Is monte carlo retirement planner too conservative?

[vagavince]

I recently started using Empower's retirement planner to check my plan and do budgeting. I understand it use monte carlo instead of market history.

Even at 2% withdraw rate its still showing only 90% chance of success. How to use this number in my planning? I thought 2-3% is already very conservative? Should I just ignore it or what should I interpret or act with this percentage of success

[jimmyshutter]

I don't give much thought about Empower's success rate since every other retirement planner says differently for me. My thoughts are they may be trying to get people to sign up for their financial advice services.

Also, most of us will adjust withdrawals if/when needed. "Rich, Dead, Broke" gives options to adjust spending which I find very useful and it really shows how having just a little flexibility in spending
 can make a huge difference in success (Firecalc also has a few options for withdrawals). There are a lot withdrawal strategies for not running out of money and I'll likely use the VPW spreadsheet since it seems the simplest to me.

[Ron Scott]

I’m guessing you’re in your mid-40s and are FI. I’m assuming you will follow the typical withdrawal strategy most people here do and not automatically inflate each year…play it by ear and try not to grow it on automatic.  A SWR to me might be a high of 3%.

Empower seems a bit conservative. But it’s not crazy and if they aren’t jumping on the “Everything will always stay the same with market returns” bandwagon, GOOD.

Betting the rest of your life on 4-5% with an uncertain future, or having to drive for Uber or something just as fun if you get scared when you’re in your 70s, isn’t my cup of tea.

[Just Joe]

We have Empower through our employer as well. Perhaps they want you to give them more money to manage? Save harder!

[joemandadman189]

i tried out the personal capital advisors for 6 months and their retirement planner (now empower) were quite conservative - the CFP i worked with had me create a financial plan that scaled down future costs to be in todays dollars. After all that work we basically had a savings plan that matched my own. but their simulator was very conservative

[flyingaway]

Nobody can predict the future accurately. Just take what is good enough, not 100% good.

[moof]

Predicting the future is perilous, but you gotta do something, right?

Historical backtesting shines when you want to think back and see what would happen if you tried to survive the great depression, WWII, 60's stagflation, or a host of other calamities that you know something about.  The percent survival rate is a bit misleading for a few good reasons however.

1)  Blindly following the 4% rule like the Trinity study did would akin be to being dropped out of a time machine with a briefcase full of cash in some random year.  You go and buy your portfolio and start living off it.  You pay no attention to how you got the money, what the current headlines are, or even care one lick if the stock market is in a crazy state.  Instead we tend to save until we hit a trigger, which is a very set of circumstances leading to a different chance of success.

2)  There is a survivor bias.  We mostly look to the USA for data which has never had a true catastrophe like WWI or WWII destroy it.  Using the 4% rule while living in various parts of Europe would historically been very different, though it was not like you would have had drastically different results being being a worker vs. retiree.

3)  With only a handful of big drops historically, it is not clear if future boom/bust cycles are proportional to history.  Basically there are a handful of circumstances that are unlikely to fully predict likely future market and society behavior.

Monte Carlo tries to fix some of this by randomizing the returns so you can try out all sorts of scenarios for returns with more than just the last ~140 years of returns.  But it has all sorts of other limitations.

1)  Who chooses the variance on the returns?  Who chooses the average returns?  Some of these are obscured from review if you plug your data into a random calculator, or hand it over to a CFPA who charges you to plug it into theirs.

2)  The stock markets have a tendency to revert to the mean, basically a really bad previous year makes the odds low that you'll get another really bad year after.  A strict Monte-Carlo just pulls random returns from thin air ignoring such behaviors.

3)  Some events that get spit out of a Monte Carlo simulation like multiple years of massive negative returns would translate in reality to a complete societal collapse, and are thus impractical to plan for let alone consider as a reasonable failure in your planning.

[Ron Scott]

Predicting the future is perilous, but you gotta do something, right?

Historical backtesting shines when you want to think back and see what would happen if you tried to survive the great depression, WWII, 60's stagflation, or a host of other calamities that you know something about.  The percent survival rate is a bit misleading for a few good reasons however.

1)  Blindly following the 4% rule like the Trinity study did would akin be to being dropped out of a time machine with a briefcase full of cash in some random year.  You go and buy your portfolio and start living off it.  You pay no attention to how you got the money, what the current headlines are, or even care one lick if the stock market is in a crazy state.  Instead we tend to save until we hit a trigger, which is a very set of circumstances leading to a different chance of success.

2)  There is a survivor bias.  We mostly look to the USA for data which has never had a true catastrophe like WWI or WWII destroy it.  Using the 4% rule while living in various parts of Europe would historically been very different, though it was not like you would have had drastically different results being being a worker vs. retiree.

3)  With only a handful of big drops historically, it is not clear if future boom/bust cycles are proportional to history.  Basically there are a handful of circumstances that are unlikely to fully predict likely future market and society behavior.

Monte Carlo tries to fix some of this by randomizing the returns so you can try out all sorts of scenarios for returns with more than just the last ~140 years of returns.  But it has all sorts of other limitations.

