Much ink has been spilled on the 4% rule, including here on *The Best Interest*.

The short and sweet definition? The 4% rule is a retirement strategy that suggests withdrawing 4% of your portfolio’s value annually, adjusted for inflation, to ensure your savings last for a 30-year retirement.

If you’d like to dive deeper on some nuance of the rule, read this: Updated Trinity Study and the 4% Rule.

And if you’d like to avoid the common mistakes of using the 4% rule, read this: You’re Probably Using the 4% Rule All Wrong

Today, though, I want to show you some compelling data about the optimism and conservatism built into the 4% rule.

Ultimately, you’ll see that the 4% rule comes with major risks.

## All Models Are Wrong

All models are wrong?

No, not that kind of model. I’m talking about *numerical *models. The idea that you can use numbers and figured to represent or simulate reality. Your numbers will never, ever be a perfect representation of reality – you can’t predict the future. *All models are wrong!*

*…but some are useful.* Models are used all over modern society. One hopes that their models can “bound” reality, providing bookend scenarios to how reality might shape up (good vs. bad, optimistic vs. conservative, etc). We can use models to explain how particular variables will or won’t affect reality.

Weather forecasting is a terrific example. Meteorologists use numerical models to simulate the atmosphere by solving complex mathematical equations based on physical laws, such as the conservation of mass, energy, and momentum. These models take in data from satellites and weather stations, then simulate how the atmosphere will evolve over time.

They’re **never** perfect. In that way, they’re **always** wrong. But they’re certainly useful.

Further reading: *The Madness of Forecasting*

## Retirement Forecasting

It’s a fool’s errand to predict the future performance of investing markets. But the 4% rule provides a numerical “bookend,” giving direction to retirees.

The hardest pill to swallow about the 4% rule is that **we know it’s wrong**. *We just don’t know which direction it’s wrong in. *

Most likely, as you’ll see below, it’s too conservative. *It’s important for retirees to internalize that truth!*

But there’s a possibility the 4% rule is too aggressive. *And that’s a scary possibility. It means you might run out of money in retirement. Yikes. *

## Visualizing the 4% Rule

I used the terrific 4% rule visualizer from *Engaging Data* to create the charts you’ll see below. I recommend playing around with that tool yourself.

As with any modeling, the input assumptions are vitally important. I assumed:

- We’re investing in
**a diversified 60% stock, 40% bond portfolio.***Yes, this varies slightly from the 4% rule’s original assumption of a 50/50 portfolio*.*But 60/40 is more in line with retirement best practice.* - I assumed a
**30 year retirement timeline**. To make the math easy, I assumed a $1 million starting portfolio. Thus, a 4% withdrawal in Year 1 is $40,000, and each future withdrawal is adjusted up for inflation. - I assumed that
**taxes**and**investment fees**arein annual spending**included***.*This is a key assumption, but one that’s often overlooked. In other words:- Are you withdrawing 4%, and
*then*paying taxes on those withdrawals (perhaps another 0.4%), and then paying investment fees on those assets (perhaps another 0.2 – 1.5%)? That’s actually more like a ~4.5+% annual withdrawal. - Or, are you withdrawing 4%, which
*includes*the requisite taxes and fees. Perhaps you’re only “netting” ~3.5% to spend on your lifestyle, and the other 0.5% pays taxes and fees. - I’m using the
**second**scenario, not the first.

- Are you withdrawing 4%, and
- While the
*Engaging Data*website does look at market history back to 1871, I don’t care too much about anything before World War 2. The American and global economic systems were far different then. I can’t cut that data out of my results, but**I recommend focusing on 1950 onward.**

Of course, it’s worth noting that all tools have limits, and Engaging Data is no different. *“Stocks, bonds, and cash”* is a simple retirement portfolio. There are many factors and sectors in stocks, various bonds, alternatives and private investments, all available in inexpensive ETFs.

