Stop worrying about the 4% rule

[CanuckExpat]

Regarding CAPE values and affect on sustainable withdrawal rates, Kitces has done some good research and has several interesting articles on the topic, including Shiller CAPE Market Valuation: Terrible For Market Timing, But Valuable For Long-Term Retirement Planning. Mad Fientist I believe attempts to distill much of those articles, including the affect of valuations on his page Safe Withdrawal Rate for Early Retirees. The latter is a good article, but I think it is worth digging into the original Kitces work, though it is time consuming. This thread is long and I don't remember the contents, so sorry if this has been posted already.

It has been sometime since I read Kitces work on CAPE and withdrawal rates, but from what I recall, they have some predictive value over moderate (10 year  time periods) but neither short nor long time periods. The MadFI page has a tool (sign in required) that lets you estimate an updated withdrawal rate based on current CAPE values.

It's interesting work, but I didn't see anything that would make me immediately change my plans. The time periods are too removed for me to think of it having immediate actionable consequences, but you can draw your own conclusion. I think it comes to the advice mentioned earlier:

The key points of the trinity study are that saving 20-30 times your annual expenses should result in your stash lasting 30 years.

When it comes to portfolio theory it should be as simple as stating that a diversified portfolio (across asset classes and within asset classes) is a good idea. It's also a good idea to utilise a stock heavy portfolio because the no 1 risk to portfolio failure tends to be inflation which stocks protect against.

Based on these points a portfolio of 50/50 through to 90/10 (international stocks/domestic bonds) that is 20-30 times your annual expenses means that you have placed yourself in a statistically likely position for your retirement to be a success from a financial perspective. That is the best that you are going to get. Trying to mimic the best past performance or protect yourself perfectly is not possible as we can't predict the future.

The past can only offer broad general principles to follow. It can't offer completely detailed mathematically exact criteria to follow.

If it brings you joy to save 20x your expenses and retire, do it.
If it brings you joy to save 30x your expenses and retire, do it.
(Similar for anywhere between 20x-30x)
If it brings you joy to save less and retire, I would say be cautious and hesitant, but it's your life.
If it brings you joy to save even more and retire, I'd say you saved a lot, good for you, no harm done and it's your life.

The cautions about your personal rate of inflation, or misunderstanding your long term expenses are good ones and worth noting, but they fall out of the realm of what can be answered with historical stock and bond returns.

[TomTX]

He doesn't even need to search.  If he just read the thread we're current posting in, and some of the links it contains, he wouldn't have to repeat the same tired misconceptions over and over again.

You must be new here

Dude, I've been caught in the September that Never Ended since AOL first connected to the actual internet.

Read.
The.
Thread.

https://forum.mrmoneymustache.com/investor-alley/stop-worrying-about-the-4-rule/

[DavidAnnArbor]

A couple of years Kitces did just that for hypothetical retirees who pulled the trigger in 2000 and 2008, after the two more recent crashes and compared the performance of these portfolios so far to how things were looking for retirees in 1929, 1935, and 1966 (some of the worst years to retire in historically) after the same length of time.

This isn't ideal, but it does a good job of harvesting some information we do have about more recent retirements, rather than pretending we have absolutely no idea what the outcome for the 1987-2017 retiree will look like until January 1st 2018.
https://www.kitces.com/blog/how-has-the-4-rule-held-up-since-the-tech-bubble-and-the-2008-financial-crisis/

Wow that's encouraging to read. Thank you for posting that.

[Retire-Canada]

If it brings you joy to save even more and retire, I'd say you saved a lot, good for you, no harm done and it's your life.

I'd say there could be plenty of damage done by over saving that people either ignore or is long-term enough they don't appreciate until it's too late...health and relationship damage would be two examples. It's easy to focus on earning money to secure your family's future only to realize later on you've destroyed the family by not being there for decades in order to earn all that money. Same goes for health.

That's why I think it's pretty critical for FIRE success to open your mind to factors that are not numerically based and can't be spreadsheeted. These are also factors that our societal programming to work, work, work, spend, spend, spend....is designed to have you ignore so the risk is high you'll do just that!

