Ten things the “Makers” of the 4% Rule don’t want you to know (SWR Series Part 26)

Welcome back! It took me two weeks again to put together a full blog post. That’s what early retirement does to you! Especially while we travel – pretty much permanently between now and December – I figure I will scale down the blog frequency to every 2 maybe 3 weeks. I hope people don’t mind. If two weeks is what it takes to come up with quality content then so be it!

CruiseSunset
Safe Withdrawal What? I’m on vacation here! I will finish the post next Wednesday!

For today’s post, I thought it was time to add another installment to the Safe Withdrawal Rate Series. 25 posts already! What have I learned after so many posts? Well, I started out as a skeptic about the so-called “4% Rule” and I thought it might be the time to poke a little bit of fun at the “makers of the 4% Rule.” Just to be clear, this post and the title are a bit tongue-in-cheek. Obviously, the “makers” of the 4% Rule, the academics, financial planners and bloggers that have popularized the rule aren’t part of any conspiracy to keep us in the dark. Sometimes I have the feeling they are still in the dark themselves! So here are my top ten things the Makers of the 4% Rule don’t want you to know…

1: We actually mean “4% Rule of Thumb

As I said before, here on the blog and in various podcasts, there is no one-size-fits-all solution to safe withdrawal strategies. Suggesting that we should follow a 4% Rule is about as ludicrous as suggesting that we should all wear size 10 shoes. No, I should correct myself; it’s actually more ludicrous than prescribing size 10 shoes for everyone! Different folks have different shoe sizes, but at least during my adult life, my shoe size hasn’t changed. It’s always been 14, hence the name Big Ern! Not so for safe withdrawal rates! There is both idiosyncratic variation across people, i.e., person A will potentially require a very different withdrawal rate from Person B, just like they have different shoe sizes. But sustainable withdrawal rates can and should also vary over time even for one single individual depending on asset price valuations, i.e., bond yields and equity price-earnings ratios, see Part 3 (Equity Valuation) and Part 18 (CAPE-based rules) for more details! At the bottom of the 2009 bear market, you’d have to be crazy to withdraw only 4%. A 6% initial withdrawal rate would have been more appropriate! But almost 10 years into this economic expansion and bull market, I like to operate a bit more cautiously, especially looking at today’s high CAPE ratio! Boy, am I glad my shoe size doesn’t fluctuate that much over the business cycle! I would have spent a fortune on shoes over the years!

2: The 4% Rule is likely way too conservative for many early retirees

Just for the record, I have come across a lot of case studies where a 4% withdrawal rate would have been way too conservative. Even with today’s expensive equity valuations, i.e., a Shiller CAPE above 30! I did a total of ten case studies for readers of my blog who volunteered their early retirement plans and finances and I was positively surprised by how well positioned a lot of early retirees were. Many expect sizable pensions and Social Security benefits. They also tended to be a little bit older, usually in their 40s or early 50s and they have the advantage of a slightly shorter retirement horizon. Add to that the fact that pension and Social Security benefits start only around 10-20 years into retirement (as opposed to 30-40 years for a lot of extreme early retirees) and you get a withdrawal rate higher than 4%. So, this goes back to point #1: there cannot be one simple unique rate and it means that there could be retirees who will vastly over-accumulate (!) wealth if they don’t do their homework and instead follow the naïve 4% Rule! It’s a failure of the 4% Rule, maybe not quite as dramatic and traumatic as running out of money in retirement. But for many, working several more years targeting a 4% Rule (25x expenses) when, say, 4.8% (20.8x expenses) would have sufficed is a failure of sorts as well!

3: We conveniently ignore expense ratios, transaction costs, taxes, etc.

True, a lot of folks in the FIRE crowd can probably avoid federal taxes if they stay within the $24k standard deduction for their ordinary income and around $77k for dividends and long-term capital gains. But how about state taxes? Oh, you live in a state with a 4% flat tax? Well, your gross safe withdrawal rate might still be 4% but you’re left with only 3.84% net of state taxes. And did I tell you that many of the simulations rely on market index returns, so let’s subtract another 0.05% to account for the expense ratio of your ETFs. (side note: In my simulations, I subtract an estimated 0.05% expense ratio). The 4% Rule is now down to 3.79%. To generate $40k per year in consumption we’d need a $1,055,000 net worth, quite a bit more than 25x expenses! And I’m not even subtracting the 1% or so for actively managed funds and/or financial advisors because then the 4% Rule would be completely toast. But even the for-fee Robo-advisors, charging around 0.25% per year (in addition to the ETF expense ratios!) would make a sizable difference.

