I saw a couple of mentions about higher SWR being achievable when the withdrawal rate was adjusted during bad years (10% less). I haven't been able to find any good posts or articles that specifically explain this strategy, and it doesn't seem firecalc supports this option. Can someone link me to a good article?
I was thinking that during the early years of being ERE, I would be willing to tolerate up to 15% reduction in my withdrawal rate. For example, if I was living on just $6,000/year, a 15% reduction would be only $900, which I could cover with a couple weeks of work at just about anything.
Seems like this strategy would allow for much earlier, partial (based on the vagaries of the economy & the performance of your portfolio) retirement extreme. (EPRE anyone?)
What I want to know is what the multiplier will be, based on a high tolerance for withdrawal reductions in the first 15 years. I imagine it would be somewhere between 5-7% SWR, or 14-20X..
Flex-SWR for earlier ERE?
I think I found a good article here. Basically it's just how much risk you want to take on. 7% in this example had a 57% chance of depletion. The lower your SWR, the lower the chance of failure. I'm interested in the same thing myself but it's pretty risky. Semi-ERE seems like a more viable option (i.e. work for 3 months a year)
http://www.fa-mag.com/news/new-research ... -8674.html
http://www.fa-mag.com/news/new-research ... -8674.html
The closest thing I've seen so far is: http://www.fpanet.org/journal/HowtoAchi ... rawalRate/
It only goes up to 40 years, and most of the SWR variations it mentions seem to be downhill, with gradually decreasing incomes.
Example variation:
4.8% SWR is good for 40 years if you follow these rules:
First year's income = 4% withdrawal
Subsequent years = (1 - 0.04) * Previous year's income + CPI
Basically, decrease the income by 4% every year but keep up with inflation otherwise.
In my opinion, I think that ERE folks would prefer more of a curve on their income. Something that starts off low between 30-40 while they are young and tolerant, rises somewhat for their prime years of 40-69, and then decreases again after 70+.
It only goes up to 40 years, and most of the SWR variations it mentions seem to be downhill, with gradually decreasing incomes.
Example variation:
4.8% SWR is good for 40 years if you follow these rules:
First year's income = 4% withdrawal
Subsequent years = (1 - 0.04) * Previous year's income + CPI
Basically, decrease the income by 4% every year but keep up with inflation otherwise.
In my opinion, I think that ERE folks would prefer more of a curve on their income. Something that starts off low between 30-40 while they are young and tolerant, rises somewhat for their prime years of 40-69, and then decreases again after 70+.
I would think you could use the export option (on the investigate tab) in firecalc to get the data that you need to put together your own custom spreadsheet to test various variable withdrawal rate scenarios. You can do one run with 0% inflation and $0 withdrawal amount to get the annual return for each period, then do another run with 0% return and $0 withdrawal amount and inflation turned back on to get the inflation amounts. That should give you all the data you'd need.
theanimal, that's an interesting article. Utility maximization is pretty cool stuff.
It's interesting that all the example retirees have a baseline guaranteed income. Everything above that is basically gravy, so it makes sense to be pretty aggressive with that.
It's hard to get more basic than $500/mo, which would require $200k @ 3% SWR for perpetuity. Then any funds above that could be treated specially, and it would be possible to pursue a higher SWR to maximize utility on the extra investments.
So one way of looking at it would be to establish the absolute minimum income that you could conceivably live with, and then anything above that would be done with a more aggressive SWR?
$300/mo -> $120k, $400/mo -> $160k, etc
It's interesting that all the example retirees have a baseline guaranteed income. Everything above that is basically gravy, so it makes sense to be pretty aggressive with that.
It's hard to get more basic than $500/mo, which would require $200k @ 3% SWR for perpetuity. Then any funds above that could be treated specially, and it would be possible to pursue a higher SWR to maximize utility on the extra investments.
So one way of looking at it would be to establish the absolute minimum income that you could conceivably live with, and then anything above that would be done with a more aggressive SWR?