Bill Bengen’s recent research over the past several years shows that SWRs of 4.7-4.8% are achievable with more diversified portfolios. Better results are possible with further knobs to turn.

#### Caveats Apply…

I’m beating a dead horse here in the world of retirement planning. There’s an asterisk on everything. But, to pre-empt any guff, let’s be clear:

- The 4% rule is
*way*more rigid than how any normal human would spend in retirement. Lifestyle variations, taxes, inflation rates, healthcare costs, and the potential for longer retirement periods make any withdrawal rule an imperfect one-size-fits-all strategy. - The 4% rule doesn’t account for Social Security, which I think is a huge mistake.
- The 4% rule uses past market results to make future retirement decisions –
*it’s a numerical model!*We know the risks involved there.

## Results – How Does the 4% Rule Hold Up?

In short, the 4% rule is like playing rec league basketball with Lebron James on your team. You’ll win, and it will be overkill.

Out of the 123 unique 30-year periods we can observe, only **one** of them leads to “failure.” It ran out of money in Year 28 (barely a failure, at that).

The **median **result not only supported our retiree’s lifestyle, but also left them with **$2.8 million at death**. Again – that’s the *median* result. 30 years worth of withdrawals, *and still* another $2.8 million leftover. That’s overkill!

We’ve made a trade-off using the 4.0% rule. That trade-off is: *in order to avoid a ~1% chance of retirement failure, are you willing to accept the 50% chance that you underspend in retirement *so severely* that you end up with 3x the assets at death as when you retired?”*

That’s what we’re talking about here. *Severe *underspending. Severely *not* enjoying the fruits of your labor. It’s worth thinking about that trade-off. Personally, I don’t think it’s worth it. I don’t mind increasing my “failure odds” above 1% if it means I get to spend a bit more.

Another crazy stat: let’s compare the 90th percentile result against the 10th percentile result. These two scenarios are equally likely, one being on the good side of fortune and the other on the bad side:

- 90th percentile: our retiree dies with $6M after 30 years.
- 10th percentile: our retiree dies $800K after 30 years

The 10th percentile result is pretty close to, *“I died with as much money as I started retirement with.” *

The 90th percentile result is, *“I died with 6x more money than I started retirement with.”*

And they’re equally likely to happen. Wild!

While it’s important to acknowledge the one failure of the 4% rule (albeit after 28 years of withdrawals), it’s hard to walk away thinking anything other than, **“The 4% rule is overkill.”**

## Adjust the Rate, Adjust the Timeline

So let’s play around with the numbers a bit.

- Let’s extend the timeline beyond 30 years.
- And let’s toggle the withdrawal rate higher.

But we should first ask – what’s an appropriate “failure” metric? The answer, by nature, will be completely arbitrary. After all, we’re using a numerical model that we *know* cannot be correct. Nevertheless, I vote for the following:

- If more than 20% of test retirements fail, then I think our scenario is too aggressive, too risky.
- If the median result spends down more than half of our retirement nest egg (e.g. dies with less than $500K after starting with $1M), then I think our scenario is too risky.
*Why?*Because retirement research clearly shows that retirement failure is a slippery-slope/accelerating problem. If our median result is down 50% after 30 years, then many of those individual scenarios would accelerate toward near-term failure.

#### Timeline Failure

If we keep our withdrawal rule at 4%, but extend the timeline out to **53** **years** (!!), then we reach ~20% of our test retirements failing.

The median scenario ends with $5.5M. The *best* timelines here finish with upward of $65 million.

We’re saying that a 40-year old can retire on the 4% rule and reasonably (~80% odds) expect to still have assets at age 90.

That leads me to a similar takeaway as before: **the 4% rule has historically skewed toward overkill! **

#### Withdrawal Rate Failure

Let’s go back to a 30-year timeline, but let’s now dial up the withdrawal rate above 4.0%. When do we reach one of failure criteria?

