New with a couple of question

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Post Reply
Erich
Posts: 9
Joined: Sat Jun 09, 2018 11:04 am

New with a couple of question

Post by Erich » Sat Jun 09, 2018 11:11 am

1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
2. Dumb math question, Within the equity portion of the 3 fund portfolio is the international section a percentage of stocks only or the entire portfolio?

So if you have 100K would it be 20% of 100K or 20% of 60K in a 60/40 portfolio?
Last edited by Erich on Sat Jun 09, 2018 11:50 am, edited 2 times in total.

mhalley
Posts: 6350
Joined: Tue Nov 20, 2007 6:02 am

Re: New with a couple of question

Post by mhalley » Sat Jun 09, 2018 11:46 am

Kitces has a good article on sor risk here
https://www.kitces.com/blog/understandi ... d-decades/
The percent of intl is the percent of equities. It would be 20% of 60k in the 100k portfolio. 42k domestic stock, 18k intl, 40k bonds.

Erich
Posts: 9
Joined: Sat Jun 09, 2018 11:04 am

Re: New with a couple of question

Post by Erich » Sat Jun 09, 2018 11:51 am

Thank you.
Wouldn't 20% of 60K in equities be 12K?

livesoft
Posts: 63077
Joined: Thu Mar 01, 2007 8:00 pm

Re: New with a couple of question

Post by livesoft » Sat Jun 09, 2018 11:57 am

Answer 1: Read this series:
https://earlyretirementnow.com/2016/12/ ... t-1-intro/

Answer 2: It would be whatever the person said it was. Basically, people report it any which way, so ask them when you see it reported.
Wiki This signature message sponsored by sscritic: Learn to fish.

Erich
Posts: 9
Joined: Sat Jun 09, 2018 11:04 am

Re: New with a couple of question

Post by Erich » Sat Jun 09, 2018 11:59 am

livesoft wrote:
Sat Jun 09, 2018 11:57 am
Answer 1: Read this series:
https://earlyretirementnow.com/2016/12/ ... t-1-intro/

Answer 2: It would be whatever the person said it was. Basically, people report it any which way, so ask them when you see it reported.
Thank you.

chevca
Posts: 1948
Joined: Wed Jul 26, 2017 11:22 am

Re: New with a couple of question

Post by chevca » Sat Jun 09, 2018 12:06 pm

Erich wrote:
Sat Jun 09, 2018 11:51 am
Thank you.
Wouldn't 20% of 60K in equities be 12K?
Yes.

User avatar
Sandtrap
Posts: 5500
Joined: Sat Nov 26, 2016 6:32 pm
Location: Hawaii😀 Northern AZ.😳 Retired.

Re: New with a couple of question

Post by Sandtrap » Sat Jun 09, 2018 12:07 pm

Erich wrote:
Sat Jun 09, 2018 11:11 am
1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
2. Dumb math question, Within the equity portion of the 3 fund portfolio is the international section a percentage of stocks only or the entire portfolio?

So if you have 100K would it be 20% of 100K or 20% of 60K in a 60/40 portfolio?
1. What others are doing re: sequence of returns risk.
https://www.bogleheads.org/forum/view ... p?t=236665
3. Google "Kitces" and other forum archives on "retirement red zone".
4. Remember that "personal and financial Black Swans" can sink a ship faster than "sequence of returns risk", so also prepare for that.

aloha
j

Ben Mathew
Posts: 225
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: New with a couple of question

Post by Ben Mathew » Sat Jun 09, 2018 1:41 pm

Erich wrote:
Sat Jun 09, 2018 11:11 am
1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
The sequence of returns matters--you want high returns early and low returns later rather than the reverse. But it's not useful to think about the sequence as a separate source of risk unless you know what the overall return will be, and therefore have some confidence that a poor return today will be followed by a high return later and vice versa. If returns are independent across time, then it's completely useless to think about sequence risks. There is no sequence of return risks, just return risk. Now we know empirically that returns across time are not perfectly independent. But it's a close enough approximation for most people. If valuations get high (like now), just assume that it's the new normal and plan accordingly--low future returns based on low current yields, but no correction to reduce valuations. If valuations get low, again assume that it's the new normal--high future returns based on high current yields, but no correction to increase valuations. There's a chance you can do do better by predicting valuation corrections and responding appropriately, but the timing can get tricky. Most people should probably not be doing this.