1)  Who chooses the variance on the returns?  Who chooses the average returns?  Some of these are obscured from review if you plug your data into a random calculator, or hand it over to a CFPA who charges you to plug it into theirs.

2)  The stock markets have a tendency to revert to the mean, basically a really bad previous year makes the odds low that you'll get another really bad year after.  A strict Monte-Carlo just pulls random returns from thin air ignoring such behaviors.

3)  Some events that get spit out of a Monte Carlo simulation like multiple years of massive negative returns would translate in reality to a complete societal collapse, and are thus impractical to plan for let alone consider as a reasonable failure in your planning.

This is very well put.

How would an historical database analyzed with a Trinity-like model predict your odds of dying from smallpox?

Why do we think an historical database of market returns for the United States would be more predictive than returns from China, India, Russia, South America, Mexico, Europe, or other economies?

People like to say the historical record is so helpful because it shows we can survive a great depression and two world wars. Of course global economies were not as interconnected as they are today. Now that they are, will another great depression be the same? Will World War III likely have the same aftermath as the previous two?

If you’re 40, how confident are you in the ability of western democracy and capitalism to survive in a form similar to what created returns in the past?

What influence will AR, AI, and quantum computing have on future economic prospect for average individuals?

What will happen if large institutional investors develop and deploy software models that allow them to suck more and more of the profits from investing away from individuals?

What will happen if the cost of coping with climate change becomes one of the largest government expenses in the world?

WE DON’T KNOW.

Statisticians DO KNOW the very real limitations of historical data:

1. Changes in technology, social structures, and world relations can render historical data less relevant as we move further into the future. Predictions based solely on past trends can’t automatically account for such changes.

2. Simple models like Trinity rely on the assumption that underlying relationships between variables remain stable over time. This assumption might not hold true in the future, leading to misleading predictions.

3. Using datasets that address snippets of time like 100 years or less can lead to models that overfit the data and perform poorly on unseen examples. This means your predictions might be accurate for the historical data but fail to generalize to the future.


You can believe anything you choose to, but the earth isn’t flat, there is no Santa Clause, and Trinity can’t predict investment returns in the future. 

So what do we do? I don’t know; maybe be conservative, flexible, cross your fingers, and come up with a Plan B.

[LD_TAndK]

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

[Wolfpack Mustachian]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

These speculations often bring up ideas where FIRE is of small importance in the grand scheme. As has been said many times, life events outside of cash saved are much more likely to cause FIRE to fail: relationship challenges/divorce, a major medical emergency in the US, etc. In most of these situations, having 1.3 million vs 1 million for the standard 40k withdrawal mentioned... I'm not seeing it as a silver bullet.

Everyone has to manage their risk in a way they see fit. I'd rather manage mine by having money in the budget that I could easily cut in rough years but that I want to spend - mine is for big vacations. I want to have easy avenues to post retirement work that I could pick up ready, especially early in retirement, in case I see things going bad and i need to pivot. With those, I view their resiliency as greater than saving up a lower withdraw rate and see targeting 4% as reasonable under those circumstances.

[Ron Scott]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

These speculations often bring up ideas where FIRE is of small importance in the grand scheme. As has been said many times, life events outside of cash saved are much more likely to cause FIRE to fail: relationship challenges/divorce, a major medical emergency in the US, etc. In most of these situations, having 1.3 million vs 1 million for the standard 40k withdrawal mentioned... I'm not seeing it as a silver bullet.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

[simonsez]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

These speculations often bring up ideas where FIRE is of small importance in the grand scheme. As has been said many times, life events outside of cash saved are much more likely to cause FIRE to fail: relationship challenges/divorce, a major medical emergency in the US, etc. In most of these situations, having 1.3 million vs 1 million for the standard 40k withdrawal mentioned... I'm not seeing it as a silver bullet.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Guy/Mr. Scott,

You can't predict specific points of the future with any high degree of precision but knowing something about the past can inform you about a likely range of outcomes moving forward and may help with planning.

Most (all?) of us absolutely do make some type of assumption that will carry into the future with regard to our financial selves.  Anyone who invests in anything is doing so under the assumption that it will be worth more (or has or will have positive cashflow, dividends, etc. including non-pecuniary/intangible value/utility) at a certain point in the future.  If you truly thought that past market returns meant zilch, you'd have no basis for ASSUMING that returns of any kind in any market with a medium to long-term lens would be positive (or at least a >50% of being in the black) - thus, why invest at all?  How would you ever operate a business if you did not assume there would be a market for your product/service in the future based on knowledge you gained in the past?

Even someone who works every year of their adult life until they drop dead, saves zero in any non-cash market and just tucks any unspent cash into the proverbial mattress is still using assumptions about the value of cash.  What is that behavior based on if not assumptions made due to history?

Assumptions seem baked into just about every kind of financial planning for the future no matter what your portfolio holds.  If you hold X, you think X will be worth it for some reason that you read about or learned in the past and are applying it to the future, no?

I'll be asinine and say that VTSAX or whatever similar holdings will be higher 50 years from now than it is today and I am basing that off the performance of other similar instruments in the past 100 years. 

So what do we do? I don’t know; maybe be conservative, flexible, cross your fingers, and come up with a Plan B.

