of the zeitgeist and the idiosyncracies ("genius"?) of the person "inventing" it. (Idiosyncracy can somewhat make up for zeitgeist and produce something completely out of the left field. Conversely, if the zeitgeist is already strong, you'll often have a situation of multiple discovery) Once the new framework is solidly entrenched, whatever breakthroughs often seem trivial in retrospect.

Guess who wrote this?
The price of anything is the amount of life you exchange for it.
Vicki Robin in YMOYL? No, Henry David Thoreau in Walden. What YMOYL did was to put a number on it, but it required an accountant to do it. A naturalist (like Thoreau) grasped the idea qualitatively. He would certain have understood the idea of quantitative life-energy, but he did not invent it because he didn't inherently posses the [accountant's] framework to do so.

People have often complained that the ERE book lacks references. My defense is that in writing the book, it quickly became clear to me that none of these ideas are really new. The only thing that's new is how the ideas are put together and what framework they're presented in. If I may be so prejudiced, the difference might come down to the difference in how an engineer (build solution with existing components) solves a problem compared to a scientist (explore the full component parameter space).

Fish
Posts: 488
Joined: Sun Jun 12, 2016 9:09 am

Re: (The Absence of) ERE in Academic Literature

Post by Fish » Thu Mar 21, 2019 9:34 am

In this post from Dec 2016, an investment advisor appears to independently (re)discover the shockingly simple math:

https://www.quora.com/How-can-I-make-a- ... Steve-Dwek

He has a remarkably similar thought process as Jacob in ERE chapter 7: First, determine how long one’s savings will last in retirement, given spending and investment returns (7.10). Second, determine how long it will take to accumulate that sum of money (7.15). Last, use longevity to connect the two equations (7.16).

Now, there is no mention of savings rate in the link, but I thought the similarity was impressive. The math is obvious if one knows what question to ask. Framing it in terms of SR (and considering the entire domain) is certainly an innovation, but the basic concept doesn’t require much creativity. It supports Jacob’s multiple discovery explanation above.

Post Reply