A crowd-sourced Financial Independence Retire Early simulator. Presented by the developer, Lauren Boland.
Hosted by the Chicago Bogleheads chapter. Recorded on April 15 2021.
Chapter meetings are listed in the Bogleheads blog calendar. You can add this calendar to your Google account. Notifications are also sent from an email subscription list. See the blog for more information.
Transcript
Bogleheads® Chapter Series - cFIREsim Demonstration
[Music]
Welcome to the Bogleheads Chapter Series. This episode was hosted by the Chicago Virtual Chapter and recorded April 15, 2021 It features Lauren Boland, the developer of cFIREsim. Bogleheads are investors who follow John Bogle’s investing philosophy for attaining financial independence. This recording is for informational purposes only and should not be construed as investment advice.
Lauren Boland: All right. Well thank you for inviting me, Jim, and whoever else was involved in that email chain. My name is Lauren. I'm the developer and creator of this site called cFIREsim. All right, so I'm the developer and creator of this sort of retirement back testing calculator called cFIREsim. Sort of a mouthful, marketing and SEO [Search engine optimization] are not my thing.
I developed this in 2013 actually as a side project, and it's sort of been this passion project of mine.I have a very loose agenda here. There are a lot of things on this calculator to go over. I'll try to hit at least every one a little bit. We can go into more depth if people have questions on specific things.
And always,one of the most important things about my site is the first word, crowdsourced. I really like to take other people's input. When I was developing this thing, the whole reason I developed it was because of other people's input. And I’ll get into that a little bit on the history. So at the end if your question isn't answered, my email is at the top of the website. You can feel free to contact me anytime. Plenty of people do.
So here's sort of a rough agenda. I'm going to go over, just sort of introduce myself a little more. I talk about how I develop this project and go over a little bit of my sort of development practices, which I won't get too much into the tech details or tech nerdery, but I have a little bit of information out there, if you're the kind of person that's interested. Then I'm going to go over basic functionality on the site and some of the more advanced features. My favorite part of the site is the different spending plans.
And then I'm going to go over some fictional scenarios to show you how someone would use it given your situation, whether you're paying for two kids college at two different times, or maybe you're still saving for a retirement, or maybe you're going to have a part-time job in retirement. All those things can be done on this tool, and it's kind of useful. Then at the end I can talk about any sort of questions.
So before I go anywhere, I guess I'll introduce myself. Again, my name is Lauren. I’m a software developer by trade. Me and my wife live in the suburbs of DC. We have two kids and I've been interested in the retire early, the FIRE movement probably since 2008 or 2009 when Your Money or Your Life was really the only resource. I was really involved in the internet when it became more of a blogging thing, with Mr. Money Mustache and various other people. And now that sort of movement is quite rampant out there, and I'm happy to see it. At some point in my life I was a moderator on Mr. Money Mustache forums; was a moderator on the Reddit financial independence site. So I’ve been around.
So a little bit of history about cFIREsim. Really this project was born out of conversations I had with people on earlyretirement.org. So for those of you who don't know--I mean I think most of you would--but earlyretirement.org is the forum website that happens to be attached to a website called FIRECalc. FIRECalc, if you look at it today, is really a sort of retirement calculator that looks like it was written in the ‘90s and I believe the data is updated yearly still. But the reason it hasn't been updated--and not to diss them--honestly because I like their project, in general. But the reason it hasn't been updated is because the person sold it. He retired and is living on a boat as far as I know, so bravo to him. But the folks at earlyretirement.org get a lot of clicks for having that site up there, so it's good.
But I'd say around 2011 or 2012, I was hanging around on those forums and it was pretty clear that there were people asking for new features on this website. No one was there to make those new features. So as a lot of people sort of screaming into the void--I was not a software developer at the time--I was trying to learn it, and I thought that hey this would be a great project.
One of the things that was useful for me is that FIRECalc outputs a csv file at the end, and it's kind of a mess, a lot like CPR sims; csv but it's essentially a data dump of every single iteration of the simulation. So I figured if I looked at this long enough I could reverse engineer it and figure out what's going on. And turns out it was right. I think it took me a month of nights and weekends of just going through the steps to figure out what was going on. How they were applying inflation and doing all the calculations. But I did reverse engineer it. And I started putting some code out there. I took a lot of input from folks on that forum, and from folks, and on credit and other places during that time.