4: Flexibility is overrated

The standard response of the makers of the 4% Rule to the potential failures: just be flexible. Who knew it was that simple? We will certainly be flexible and adjust our withdrawals and/or take one some side gigs if when we go through the next equity bear market. Think about this as starting with “FatFIRE” and scaling down our budget to FIRE for a while. Or maybe even “LeanFIRE.” But I don’t have any illusion that this is going to easy. I did some simulations in Part 23, Part 24, and Part 25 to check how long one would have to be flexible if we were to go through something comparable to the bad equity bear markets of the past. The implicit (sometimes explicit) claim by the 4% Rule makers is that you have to be flexible only for as long as the stock market is down. Not true! I knew that the “flexibility” had to last longer than the equity bear market thanks to, you guessed it, Sequence Risk (see Part 14) but I was shocked that the lower withdrawals lasted for 15, sometimes more than 20 years! I would still call that a failure!

5: Some of the simulated cohorts that didn’t run out of money still would have had a very scary and turbulent retirement!

A side effect of that flexibility mantra (see #4 above) is that we’d also potentially overreact to market fluctuations and lower our withdrawals, tighten the belt, and/or get a side gig in retirement even though in hindsight it wouldn’t have been necessary. Look at it through this analogy: Imagine you are flying with “Trinity Air” which has a slightly spotty safety record. Its success rate is 97%, which means that out of 100 flights only three would crash. What makes us think that the 97 “successful” flights would have had a smooth ride? What if out of the 97 flight that make it, an engine blew up in 30 of them? Or smoke in the cabin? Or the pilot shouting “we’re all gonna die” into the intercom? Sure, in the end, you made it, but it likely wasn’t a pleasant ride. The analogy of “we’re all gonna die” scenario would be the numerous cohorts that saw their portfolio value drop precipitously (e.g. 1972). They might have been “flexible” for years, even a whole decade which might involve going back to work and/or substantial cuts in the retirement budget. Only to find out in hindsight that all that hassle wasn’t even necessary and the “inflexible” 4% Rule would have worked after all. (See parts 23-25 of the SWR series again for the simulations!) Quite intriguingly, nobody else in the retirement research world (at least to my knowledge) seems to want to research this issue of type 2 errors (false alarms) very much. I guess they not only don’t want you to know this, they don’t even want to know themselves…

6: The 4% Rule works all the time! But only if you assume that the retiree had perfectly timed the bond duration decision over time

If you read this and wonder what the hell does it even mean to “time the bond duration decision” then sorry to tell you, you’ve already lost! The 4% Rule would have failed for you. You see, a lot of retirement researchers use a 60/40 portfolio as their baseline for their standard 30-year horizon. (Just as an aside, 40% bonds is way too high for a long horizon in early retirement, see Part 2 of the series). There is relatively little ambiguity about the equity portion. Normally people use a U.S. Total Stock Market index. I use the S&P500 with very similar results. But on the bond side, results vary wildly depending on whether you use short-term bonds (e.g., 3-month T-bills) or longer-term government bonds (e.g., 10-year Treasury Bonds). Long duration bonds did well in the Great Depression, Dot-Com bust and Great Recession (all demand shocks with low inflation or even deflation = good for bonds, i.e., bonds offered great diversification) but not so well during the inflationary 1970s and early 80s. So, some retirement researchers who tout the success of the 4% Rule have very sneakily made that little switcharoo: assume that the retiree had short-term bonds during the 1970s. But the folks who retired in 1929 and 2000 had the magical foresight of using long-duration bonds during the 2000s.

SWR-Part26-Table01
Failsafe withdrawal rates for three cohorts impacted by major bear markets.

But, unfortunately, for the rest of us who don’t have the perfect foresight, we have to make a decision about the bond duration ahead of time without knowing how exactly bonds will perform in the next recession and bear market!

7: We routinely confuse nominal and real numbers!

I frequently hear and read stats like “the average retiree with a $1m initial portfolio using the 4% Rule would have ended up with $2.7m after 30 years!” Or “90% of the retirees using the 4% Rule would have ended up with more than their initial capital after 30 years!” Maybe you’ve heard those, too? Impressive stats for the 4% Rule, right? Maybe, maybe not! You see, the folks who coined those phrases don’t want you to know that those numbers are completely meaningless. They are in nominal dollars! $2.7m might be a lot if inflation was close to zero for 30 years. $2.7m might be a lot less impressive if the 30-year time span included the inflationary 1970s and early 80s where the dollar lost almost 80% of its value.