Without making any financial assumptions about the future, what exactly does this mean?

[mathlete]

Just a few comments.

1. A monte carlo simulation, even if not based on market history, should at least be calibrated by market history. In other words, if the long term average of stocks and bonds is X%, that should probably be reflected in the baseline scenario that the simulation works off of.

2. Monte Carlo models are useful because they allow you to look at a range of outcomes. Maybe Empower has all of that visible on the back end, but all you're seeing is the 10% failure rate number. It would be nice to know, of that 10%, how many simulations failed before 25 years? It's probably a lot smaller of a number.

3. A retirement brokerage might be incentivized to lower the success rate, encouraging you to save more. This gets them more assets under management, and thus, makes them more money. Put more charitably, it encourages people to save more, which they should be doing anyway.

4. Monte Carlo models are actually pretty easy to build! You can make one in Excel in 30 minutes. I strongly encourage you to check that out if you're interested. All you need is an assumption about the historical rate of return of the markets, the historical volatility of the markets, knowledge of the RAND() function and the lognormal.inv function in Excel, and you're good to go!

[Wolfpack Mustachian]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

These speculations often bring up ideas where FIRE is of small importance in the grand scheme. As has been said many times, life events outside of cash saved are much more likely to cause FIRE to fail: relationship challenges/divorce, a major medical emergency in the US, etc. In most of these situations, having 1.3 million vs 1 million for the standard 40k withdrawal mentioned... I'm not seeing it as a silver bullet.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Criticisms like this definitely do not miss the mark. The fact is, the same style examples come up often because they are significant paradigm shifts, so explaining why they actually don't really change the reality of things or are so extreme they make FIRE irrelevant in importance by comparison illustrate that they're not something to be afraid of .

Beyond that is your overarching point which, if I may sum up, seems to be there's no way to guarantee or even really know, so, *shrug*. Ultimately, you are correct that we can't know for sure. However, what would you say you'd do? Pick a specific withdraw rate? What percentage? Why? With what guidance? I believe I've seen you say 3% but why? Is it just because it's more conservative than 4%?

We either take a leap of faith and retire or work until we die. If we retire, we either base it on some line of thought or study or we just wing it. I honestly have no idea what you're arguing to do other than don't retire at 4% with no contingency which literally no one I've seen has argued for. Beyond that, other than arguing that we  can't know for sure, which is simultaneously inarguably true and, imo, extremely unhelpful, I'm totally lost.

[Ron Scott]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Beyond that is your overarching point which, if I may sum up, seems to be there's no way to guarantee or even really know, so, *shrug*. Ultimately, you are correct that we can't know for sure. However, what would you say you'd do?

i am glad you realize we can’t know future returns. Attempts like Trinity, that study the entrails of the 20th century, fail from the get-go—in their most basic assumptions. But also realize people like me, who simply call out their failure, have no responsibility whatsoever for finding a resolution.

I do not believe in a supernatural god that promulgates rules for living, yet I have a moral code I strive to live by. We do not need to force beliefs to live productively or explain what we don’t know.

I have no magic wand. I personally assume zero real growth in returns in a 60-40 portfolio and spend even less in my own retirement. (We need some kind of plan, right?) I will recommend a very conservative approach in discussions but never claim to back up such a recommendation with proof of fitness for use. I owe nothing.

[Wolfpack Mustachian]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Beyond that is your overarching point which, if I may sum up, seems to be there's no way to guarantee or even really know, so, *shrug*. Ultimately, you are correct that we can't know for sure. However, what would you say you'd do?

i am glad you realize we can’t know future returns. Attempts like Trinity, that study the entrails of the 20th century, fail from the get-go—in their most basic assumptions. But also realize people like me, who simply call out their failure, have no responsibility whatsoever for finding a resolution.

I do not believe in a supernatural god that promulgates rules for living, yet I have a moral code I strive to live by. We do not need to force beliefs to live productively or explain what we don’t know.

I have no magic wand. I personally assume zero real growth in returns in a 60-40 portfolio and spend even less in my own retirement. (We need some kind of plan, right?) I will recommend a very conservative approach in discussions but never claim to back up such a recommendation with proof of fitness for use. I owe nothing.

We're all on an anonymous internet forum - of course you owe nothing. That being said, the easiest thing in the world is to point out obvious truths (we can't know the future) while using them to try to pick apart attempts at actually doing something useful - looking for guidance on when you can retire. A decision has to be made about when or if to retire. People using the Trinity study or the Monte Carlo planner or whatever are at least trying to make a decision that we all have to make with data. It may have flaws and one study may be better than another, but they're good starting points.

The belief that we can't know for sure so we should just "be conservative" with only a generic "we don't know the future" as a reason is a perspective that I feel should be pushed back on hard, especially on forums like this. Everybody must make their own decisions. That being said, you have deflected when asked for specific reasons for your perspective and only provided criticisms that are so generic that they could be applied to an argument for not planning for anything ever. I feel a perspective like this, compared to a perspective looking at data to try to make an informed decision, is a perspective that has little credibility.