There's a couple people who are either on this call or be or will be listening that were a big part of that. I've never actually heard his name said out loud but Jerome Moisand, is a boglehead and in your group, that was a big help back then, and he’s the user named siamond that I spoke to quite a bit, about 10 years ago. So if you're out there thank you for your help.
And anyway I put this project out there in 2013. And personal finance and FIRE is sort of my thing and I actually used it as a way to shift my career. Now I am a software developer by trade, so it's great to be paid for something that you like, but real quick, going over general development practices of how this tool works, or how I work on this tool. For any of you nerds out there I host this on a platform called digitalocean, which I believe just had an IPO recently. So go digitalocean!
I'm not an advisor. Don't take my word for that.I viewed websites written in python and a framework called django, has a MYSQL database and I use a lot of docker containers, and of course there's some javascript written in there for the form magic that you see. In the past I've written it in php. Then I rewrote it completely in AngularJS. Then I rewrote it completely in python. So it's gone through a few iterations.
One of the things I know a lot of people are concerned about is testing. How do I go about testing this? When I add a new feature, does it change the outputs? This is kind of a tricky question. Some of the time it does change the outputs. And the question is whether or not that's the intended thing. I talk to a lot of people when I make changes that actually affect the outputs. A very big example of this is--I would say sometime in 2014. I'll provide a link that we can attach to this video or the thread in the bogleheads forum. But sometime in 2014 or so, someone came up to me in the forums and said,” hey, how do you calculate the bonds on your site?”
And I gave them the formula that I've been using. It's based on the GS10 long interest rate. That's publicly available data that's gone back for 150 years. They came to me and they offered some information and we sort of hashed out the mathematical formula. Well it turns out that the FIRECalc and a lot of different sites were calculating the bond rate incorrectly. We went over it as a group and came up with a slightly different formula, which actually affected some of the success--I would say that there's probably a two or three percent difference in success rates between cFIRSsim and FIREcalc, given the amount of bonds you have in it.
So this is an interesting point in testing where it's just going to be different. I do have lots of unit tests. And for those who are not software developers, unit tests are pre-built tests that have predetermined outcomes,and every time you make a change in your software you can just run them immediately and see what happens. If those predetermined outcomes are different it will let you know. So I do keep an eye on that.
Honestly the biggest thing that I rely on is user testing. I would say almost 95% of the bugs that happen on my site are just UI bugs. I was mentioning to Jim before everyone came on that javascript is the programming language that lets you kind of mess around with the UI. That's not my day job. I'm not used to doing anything with javascript, so sometimes little UI bugs get in there.
And lastly, I've mentioned this before but I always take advice from users. I'm always looking for new features or quality of life things. I don't want to make this thing too complicated in the end. so I reserve the right to shoot things down, but over the years I've added a lot to the site based on suggestions. And it's been great.
All right. That is basically some of the back end history of this. and I'm going to go through some of the basic functionality on the site and see if any questions arise. So I'm going to refresh this. Just in case, when you arrive here on the site you're presented with what I call the giant form.There's a lot of inputs here. You can open it up--depending on what you select, for even more inputs--but there's this basics section that I'll go over in the beginning.
A little bit of the back end. I go off of Robert Shiller's data set which is pretty common. Robert Shiller's data set has sort of like the S&P 500 equivalent stock index for the last 170 years, and also has the bond rates. And I also have a separate data source for gold. One thing that I would say is lacking in this program is historical cash data. I've never been able to find good historical cash data that goes back all the way to 1871. There's quite a few that go back to 1900 and I've never bothered to jam those together. That being said, I offer the ability for you to put in how much cash, you think how much growth you think cash will have in your portfolio. Maybe this is a lesser ten to two feature because I don't have a lot of cash
But, so the basics here are, I would say, pretty common for a lot of retirement calculators. You have a retirement year. You have a retirement end year, better known as your mortality. You have a check here for what kind of data you want to use. I almost always use all historical data, but some people like to really nitpick on the worst periods of history and see if their portfolio would have survived. Which is useful but I mostly like to see everything.
And then have your portfolio value as it stands now, which is important. And then what your yearly spending is now in today's dollars. So I'm not going to assume people's knowledge here, but everything on this page is essentially in today's dollars, which means you don't want to inflate these numbers for the future. If you're going to retire in 30 years and right now you're spending $40,000 a year, sure maybe in 30 years that's going to be $70,000. Don't put that in this calculator, everything is in today's dollars. I personally find it easier to wrap my brain around when everything's in today's dollars. And everything does get adjusted by inflation as you'll see on the output.
