Bernicke Model vs Constant Spend Level

[marko]

Playing around with FireCalc today to kill some time.

Ran a 'Constant Spending Level', checked "Determine Spending Level" and got $X

Then I ran Bernicke model (no other changes...everything else stayed the same), with my current age--69 and got almost 2 $X as an amount I can spend.

Does this make sense? I appreciate that Bernicke takes into account a lower spend as one ages but double? Struck me as odd.

 

[Dtail]

At the default levels, it is around 20% higher, so your numbers might be high.
OTOH, you are starting at age 69 and the Bernicke model starts at a younger age, so less differential.

 

[ProspectiveBum]

I think the answer is that the Bernicke model assumes that your spending will drop 2% annually starting at age 56, so if you enter an older age, it takes your starting spending number, and drops it 2% per year from 56 to your current age, and uses that as your current “starting” spend number. So instead of spending $60+K, it only has you spending ~$37K in the first year. If you use Constant Spending and the Bernicke model, and enter $60k as your starting spend, it’ll give you the line graph of your annual spending, and you can see how it’s discounting the figure.

 

[audreyh1]

56 eh? My spending started climbing quite a bit at that age as we became more comfortable with spending and gifting more.

 

[Dtail]

Yeah 56 appears to be a bit young for spending less, as many early retirees make a special effort to reap the rewards that they usually worked and saved so hard for.
Additionally, for those that end up facing LTC issues, there can be quite an uptick in spending.

 

[NW-Bound]

Bernicke based his model on the US Bureau of Labor Statistics (BLS) Consumer Expenditure (CE) Surveys, which shows that households of 75+ spend 2/3 that of households of 55-64. This ratio persists over the years when the CE was taken.

One point that may be missed is that the older households have 1.6 persons on the average, while the younger households have 2.1 persons. Fewer people, fewer mouths to feed. Fewer airplane tickets to buy. This is offset by one less SS check, but of course that is on the household income side, not the expense side.

 

[Dtail]

Bernicke based his model on the US Bureau of Labor Statistics (BLS) Consumer Expenditure (CE) Surveys, which shows that households of 75+ spend 2/3 that of households of 55-64. This ratio persists over the years when the CE was taken.

One point that may be missed is that the older households have 1.6 persons on the average, while the younger households have 2.1 persons. Fewer people, fewer mouths to feed. Fewer airplane tickets to buy. This is offset by one less SS check, but of course that is on the household income side, not the expense side.

Yeah but what about between the ages of 64 and 75?
Separately if the above is true, then why would Bernicke start the decline of spending at age 56?

 

[NW-Bound]

Yeah but what about between the ages of 64 and 75?
Separately if the above is true, then why would Bernicke start the decline of spending at age 56?

The CE Survey has the intermediate ages of 65-74 in one group. This age group has 1.9 persons/household, and spends 88% that of the 55-64 group.

One thing from the CE Survey that people do not talk about is that the 65-74 and 75+ have only 67% and 45% the income of the young 55-64 group.

Obviously, the lower household income with age reflects the loss of earned income, and also one of the SS checks. This could impact the spending.

Now, many posters here have a large stash, such that the loss of SS does not impact their spending potential as much. Still, one usually spends less than two.

About Bernicke starting the spending decline at 56, perhaps he did a curve fit to have a smooth transition through the 3 broad age ranges. I have not read his original article.

 

[GMack64]

Hi all, just joined -- first post. Happy to be here.

Yes, I'm sorta-kinda hijacking the thread I suppose, but the question I have is pretty related to the thread topic. Have been running several scenarios in FIRECalc. The description on the Bernicke model says that it assumes a "2 to 3%" reduction in spending per year starting at age 56 (or whatever age you game the calculator to think). When I look at the actual results, by my math it looks like an even more aggressive yearly reduction of roughly 3.4%. Therefore, it strikes me that it's overly optimistic about your portfolio survival chances.

Anyone want to tell me what my error is here? Cheers.

G.

 

[Dtail]

Hi all, just joined -- first post. Happy to be here.

Yes, I'm sorta-kinda hijacking the thread I suppose, but the question I have is pretty related to the thread topic. Have been running several scenarios in FIRECalc. The description on the Bernicke model says that it assumes a "2 to 3%" reduction in spending per year starting at age 56 (or whatever age you game the calculator to think). When I look at the actual results, by my math it looks like an even more aggressive yearly reduction of roughly 3.4%. Therefore, it strikes me that it's overly optimistic about your portfolio survival chances.

Anyone want to tell me what my error is here? Cheers.

G.

Welcome to our forum.
Don't know the answer, but don't know one person who actually uses the Bernicke model for spending.

 

[Jerry1]

Welcome to our forum.
Don't know the answer, but don't know one person who actually uses the Bernicke model for spending.

I didn’t use the model, but I agree with the concept. Therefore, by using a constant spending model, I felt I was being even more conservative than just putting in conservative amounts for spending.