Erich
Posts: 9
Joined: Sat Jun 09, 2018 11:04 am

Re: New with a couple of question

Post by Erich » Sat Jun 09, 2018 2:05 pm

Ben Mathew wrote:
Sat Jun 09, 2018 1:41 pm
Erich wrote:
Sat Jun 09, 2018 11:11 am
1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
The sequence of returns matters--you want high returns early and low returns later rather than the reverse. But it's not useful to think about the sequence as a separate source of risk unless you know what the overall return will be, and therefore have some confidence that a poor return today will be followed by a high return later and vice versa. If returns are independent across time, then it's completely useless to think about sequence risks. There is no sequence of return risks, just return risk. Now we know empirically that returns across time are not perfectly independent. But it's a close enough approximation for most people. If valuations get high (like now), just assume that it's the new normal and plan accordingly--low future returns based on low current yields, but no correction to reduce valuations. If valuations get low, again assume that it's the new normal--high future returns based on high current yields, but no correction to increase valuations. There's a chance you can do do better by predicting valuation corrections and responding appropriately, but the timing can get tricky. Most people should probably not be doing this.
"If valuations get high (like now), just assume that it's the new normal and plan accordingly--low future returns based on low current yields, but no correction to reduce valuations. If valuations get low, again assume that it's the new normal--high future returns based on high current yields, but no correction to increase valuations. There's a chance you can do do better by predicting valuation corrections and responding appropriately, but the timing can get tricky. Most people should probably not be doing this."

Good points. No one will know how the markets are going to react in the net 50 years. So, My follow up question to that would be define "plan accordingly" based on low and high valuations.

Ben Mathew
Posts: 225
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: New with a couple of question

Post by Ben Mathew » Sun Jun 10, 2018 2:11 pm

Erich wrote:
Sat Jun 09, 2018 2:05 pm
Ben Mathew wrote:
Sat Jun 09, 2018 1:41 pm
Erich wrote:
Sat Jun 09, 2018 11:11 am
1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
The sequence of returns matters--you want high returns early and low returns later rather than the reverse. But it's not useful to think about the sequence as a separate source of risk unless you know what the overall return will be, and therefore have some confidence that a poor return today will be followed by a high return later and vice versa. If returns are independent across time, then it's completely useless to think about sequence risks. There is no sequence of return risks, just return risk. Now we know empirically that returns across time are not perfectly independent. But it's a close enough approximation for most people. If valuations get high (like now), just assume that it's the new normal and plan accordingly--low future returns based on low current yields, but no correction to reduce valuations. If valuations get low, again assume that it's the new normal--high future returns based on high current yields, but no correction to increase valuations. There's a chance you can do do better by predicting valuation corrections and responding appropriately, but the timing can get tricky. Most people should probably not be doing this.
Good points. No one will know how the markets are going to react in the net 50 years. So, My follow up question to that would be define "plan accordingly" based on low and high valuations.
All assets have an engine inside them that drives returns. For bonds, it's the interest rate paid by the bond. For stocks, it's the profit generated by the company (earnings). For property, it's the rent generated (minus costs). For gold, it's nothing--there is no engine. If valuations stay constant, then the return paid by the asset equals the return generated by its internal engine. In the short run, a lot of the returns is determined by valuation changes. In the long run, the earnings generated by the engine starts playing a bigger and bigger role. That's because earnings are cumulative, but valuation changes are not. Valuation changes still matter over long horizons, just not as much as in short horizons.

Valuation changes are important, but they are hard to predict and equally likely to help or hurt. So I just focus on the internal engine of the asset as my best guess of what the asset will yield. For bonds, I look at the real interest rate. For stocks, I look at the earnings yield (reciprocal of the price earnings ratio). This is the current yield put out by its internal engine. The total return over my horizon may be higher or lower based on valuation changes. The engine might also speed up (growth firms) or peter out (dying firms). But for the stock market as a whole, I'm comfortable taking today's earnings as my best guess about the real return I can expect from it. For example, Vanguard's Total Stock Market (VTSAX) has an price/earnings ratio of 20.5. So it's earnings yield is 1/20.5= 4.9%. So this becomes my best guess for the real return of U.S. stocks as a whole. If you want to be conservative, knock down 0.5% or 1% to get 4.4% or 3.9%.