[steveo]

Most tests of safe withdrawal rates use US-only data because that's the longest historical dataset in the public sector. Conceptually including investments across multiple countries as part of an international index provides greater diversification across imperfectly correlated assets, which would reduce sequence of returns risk and hence increase the minimum safe withdrawal rate above what has been calculated using data the USA alone. However, there are no public domain datasets of stock/bond performance for other countries that go back far enough to test this concept with any sort of statistical rigor.*

The Dimson, Marsh, and Staunton dataset does have that type of historical performance data for 20 plus countries that have accounted for the vast majority of world equity valuations for the past century, so someone who had access to it could run this test. Instead it seems to only get used whenever some idiot wants to write another "the 4% rule doesn't work in countries other than the USA" article about the groundbreaking finding that the 4% rule would not have protected retirees while whose entire countries were leveled to the ground during world war II.

*And the concern even if there were is that global markets are more integrated now than in the past, so the performance of US, German, and Brazilian stock market indices are likely more correlated than historical data would indicate, meaning we may well get less of a reduction in volatility from international investing than historical data would show, giving folks a false sense of security.

All of these discussions again miss the point. Runewell still seems to think that he can utilise the data that we currently have (historical data) to somehow optimise his future portfolio. I suggest Runewell read a bunch of stuff from William Bernstein.

https://www.amazon.com/Deep-Risk-History-Portfolio-Investing/dp/0988780313/ref=sr_1_3?s=books&ie=UTF8&qid=1500682456&sr=1-3&keywords=william+bernstein+investing
https://www.amazon.com/Skating-Where-Puck-Was-Correlation/dp/0988780305/ref=sr_1_8?s=books&ie=UTF8&qid=1500682456&sr=1-8&keywords=william+bernstein+investing

I got these books for free based on a recommendation on this board and they are really good.

You need to stop looking at the available data as somehow more predictive of the future than what it is. For instance as I think has already been mentioned gold based on the historical data-set looks and sounds like a great asset class to increase your potential SWR. The problem is that in the past you probably couldn't have actually bought gold to add that to your portfolio and now it will probably work completely differently than what it did in the past.

Markets are now also much more integrated as you state. So international stocks may have been less correlated in the past compared to what they are now.

It honestly comes down to not being able to make perfect decisions because you can't predict the future. You need to look at the broad principles coupled with some rational reasoning and then make smart informed decisions on the size of your stash and your portfolio. The good thing is that you also don't have to be perfect.

Micro analysing cape or P/E ratios or historical data to try and figure out the specific reason why portfolios fail or trying to get to a ridiculously low WR (unless you are working because you love it) all miss the point that we have imperfect information because we are to a degree trying to predict the future.

[steveo]

If it brings you joy to save even more and retire, I'd say you saved a lot, good for you, no harm done and it's your life.

I'd say there could be plenty of damage done by over saving that people either ignore or is long-term enough they don't appreciate until it's too late...health and relationship damage would be two examples. It's easy to focus on earning money to secure your family's future only to realize later on you've destroyed the family by not being there for decades in order to earn all that money. Same goes for health.

That's why I think it's pretty critical for FIRE success to open your mind to factors that are not numerically based and can't be spreadsheeted. These are also factors that our societal programming to work, work, work, spend, spend, spend....is designed to have you ignore so the risk is high you'll do just that!

The closer I get to FIRE the more I start thinking about the factors that are important to a successful retirement  that don't include money.  How am I going to spend my time ? Will I travel ? Will I downsize my house and if so when ? Will I work part time for a year or two to pay for some bigger expenses (for instance painting the house or a new car or a holiday) ?

[Retire-Canada]

The closer I get to FIRE the more I start thinking about the factors that are important to a successful retirement  that don't include money.  How am I going to spend my time ? Will I travel ? Will I downsize my house and if so when ? Will I work part time for a year or two to pay for some bigger expenses (for instance painting the house or a new car or a holiday) ?