To the OP, you could spend time on figuring out the difference m but that makes me wonder if you’re trying to justify retirement too early. Unless you’ve been let go or have health issues, retirement is no time to be planning on a tight (unrealistic) assumption. As I said, I agree I’ll probably spend less per year as I age, but wouldn’t base my retirement on it. There’s already so many unknowns that adding one more “will I spend less every year” into the planning doesn’t seem prudent to me. I just looked at the Bernicke model and thought, we’ll, if that happens, things will be better than I’m planning. Bernicke, imho, interesting but not too useful.

 

[GMack64]

Welcome to our forum.
Don't know the answer, but don't know one person who actually uses the Bernicke model for spending.

Thanks. I wondered that too, since it feels like a pretty aggressive reach as an assumption.

 

[MichaelB]

Hi all, just joined -- first post. Happy to be here.

Yes, I'm sorta-kinda hijacking the thread I suppose, but the question I have is pretty related to the thread topic. Have been running several scenarios in FIRECalc. The description on the Bernicke model says that it assumes a "2 to 3%" reduction in spending per year starting at age 56 (or whatever age you game the calculator to think). When I look at the actual results, by my math it looks like an even more aggressive yearly reduction of roughly 3.4%. Therefore, it strikes me that it's overly optimistic about your portfolio survival chances.

Anyone want to tell me what my error is here? Cheers.


G.

Welcome to our forum. Why not stop by here and tell us a little about yourself.

Regarding your question, are you looking at nominal or inflation adjusted numbers?

 

[GMack64]

Welcome to our forum. Why not stop by here and tell us a little about yourself.

Regarding your question, are you looking at nominal or inflation adjusted numbers?

I left the default of using CPI in the calculator to account for inflation.

 

[MichaelB]

I left the default of using CPI in the calculator to account for inflation.

How are you calculating the 3.4% you see at the real spending reduction?

 

[marko]

I didn’t use the model, but I agree with the concept. Therefore, by using a constant spending model, I felt I was being even more conservative than just putting in conservative amounts for spending.

To the OP, you could spend time on figuring out the difference m but that makes me wonder if you’re trying to justify retirement too early. Unless you’ve been let go or have health issues, retirement is no time to be planning on a tight (unrealistic) assumption. As I said, I agree I’ll probably spend less per year as I age, but wouldn’t base my retirement on it. There’s already so many unknowns that adding one more “will I spend less every year” into the planning doesn’t seem prudent to me. I just looked at the Bernicke model and thought, we’ll, if that happens, things will be better than I’m planning. Bernicke, imho, interesting but not too useful.

As noted in my original post, I was killing time and just playing around. I've been successfully REd for 16 years now. Thanks for the thoughtful insight however....hopefully someone can make use of it!

 

[GMack64]

How are you calculating the 3.4% you see at the real spending reduction?

I looked at the spending graph. Kinda have to find where the spending line crosses the lines identified with $ so that you can get a decent idea of the numbers. The pattern I found was that after 15 years, you were spending 60% of your initial spend.

x^15 = 0.6
x = 0.6^(1/15)
x = 0.9665

 

[Jerry1]

As noted in my original post, I was killing time and just playing around. I've been successfully REd for 16 years now. Thanks for the thoughtful insight however....hopefully someone can make use of it!

My bad. I was thinking the OP was GMack64.

 

[GMack64]

My bad. I was thinking the OP was GMack64.

Yep, I'm pretty sure we are all on the same page. That Bernicke assumption feels awfully risky to me as well. I'm just trying to make sure that I'm interpreting the intention of the calculator tool correctly.

 

[marko]

My bad. I was thinking the OP was GMack64.

No prob!

 

[Markola]

Don't know the answer, but don't know one person who actually uses the Bernicke model for spending.

Ha, I’ve been on here for several years and haven’t seen two people use the same spending model. There are just too many variables from person to person.

 

[Dtail]

Ha, I’ve been on here for several years and haven’t seen two people use the same spending model. There are just too many variables from person to person.

Yeah that's probably true.
In general IMHO, the retirement calculators are best used as indicators on whether one is in a good place as to being able to retire and is one's retirement in good shape along the way.
As for using them as actual withdrawal strategies, not so much.

 

[tb001]

The problem with using this model, IMO, is that while it seems to hold for most people, as soon as you throw the need for ltc or a full time aide into the mix, it goes out the window.

 

[Jerry1]

The problem with using this model, IMO, is that while it seems to hold for most people, as soon as you throw the need for ltc or a full time aide into the mix, it goes out the window.

I think that problem exists for the constant spending model too. Probably all models. LTC or major health related events need to be addressed. In the Bernicke or constant spending model, I would want to know that I’d have a pretty significant sum left in my later years that would essentially be my LTC insurance. If I don’t spend it, it would be my estate/inheritance.

 

[Dtail]

I think that problem exists for the constant spending model too. Probably all models. LTC or major health related events need to be addressed. In the Bernicke or constant spending model, I would want to know that I’d have a pretty significant sum left in my later years that would essentially be my LTC insurance. If I don’t spend it, it would be my estate/inheritance.

It also depends if one is willing to or can financially go the Medicaid route for LTC.