(It's because of this approach that I don't hold gold--my best guess about the future is 0% because there is no engine. All returns come from valuation changes, and I have no idea what that will be.)

Once you have estimates for the real returns of assets, you simply adjust your savings rates, asset allocation, retirement age, and retirement draw until you have a combination that will leave you with positive assets at the end of life. I do this with a spreadsheet. I do all of my analysis in real (inflation adjusted) terms because nominal dollars decades from today will be very misleading. It's hard to wrap your head around what $100K means decades from now.

Of course, you can't know today how the next fifty years will play out. Your best guess will almost certainly be way off. You may be hit by poor returns, low earnings, and unexpected expenses. The key is to keep adjusting your plan. As life unfolds, keep tweaking. Over the course of your life, you will have tweaked your way over to a pretty good plan that reflects the actual circumstances that you found yourself in. It won't be the perfect plan because you never knew the future. But it will be a pretty good plan because you kept adjusting as it slowly revealed itself.

Erich
Posts: 9
Joined: Sat Jun 09, 2018 11:04 am

Re: New with a couple of question

Post by Erich » Sun Jun 10, 2018 3:43 pm

Ben Mathew wrote:
Sun Jun 10, 2018 2:11 pm
Erich wrote:
Sat Jun 09, 2018 2:05 pm
Ben Mathew wrote:
Sat Jun 09, 2018 1:41 pm
Erich wrote:
Sat Jun 09, 2018 11:11 am
1. What is the best way to deal with sequence of return risk over a possible 45 to 50 year retirement?
The sequence of returns matters--you want high returns early and low returns later rather than the reverse. But it's not useful to think about the sequence as a separate source of risk unless you know what the overall return will be, and therefore have some confidence that a poor return today will be followed by a high return later and vice versa. If returns are independent across time, then it's completely useless to think about sequence risks. There is no sequence of return risks, just return risk. Now we know empirically that returns across time are not perfectly independent. But it's a close enough approximation for most people. If valuations get high (like now), just assume that it's the new normal and plan accordingly--low future returns based on low current yields, but no correction to reduce valuations. If valuations get low, again assume that it's the new normal--high future returns based on high current yields, but no correction to increase valuations. There's a chance you can do do better by predicting valuation corrections and responding appropriately, but the timing can get tricky. Most people should probably not be doing this.
Good points. No one will know how the markets are going to react in the net 50 years. So, My follow up question to that would be define "plan accordingly" based on low and high valuations.
All assets have an engine inside them that drives returns. For bonds, it's the interest rate paid by the bond. For stocks, it's the profit generated by the company (earnings). For property, it's the rent generated (minus costs). For gold, it's nothing--there is no engine. If valuations stay constant, then the return paid by the asset equals the return generated by its internal engine. In the short run, a lot of the returns is determined by valuation changes. In the long run, the earnings generated by the engine starts playing a bigger and bigger role. That's because earnings are cumulative, but valuation changes are not. Valuation changes still matter over long horizons, just not as much as in short horizons.

Valuation changes are important, but they are hard to predict and equally likely to help or hurt. So I just focus on the internal engine of the asset as my best guess of what the asset will yield. For bonds, I look at the real interest rate. For stocks, I look at the earnings yield (reciprocal of the price earnings ratio). This is the current yield put out by its internal engine. The total return over my horizon may be higher or lower based on valuation changes. The engine might also speed up (growth firms) or peter out (dying firms). But for the stock market as a whole, I'm comfortable taking today's earnings as my best guess about the real return I can expect from it. For example, Vanguard's Total Stock Market (VTSAX) has an price/earnings ratio of 20.5. So it's earnings yield is 1/20.5= 4.9%. So this becomes my best guess for the real return of U.S. stocks as a whole. If you want to be conservative, knock down 0.5% or 1% to get 4.4% or 3.9%.

(It's because of this approach that I don't hold gold--my best guess about the future is 0% because there is no engine. All returns come from valuation changes, and I have no idea what that will be.)