[Ron Scott]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Beyond that is your overarching point which, if I may sum up, seems to be there's no way to guarantee or even really know, so, *shrug*. Ultimately, you are correct that we can't know for sure. However, what would you say you'd do?

i am glad you realize we can’t know future returns. Attempts like Trinity, that study the entrails of the 20th century, fail from the get-go—in their most basic assumptions. But also realize people like me, who simply call out their failure, have no responsibility whatsoever for finding a resolution.

I do not believe in a supernatural god that promulgates rules for living, yet I have a moral code I strive to live by. We do not need to force beliefs to live productively or explain what we don’t know.

I have no magic wand. I personally assume zero real growth in returns in a 60-40 portfolio and spend even less in my own retirement. (We need some kind of plan, right?) I will recommend a very conservative approach in discussions but never claim to back up such a recommendation with proof of fitness for use. I owe nothing.

We're all on an anonymous internet forum - of course you owe nothing. That being said, the easiest thing in the world is to point out obvious truths (we can't know the future) while using them to try to pick apart attempts at actually doing something useful - looking for guidance on when you can retire. A decision has to be made about when or if to retire. People using the Trinity study or the Monte Carlo planner or whatever are at least trying to make a decision that we all have to make with data. It may have flaws and one study may be better than another, but they're good starting points.

The belief that we can't know for sure so we should just "be conservative" with only a generic "we don't know the future" as a reason is a perspective that I feel should be pushed back on hard, especially on forums like this. Everybody must make their own decisions. That being said, you have deflected when asked for specific reasons for your perspective and only provided criticisms that are so generic that they could be applied to an argument for not planning for anything ever. I feel a perspective like this, compared to a perspective looking at data to try to make an informed decision, is a perspective that has little credibility.

Yeah, well shooting the messenger doesn’t work in the virtual world so…

Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)

I also think society can tolerate A LOT of people falling on bad times with no end in sight and even a large majority in that position for certain classes. But I don’t think it can tolerate a majority of the majority in this situation. So, I try to spend a smaller % of my portfolio that I think others do. Kinda like the joke about not having to outrun the bear.

Finally, I don’t think highly of the strategy to spend as little as you can—to retire as soon as you can—so you can spend as much you can. FI is the key to me, not RE. If I die with money in the bank…whatever.


All that said—I AM NOT RECOMMENDING ANY APPROACH TO YOU OR OTHERS. I do my own thing and like it. And I don’t feel as if I owe you a recommendation for living your life or saving your money.

[moof]

...
You can believe anything you choose to, but the earth isn’t flat, there is no Santa Clause, and Trinity can’t predict investment returns in the future. 

So what do we do? I don’t know; maybe be conservative, flexible, cross your fingers, and come up with a Plan B.

Indeed, nothing can predict the future.  I personally prefer historic data as I have an easier time thinking back to what actually happened to calibrate how badly stressed and panicky I would be if a particular time in history occurred again.  My few brushes with Monte Carlo sims have left me annoyed with how opaque they were, and basically blew off the results as being unhelpful.  A third option is to use a conservative flat growth number like some planners do, which is likely very helpful for the statistically challenged, but decidedly not for me. 

People have been retiring and investing far longer than I have been alive, and the ones who don't go nuts generally do OK.  Avoid leverage.  Avoid concentrated risks (single stock portfolios).  Avoid charlatans (Madoff, Bitcoin, etc).  Use a modest withdrawal rate (~3-5% depending on your circumstances).  Live modestly, and flexibly.  So on and so forth.

Rich, Broke, or Dead https://engaging-data.com/will-money-last-retire-early/ is likely the most relevant to point to for my own thinking.  We have an absolute risk of death, so as long as I am unlikely to be broke before my time is up, I'll sleep well at night with my current plan.  My portfolio can't fix the world issues, so I don't try and think that it will.

[mathlete]

I built a MC model and simulated 1,000 50 year scenarios for a variety of inputs.

Assumptions:

Growth Rate - I flexed this between 6% and 9.5%
Market Volatility - Held this constant at 15% - I think this is a reasonable / conservative number based on history
Inflation - Assumed a constant 2.5%. A more robust model would have simulated inflation with some degree of correlation with market returns
Withdrawal Rate - I flexed this between 3% and 4%

Taking the above into consideration, the smallest implied real growth in portfolio values was 3.5%. The largest was 7%. 7% is right at the 150 year average of the US stock market.

Here are the results!

Spoiler: show

With the most aggressive withdrawal assumption, the most pessimistic returns assumption, and the longest time horizon, the success rate is a coin flip. Not bad!

Any good financial planner will hopefully run something like this for you. But while we tend to hyperfocus on the market returns we'll realize in retirement, IMO, the true risk is stuff like, "you get cancer and meet your out of pocket max 3 years in a row."

A financial planner would hopefully get you thinking about things like that, too.

[Metalcat]

I built a MC model and simulated 1,000 50 year scenarios for a variety of inputs.