That makes sense. When you are a decade out it doesn't make sense to sweat the details, but as you get closer to pulling the trigger you are motivated to work out your FIRE plan in finer detail. I'm doing the same thing.

Congrats on getting close. That's a wonderful feeling. :)

[DavidAnnArbor]

The closer I get to FIRE the more I start thinking about the factors that are important to a successful retirement  that don't include money.  How am I going to spend my time ? Will I travel ? Will I downsize my house and if so when ? Will I work part time for a year or two to pay for some bigger expenses (for instance painting the house or a new car or a holiday) ?

That makes sense. When you are a decade out it doesn't make sense to sweat the details, but as you get closer to pulling the trigger you are motivated to work out your FIRE plan in finer detail. I'm doing the same thing.

Congrats on getting close. That's a wonderful feeling. :)

Me too I'm in the same boat.

[maizefolk]

Micro analysing cape or P/E ratios or historical data to try and figure out the specific reason why portfolios fail or trying to get to a ridiculously low WR (unless you are working because you love it) all miss the point that we have imperfect information because we are to a degree trying to predict the future.

Is the 4% rule immune from this criticism?  People have argued that the US market has also changed over time.  Therefore isn't trying to predict the adequacy of future withdrawal rates merely another flawed attempt to predict the future usinh imperfect information?

Which people have argued that the US market has changed over time? (I'm assuming you're referring to systematic changes, not "in some years stock prices go up, but in other years stock prices do down.)

The 4% rule is certainly imperfect. For example, if the USA (or whatever country you're FIRE-ing in) is in a war and loses, with significant destruction of infrastructure and loss of life, the 4% rule probably will not allow you to maintain your lifestyle without adjustment. Predicting whether or not such a war will occur in the next 50 years isn't something anyone can do with any degree of certainty.

The advantage of the 4% rule (and historical backtesting generally) is that it requires fewer assumptions* and uses few variables which means that while predictions may not be precise there is less risk of systematic errors which would produce extremely misleading results. Complex models are much more prone to overfitting, which DOES produce systematic errors, combined with misleadingly high precision of prediction.

A general rule of thumb is that you should have no more than one variable for every 10 or 15 cases in your data set. So a model to explain what happened in 15 elections should ideally contain no more than one or two inputs. By a strict interpretation, in fact, not only should a model like this one not contain more than one or two input variables, but the statistician should not even consider more than one or two variables as candidates for the model, since otherwise he can cherry-pick the ones that happen to fit the data the best (a related problem known as data dredging). If you ignore these principles, you may wind up with a model that fits the noise in the data rather than the signal.

To put the above quote in context, historical stock market data goes back ~150 years at this point. That means we have only something like 5 datapoints from non-overlapping 30 year retirements to examine.

*Essentially just that the future is more likely to be similar to the past than to be completely and totally different from the past.

[Telecaster]

But even "inflation" can be considered a misleading indicator.  I agree that technically CPI is included in the 4% rule, and if you aren't buying new iPhones and a Tesla, maybe that is a good enough benchmark, but it is a broad basket of goods and also makes value judgements for you about tangible cost discounted by added features.  So I'd argue that a 30 and 40 y.o. experiences higher inflation (compounded if you are raising kids) than what the CPI reports.  Of course, with the reported figure being so low for so long, Americans have been able to ignore this, but go to other countries where their prices in their local currency are 'outrageous'...  We are very fortunate to enjoy a strong dollar and the benefits of importing cheap goods. 

While you're employed, your income generally rises with inflation and when you are collecting Social Security, that adjusts for COL, but ER'd folks are SOL :).

That's easy enough to deal with.  Just start with a lower SWR.

[steveo]

Micro analysing cape or P/E ratios or historical data to try and figure out the specific reason why portfolios fail or trying to get to a ridiculously low WR (unless you are working because you love it) all miss the point that we have imperfect information because we are to a degree trying to predict the future.

Is the 4% rule immune from this criticism?  People have argued that the US market has also changed over time.  Therefore isn't trying to predict the adequacy of future withdrawal rates merely another flawed attempt to predict the future usinh imperfect information?

