The Safe Withdrawal Rate Series

As suggested by several readers now, here’s a new “landing page” for everyone interested in my Safe Withdrawal Rate Series, which has now grown to 50+ posts. What is a reader new to this topic supposed to do? Should you read from the beginning to the end? That seems intimidating! It almost feels like there will be a quiz at the end of the series. No worries, there isn’t! 🙂

But seriously, just because I wrote this comprehensive series in this order doesn’t mean that one should read it in that order. Keep in mind that this whole series was a learning experience for me as well, so reading parts 1 through the end will seem a bit jumpy at times. In fact, a lot of the posts came about because readers suggested I check out certain sub-topics and all that came about in random order. If I were to write this series again today, from scratch, with everything I know today, or if I were to write a book I would obviously structure this very differently. It would be unfair to new readers to make them go through the entire series! This convinced me to write this new page, so newcomers feel less intimidated starting their own thought process on Safe Withdrawal Rates!

How do I even get started with the Safe Withdrawal Rate Series?

Great question! If you’re completely new to FIRE, maybe start with The Basics of FIRE as an intro to the concept of Financial Independence and Early Retirement.

If you have already graduated from the FIRE basics, you will likely look for insights on Safe Withdrawal Strategies. When I was at that point, I wrote Part 1 – Introduction: This post sets the stage and explains the similarities and differences between my study and the Trinity Study. I replicate some of the Trinity Study results but also generalize the results to make them applicable to the FI/FIRE crowd; most importantly, I look at longer retirement horizons. If you have already read the Trinity Study, you are already skeptical about the simple naive 4% Rule and you like technical stuff, then sure, start with this one. But to make this series more accessible, palatable, and enjoyable to a wider audience, I also wrote some high-level posts without much technical and math mumbo jumbo. Maybe you read those first to get an appetite:

  • Without diving into any math, you might enjoy these two as your first post: Part 26 – Ten Things the Makers of the 4% Rule Don’t Want you to Know and Part 50 – Ten things the “Makers” of FIRE don’t want you to know. As the titles suggest, these are tongue-in-cheek posts where I debunk some of the myths surrounding FIRE and the 4% Rule.
  • If you attended the April 2018 CampFI in Virginia where I was one of the presenters, you’ll recognize a lot of the material in Part 27 – Why is Retirement Harder than Saving for Retirement?: Another more philosophical, not-so-technical post, perfectly suited for folks trying to get started. Probably you’ve heard about MMM’s Shockingly Simple Math and JL Collins’ Simple Path to Wealth. But the simplicity of accumulating assets can’t be so easily extrapolated into the withdrawal stage. That phase is much more complicated and the reason why I have written 30+ parts on this topic!
  • Part 32 – You are a Pension Fund of One (or Two). Compare and contrast our personal challenges to those of large corporate and public pension funds. It turns out that I found many aspects of running my own (early) retirement finances, essentially a pension fund with two beneficiaries, that are much more complex than running a pension fund for thousands of beneficiaries!
  • New to the site here? Do you think that rigor and math don’t belong in retirement planning? Why not just wing it? In Part 46, I detail my thought process and why I like to perform my careful, customized, and rigorous retirement withdrawal simulations, not despite but exactly because we face so many uncertainties in retirement. But we don’t have to overdo the precision either! In the follow-up post, Part 47, I go through a few case studies and show when it’s OK to wing it and when it’s time to worry about those nitty-gritty details of the withdrawal rate analysis.
The Safe Withdrawal Rate Series

OK, after the intro, what else would you want to read? Maybe you found interesting sub-topics and relevant links already when reading through Part 26 or Part 50. But I also present all the remaining posts as they fall into several major themes. And don’t worry, I also give a few suggestions on what to read when at what stage in your FIRE journey, later in the summary below…

Introduction to Sequence of Return Risk

Why are we even worried about running out of money in retirement at all? Over the long haul, stocks should return much more than 4%. And a retirement horizon of 30 years for a traditional retiree and 50+ years for an early retiree certainly qualifies as long-run, right? Well, not so fast! What I learned from my historical simulations is that you can still run out of money even if your 30-year or even 50+-year average return was way above your withdrawal rate. If returns were bad enough initially and you keep withdrawing through the bear market there is a chance you deplete your portfolio so severely that even the subsequent bull market and recovery will not save you. The sequence of returns matters. Low returns early on are poison to your retirement finances. It’s like the opposite of Dollar-Cost Averaging; you sell more shares when prices are down!

