question about safe withdrawal rate (support, single, wife, 401k)

[Wile E. Coyote]

Originally Posted by mathjak107

No , I’m am arguing you can do what you are doing and get an accurate picture for the reason I explained .

ACTUAL Yearly Inflation and yearly interest rates must be dealt with year by year ..you will spend more in years with higher negative real returns changing each years balance in ways simply dividing will not account for

I cant explain it to you any better .

If you want to learn why you are wasting your time learn why sequences matter otherwise this conversation is silly.

If the first 15 years of a 30 year retirement are at 6% inflation and spending is adjusted each year by 6% if years 15-30 are at 3% and spending adjusted accordingly the average inflation is 3% .

But balances will be very different calculating the first 15 years at 3% and the last 15 at 6% even though it is the same 3% average

I do understand that. You do have to deal with reality year by year. I would obviously be doing my own math (not relying on a calculator I have no idea what data is in it at any given point in time).

[mathjak107]

Which goes back to what a safe withdrawal represents because no one can predict future rates , inflation , markets or the order .

It is based on the worst data sets and sequences we have had .

It is the most you can safely take based on the worst we have had .

Then anything better just means a better outcome and more money…

This is why trying to use some made up interest rates or average inflation number is going to be far off ….you can see in those firecalc results the huge span between best and worst outcomes

[kavm]

Originally Posted by Wile E. Coyote

I do understand that. You do have to deal with reality year by year. I would obviously be doing my own math (not relying on a calculator I have no idea what data is in it at any given point in time).

I share this sentiment. I understand the point Mathjack107 is making as well but, for me, it is important to know the model - what is is modeling and what not - to make decisions. Historical patterns are not without value but past is not necessarily reflective of the future. While pattern of past results reflects the values of the random variable, how reflective it is about the future is not always obvious. It is more useful that the history of coin toss results but its value can be overstated. Sometimes simpler models are more robust.

Not saying Mathjack107 doesn’t have a point but different people approach the decision making differently.

[mathjak107]

Originally Posted by kavm

I share this sentiment. I understand the point Mathjack107 is making as well but, for me, it is important to know the model - what is is modeling and what not - to make decisions. Historical patterns are not without value but past is not necessarily reflective of the future. While pattern of past results reflects the values of the random variable, how reflective it is about the future is not always obvious. It is more useful that the history of coin toss results but its value can be overstated. Sometimes simpler models are more robust.

Not saying Mathjack107 doesn’t have a point but different people approach the decision making differently.

The calculators are all based on math not what was or what will be .

Think of it as building a house in a hurricane area ….

If you know the worst winds and tides and floods you can build a house to weather the worst conditions ever seen and then build in a fudge factor .

That is a pretty good place to start and odds are likeLy higher than 90% of standing up well .

That is what a safe withdrawal rate is based on . It is based on the worst combinations of returns , rates , inflation and sequences… the Likes of which we have not seen since 1966 for a retiree .

Not 2000 or 2008 came close.

When using a typical spread sheet you can take the same parameters and just change the order they come in and you can run out of money 15 years sooner between the best and word order.

That is a big margin of error so far off it is really a bad way to do it .

One of the best times we had were 1987 to 2003 …that is 17 years averaging almost 14% …

But put that return in a spread sheet and you could be in a real pickle if spending down .

Because by taking those 17 years and just reversing the order the outcomes came in at are amazing

in the wrong sequence the best of returns will fail .

drawing money out of our nest egg changes the results of even the biggest bull market returns . drawing out money is no different in flat or down markets to a trader having a string of loosing trades. the end result is the same.

while the 17 year period of 1987 to 2003 had the s&p averaging over 13% a year the returns a retiree saw werent even close to that because of the sequence in which the gains and losses came.

three years of negative returns for retires drawing out money right at the beginning negated one of the greatest bull markets ever.

they took a 100k portfolio and an almost 14% average return for 17 years and ran a few different sequence simulations on it.

using the rule of thumb of having an average long term return of 7% from a 50/50 mix which would allow 4% to be drawn and 3% to grow the portfolio by inflation they increased the withdrawl rate to match an almost 14% return and 4% inflation . they left 4% to grow the nest egg by the inflation rate and in this case were able to take a 10% withdrawl rate to keep the same ratio.

the results were mind blowing.

the balance ranged from a high of 76,629 left over to a low of minus 187,606 depending on the sequence of gains and losses.
.
thats amazing, because the return was a whopping average of almost 14% a year over that time frame and the only thing they changed was the order of the actual gains and losses.

it shows you that besides market rate risk and interest rate risk the biggest risk we take of all is sequence risk. it also shows you there is no such thing as a long term average return thats meaningful when it comes to the de-accumulation stage of our lives.

looking at past performance of funds and seeing the 6 % or 10% average return for a 10 year period is meaningless for predicting how you may do going forward. you cant say if i had this fund the last decade i would have been just fine. nope ,the order of events happening may still have wiped you out even though the average return says you should have been able to pull 4% a year and had more now then you started with.

if only it was that easy .