$300/mo -> $120k, $400/mo -> $160k, etc
-
jacob
- Site Admin
- Posts: 17112
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
"What I want to know is what the multiplier will be, based on a high tolerance for withdrawal reductions in the first 15 years."
You're making the mistake of aiming for a precise answer to a problem which does not have an accurate solution.
You can only figure this out using historical data. However, historical data is not future data. This means that any such calculation (which can be done in a spread sheet using inflation and return data per year --- use all sequences) is conditioned on data which may be obsolete.
In short, I believe looking to tweak SWRs to find a more optimal number creates a mindset that lets in risk. Rather you should be thinking about the following things
1) What can possible go wrong between the years of 2013 and 2070 that will destroy capital? (WWIII, hyperinflation, pandemic, superdepression, global crop failure)
2) Any possible reason why futures economic growth rates will not be the same? (resource depletion, climate chaos, population pressures, diminishing returns, demographic pressures...)
3) What you will do if that happens?
So that's the thinking. In terms of actual doing, gain experience actually getting paid from investment. Coupled with the above considerations is a better learning experience than any firecalc simulation.
Do not trust equations when it comes to economics and finance!!!
You're making the mistake of aiming for a precise answer to a problem which does not have an accurate solution.
You can only figure this out using historical data. However, historical data is not future data. This means that any such calculation (which can be done in a spread sheet using inflation and return data per year --- use all sequences) is conditioned on data which may be obsolete.
In short, I believe looking to tweak SWRs to find a more optimal number creates a mindset that lets in risk. Rather you should be thinking about the following things
1) What can possible go wrong between the years of 2013 and 2070 that will destroy capital? (WWIII, hyperinflation, pandemic, superdepression, global crop failure)
2) Any possible reason why futures economic growth rates will not be the same? (resource depletion, climate chaos, population pressures, diminishing returns, demographic pressures...)
3) What you will do if that happens?
So that's the thinking. In terms of actual doing, gain experience actually getting paid from investment. Coupled with the above considerations is a better learning experience than any firecalc simulation.
Do not trust equations when it comes to economics and finance!!!
-
secretwealth
- Posts: 1948
- Joined: Mon Jun 27, 2011 3:31 am
+2 with Jacob.
I tried to figure out how a variable SWR could make me retire earlier--but at the end of the day, no one knows what's going to happen. I think it makes sense to make upside/downside models for various scenarios and plan accordingly with the flexibility to change plans if opportunities or tragedies arise. But that's much less about numbers and more about mindset.
I tried to figure out how a variable SWR could make me retire earlier--but at the end of the day, no one knows what's going to happen. I think it makes sense to make upside/downside models for various scenarios and plan accordingly with the flexibility to change plans if opportunities or tragedies arise. But that's much less about numbers and more about mindset.
You are right, definitely there's only so far that "optimizing SWR" can really go, at the end of the day what's needed regardless is to save up more than just a sufficient amount - a buffer for worse case scenarios is a must.
However, if firecalc.com is useful to us, then I see definite room for improvement to support more ERE-based assumptions.
Currently we have three major milestones on the path of ERE:
Milestone #1: FI (e.g. $85k @ 6% dividend interest, see ERE blog post from 2009)
Milestone #2: 4% SWR (e.g. $150k from $6k * 25)
Milestone #3: 3% SWR (e.g. $200k from $6k * 33.3)
I don't think it would hurt to have another milestone that's based on a semi-ERE variable SWR, since retiring so young still leaves a willingness to do some small amount of work when it might be necessary.
However, if firecalc.com is useful to us, then I see definite room for improvement to support more ERE-based assumptions.
Currently we have three major milestones on the path of ERE:
Milestone #1: FI (e.g. $85k @ 6% dividend interest, see ERE blog post from 2009)
Milestone #2: 4% SWR (e.g. $150k from $6k * 25)
Milestone #3: 3% SWR (e.g. $200k from $6k * 33.3)
I don't think it would hurt to have another milestone that's based on a semi-ERE variable SWR, since retiring so young still leaves a willingness to do some small amount of work when it might be necessary.