Assumptions:

Growth Rate - I flexed this between 6% and 9.5%
Market Volatility - Held this constant at 15% - I think this is a reasonable / conservative number based on history
Inflation - Assumed a constant 2.5%. A more robust model would have simulated inflation with some degree of correlation with market returns
Withdrawal Rate - I flexed this between 3% and 4%

Taking the above into consideration, the smallest implied real growth in portfolio values was 3.5%. The largest was 7%. 7% is right at the 150 year average of the US stock market.

Here are the results!

Spoiler: show

With the most aggressive withdrawal assumption, the most pessimistic returns assumption, and the longest time horizon, the success rate is a coin flip. Not bad!

Any good financial planner will hopefully run something like this for you. But while we tend to hyperfocus on the market returns we'll realize in retirement, IMO, the true risk is stuff like, "you get cancer and meet your out of pocket max 3 years in a row."

A financial planner would hopefully get you thinking about things like that, too.

Yup.

Literally every time the 4% rule gets debated here, I make the exact same point every single god damn time, that personal risk radically outweighs market risk.

There is SO MUCH that can happen in ones personal reality that can and will impact finances well beyond any variables that can be captured in MC models.

That said, no one with an ounce of brain power assumes that MC models actually predict the future, so arguing against that is silly.

They are fantastic models for examining the impacts of the variables that they can capture, and wonderful starting points for people to make real life decisions.

Which is, y'know, exactly how everyone here actually uses them when it comes down to it.

[By the River]

<snip>...
Rich, Broke, or Dead https://engaging-data.com/will-money-last-retire-early/ is likely the most relevant to point to for my own thinking.  We have an absolute risk of death, so as long as I am unlikely to be broke before my time is up, I'll sleep well at night with my current plan.  My portfolio can't fix the world issues, so I don't try and think that it will.

I agree that Rich, Broke, or Dead is a great tool and I've sent it to a few people who have talked finance/investing.  However, it is misleading in once sense: BROKE.   If we run out of savings, we will still have a pension (mine) and 2 social security payments each month.  The total of those 3 should basically equal average US household income.  Thus, its Rich, (Like many people), or Dead.   Its not living in a van down by the river eating grubs level broke.

[mathlete]

<snip>...
Rich, Broke, or Dead https://engaging-data.com/will-money-last-retire-early/ is likely the most relevant to point to for my own thinking.  We have an absolute risk of death, so as long as I am unlikely to be broke before my time is up, I'll sleep well at night with my current plan.  My portfolio can't fix the world issues, so I don't try and think that it will.

I agree that Rich, Broke, or Dead is a great tool and I've sent it to a few people who have talked finance/investing.  However, it is misleading in once sense: BROKE.   If we run out of savings, we will still have a pension (mine) and 2 social security payments each month.  The total of those 3 should basically equal average US household income.  Thus, its Rich, (Like many people), or Dead.   Its not living in a van down by the river eating grubs level broke.

You can adjust the "spending" input by subtracting out the annual pension / social security income you expect and it'll give you more useful results ;)

[Wolfpack Mustachian]

Here's the thing. If I get smallpox and die already FIRED, my spouse's success rate probably goes up because I'm not spending money. If the US goes through a WWII style event where we're invaded and our infrastructure is destroyed like France or we're getting air raided like Britain, I'm more worried about my families survival than FIRE. If AI is a technological game changer, I'm probably going to lose my job like millions of other Americans and arguably could be better off with money invested in index funds of companies that are even more profitable because AI has made them more efficient.

Christ, if World War III destroys the American economy, I'll no longer be retired, and I'm fine with that. I'm not building a retirement to survive every possible catastrophe.

Guys—

Criticisms like these miss the mark entirely. The issue isn’t WWIII or smallpox.

Here’s the thing: The events of the past, including the market returns realized then, are just that…the past. You can’t pick one aspect of the past—like returns—and assume it will continue into the future.

Thinking you can isolate one aspect of the past hundred years and predict it 50 more years into the future is asinine. It’s Santa and the Easter Bunny.

Beyond that is your overarching point which, if I may sum up, seems to be there's no way to guarantee or even really know, so, *shrug*. Ultimately, you are correct that we can't know for sure. However, what would you say you'd do?

i am glad you realize we can’t know future returns. Attempts like Trinity, that study the entrails of the 20th century, fail from the get-go—in their most basic assumptions. But also realize people like me, who simply call out their failure, have no responsibility whatsoever for finding a resolution.

I do not believe in a supernatural god that promulgates rules for living, yet I have a moral code I strive to live by. We do not need to force beliefs to live productively or explain what we don’t know.

I have no magic wand. I personally assume zero real growth in returns in a 60-40 portfolio and spend even less in my own retirement. (We need some kind of plan, right?) I will recommend a very conservative approach in discussions but never claim to back up such a recommendation with proof of fitness for use. I owe nothing.

We're all on an anonymous internet forum - of course you owe nothing. That being said, the easiest thing in the world is to point out obvious truths (we can't know the future) while using them to try to pick apart attempts at actually doing something useful - looking for guidance on when you can retire. A decision has to be made about when or if to retire. People using the Trinity study or the Monte Carlo planner or whatever are at least trying to make a decision that we all have to make with data. It may have flaws and one study may be better than another, but they're good starting points.