So initial spending, spending plan, is something I'll go over later. There's quite a bit in here, quite a bit of complexity in there. But right now this is inflation-adjusted spending which will adjust this initial value up and down based on historical inflation, which is a pretty common feature. And I think some of these calculators-- here's sort of a drop down for what kind of inflation type you have--so this is CPI. This is historical inflation. You can do your own flat rate inflation if you really want to. Some people like to use this for bad scenarios where they're, like oh maybe inflation's going to be 4% for the rest of eternity, or maybe it's going to be 2%. I'm not sure, but I tend to use historical myself.
Those are the basic things that you need to run this. In general everything else below here is adding details to make it more of a fine-grained simulation. So like I mentioned before about cash, here's your portfolio area. Important to note that this is essentially the portfolio that's going to be throughout your simulation. Now you may ask, I'm going to change my portfolio. Maybe I'm going to get more aggressive, or maybe I'm going to get more conservative as I get closer to retirement. You can do that. I do have the functionality in there to provide what's called a glide path, which takes your allocation at one point and moves it slowly to another point over a certain amount of years.
So the default here is 75% equities/ 25% bonds. I think this is a pretty standard default for a lot of retirement calculators. If you want to do those glide path type things, you would uncheck this keeping the allocation constant. It will reveal a new panel here. And so in this particular situation, this person is retiring in 2021, which is this year. There are some financial gurus that seem to think you if you start off in a conservative portfolio and slowly move to a more aggressive one during retirement, that that's a viable thing. So I wrote in here, we're starting off with a 50/50 portfolio. And here is the start year of the glide path and here's the end year of the glide path. So over a 10-year span it's going slowly ratchet it up to a 90/10 portfolio.
I would say one of my personal to-do lists for features would be to add the ability to do more than one of these. I think it's pretty common for people to change their allocation throughout their investment careers, multiple times. So I wish I could do this twice or three times but right now you have the ability to do it once.
For simplicity's sake I'm going to hide this for now. But this can be pretty useful in trying to model different risk scenarios. So I'm going to click this button and make that thing disappear. You'll notice that a lot of things on this website appear and disappear based on your inputs. Hopefully that's intuitive. I haven't found a lot of people who have a problem with that. So hopefully it's not too bad.
Rebalance annually is an interesting thing. I don't know your own personal views, but I think most folks that are interested in planning out their own retirement are rebalancing annually or twice annually. I mean maybe people are doing it every quarter. I don't know. All this means is it's rebalancing your funds back to this portfolio every year. Based on the simulation, if you don't do that, it'll just let the simulation drift. It'll let your equities go up and up and up; or your bonds go up and up and up during certain times in history, and it'll never rebalance. That's certainly an interesting thing to simulate. You can try it out on your own and see the effects of that.
And lastly, one thing I haven't mentioned here is fees. This is essentially the average fees across your entire portfolio. I know that that's not ideal. People have different fees and different accounts because they have different funds in them. For the purpose of simplicity I put this in here as just an overall average on your portfolio. This is some sort of average of Vanguard funds that I found out there. So I put it in there.
All right. Below portfolio we start getting into some of the inflows and outflows that you can add to your simulation. And starting with the most popular, at least in the United States--here I do have several people from Australia and the UK that visit this site and make use of other adjustment types. But since I'm in the US and I will be collecting Social Security, hopefully, I added this.
So we have two different Social Security fields: one for yourself, one for your spouse. By popular request I’ve made it so that you can change this from a monthly to an annual number. I used to have everything annually, but most people think in monthly terms when it comes to Social Security. How much am I getting each month? So whichever you put in there it will do the right thing on the back end of the simulation.
Of course you can model different start years. I know lots of people try to figure out whether it'd be worth it to take Social Security early or take it at the full retirement age. You can certainly do different simulations based on the start year and of course there--I used to not have an end year in here. I used to just assume hey this is going to be you're collecting this for the rest of your life-- but lots of people want to model their spouse's deaths so I added an engine in here and you can model different people's deaths and changing of values here.
So this is essentially one example of what I call an adjustment. It's a general term for the inflows and outflows of money in your portfolio each year. In the simulation, a pretty common theme and I'll show you right here, this is the generic, the general adjustment section. So you can put up to, I believe I've allowed, up to 10 adjustments in here. There's no reason I couldn't increase it, I just have kept it at 10. But this is essentially where you would put your pensions. Where you would put different spending things that would happen during retirement. Maybe you're going to retire and then you're going to start paying for your kids' college.