-
secretwealth
- Posts: 1948
- Joined: Mon Jun 27, 2011 3:31 am
Your expenses are 6k per year? If so, I think $85k could be enough, provided:
1. You have a strong stomach.
2. You're willing to go back into the workplace if need be.
3. You have a strong stomach.
Shoving that $85k in dividend yielding stocks, you'd need a 7.1% dividend yield to pay for your $6k living expense. With large-cap dividend yielders, dividends tend to go up higher than the rate of inflation, so investing in these companies would ensure that you'd have a lower chance of running out of money even if your current rate of return is at or slightly above your expense needs.
The problem is dividend yielders don't yield 7.1%. You could buy a bunch of AT&T which is yielding near 5% now (if you wait for a dip you might get it at about 6%), and they have been raising dividends since before some forumites were born. If you buy on a dip and lock in 6% and you are willing to sell some stock to cover the 1.1% differential if it goes up, this might cover you. But as you can see, we're skating on thin ice.
SDIV, a dividend ETF, yields a bit above 7.1%, but it doesn't have much of a track history and Global X isn't the best performing money manager out there. Plus there's no way to know if there are dividend increases in the future.
Another option would be to mimic George's leveraged income experiment, which has yielded far more than 7.1% for a long time. It's uncertain (life is uncertain), but could work.
I'd kind of like to see someone do an ERE with an ESWR (that's "e" for extreme, of course.) While it might not work permanently, it'd provide another option to the work 40years and die or live extremely frugally and retire in 5 years options that we currently focus on so much in the forum
1. You have a strong stomach.
2. You're willing to go back into the workplace if need be.
3. You have a strong stomach.
Shoving that $85k in dividend yielding stocks, you'd need a 7.1% dividend yield to pay for your $6k living expense. With large-cap dividend yielders, dividends tend to go up higher than the rate of inflation, so investing in these companies would ensure that you'd have a lower chance of running out of money even if your current rate of return is at or slightly above your expense needs.
The problem is dividend yielders don't yield 7.1%. You could buy a bunch of AT&T which is yielding near 5% now (if you wait for a dip you might get it at about 6%), and they have been raising dividends since before some forumites were born. If you buy on a dip and lock in 6% and you are willing to sell some stock to cover the 1.1% differential if it goes up, this might cover you. But as you can see, we're skating on thin ice.
SDIV, a dividend ETF, yields a bit above 7.1%, but it doesn't have much of a track history and Global X isn't the best performing money manager out there. Plus there's no way to know if there are dividend increases in the future.
Another option would be to mimic George's leveraged income experiment, which has yielded far more than 7.1% for a long time. It's uncertain (life is uncertain), but could work.
I'd kind of like to see someone do an ERE with an ESWR (that's "e" for extreme, of course.) While it might not work permanently, it'd provide another option to the work 40years and die or live extremely frugally and retire in 5 years options that we currently focus on so much in the forum
secretwealth,
Whoops, you're right, it needs to be 7% interest. I was referencing http://earlyretirementextreme.com/early ... 85000.html by the way.
Will check out that experiment, thanks!
Whoops, you're right, it needs to be 7% interest. I was referencing http://earlyretirementextreme.com/early ... 85000.html by the way.
Will check out that experiment, thanks!
-
George the original one
- Posts: 5406
- Joined: Wed Jul 28, 2010 3:28 am
- Location: Wettest corner of Orygun
-
secretwealth
- Posts: 1948
- Joined: Mon Jun 27, 2011 3:31 am
I like BDCs a lot. If the leverage gets you to a net yield of ~17% and you can sustain it for a year or two and invest the excess into conventional dividend growth companies as George suggests. If I have time I'm going to model this approach out--sounds kinda appealing, provided you time the market right and the apocalypse doesn't happen beforehand.
-
George the original one
- Posts: 5406
- Joined: Wed Jul 28, 2010 3:28 am
- Location: Wettest corner of Orygun
-
secretwealth
- Posts: 1948
- Joined: Mon Jun 27, 2011 3:31 am