According to my calculations, for retirees with an initial $1m portfolio using a 60/40 portfolio the average and median final portfolio values were $1.47m and $1.14m in real, inflation-adjusted dollars. Still impressive but only about half as impressive as when you bungle nominal and real numbers! The probability of preserving the initial portfolio value drops from 90% to only 58% over the entire sample. And that probability drops to exactly 0% conditional on a Shiller CAPE above 30!

8: For the next 12 years, we can conveniently ignore another potential failure of the 4% Rule!

Proponents of the 4% Rule always point out that we now have data on the dot-com bust and Global Financial Crisis and there are still no failures despite the pretty atrocious equity performance between 2000 and 2009. That means the 4% Rule must be pretty robust, right? Maybe it is but I wouldn’t bet my retirement on it. So far we can only gauge the success or failure of cohorts that started retirement up until 1988 for a 30-year horizon. The jury is still out on the cohorts since then, especially the year 2000 cohort. Michael Kitces pointed out that this cohort has actually recovered its initial portfolio value already, but as I pointed out in the SWR Series, Part 6, that’s in nominal dollars (file that one as another example of #7 above: folks confounding real and nominal numbers!). In real dollars, that cohort is still severely underwater. Time will tell if they make it all the way to the year 2030 (and 2040, 50, etc. for early retirees). If we assume that the stock market keeps chugging along with double-digit returns for another twelve years everything will be OK. But I’m not that hopeful; even a short recession and shallow bear market over the next few years and the year 2000 cohort will be in trouble!

9: We won’t have to rely on the 4% Rule ourselves

Do you ever have the suspicion that the folks on TV marketing the newest exercise equipment or diet pill or whatever product have never actually used that product themselves? They probably looked this fit and slim and healthy way before that product ever existed. Could that also be true about the “Makers of the 4% Rule?” Who are these “makers” anyway? They are the academics, the financial planners and quite a few of the FIRE loggers. The tenured professors have their nice lifetime employment guarantee and cushy university pensions. The financial planners are probably far from retirement themselves and once they retire they might still have some side income as a senior partner of their advisory business. And the bloggers that market the 4% Rule most aggressively are probably raking in more money from their blog than they need to withdraw from their portfolio. I doubt that any of the “Makers of the 4% Rule” actually fit the model assumptions of the Trinity Study: a guy or a gal who has a pile of money invested in stocks and bonds and tries to make that pile last for 30+ years without much (or any?) additional cash flows. Of course, just because the “makers” don’t fit the model assumptions of the Trinity Study doesn’t mean they can’t write credibly about the challenges of the safe withdrawal math. But sometimes when I read “oh, don’t worry about the 4% Rule” and “yeah, just be flexible” as the “solution” to a 4% Rule failure it has the ring of a “why don’t they eat cake” attitude.

10: The “small-cap value bias” may or may not work in the future

Pssst, do you want to know a secret? All of this talk about running out of money, 4% Rule, etc. is completely moot! If you had only invested your equity portfolio with a tilt toward small-cap and value stocks you would have done so much better! 5.25% would have been the failsafe rate for all 30-year retirement windows! (assuming a 30-year horizon, 60/40 portfolio and the 60% small-cap and 60% value tilt using the well-known Fama-French factors) That’s right, you could have not just saved the 4% Rule but withdrawn a cool additional 30%, for life (well, a 30-year life), even during the tumultuous 1930s and 1970s!

The only problem: Who knew that in advance? If you had bet on the opposite side (large-cap growth stocks) your safe withdrawal rate would have been below 3%! The advice “small-cap value would have saved the day” is about as dubious and suspect as telling yourself that Amazon and Google beat the index over the last few years. Quite intriguingly, Bill Bengen (one of the “Makers of the 4% Rule) now tours around and does just that: he advocates for upping the safe withdrawal rate to 4.5% if you shift to small-cap and value! But there is no guarantee that those styles will continue to outperform. Especially after Eugene Fama won the Economics Nobel Prize in 2013 and the advent of lots of ETFs targeting value and size tilts. Once you go mainstream – and you can’t get much more mainstream than the Nobel Prize – any anomaly will eventually be arbitraged away! Sure enough, small-cap value has not outperformed the overall market since 2004! As much as I appreciate Bengen’s early work on the topic, I think his most recent musings should be taken with a grain of salt!