The belief that we can't know for sure so we should just "be conservative" with only a generic "we don't know the future" as a reason is a perspective that I feel should be pushed back on hard, especially on forums like this. Everybody must make their own decisions. That being said, you have deflected when asked for specific reasons for your perspective and only provided criticisms that are so generic that they could be applied to an argument for not planning for anything ever. I feel a perspective like this, compared to a perspective looking at data to try to make an informed decision, is a perspective that has little credibility.

Yeah, well shooting the messenger doesn’t work in the virtual world so…

Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)

I also think society can tolerate A LOT of people falling on bad times with no end in sight and even a large majority in that position for certain classes. But I don’t think it can tolerate a majority of the majority in this situation. So, I try to spend a smaller % of my portfolio that I think others do. Kinda like the joke about not having to outrun the bear.

Finally, I don’t think highly of the strategy to spend as little as you can—to retire as soon as you can—so you can spend as much you can. FI is the key to me, not RE. If I die with money in the bank…whatever.


All that said—I AM NOT RECOMMENDING ANY APPROACH TO YOU OR OTHERS. I do my own thing and like it. And I don’t feel as if I owe you a recommendation for living your life or saving your money.

I don't feel like I have attacked the messenger - just the message.

What gets me about your responses is that on the one hand, you clearly are, as you say focused on FI not RE. That's a completely justifiable perspective.

On the other hand, you say you're "NOT RECOMMENDING ANY APPROACH TO YOU OR OTHERS." That would be fine if it were true. However, you are and have been firmly and vehemently recommending an approach to others - recommending that they do not use or put validity in the 4% "rule". That's fine, again, if you have data to back it up rather than gut feelings and general "societal norms."

I'm an engineer by training. I am fine and completely used to people shooting down my ideas or thoughts with data to back it up. I am also used to management shooting down ideas without data and just gut feelings, and that's annoying.

Again, you are certainly and have in the past ardently expressed an opinion on how people should approach early retirement in regards to discounting a guide that many have found useful and one that has an attempt at data to back it up. I feel it's disengenous to do that so strongly, and then when asked for what you choose to do and what data you have to back up why you choose to do it, for you to jump to "I don't owe anyone anything" or the vibe of I'm not going to tell others how to live their lives. I'm glad that you put down what you actually use (a 0% real rate of return as best as I can tell from your post) - I don't recall seeing you post this before. That's something that can be discussed, and people can read your posts knowing where you're coming from. Planning for this is much more conservative than I would use, but again, thanks for providing what you use.

[maizefolk]

Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)

How are you estimating "years remaining"? Since you've expressed that you feel historical data has no use in trying to guess what the future holds, things like life expectancy tables aren't helpful, those are calculated based on the ages of people dying now and the future could be different.

Who can say what breakthroughs AI + CRISPR + (insert favorite technology here) might hold in the next several decades? Particularly with a whole generation of newly minted tech billionaires who seem to be investing greater and greater sums into longevity research each year.

Alternatively who can say what new diseases might emerge from a growing and increasingly dense population in contact with growing number of domestic animals? And with this new post-pax american era, wars and nuclear wars could certainly happen. Obviously even a small nuclear war is quite bad for life expectancy.

Your "years remaining" variable could be anything from 0 to "until the last star burns out 100 trillion years from now."

It's easy to point to the fact that the future may (and almost certainly will) be different from the past. It is surprisingly hard to go through one's life without using the past to make some baseline predictions about the future.

[Ron Scott]

Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)

How are you estimating "years remaining"? Since you've expressed that you feel historical data has no use in trying to guess what the future holds, things like life expectancy tables aren't helpful, those are calculated based on the ages of people dying now and the future could be different.

Who can say what breakthroughs AI + CRISPR + (insert favorite technology here) might hold in the next several decades? Particularly with a whole generation of newly minted tech billionaires who seem to be investing greater and greater sums into longevity research each year.

Alternatively who can say what new diseases might emerge from a growing and increasingly dense population in contact with growing number of domestic animals? And with this new post-pax american era, wars and nuclear wars could certainly happen. Obviously even a small nuclear war is quite bad for life expectancy.

Your "years remaining" variable could be anything from 0 to "until the last star burns out 100 trillion years from now."

It's easy to point to the fact that the future may (and almost certainly will) be different from the past. It is surprisingly hard to go through one's life without using the past to make some baseline predictions about the future.

Hard to do, of course…

I wing it on the upside: 95ish. Could be longer I suppose.

But if you're intimating that using decades old longevity tables to predict longevity today is a bad idea, I couldn’t agree more.

Like the Boy Scouts say: Be Prepared!

[maizefolk]

But if you're intimating that using decades old longevity tables to predict longevity today is a bad idea, I couldn’t agree more.

I'm saying the data we have on longevity today is no more or less useful for predicting any of our individual future lifespans than the stock market data we have today is for predicting any of our personal stock market returns in the future.

[Ron Scott]

But if you're intimating that using decades old longevity tables to predict longevity today is a bad idea, I couldn’t agree more.

I'm saying the data we have on longevity today is no more or less useful for predicting any of our individual future lifespans than the stock market data we have today is for predicting any of our personal stock market returns in the future.