At **4.85%**…

Using a 4.85% withdrawal rate, we see that:

- 20% of our test cases fail before 30 years, the earliest of which being after Year 20.
- Our median case still has $1.7M after 30 years –
*more than we started with!*

As a reminder, 4.85% equates to **21% more spending every year** than the 4% rule. That’s a big difference in lifestyle.

#### Let’s Get the Median Below $1M

I want to press our luck and push the limits. **What withdrawal rate does it take so that our median result has a retiree with less than $1M at death?** In other words, I want to see more than half of our simulations outspending their investment growth.

The answer: a withdrawal rate of **5.2%**

In this case, about 28% of our scenarios outright fail, and another 22% finish between $0 and $1M.

#### Let’s Get Failure to 50%

One more experiment: at what withdrawal rate do 50% of scenarios outright fail? Clearly this is too much risk to bear. But it’s helpful to use numerical models to “define your limits,” and this is just that.

The answer: a withdrawal rate of **6.25%**

You’d never want to start retirement knowing that you have a 50% chance of go broke prior to death. But it’s worth understanding where the limits of withdrawal rates lie.

## “But I’ve Read About 3.5%, 3%, and Lower Rules…”

Yes, some conservative retirement commentators combine multiple factors that result in low withdrawal rates like 3.5%, 3.0%, and less.

Quite simply, I think 3.5%, 3.0%, or lower withdrawal rates are simply beyond the pale. Those commentators are suffering from the crushing costs of conservative retirement planning.

Simply look at the 3.5% withdrawal rate chart below:

3.5% is a recommendation that someone chronically underspends their potential, knowing that anyone who would have done so in the past would have **at worst** died with the same $1M they started with (for 3.5% withdrawal rate) or **at worst** died with $1.6M (for 3.0% withdrawal rate).

It’s like driving at 40 miles per hour on the Interstate. *“But I want to be safe!”,* they clamor as normal traffic wizzes around them. Their obsession with safety causes more harm than good.

## What’s Jesse’s Answer?!

My big takeaway from this fun experiment: **I plan on starting higher than 4.0%, and adjusting as I go. **

On one hand, I would hate to start at 4.75%, then live through an “unlucky future” and ultimately run out of money.

But I would equally hate to start at 4.0%, then live through a “normal” (or better) future, and ultimately end up with many multiples of my original nest egg at death.

The “adjust as I go” takes this into account. If I need to be more conservative for a few years, I will be. If I can press on the gas for a few years, I’ll do that too.

Quite simply, **the biggest risk of the 4% rule is underspending your retirement potential**. **And it’s biggest flaw is its rigidity. **

It’s a numerical model. And a helpful one at that. But it’s not real life.

Thank you for reading! If you enjoyed this article, join **8500+ subscribers** who read my 2-minute weekly email, where I send you links to the smartest financial content I find online every week. You can **read past newsletters** before signing up.

-Jesse

*Want to learn more about The Best Interest’s back story?* Read here.

Looking for a great personal finance book, podcast, or other recommendation? Check out my favorites**. **

Was this post worth sharing? Click the buttons below to share!

Good article Jesse. I too have researched the 4% “rule” extensively. A huge oversight in its calculation is using index returns historically. Fund fees, trading costs, and taxes were much higher long ago, so the actual investor returns would have been 1-2%, or more, lower annually. Using some estimate of true net returns would have led to a lower SWR result. Even though those costs now are much lower, there is a theory that expected returns would be lower since they don’t have to compensate for the costs.

Lastly, covering more of your spending with SS, Tips ladders, and annuities, leaving a smaller portfolio balance to finance remaining spending will likely further increase annual spending, though lower remaining balance.

Just some observations. Thanks for going against the 4% grain.

Hi Dan – thanks for that intriguing feedback. Those are good facts to keep in mind, and an interesting (and logical) conclusion re: lower expected returns as compensation for lower frictional costs.

Best,

Jesse