So my kids are going to be in college in 2030, and then I've got another kid in college 2033. So I would model a couple different spending things. Good old in-state tuition in Virginia right now is roughly $15,000 a year. So this is a general thing that you can do with these adjustments. I have pension availability in here, any sort of income and savings which could be anything. It could be you've got a part-time job, you're renting a house out and you're a landlord,so this is your rent.
You can also make this a one-time thing. So imagine you're going to sell your house. You can uncheck this recurring box which will sort of gray out the end year. And this particular adjustment will only happen in 2030. So these kinds of things are pretty useful. Some of the inputs in here might look familiar, you know amount per year, that's pretty simple.
This label here I find useful mostly because there's an ability to come back to your simulations later. So real quick I'm just going to run this and ignore the output, because I can click this link and it will reload the exact same things, exact same inputs, that you entered before. I'll talk a little bit more about that when I'm on the output page. I apologize for jumping back and forth. So this label is important.
This is also important. Some people have been known to mess up and put the wrong thing in here. Whether they meant spending and they put savings, and they're like why do I have so much money at retirement. Well you said your kids college was actually a savings or income stream instead of a spending stream. So that's an interesting thing that you need to look out for.
And one sort of option in here that might be a little different for you is this value here. I wasn't really aware of this. This is where crowdsourcing comes in. I wasn't really aware of this because I don't have a government pension. I know that there are other--my wife has a private pension--and actually I think this applies to it, but I don't pay much attention to her pension and stuff.But this little check box essentially is for pensions.
So a lot of pensions will have this set up where they'll tell you hey you're going to get $5,000 a year once you start collecting your pension. That value is often frozen and it's always going to be $5,000 until you start collecting, and then only after that will it be adjusted for inflation. So if you have a pension that's like that you'll want to click this button and have this clicked inflation adjusted and it will stay at $5,000 until you start collecting it.
Now that may seem intuitive but anything on here that you select inflation adjusted for, that's in the future. For instance, so I just unchecked this and this is nine years in the future. What's going to happen is even though you're not taking this pension until 2030, this adjusts that value for inflation every single year. So nine years down the road this might suddenly be $7,000. That wouldn't normally be a problem for a lot of things, but like I said, pensions have this knack for being frozen in time until you start actually taking them. So this is a worthwhile thing.
I forgot to mention at the very beginning of this--some of the things at the top of my site. Fairly recently I've started writing tutorials that describe some of this functionality. So if you miss something on here, and you don't want to go back through the video, you can certainly check out this tutorial section. I'm going to click here real quick right now. I only have information about the basics section.
And yeah, what the output looks like. But I definitely have--I've already written articles for spending plans--I just haven't bothered, I haven't gotten them up yet. And I definitely am going to talk more about some of these things. I'm going to quickly scan over essentially the information that I'm going to put in these in this demo. So hopefully that's helpful.
And go back here. Okay, so adjustments. Recurrent. The recurrence button I mentioned, does this occur over a span of years or is it just once. And then whether or not this is inflation adjusted or not is pretty straightforward. Just like above you can choose the inflation type. It can either be the historical CPI or it can be a flat rate. I actually know that a lot of pensions have a flat rate adjustment so that can be pretty useful.
And let's see, I believe that is essentially a quick scan through of all available inputs without touching the spending plan. So the spending plan has lots of different things that show up. After you click the appropriate spending plan, and I'm going to go over that in just a moment.
Firstly though, I'm going to refresh this page, have the default scenario, and we're going to talk about the output page. So this is a pretty standard situation. In fact, I'm pretty sure this is the exact same inputs that FIREcalc does.
This is a 30-year scenario. It may look like it's 29 years but it counts this year, and if it's 30 years with a million dollars and $40,000 spending per year this is the classic 4% rule. I'm going to hit enter or hit this button up here and this lovely rainbow graph is what you would see when you hit run simulation.
There's a lot going on on this page. I've tried really hard on this site to put as much information out there as I can without making it too confusing. However I do understand that because I'm the one who wrote this, I know where everything is, and it might look a little wild to someone who hasn't checked it out. So I'm going to go over a couple of the sections here.
So the first thing you're going to see is this really big wild rainbow graph. So what this is is essentially every single simulation of a 30-year span over the course of history starting in 1871. and what I mean by that is-- let me see if I can, I think I wrote a pretty nice graphic of this--I don't know I got rid of the graphic, that's a bummer.