Yeah, not following the logic on that one LOL.

[Telecaster]

Yeah, not following the logic on that one LOL.

He's saying (correct me if I'm wrong @maizefolk) that you can make a reasonable guess how long you are going to live based on simple observations, with the understanding there are and could be outlying factors (which is part of the definition of reasonable).  Similarly, you can make a reasonable guess what the stock market might do over some reasonably long period of time, with the same caveats. 

[maizefolk]

Yup, said it more clearly and directly than I could manage. Thank you, Telecaster.

[GilesMM]

...


Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)


...

For a 30 year retirement, that is 3.3% first year withdrawal which is not radically conservative. A lot of Bogleheaders like to use 3-3.5% to be safe.  Your approach would increase withdrawal rates with age, which is appealing and similar to the VPW approach favored by BHers and a great way to make sure you actually get close to spending it all instead of dying with too much.

[nereo]

...


Anyway, I think I’ve been pretty clear on this forum that my approach is more grounded in societal norms of behavior than playing number games with past data. I have no idea if I’m right—but it seems to me that government leaders would come under tremendous pressure to fix a situation in which real rates stay below 0% for significant period of time. So I base “S”WR on zero returns…or portfolio divided by years remaining…however you like to think about it. (You may think that’s a bad idea and I don’t care.)


...

For a 30 year retirement, that is 3.3% first year withdrawal which is not radically conservative. A lot of Bogleheaders like to use 3-3.5% to be safe.  Your approach would increase withdrawal rates with age, which is appealing and similar to the VPW approach favored by BHers and a great way to make sure you actually get close to spending it all instead of dying with too much.

…. I think it’s important to point out here that Bogelheads is generally viewed as being financially conservative, if not radically so. If your frame of reference is a forum that’s much more financially conservative than the mean, you can use it as evidence that the consensus there makes a plan not conservative.

Note; I’m not arguing for or against a starting 3.3% WR here. That may be perfectly reasonable or mind boggling Lu conservative depending on their specific circumstances. Just that - as a starting point - you are in deeply conservative territory to begin.

[Laserjet3051]

I've spent a lot of time generating monte carlo simulations for a 20-25 year retirement where my total portfolio balance is expected to be less than the target value I am/was aiming for.  That said, and considering all of the uncertainties of the future, I am comfortable using a 4% SWR for my retirement modeling. If it turns out that 4% isnt working, I will deploy various strategies including VWR, adapting lifestyle/expenses, and/or generating additional income streams.

One part of my retirement strategy includes minimizing the probability for needing assisted living  and/or nursing home facilities through my last days. The thinking is that if I can still deadlift 300 lb working sets at 87, I should be able to also get out of bed myself each morning without help and put a kettle to boil for tea.  This plan aims to reduce medical and assisted living/long term care costs in retirement, though admittedly is far from guaranteed.

[Metalcat]

I've spent a lot of time generating monte carlo simulations for a 20-25 year retirement where my total portfolio balance is expected to be less than the target value I am/was aiming for.  That said, and considering all of the uncertainties of the future, I am comfortable using a 4% SWR for my retirement modeling. If it turns out that 4% isnt working, I will deploy various strategies including VWR, adapting lifestyle/expenses, and/or generating additional income streams.

One part of my retirement strategy includes minimizing the probability for needing assisted living  and/or nursing home facilities through my last days. The thinking is that if I can still deadlift 300 lb working sets at 87, I should be able to also get out of bed myself each morning without help and put a kettle to boil for tea.  This plan aims to reduce medical and assisted living/long term care costs in retirement, though admittedly is far from guaranteed.

I always get a good chuckle reading how able-bodied folks talk about health and aging. It entertains me.

You're on the right track, but it's about so much more than staving off illness. Last year my mom and I went through massive rehab relearning how to walk at the same time. I had a broken femur (and I have a complex genetic condition) and my mom had a massive brain bleed and lost a 5th of her brain. Just coincidence that these happened at the same time. I was 40, she was 64.

Anyhoo, being healthy and fit doesn't guarantee anything about preventing serious, life altering, function-reducing conditions, but it sure does have a massive impact on how well you survive them.

For me, they won't even do the surgery I had on someone as old as I am, they just leave them to get more and more crippled until they have to step in with sub-optimal treatments. However, because of my extraordinarily healthy lifestyle, I was able to have major reconstructive surgery because my risk profile was so much lower than a typical person in their 40s.

My mom was projected to be stuck living in a rehab facility for 6+ months, to be reassessed at that point if she would have to be moved to assisted living or be able to go home with supports. She was instead released in a matter of weeks because she progressed so quickly and requires zero home supports.

This was because she had spent the 20 years before then religiously doing Pilates and had exceptional core strength and balance for someone her age. Lack of balance is one of the major problems after stroke damage, and a huge limiting factor to independent living, and core strength is extremely difficult to build back up if it was lacking to begin with.

There are two types of health, there's health you can control and health that you cannot control. I call the health you can control "elective health." This is what you can modulate through lifestyle. The rest, which you cannot control is "compulsory health" because you have no choice but to experience it.