So, no series on Safe Withdrawal Rates would be complete without at least some treatment of Sequence of Return Risk (sometimes also called “Sequence Risk”). That’s what I do in Part 14 and Part 15. I go through some numerical examples, both with “made up” return data and actual historical simulations to showcase how bad Sequence Risk can be. None of this is required reading for a successful (early) retirement but it’s a nice reference to have!

A question I keep getting: Is there ever a time when we can stop worrying about Sequence Risk? Unfortunately, Sequence Risk will be with you for longer than the often-quoted 5-10 years. See Part 38 for details.

One “solution” I hinted at in 2017: If a saver and retiree could team up so that their aggregate cash flows look like a Buy-and-Hold investor, we could certainly eliminate Sequence Risk. See Part 53.

Also, you might check out my appearance on the ChooseFI podcast. It was my first podcast appearance ever and I talk with Brad and Jonathan about Sequence Risk and the pitfalls of the 4% Rule. Also, check out my podcast appearance on Hack Your Wealth, talking about early retirement, the sequence of returns, and FIRE. Part 1 and Part 2. Also available as a Youtube video! All really low-tech with no math degree required! 🙂

The effect of Supplemental Cash Flows (Social Security, Pensions, etc.) on Safe Withdrawal Rates

This is one of the key reasons why Safe Withdrawal Rate Rules are really only Rules of Thumb! We each need a personalized financial analysis to determine our safe withdrawal rate. A 30-year-old early retiree with relatively modest expected Social Security benefits many decades in the future has a lower safe withdrawal rate than a 50-year-old early retiree expecting generous Social Security benefits in only a little bit over a decade. How much of a difference does that all make? Well, I studied this in great detail, once in a really early post, Part 4, and then again in Part 17.

You can also use a tool I provide in Part 56 to evaluate the net present value (NPV), internal rate of return (IRR), and crossover (to profit) points of annuities, pensions, and Social Security strategies. I do this all from an actuarial point of view, independent of the withdrawal rate analysis.

Why the Trinity Study success/failure probabilities don’t easily apply to you and me

My main “beef” with the Trinity Study? Even if your personal financial situation exactly fits the model assumptions of the Trinity Study – a 30-year horizon and no additional cash flows – I’d not take the Trinity Study success probabilities very seriously, for the following reasons:

  • Part 2 – Capital Preservation vs. Capital Depletion: Do you want to leave a bequest? Do you want to leave your entire principal intact instead of exhausting your nest egg? Then the 4% Rule may no longer be sustainable because it was designed with capital depletion in mind. This post looks into how much of a haircut it takes to make the 4% Rule water-tight again when you have a more aggressive final net worth target than $0.00! This post is also relevant for everyone planning a longer horizon than 30 years. Remember, the Trinity Study would consider a final net worth of $0.01 after 30 years a success. Not very useful if you have a 50-year horizon!
  • Part 3: Equity Valuation: After a 10-year-long equity bull market and high equity valuations (and low bond yields), you will face a higher failure probability with the 4% Rule than the Trinity Study may suggest. But does that mean that I can’t retire at all? Do I have to wait until equity valuations “normalize” again? Not at all, I show that with a withdrawal rate in the 3.25-3.50% range, you would have survived even during the most catastrophic historical market conditions!
  • Part 6 – A 2000-2016 case study (Welcome to the Potemkin Retirement Village): Maybe you’ve heard/read that the year 2000 retirement cohort is doing just fine even in light of the deep bear markets in 2001-2003 and 2007-2009. That’s not exactly true and I got the numbers to prove it! 
  • Part 22 – Can the “Simple Math” make retirement more difficult?: This is a post very relevant to folks in the FIRE community. If we all pull the plug and (early-)retire the minute we reach a certain savings target, e.g., 25x annual expenses we could potentially face a much higher failure rate than what’s quoted in the Trinity Study. The Trinity Study calculates the unconditional probability of having randomly retired over the last century or so. But retiring conditional on reaching a savings target, often after a long bull market run, is a completely different ballgame. This is clearly related to Part 3, see above, but it’s still a new and unique angle!