So what you have to imagine is this data goes back to 1871. So it's going to take your inputs and it's going to go to 1871 and put them in as a theoretical million dollars in a theoretical $40,000 spending and it's going to run the data from 1871 to 1901 and it's going to see how your portfolio would have done. Then it's going to start a new simulation in 1872 and go to 1902, and so on and so forth until you get all the way to the 2020 data, which if I try real hard I can find it in here.
But this results in sort of this cool graph. And you can kind of get a sense of what's going on here. There are periods, if you hover over this it sort of hovers on an entire simulation. So right here, it's kind of hard to see in the blues, but that blue line dips down really far. This was not a good period in time. In fact this is the classic beginning of the stagflation period in the late ‘60s and early ‘70s.
So what this represents is your portfolio over time. Now like I mentioned before this is today's dollars so it's adjusted to make it easier on your brain to figure out what this is worth. So if you started with a million dollars, this is worth roughly six times as much. However in nominal dollars this is probably way more. I'm just going to make a speculative guess and say this is like 11 million dollars, you know double that or more in nominal dollars.
Now how do you figure that out? You could go into the csv up here. I guess I will open this up-- this is sort of a stream of consciousness--so if you click on this button you're going to get this long named csv file. I don't even have a word on here. I don't even have Excel on here. Where'd my screen go? All right it's going to load it up, and this is not going to be very easy to show people, but this is a very long file and it has every year of every 30-year period in this simulation. So if you go to the top, it says the cycle start year is 1871. So this is the 30 years starting in 1871. And it has your starting portfolio. It has the adjust inflation adjusted portfolio. Has the spending and the inflation adjusted to spending.
So for this particular one, which is the one I think I was highlighting, if we look at the ending portfolio, and this is just a quick tangent ending portfolio. Right here, what was the ending portfolio of this one? 3.2 million, and the inflation adjusted was higher. Gotcha. Anyway this is a bit of a tangent, but good to know that you can look at all the numbers there.
This here is a big ugly link that will allow you to get back to your same inputs. A lot of people ask me why I don't have a login system so that you can save your data. I have taken a hard stance on that because I don't want to deal with keeping the data anonymized. To be honest, right now as it stands I have a database full of people's inputs. I have no idea who put them in and I'm okay with that. I'm not in the business of selling data. I don't make any money from this site. This is a 100% passion project and I want to keep it that way. So I anonymize the data, and I put it into here as a link, so you can get back to it rather than have a login system that's tied to your email. Maybe that makes you feel good, maybe that makes you feel like this tool is clunky. I'm not sure.
But okay. Back to this chart real quick. As I'm scrolling through here, if you look below down here--I can't point to it while being on the screen--but if you look below it says that the year is 1953. The cycle start year of this particular simulation was 1935, and at this point in time the portfolio is 1.3 million. So as you go through here you can get a good idea of what happened in history.
Very similar to this chart is right next to it, the inflation-adjusted spending over all the simulations. Now a lot of people get confused when they first see this, and they say, why is it a flat line? Well, because you chose $40,000 as a spending point and because you chose it to be just straight up inflation adjusted. It's $40,000 every single year. And it sees that there. You see it there.
If we choose other spending plans which are variable, which we will go over in a bit, they definitely go wild over here. And we'll talk about that. Okay, there are a lot of numbers here, and without getting too verbose I'm going to talk about the most important ones and then let you figure it out after that.
I think the most important thing is this little section here right below the portfolio chart.This is the success rate. Success on this tool is defined as you still having any money in your portfolio when the simulation is over, or when you're gone. You can have one dollar in your account and it will be a success. If you look on this chart--there are a couple of these--this is the very last year of this portfolio. It started off with a million dollars. It's down to $50,000, but that person made it to the end. That's a success.
So out of a possible 121 different 30-year periods in history, this particular setup would have only failed five times. This is essentially the definition of the 4% rule that is often cited. The 4% rule is not 100% successful. It's 95% successful, which for most people is good enough.
I'm not going to tell you what success rate is desirable. You can look and read a lot about that topic. One of the recent papers that sort of has blown my mind recently is Michael Kitces, a financial planner, has written a paper that essentially says that if you do a monte carlo simulation, which is different than this--this is just historical data--if you do a monte carlo simulation, which is randomized data, and you have a 50% success rate for retiring. That if you did that every single year, you'd be fine, as long as it was 50% every single year. Now, do I feel comfortable with 50%? Probably not. But that's up to you to decide. Some people need this to be 100%. It's totally how your brain works.