We hope you enjoyed today’s post! Please check out the other posts in this series and leave your comments and suggestions below!

Picture Credit: pixabay.com

69 thoughts on “Ten things the “Makers” of the 4% Rule don’t want you to know (SWR Series Part 26)

  1. Hey ERN. Another good article, but it isn’t clear to me what you recommend as a safe withdrawal rate strategy. You have knocked a lot of holes in several strategies,but as to what you recommend, it isn’t clear to me.

    Liked by 1 person

    • For me personally, 3.25% would be the fail-safe initial rate, considering a very long horizon and no additional supplemental income (pension, Social Security). I can walk that up a bit when taking into account future expected benefits. Maybe to 3.5%. 🙂

      Like

      • ERN,
        How much faith/credibility do you place in a lot of these Monte Carlo retirement calulator simulations which generally recommend a 3.2% withdrawal rate to get 90% confidence of not outliving your assets within 50 years in “below market average” conditions (e.g. Fidelity Investments calculator, Vanguard has a simplier calculator that yields the same results).

        Liked by 1 person

        • I’m not a huge fan of MC simulations. Too much depends on expected returns. Then there is also the issue that MC in the most basic structure can’t adequately model the mean reversion of equity returns. I think MC would be nice to.supplement historical simulations but I’d never use it just by itself.

          Like

  2. I have long considered it the “4% Rule…of Thumb”.

    Thank you for all of the excellent commentary. Perhaps this has been previously covered (or impossible to cover), but I have wondered to what degree a volatility dampening strategy (like “Dual Momentum”) might have on the SWR. I would expect that if we could reduce volatility, the SWR should be higher, but perhaps I am mistaken). Perhaps this could be tackled in some mathematical way.

    Thanks again.

    Liked by 1 person

    • I played around with that and it looks hopeful. Momentum would have afforded you a higher failsafe rate. But something tells me that I shouldn’t rely on a simple mechanical momentum rule in the future. I’d be afraid that when everyone is using this it’s bound to fail in the future.

      Liked by 1 person

  3. Another great article and my favorite part is hearing “You’re all going to die” over the intercom at Trinity Air! Well, that makes me think that you should add a Number 11, which is that the 4% rule people have not factored in the Travel Hacking and amazing savings during retirement, especially for credit card points on Trinity Airlines. They can take you very far, for next to “nothing”!

    Liked by 1 person

  4. Karsten – I am really enjoying this series and everything you post. Thank you for your work. My wife and I have been retired for 4 years, (early 50’s) and are using an income approach (dividends from individual securities and interest) to cover our household expenses. This approach is not for everyone as it takes a bit of initial due diligence and ongoing monitoring of Dividend Payer cash flows. The largest potential concerns (which you have covered in earlier posts) that we have about potentially outliving our savings, are about large unforseen expense increases or unforseen expenditures. This would include larger than expected increases in healthcare expenses, and costs for long term care (if one of us develops dementia, our annual expenses could increase significantly), and other large unforseen expenses. We have budgeted $ for these potential expenditures – but the actual costs may be paid far in the future. Your writings about the need to consider the impact of inflation on expenditures and withdrawals is spot on!

    Liked by 1 person

    • Awesome! Thanks for sharing! Health expenses and also unforeseen home repair expenses are clearly on my mind. Hard to budget and plan for those!
      What average dividend yield do you target right now? Just curious. I’m trying to build up our own little dividend portfolio: high dividend payers but also with a quality screen…

      Like

      • Some of the stocks that I am watching, such as ITW, HON and INTC, and NEE are yielding 2 to 2.6%. The yields on these names have been below 3% for many years. But the growth in the dividends per share for these names has been fairly high (as a result of dividend increases and share repurchases). Will this growth continue at the same recent levels? Not sure, given the uncertainties with the trade negotiations. There are other high quality dividend payers such as PEP, XOM, O, and PSA, and AEP that are trading at attractive yields north of 3%. I try to keep our portfolio diversified (no individual stock is greater than 3.5% of the total portfolio income – and no industries represent more than 10% of total portfolio income). I stay away from some financial industries (commercial banks, investment banks, mortgage REITS, insurance companies), due to their histories of dividend cuts. Also, I do not invest in stocks that have had dividend cuts during the last 20 years (which captures the 2000-01 and 2008 to 2010 recessions).