The better your elective health, the more robust your system is to handle whatever compulsory health problems arise. And this is where A LOT of quality of life comes into play in senior years.

By contrast to myself and my mom, my dad had never in his life had a serious health issue except for a very minor stroke at 65 and a knee injury from sports.

However, he was a smoker for most of his adult life, an alcoholic for all of it, and has never regularly exercised. His diet would be considered very healthy if he didn't get at least half of his calories from wine. His compulsory health is exceptional, his heart is in miraculously good shape, he has none of the usual lifestyle diseases one would expect. He basically won the genetic lottery when it comes to lifestyle disease resistance.

However, because his lifestyle is so poor, he has some weakness in his lower body mechanics and has some compression of his blood vessels in his legs. This is similar to sciatica, except it's the blood vessels being compressed instead of the nerve.

The consequence is that he can't walk more than ~250 meters/yards before his legs lose all strength and he'll fall down. He can barely shuffle those 250m as it is.

There is nothing anyone can do for this aside from PT to try and strengthen his muscles and improve his mechanics. He won't do that though, he's never exercised and isn't going to start, and even then, he would be fighting a massively uphill battle to repair the weakness in his structures. It's hard to undo a lifetime of all-day-sitting damage in a senior.

So my mom who has MS and severe brain damage actually has more function that my father who has virtually no health issues whatsoever. And if when some compulsory health issues does finally hit him, his ability to recover or cope will be extremely compromised. If he is ever hospitalized, I doubt he will ever return home.

I find that able-bodied folks have this very strange binary concept of "being healthy" vs "being sick" when really, it's an infinitely complex interaction between compulsory and elective health. And frankly, most "able-bodied" folks aren't all that healthy.

In virtually every single scenario short of death, your elective health will have MASSIVE impacts on your quality of life, no matter what happens to your body along the way.

It's not *just* about preventing disease, how your body handles damage of any kind is just as, if not more important than preventing damage.

I don't mean this as a post to pick on *you* specifically, your post just printed me to want to share. I'm just sharing my perspective as someone who contemplates what "health" means a lot, and who has years and years of healthcare experience working with patients of all ages and all manifestations of "health" and "illness."

And I'm now in the process of specializing as a therapist who helps people with health issues live their best lives.

[ToughMother]

There are two types of health, there's health you can control and health that you cannot control. I call the health you can control "elective health." This is what you can modulate through lifestyle. The rest, which you cannot control is "compulsory health" because you have no choice but to experience it.

The better your elective health, the more robust your system is to handle whatever compulsory health problems arise. And this is where A LOT of quality of life comes into play in senior years.
...
And I'm now in the process of specializing as a therapist who helps people with health issues live their best lives.

@Metalcat - THIS RIGHT HERE! Thanks for this distinction between elective and compulsory health. Like you, I have more than a few compulsory health issues to experience, but I'll be using my FIRE time to kick ass on the elective health component to increase my opportunities to weather the various health challenges that come my way.

Your family examples were illustrative and looking forward to hearing more about your practice. Thanks.

[Metalcat]

There are two types of health, there's health you can control and health that you cannot control. I call the health you can control "elective health." This is what you can modulate through lifestyle. The rest, which you cannot control is "compulsory health" because you have no choice but to experience it.

The better your elective health, the more robust your system is to handle whatever compulsory health problems arise. And this is where A LOT of quality of life comes into play in senior years.
...
And I'm now in the process of specializing as a therapist who helps people with health issues live their best lives.

@Metalcat - THIS RIGHT HERE! Thanks for this distinction between elective and compulsory health. Like you, I have more than a few compulsory health issues to experience, but I'll be using my FIRE time to kick ass on the elective health component to increase my opportunities to weather the various health challenges that come my way.

Your family examples were illustrative and looking forward to hearing more about your practice. Thanks.

If you check out my journal, we talk about exactly this stuff pretty much daily.

[Ron Scott]

The health discussion is interesting. Nowadays, most older people die of heart disease or a cancer. Then comes lower respiratory disease/flu/pneumonia.  (Men have a much higher incidence of dying from unintentional injuries too.)

A few thoughts:

1. Most people know about healthy living habits that lead to longer, more satisfying lives. No smoking, weight management, healthy foods, sleep, exercise, etc. Walking the talk is harder and developing the right habits is important.

2. Ongoing health monitoring—physicals, heart tests, etc.—is critical and becoming educated about the right regimen for yourself is important.

3. While assisted living and nursing care can be a drag, alternatives to staying alone—with or without visiting health support services—can be really good. Daily socialization and activities can make a world of difference in living a satisfying vs. depressing old age. There are a number of ways to approach this, often depending on $$. My mother is in her 90s and is now in an independent living place in Long Island NY. Her total annual budget—all in including personal incidentals—is about $78,000 a year. Not cheap but no Taj Mahal either.  Run by nuns.

4. Attitude gets more important as you age. Rolling with punches = good; Cranky = not so much. Better to address this before it’s too late.

5. Don’t watch TV news, especially those who broadcast news for many hours a day, like CNN, Fox et al. Their MO is to keep you angry and afraid—which their studies show keep you watching, so they can sell toothpaste.


What else?