This number here is what your portfolio was at retirement for this particular scenario. Because I retired right now in this scenario it's just exactly what you put in as an input for your portfolio. Beyond that, this entire left side is all talking about portfolios. This entire right side is talking about your spending.
I have all sorts of statistical numbers in here. I’ve got the average ending portfolio. I’ve got the yearly with average yearly withdrawals. I’ve got the average total withdrawals. I've got the median, the standard deviation, which is sort of the volatility. I've got the highest value, the lowest. And then by popular demand I put in the lowest 10% , in the lowest 5 %, just to see what the slices of the statistical pie are.
Here for these particular simulations, you'll notice that a lot of these numbers are the same. That's not a bug. When you have very flat data it tends to end up like that. Like I said before, the spending is $40,000 flat every year, inflation adjusted. So these numbers are all going to be the same.
Just like over here. I'm not going to go over this entire chart for spending, but I will say that I've split it up into quarters. And the first five years, these are both statistics that are very often cited in financial planning articles, or people, even on the bogleheads forums, will say the first five years are the most dangerous part of retiring.
So I’ve included statistics on spending for these different time periods. Now someone asked me recently, do these quarters mean only the retirement period. And the answer is yes. If I wrote that I was retiring in 2031--so 10 years of pre-retirement simulation--this information would only be including the retirement period. I’ll let you look at some of these later. I don't want to get too verbose on this page.
The last thing I’ll talk about is this. I hide this a little bit because it's long and most people don't want to see it. But if you click this “show” button it brings up an extra chart. Now these are the single years where the portfolio was the lowest out of all the simulations. It tells you what the portfolio was now if you look at five of these failed. So of course we have five zeros.
But look at this guy. I didn't see it before, but in a starting year of 1967, in year 28 it got down to $4,000. Actually this makes sense because right here, starting here in 1967, the very next year it goes to zero. These are just individual data points. People like to see how low it goes before failure. If you happen to have a simulation that has 100% success it's still interesting to look at how low it has gone in the past. So keep this little section in mind if you're the kind of person that likes to look at the very edge cases of things.I don't usually look at this very often but to each their own.
Okay I've gone over this. Another basic functionality to note is if you look up here I have these two little tabs. When you submit the form it doesn't refresh the page. It opens a little tab here, and puts the information here. If I click back onto the inputs tab I see my same inputs that I had before. This is just the regular default data.
Now what you can do, is you can do things like this. The retirement end year is 2050 in this case. What if it was 2055? I'm going to leave everything the same, but what if it was a longer period of time. What if I think I'm going to live till I’m 95.
If you notice up at the top here it loaded a new tab called Sim2. This shows the statistics for what I just put in. It failed another extra couple times because it's trying to eke out another five years of retirement. But most importantly what I wanted to show you is that you can click between these tabs and essentially have a bunch of different simulations out at once whichever tab you were last on.
When you click on the inputs tab it will load the inputs for that simulation. It will tell you that it loaded it. And if you look, this is the input that I used for Sim tab 2. If I go back to 1 and then go back to the inputs, it will load it from Sim tab 1, which was 2050.
It does that for everything on this page so it's kind of useful if you're iterating through and trying to find a very specific time that you know how many years you can retire or things like that. So this is a very useful feature. I personally like it. You can also hit this little red X, which I hope is intuitive, and it will make it go away.
So let's see. One last thing. I guess I missed this. So I haven't gotten to do a lot of these yet, but if you click this button, open investigation options, it opens this little extra spot down here.Once upon a time this was a very popular thing on the site. And I can't remember if FIREcalc does something similar, but essentially what this does, if you click this little slide thing it's going to look for the maximum initial yearly spending for this amount of success. This is actually a typo. I should just delete this, and just say for this amount of success.
So if you type in 95 here it's going to run the simulation many times to figure out what is the maximum number of spending here to get 95% success. Oh I unchecked it right before I hit enter. Let's go back. This one takes a little bit longer because it actually runs multiple scenarios in a row, so I would not worry too much about it spinning.
Okay this looks a lot like the other simulations. But down here I write this little note, just says investigation was complete. Then it tells you the maximum initial spending is $40,507 exactly and it results in 95.04%. I would say that this is accurate within ten dollars. Usually. just the way that I'm iterating through the thing, it makes it really inefficient to find it down to the one dollar. I'd rather it spit out something very close.