Calculate your own personalized(!) Safe Withdrawal Rate with a free Google Sheet

What does all of this Safe Withdrawal research really mean for me personally? Nobody wants to read just a bunch of academic-style research on this topic. The problem I always had with the existing body of research, whether it’s from academics or practitioners, is that my personal financial situation looks completely different from their “model household.” I like to be able to make adjustments for my specific situation, for example…

  • A 50 or even 60-year horizon
  • I like to be able to account for additional future cash flows, both positive (Social Security, a small pension, even a little bit of blogging income) and negative (college expenses for our daughter, higher healthcare costs when older, etc.)
  • Leave a bequest for our daughter when we pass
  • Account for today’s expensive equity valuations
  • etc.

So, I decided to publish the exact same Google Sheet I use myself (though not with my precise personal data) to gauge my personal Safe Withdrawal Rate:

Link to the EarlyRetirementNow SWR Toolbox v2.0

The sheet may not be 100% self-explanatory, so please refer to Part 7 for the basic instructions. I also wrote Part 28 to go through all the enhancements added over the years! If the Google Sheet and all the options seem intimidating, my friends Jason and Eric at Two Sides of FI put together a video tutorial on YouTube explaining how to get started with this toolkit.

As a supplementary tool, I also recommend you look at Part 33 where I look at how actuarial calculations may (or may not!) help in gauging safe withdrawal rates. There’s a free Google Sheet tool (link in the post) that uses pretty much the same inputs as in the SWR Toolbox.

For the mathematics geeks, I also provide a small mathematical appendix (Part 8). But again, this is the absolute last post you want to read unless you’re really into math and want to see where those formulas in the Google Sheet came from! 🙂

Withdrawal strategies are never a one-time set-it-and-forget-it. You want to periodically monitor if your plan is still on track, especially once the market starts dropping. In Part 37, I go through an example of how I would respond to the inevitable Bear Market.

Flexibility to the rescue? Not so fast!

Why would I sit on my hands if I get unlucky in (early) retirement and my portfolio melts down year after year? If you stubbornly withdraw that same initial amount plus inflation adjustments, the way it’s modeled in the Trinity Study, it’s neither psychologically sustainable nor is it financially sound or mathematically optimal. Rather, at some point, everyone would change course: either try to generate some supplemental income or consume less for a while (or a combination of the two). Flexibility is often sold as the panacea against Sequence Risk. I’m certainly not recommending everyone to be inflexible but I’d argue that flexibility is often overrated:

  • Part 9 and Part 10: Under Guyton-Klinger Rules you’d systematically lower your withdrawal amounts in response to significant underperformance of your portfolio. And if your portfolio recovers again you ratchet up your withdrawals. Sounds very intuitive and in some places, Guyton-Klinger-style rules are sold as the solution against Sequence Risk. But what’s often ignored is that GK Rules can take years, even decades for your withdrawals (and your spending) to recover to the original level. So, while GK Rules are certainly a useful tool to prevent running out of money in retirement, there are some unpleasant and often unknown side effects!
  • Flexibility means that we’re using a dynamic and responsive withdrawal strategy. I might replace the risk of running out of money in the long term with very unpleasant short- to medium-term spending levels (see the GK research above!). In Part 11, I study what criteria I’d use to weigh different dynamic rules. Rules that take into account both your portfolio level and economic fundamentals (e.g. the Shiller CAPE) seem to make the most sense, which brings us to the next post…
  • … on one really elegant implementation of flexibility: Part 18, where you base the withdrawal amount on the Shiller CAPE Ratio. The advantage: the fluctuations in withdrawal amounts are muted relative to Guyton-Klinger. I also wrote a more in-depth post on flexible/dynamic CAPE-based withdrawal rates in Part 54.
  • Part 23: A reader suggested an intuitive way of dealing with Sequence Risk: If your portfolio value falls below a certain threshold (e.g., 70% of initial), simply reduce your spending by, say, 30% (or work to make up the difference) until the portfolio recovers. I point out two serious concerns about this procedure: 1) in the really bad recessions that may imply 20+ years of reducing withdrawals and 2) even in a not-so-bad bear market you would have experienced a “false alarm,” i.e., you would have cut withdrawals for years only to find out – in hindsight(!)- that the portfolio would have survived without flexibility.
  • The problem with “flexibility” is that it’s so ambiguous. It’s like playing whack-a-mole; you think you debunked one flexibility scheme and then people suggest another. So, in Part 24 I look at all the different “flexibility schemes” I could think of and test how they would have performed in historical bear markets. And just to be sure, I also solicited some additional reader suggestions in Part 25. Always with the same results: flexibility eliminates the risk of running out of money in 30 years, but raises the risk of long and painful cuts in withdrawals in the short-run and medium-term! It’s like squeezing a balloon!
  • The kind of flexibility I like: If you’re flexible enough to put in One More Year, you can boost your retirement security substantially! See Part 42.

Alleviate Sequence Risk through Dynamic Asset Allocation!

Part 13 – Prime Harvesting: Many readers suggested this: We can alleviate (never eliminate!!!) Sequence Risk through a smarter, non-passive asset allocation to prevent selling equities through the bear market. One such method was suggested by Michael McClung in his book Living Off Your Money. I looked at that method and actually liked it quite a bit. Again, you’re not going to prevent Sequence Risk, but this is a way to slightly alleviate it!

Though not identical, Prime Harvesting will likely look a lot like an equity glidepath in practice. The idea with a glidepath is that starting with a higher bond allocation initially gives you something of an insurance policy against Sequence Risk because you tend to mostly liquidate your bond holdings in the beginning, right when Sequence Risk is most prevalent. Then, over time, you shift more into stocks when Sequence Risk becomes less of an issue. I slightly prefer this approach over Prime Harvesting because it’s easier to keep track of. See Part 19 and Part 20! For information on pre-retirement glidepaths, please see Part 43.

How about bucket strategies? Not really a panacea for Sequence Risk either as I show in Part 48. Effectively, they have a lot of similarities with Glidepaths in retirement. But I found that the glidepath simulations looked slightly better than the bucket strategy results. In Part 55, I discuss this method with my FIRE blogging buddy Fritz Gilbert (Retirement Manifesto).

How about using a margin loan to fund your retirement to avoid liquidating assets at an inopportune time? Part 49 deals with that. Part 52 goes into more detail and looks at how much we can improve the results when we carefully time when to use leverage, i.e., when to fund retirement out of the margin loan and when to use withdrawals from the portfolio.

Special Topics

How much of a difference would it make if you don’t have a perfectly flat withdrawal profile over time? Maybe skip the inflation adjustmentsThat’s what I look at in Part 5 – Cost-of-Living Adjustments (COLA). Related to that, back in 2020 we found ourselves in a low-inflation environment. Due to the lower predicted COLA in that case, can we increase our withdrawal rate? Probably not, at least not by much. See Part 41! The opposite is true as well! In 2021 and early 2022 CPI inflation spiked. That doesn’t necessarily mean that we have to lower our SWR. See Part 51.

Mortgages and (early) retirement don’t mix! In Part 21 I go through my rationale for paying off the mortgage before retirement: Having a fixed, inflexible payment like a mortgage is poison from a Sequence Risk perspective! Accounting for homeownership in general: please see Part 57.