        Liked by 1 person

  5. Great post, big ERN. This has a treasure trove of golden advice in it. My favorite part was the example of the flights that made it but were scared to death. It remind me of the Maya Angelou quote:

    “I’ve learned that people will forget what you said, people will forget what you did, but people will never forget how you made them feel.”

    How you feel on those plane rides is what you will remember. And I bet you’d never recommend that plane ride to anyone else.

    TPP

    Liked by 1 person

  6. Karsten, what a fantastic article. I have always wanted to be “flexible” with 25x and 4% but this just makes me want to really double check those numbers and calculations. I still have some years to go so this isn’t top of mind right now (still in the acquire phase).

    I appreciate your ability to challenge the status quo of FI at times in a professional and scientific/data fueled way!

    Liked by 1 person

  7. […] Ten things the “Makers” of the 4% Rule don’t want you to know – We’ve talked about the 4% rule a lot. 4% is the most commonly cited safe withdrawal rate. That means you can safely withdraw 4% of your assets each year forever…maybe. There are a lot of other factors to consider. This is a good explanation of a few. […]

    Like

  8. Another great post. Rules of thumb or textbook answers do not cover all situations. Those of us from medicine know the truth of this. Complicated interactions often produce off the wall results. To me the 4% rule is a good place to start but not the end of thinking and planning. I appreciate your posts even though I am above the FIRE age.

    Liked by 3 people

  9. Great post. At a high level I’ve always been conscious of “hidden costs” of the 4% rule, but this is the clearest risk evaluation of it that I’ve seen. Point #3 on taxes and expense ratios was especially valuable for me, particularly when highlighting the potential oversight of state taxes after accounting for the federal deductions (I live in California) Thanks for writing!

    Liked by 1 person

  10. Very minor in the scheme of your overall post, but a note about your Colorado state tax example in #3. Colorado state taxes are calculated on your Federal taxable income, which means a couple would automatically get that $24K standard deduction (or their itemized amount if over that) as well as any other deductions that come before taxable income (and therefore wouldn’t pay Colorado state tax on that). In addition, if any of their income is from a pension or social security and they are at least 55, then they get additional deductions for that ($20K each if over 55, $24K each if over 65). So, for example, because both my wife and I are teachers, we will have $64K of income that won’t be Colorado taxable at age 55, which then bumps up to $72K at 65.

    None of that detracts from your overall point about thinking about state taxes and other expenses you might encounter, but I think stating that in Colorado your net would only be 3.81% is too low for most folks.

    Liked by 1 person

  11. I still wonder if there are strategies or options that could increase SWR by mitigating the impact of stock market plunges. E.g. protected put or collar options strategies. The market tends to climb a staircase and fall off a cliff. If one could put a floor on losses in a few disaster years like 1974, 2000-2002, or 2008 while carrying a 100% equities allocation, that would be the best of both worlds.

    An “IF” function could simulate the result of carrying a protected put. For example, the following would judge whether puts that were 10% out-of-the-money and cost 3% of the portfolio to buy would have supported the portfolio in a down year, where that year’s market performance is in cell A1:
    =IF(A1<-.1,-.1,A1)-.03

    The following would simulate returns with a collar strategy where we sell a call 10% OTM and buy a put 10% OTM at a net cost of 1% of the portfolio:
    =IF(A1.1,.1,A1)))-.01

    Liked by 1 person

  12. Karsten, Great article and I agree entirely, as must be quite obvious to you! You and i have discussed this topic in comments both in your website and mine. I started with 3.27% (based on back-tested simulation, essentially same as your 3.25%). However, I now rely on a CAPE-adjusted SWR formula, which as of today (CAPE at 32.54), stands at 2.8%. As you know, I wrote two articles on “safest” withdrawal rate for the benefit of, shall we say, “nervous” early retirees. I have finally made my peace with this!

    Liked by 1 person

    • Yup, glad we’re still on the same page! I think that the CAPE-based rules make the most sense in principle, but everything hinges on the parameters. I find the CAPE of >30 not quite as worrisome. It will drop to below 30 just by rolling out 2008-10. And with lower payout ratio as before one can shift that intercept up a bit. So, 3.25% would still be OK even with a CAPE rule (intercept 1.75% and slope 0.50).
      Thanks for stopping by! 🙂

      Like

  13. Great series!

    I think there might be a missing word in this sentence (I marked the possible location with XXX)

    Well, I started out as a skeptic about the so-called “4% Rule” and I thought it might
    XXX the time to poke a little bit of fun at the “makers of the 4% Rule.”

    You can delete my message afterward.

    H.

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s