So if you were to put this amount back in the inputs tab, it would give you the same. It should give you the same amount of success. Boom, 95.04% perfect! Really hoping I can make it through this demo without displaying a bug in real time.
Okay,all right. I've gone over most of the basic stuff, the spending plans I've yet to go over. Jim or whoever else, Keith, I don't know if we want to take any questions now. All right, we're recording. All right, spending plans. Spending plans are my favorite part of this tool. Honestly this is the thing that I mess with the most in terms of my own personal situation, and I think that it's a pretty undervalued thing.
I am not--I'm going to say this a lot--I work for a financial firm and this is not part of their work. And I know all the legal disclaimers. I am not a financial planner.
However I know that people do not spend the exact amount of money or same amount of money every year and modeling that is unrealistic. Ideally people are vacillating between a spending high and a spending low and they know roughly what their average spending is. We can do that with these different variable spending plans. VPW [Variable Percentage Withdrawal] was just asked about in the question and answer section. So I'm going to go over that one first.
So VPW is essentially a spending plan, or a withdrawal plan if you will, that they came up with on the bogleheads forums. I can't remember the username off the top of my head, I want to say it was longinvest. Anyway one of the users came up with this and the premise is pretty simple. Well the premise is simple but the calculations are complicated.
The premise is that this user would look at the standard and inflation adjusted simulations and they would see that a person who had a million dollars and spent roughly $40 000 a year would end up with a portfolio of six million dollars at the end. And they thought you know you can't take it with you, I want to spend all the money. How can I do that in a safe way?
So he wanted to find a safer way to get as close to zero at the end of life as possible. What he did was take a combination of writing his own annuity each year based on the end point of your simulation--excuse me, I'm talking too much--so he wrote out a sort of annuity formula--and then based it on whatever future value you wanted.
So these two values are new values when you select VPW. And this is sort of a rate of return that the aforementioned annuity is based on. And this future value is essentially what you want your portfolio to be at the very end. Now this number is largely ignored in the VPW. I believe it has some significance to the starting point, but it very quickly gets changed along the way.
So I'm going to, just in case I change some other things, I'm going to refresh this real quick to the defaults then I'm going to switch it to VPW and then this is the default stuff. I'm just going to run the simulation and see what happens. So you'll see that it says 100% success. Now that means that the portfolios were above zero when the person died.
Now if you look at this chart on the left, it looks a lot different than the other ones. If you look everything is spiraling down to zero, but it falls just short. So the math behind this allows you to spend it down so that you essentially end up with less money at the end of retirement.
Now if you look on the right here, this is the first time you've seen the spending graph become a variable graph. Now if you look here this might be unfathomable. This is a million dollar portfolio and at some points spending is getting to be $200.000 in a year. How can that be? Well it follows the market gains for a given year, and every single year it's calculating based on the current portfolio and how many years you have left to live, and the expected return rate.
What it thinks you're going to have at the end, so it keeps increasing that. Well it really just varies it up and down. So if you look here it goes up and down quite a bit.
But if you look at the average spending for this entire simulation it is quite a bit above 4%. But if you look on the left side of this chart the average ending portfolio is $85,000. This is as close to zero as we can get without failing before your end of life. So this is an interesting example of something crowdsourced.
You know someone on the Bogleheads forum was writing this sort of thesis, and you can find it out there, it's still a very, I think it's a very trafficked thread. But that person came to me in 2015 or 2014 and asked if I could make this happen as part of the website. And I did so. It's kind of an interesting way of thinking about spending, if you're willing to spend it all down.
And before I move on to the next one, one thing to note is you can change the future value if you're planning on spending more than normal but you want more than just zero left at the end. Maybe you want to leave $200,000 to your kids or something. You can change this future value and it tries real hard to make things happen at $200,000.
Actually it seems to not have done that. I told you I was trying to make it through the entire video without a bug, but here we go. I'll have to check that out. So if you look at some of these portfolios, they actually dip down to zero--it looks like they're failing right before the end. So that's interesting.
Anyway we'll move on to the next one. All right, so when you have inflation adjusted, of course, this is not the inflation adjusted scenario. Essentially it takes your spending and never changes it. That means over time your actual value of your money is going to go down because inflation is going to be reducing the value of it. So if you run this really quick you'll see on the right side here that inflation really changes things.