Part 29, Part 30, and Part 31 all deal with the widely held misconception that we can easily save the 4% Rule if we were to increase the dividend yield to a level close enough to your 4% annual withdrawal rate. Simply live off your dividend income, avoid ever selling your principal at depressed valuations and you shield yourself from Sequence Risk, right? It certainly sounds intuitive and I was intrigued enough to research this option and absolutely hoped that this can indeed lower Sequence Risk. People in the FIRE community call it “Yield Shield” but, alas, that label is extremely deceptive because the whole thing doesn’t work very reliably. It would have backfired really badly during the 2007-2009 bear market. Instead of shielding yourself from the downturn, you would have aggravated the Sequence Risk.

Related to the “Yield Shield,” in Part 40 I look at how your retirement budget would have looked like if we had simply lived off the dividends of an S&P 500 equity portfolio without ever touching the principal. For most retirees, this will not be a viable hedge against Sequence Risk.

How about exotic asset allocation strategies? In Part 34, I write about the impact of adding Gold. And it turns out that the safe-haven asset indeed alleviates Sequence Risk, exactly during the bad historical bear markets (1929, the 1970s). That said, I’m not too confident that some of the other heavily hyped, “sexy” asset allocation flavors like the “Permanent Portfolio,” “Risk Parity/All-Weather Portfolio” and “Golden Butterfly Portfolio” add a lot of value.

How does Rental Real Estate fit into early retirement planning? How do we model safe withdrawal and safe consumption rates with a real estate portfolio? See Part 36 of the Series!

A very important practical concern for retirees: How often do you want to rebalance your portfolio? Can the rebalance strategy and frequency compound or alleviate Sequence Risk? See Part 39.


I haven’t written much about taxes until Part 35. It’s about planning your asset location (as opposed to asset allocation), i.e., what assets belong in what account types: taxable brokerage accounts, vs. tax-free accounts (Roth IRAs, Health Savings Accounts) vs. tax-deferred accounts (IRAs, 401(k), etc.). And in Part 44, I write about some general tax optimization principles. Part 45 has a case study and an Excel Toolkit I use for tax planning and optimization.

Summary and Suggestions

Withdrawing money from your nest egg is a lot more complicated than the “simple math” and “simple path” to accumulating assets. I can’t change that. But if you spend 10-15 years saving diligently to invest six-figure or even seven-figure sums, it may also be a wise “investment” to spend a few hours to familiarize yourself with the challenges of withdrawal strategies. Depending on where you are in your FIRE journey and your personal preferences I’d suggest proceeding as follows:

  • A new reader, not even close to early retirement, just looking around: Start with Part 26 and Part 27. Don’t even sweat any of the technical and more detailed posts linked there. But if something sparks your interest, sure, check that out, too!
  • For the math geeks: Sure, add Part 1 (simulation details) and all the nitty-gritty details on Sequence Risk (Part 14, Part 15). And the mathematical appendix with formulas (Part 8) for the math pros!
  • Getting closer to retirement: Now you probably want to start looking into performing your own personal safe withdrawal simulations. I’d suggest you familiarize yourself with the Google Sheet and the two posts that explain how to use it (Part 7 and Part 28). Probably also Part 4 and Part 17 on how to treat supplemental cash flows. And getting closer to retirement in this day and age, you can’t escape the reality of expensive equity valuations so you probably want to convince yourself that PE ratios (and CAPE ratios) have an impact on sustainable withdrawal rates: Part 3
  • I’d also urge every reader close to a “Lean-FIRE” retirement to check out some of the simulations in Part 23, Part 24, and/or Part 25Can you stomach a multi-year (maybe even 20-year!) drought period with lower withdrawals? Lower than your already “bargain-basement budget?” If you don’t mind cutting your already optimized budget and/or doing a side hustle for that long then, sure, go ahead. But if you don’t, then maybe plan for a bit of a cushion, and don’t pull the plug the minute you reach 25x annual expenses!
  • For more adventurous, hands-on, and not 100% passive investors: Make sure you check out the dynamic asset allocation posts: Glidepaths (Part 19 and Part 20) and/or Prime Harvesting (Part 13). 