[kavm]

Mathjack107- I understand and appreciate the point you are making. I am fairly well trained mathematical and quantitative reasoning in presence of uncertainty. I was offered quant roles in investment banking but did not pursue them.

How one looks at these questions does, I believe, depend upon the individual and the situation. We are still in an accumulation (not merely from the stock market returns). So, the questions of withdrawal rates are not terribly relevant yet. Furthermore, our savings and their returns is likely exceed our needs. When I posted my numbers on the FIRE forum a couple of years back, I got bombarded with insistent responses suggesting I am crazy to continue working and should retire right away. My point is that we aren’t trying to thread the needle. By nature, we do not try to work out the financial decisions to a very precise level as we understand the variables involved and the unknowns. Some are heavy into financial planning, we are not. But, we have achieved a high net-worth status starting from zero - and are comfortable with our approach.

In quantitative modeling, there is always a trade off between robustness and sensitivity. A richer model explains more than a simpler model but can also mislead more badly. Richer models look great in backcasting (as they are fitted to the historical data) but are not nearly as good in forecasting the future. All this is not to say that a simplistic model but the belief in the accuracy or superiority of the richer model can mislead badly. So, individuals can differ on how much precision they care for.

This is just my humble opinion. I truly respect your perspective.

[phoebesmom]

Wile E, I get your point about just having the money in cash and then dividing it. Unfortunately, I don't have a million dollars and my money has to work for me, and for twenty years I've been investing it and it has supported me in the past, and I need it to support me into the future. Both Mom and Dad lived into their 90's, and I have very low social security, so I have to keep investing as I have done in past, and keep living off proceeds of investments. I don't want the balance to go down, I want it to stay steady. So I am more interested in income investing, and also not taking out too more than I can grow it, if that makes sense.

The money has to support me for 30 more years, if I live as long as my mother has lived. I selected 3 per cent to be more conservative and less risky than 4 per cent.

[mathjak107]

Originally Posted by kavm

Mathjack107- I understand and appreciate the point you are making. I am fairly well trained mathematical and quantitative reasoning in presence of uncertainty. I was offered quant roles in investment banking but did not pursue them.

How one looks at these questions does, I believe, depend upon the individual and the situation. We are still in an accumulation (not merely from the stock market returns). So, the questions of withdrawal rates are not terribly relevant yet. Furthermore, our savings and their returns is likely exceed our needs. When I posted my numbers on the FIRE forum a couple of years back, I got bombarded with insistent responses suggesting I am crazy to continue working and should retire right away. My point is that we aren’t trying to thread the needle. By nature, we do not try to work out the financial decisions to a very precise level as we understand the variables involved and the unknowns. Some are heavy into financial planning, we are not. But, we have achieved a high net-worth status starting from zero - and are comfortable with our approach.

In quantitative modeling, there is always a trade off between robustness and sensitivity. A richer model explains more than a simpler model but can also mislead more badly. Richer models look great in backcasting (as they are fitted to the historical data) but are not nearly as good in forecasting the future. All this is not to say that a simplistic model but the belief in the accuracy or superiority of the richer model can mislead badly. So, individuals can differ on how much precision they care for.

This is just my humble opinion. I truly respect your perspective.

It really does boil down to simple math anyone can monitor ..

All the data and facts show us that every single failure to hold 4% inflation adjusted has happened in the first 15 years of a 30 year retirement .

They were a retiree starting out in 1907,-1929 ,1937 ,1965 ,1966

So mathematically for 4% to hold inflation adjusted it simply needs to see a minimum of a 2% real return over the first 15 years .

So it is easy to monitor along the way . If 5 years in and you are below then a red flag should go up , 8-10 years in I would take a pay cut if that happened .

So monitoring the income is easy based on research by michael kitces .

Now that is the income side of things ..that is pretty predictable and just seeing that 2% real return the first 15 years provides a safe , secure , CONSISTENT income stream that can pretty much be counted on .

What can’t be predicted is the balance along the way .

That is subject to way to many unpredictable variables .

So the income stream holding its level is one thing and that is predictable mathematically…

So the way a safe withdrawal rate is designed by definition , the income stream just needs a minimal real return to stand up …it is designed to have the balance vary at the end based on the unknowns

[lovnova]

My question? Is this lecture in any way useful to the average retired person looking for financial guidance?

[mathjak107]

Originally Posted by lovnova

My question? Is this lecture in any way useful to the average retired person looking for financial guidance?

It sure is if they are counting on that pile of money they accumulated supporting them over a retirement life time .

How will you know how much you can count on without a basic understanding of what to use to calculate it.

Are you going to pull a number out of your butt if you are creating your own income stream off what you have ?

So while it is not necessary to know what the underlying reasons are that certain respected calculators and standards in the financial planning world are used . It is important to know how to arrive at that number via calculators thats do a good job of laying it out for you , like firecalc does to mention one of them or the fidelity planner

[lieqiang]

Not without basic formatting, paragraphs are a thing.