Now what's interesting here is in the early 1800s there's quite a bit of deflation. So even though you're not adjusting your spending here, your inflation adjusted value of that spending is going up. But the vast majority of modern times your the value of that money is going down quite a bit. If you look at the averages here, really you're averaging 30, you're averaging lower than $40,000 per year if you never adjust it, and I don't know if anyone's--I can't remember from our poll--who frequents earlyretirement.org, but it's a common trope on early retirement that you don't want to be eating cat food in retirement. So if your spending reduces to half of what it started you might be in trouble.
Okay. Let me go back to what's next. Percent of portfolio. Okay this is a pretty simple one on its surface but there's some nuances that a lot of people miss. So percent of portfolio pops up, one new field called the yearly spending, which is a percentage of the portfolio for that given year.So when we talk about the 4%rule--If you're not familiar, which I hope you are if you're here-- but if you're not familiar, for people out there the 4% rule is based on a study where they had portfolios of people at retirement and starting at retirement, their starting spending was 4% of that portfolio value. Then from then on out, no matter what the portfolio did, you just only adjusted your spending for inflation. So every year wasn't 4%, it was just the very beginning.
This particular spending plan, percent of portfolio every single year, it's going to take your portfolio value and then adjust your spending based on that. If you start off with a million dollar portfolio and somehow it gets to be two million dollars, well 4% of two million dollars is $80,000 and that's what it's going to set your spending to. So if you run this, you'll see on the right that the spending is wildly varied. If you would have retired in the ‘80s, your spending would have gotten up to $157,000 by using this method.
Now this says here that it's a 100% success rate. Now that is true. And that is true because no matter how high or how low your portfolio gets it's going to only take out 4%. So if you look here, the spending in this year was only $15,000. That's because the portfolio dropped quite a bit. By definition, if you're only ever taking 4% of the portfolio, it's almost never going to go to zero.
So how do we simulate these in a more realistic manner? If we go back here--I'm going to introduce this other section of the variable spending that is very important--it's a concept of a spending floor and a spending ceiling. So well, I'm going to go back to the tab real quick. If we look here we're trying to spend 4% of our portfolio. Now if you're trying to stay within the realm of $40,000 inflation-adjusted spending, you know you look at this chart and you say,”Wow, it's really, I mean it doubles or triples sometimes, and it goes down to you know major, I'm eating cat food level. How do we fight against that?”
We do that with spending floors and spending ceilings. So the idea of a spending floor--and I'm going to choose a defined spending floor value--is that the simulation will never allow the spending to go below the floor. So in this case, this is a 4% rule. I personally think like, hey I can take a 10% haircut on my spending and be okay. So I'm going to put the floor to be $36,000.
Now if we just run this as is, without doing the ceiling, if you look on the right you'll see some very distinct flat bottoms. It never goes below the $36,000 that we set and that's great. However it also goes out of control in these other places. Now maybe you want it to go out of control. Maybe you want to set a floor and when your portfolio goes crazy you can go crazy too. That's great.
Some people are trying to pay more attention to their average ending portfolio, to make sure that they can pass money along for some sort of generational wealth. So if we go back to the inputs again, what we can do is we can rein in the spending by setting a ceiling. Now like the floor, you might have guessed the ceiling makes it so the spending will never go above a certain amount. So let's say we have a $40,000 target. Let's go crazy and say if in good years we can spend up to $60,000. That'd be great.
And we're going to run that. Now this might look a little wilder because this is to scale, so when the values went up to $157,000 they looked a lot more squished. But now they're within $24,000 of each other. So if you look there's some very distinct flat tops and flat bottoms. It's bouncing around in between the spending floor and the spending ceiling.
I love this feature. I think that it's one of the more realistic things that happens in retirement. Most people will have this idea in their head of what they want to spend, and they'll have a belt tightening amount in their head that they don't want to go below. and then in good years maybe they'll spend more. This allows you to do that. I really like this feature.
So that's what that looks like. One of the things you can do is set the floor to be the value of your adjustments. So if your retirement is dependent on a lot of Social Security and pensions, and say this person has $30,000 worth of pensions and Social Security. If you can set the floor to be that, then it's a quick and easy way to deal with the floor. I don't have that same ability for the ceiling because I think it's kind of silly to be able to select your adjustments as the ceiling but maybe someone can convince me otherwise.
Wow! It's 9:30 Eastern already.
[Music]