Just for your reference: all posts are listed in chronological order

Comments or questions?

Feel free to leave comments and suggestions here or at one of the specific SWR posts. I will get a notification and will try to respond, usually within a week. Also, comments with up to two external links are allowed. More external links might cause your comment to land in the spam box! 🙂 Looking forward to your feedback!

 * * *

Update 11/15/2021: if you like to navigate across the different parts of the series, there is an easy way:

  • is this landing page
  • will get you part XY or the series, for example, the “swr01” suffix for part 1, “swr02” for part 2, etc.
  • Notice that parts 1-9 need to be entered as two digits with a “0” in front and the suffix, i.e., the “/swr” part is case sensitive, all lower letters!

275 thoughts on “The Safe Withdrawal Rate Series

  1. Do your SWRs already account for the expense ratios of funds used or do you use total returns of the asset class (meaning I need subtract expense ratios myself)?

  2. Great tool. When I change the ‘Retirement Horizon (Months)’ in Parameters & Main Results it is not adjusting the ‘Cash Flow Assist’. For example, it’s not reducing the number of lines/months and it appears the calculations of Results relies on the default of 720 months. Is this by design or is there another place I should adjust for this value? Or do I just delete lines on the Cash Flow Assist tab for the Horizon? Thanks.

    1. The number of lines is set at 720. If you enter cash flows beyond your retirement horizon, they’d be ignored in the simulations.
      It’s not recommended to delete the lines. Just enter zeros to be sure.

  3. Big Ern – thank you for doing the heavy lifting. How can I view retro active SWRs by months adjusted for historical s&p and cape (75 equities/ 25 bonds / 0 percent failure). You’ve suggested an active SWR based on cape can significantly improve retirement success. This would help to view potential swr fluctuations. Many thanks. Nik

    1. In the SWR Simulation sheet I, have a tab “CAPE-based Rule” where you can specify your CAPE SWR parameters and check how withdrawal amounts and rates would have evolved historically.

  4. Hi Big ERN,
    New to fire and was recommended your blog by a friend. I’m interested in your thoughts on how to accommodate a post-retirement major expense like tuition for children. Drinking from a firehose! I am 35 now, and based on current projections envision hitting FIRE (25x my average projected annual expenses) in ~11 years. My plan had been to work another 2 years full time after that and then retire. The issue that I’m not sure how to account for is the impact of private high school tuition + college costs for 2 kids, which would hit immediately following my retirement. I had been using Projectionlab’s simulation sandbox to model these expenses in, and it projects my net worth would keep increasing, but if the early years immediately following retirement have the greatest impact on overall risk, do I need to incorporate that into your google sheets somehow to change how I think about my threshold for retiring? Hopefully the answer isn’t “you shouldn’t retire until these expenses are past” but I don’t know how to incorporate it into my thinking of a safe withdrawal rate and whether I can, in fact, retire when I anticipated. I appreciate your guidance, and would love a reply or even a post on how to think about these kinds of short term, high cost “life is lumpy” expenses!

    1. I can sympathize with that. It’s essential to model your supplemental cash flows, like outflows for assisting kids with college and inflows like Social Security and pensions later in retirement.
      In my Google Sheet, please see Part 28 for the link and instructions, you can model those flows that go above (or below) your baseline retirement spending and see how much of a difference that would make.
      If you have a large enough nest egg, you should not wait until after those expenses are taken care of.
      Another approach for college expenses: set aside a specific pile of money dedicated to helping your kids with college expenses. Then keep that pile of cash out of your calculations and ignore the college expenses in your retirement calculations. If you experience some bad sequence risk, it doesn’t impact your retirement savings too much. And if the college savings account doesn’t entirely cover the expenses for junior, then so be it. Tell your kids to cover the difference with a college loan.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.