Once you have estimates for the real returns of assets, you simply adjust your savings rates, asset allocation, retirement age, and retirement draw until you have a combination that will leave you with positive assets at the end of life. I do this with a spreadsheet. I do all of my analysis in real (inflation adjusted) terms because nominal dollars decades from today will be very misleading. It's hard to wrap your head around what $100K means decades from now.

Of course, you can't know today how the next fifty years will play out. Your best guess will almost certainly be way off. You may be hit by poor returns, low earnings, and unexpected expenses. The key is to keep adjusting your plan. As life unfolds, keep tweaking. Over the course of your life, you will have tweaked your way over to a pretty good plan that reflects the actual circumstances that you found yourself in. It won't be the perfect plan because you never knew the future. But it will be a pretty good plan because you kept adjusting as it slowly revealed itself.
Thank you for the detailed explanation. I was very good.

Any resources where I can learn to determine real return (by real I'm guessing you mean after taxes and inflation returns) and developing the spreadsheets needed to adjust as I go? Basically, where can I learn this skill?

I've reached a point where I have saved enough based on the sale property and savings that I can most likely retire. But as you have said 50 potential years is a long time. A 4% SWR may not be so safe. According to some of the other people who have posted to my question and suggested some several websites to read, It seems 3% may be a better option. But working part time will lower that 3% hopefully leaving some for the future where I can't work or am no longer willing to work even part time. So based on what you are saying, having the skill to adjust based on data is a good way to have a strong idea (nothing is ever 100%) of how much I could withdraw without breaking the bank. I don't care as much about preserving the principal for 50 years. I care more about being 70 and the market tanking and now having to live on a lot less. If I live to 100 and I have some money left at the end, then I've made it. The reason I asked about sequence of returns is because I don't want to be in a situation where I have significant stress every time there are large market swings. I guess that comes down to accepting some ups and downs with a good enough asset allocation that allows me to sleep well at night.

Ben Mathew
Posts: 225
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: New with a couple of question

Post by Ben Mathew » Tue Jun 12, 2018 6:35 pm

Erich wrote:
Sun Jun 10, 2018 3:43 pm
Any resources where I can learn to determine real return (by real I'm guessing you mean after taxes and inflation returns)
By real, I mean inflation adjusted. But yes, if the account is taxable, you should adjust the return for taxes as well.
Erich wrote:
Sun Jun 10, 2018 3:43 pm
Any resources where I can learn to determine real return (by real I'm guessing you mean after taxes and inflation returns) and developing the spreadsheets needed to adjust as I go? Basically, where can I learn this skill?
I have made a spreadsheet which will take your assumptions and create a graph of your wealth trajectory. You can use it to adjust your planned annual savings, asset allocation, retirement age, and retirement draw till you are left with positive assets at age 100.

The nine inputs you'll need to enter into the spreadsheet are:

1. Bond return (inflation adjusted)
2. Stock return (inflation adjusted)
3. Current savings (what you have accumulated so far)
4. Current age (if you aren't saving now, age you plan to start)
5. Retirement age
6. Annual savings between now and retirement (inflation adjusted)
7. Retirement draw (inflation adjusted)
8. Starting stock allocation (today's allocation)
9. Ending stock allocation at age 100

When entering these inputs, keep in mind:

(1) For a taxable account:
- reduce the real return to account for taxes on growth in the form of interest, dividends, and realized capital gains
- increase your retirement draw to account for deferred capital gains tax due at withdrawal

(2) For traditional retirement accounts:
- increase your retirement draw to pay the deferred income tax due at withdrawal.

(3) For Roth retirement accounts:
- no adjustments needed

Since different types of accounts (traditional, Roth, and taxable) require somewhat different inputs to account for tax treatment, if you are using a mix of account types and want to be precise, you would need to create a separate spreadsheet for each account type. If you do this, you might want to add them up together with another spreadsheet to see the total wealth trajectory.

The default calculation in my spreadsheet assumes constant real savings and retirement draws. But if you want to vary savings and draws by year (for example, you are anticipating some years of low savings due to kids' college tuition or high draws due to delayed social security), you can click on the worksheet tab and change the numbers manually.

I have uploaded the spreadsheet here: Wealth Trajectory Calculator

If you have any questions or feedback on how to improve the spreadsheet, please let me know.

Post Reply