You are missing the big picture if you are doing this type of analysis. That is why I suggested you read up on portfolio theory a bit. Trying to predict the right portfolio for the future based on historical data is doomed to failure.

The big picture in relation to how much money you will need will never be completely accurate. If you though take a statistically valid and rational position and utilise the 4% rule that is probably the best that you will do. Just don't retire and think that everything is fine from this point on. You are living in the real world and you will probably need to adjust. One thing I feel confident stating is that the larger your stash the less chance you have of running out of money assuming you are using a valid portfolio.

I think trying to predict exactly future WR's or portfolio's shows the same lack of bigger picture understanding.

[dividendman]

Yeah, it generally comes down the absurdity of trying to predict the world 20 years out.

20 years ago - no smartphones
Today - smartphones are a massive industry that has spun off even more industries worth trillions of dollars, everyone has a device on their person that probably equals the computing power of all computers 20 years ago

20 years ago - no social media
today - social media is a massive industry, blah blah

20 years ago - no 9/11
today - repercussions of 9/11 are still impacting the planet

20 years ago - everyone is scared of peak oil and the world running out of oil
today - oil glut, so much oil everywhere! shale, etc.

There are many more examples.

If you have saved up 25 years of spending, in my opinion, it's not going to help to save more because you're just incapable of figuring out what the world will look like 15 years from now, never mind 25.

Also, the pace of change is accelerating, so you're even less likely to figure anything out.

[Retire-Canada]

If you have saved up 25 years of spending, in my opinion, it's not going to help to save more because you're just incapable of figuring out what the world will look like 15 years from now, never mind 25.

Agreed that predicting what will happen is impossible, however, there is one benefit I can think of to saving more than 25x annual spending. Every extra year you work past that point is one year you'll be closer to dying and that makes your chance of running out of money less since your retirement will be shorter. Of course this approach at risk mitigation has an obvious downside. ;)

[dandarc]

If you have saved up 25 years of spending, in my opinion, it's not going to help to save more because you're just incapable of figuring out what the world will look like 15 years from now, never mind 25.

Agreed that predicting what will happen is impossible, however, there is one benefit I can think of to saving more than 25x annual spending. Every extra year you work past that point is one year you'll be closer to dying and that makes your chance of running out of money less since your retirement will be shorter. Of course this approach at risk mitigation has an obvious downside. ;)

Those graphs combining the portfolio success and mortality really make the point.  People get up in arms over that tiny slice in the middle, when odds are extremely high that by the time you die, you're just fine from a running-out-of-money perspective either because your portfolio succeeded or you died young enough.  One look and some thought was enough to get me completely over "is 4% enough?" and leaning more and more towards "5% is probably plenty".

[Retire-Canada]

One look and some thought was enough to get me completely over "is 4% enough?" and leaning more and more towards "5% is probably plenty".

Yes! I'm excitedly waiting for my investments to reach 5%WR. 4%WR seems sane, but cautious. Going any lower and for me personally the trade off between opportunity cost for extra time worked plus the negative health impacts don't compute.

I'm far more worried about my health failing or death than I am running out of money once I get into the 4%WR-5%WR range.

[EscapeVelocity2020]

But even "inflation" can be considered a misleading indicator.  I agree that technically CPI is included in the 4% rule, and if you aren't buying new iPhones and a Tesla, maybe that is a good enough benchmark, but it is a broad basket of goods and also makes value judgements for you about tangible cost discounted by added features.  So I'd argue that a 30 and 40 y.o. experiences higher inflation (compounded if you are raising kids) than what the CPI reports.  Of course, with the reported figure being so low for so long, Americans have been able to ignore this, but go to other countries where their prices in their local currency are 'outrageous'...  We are very fortunate to enjoy a strong dollar and the benefits of importing cheap goods. 

While you're employed, your income generally rises with inflation and when you are collecting Social Security, that adjusts for COL, but ER'd folks are SOL :).

That's easy enough to deal with.  Just start with a lower SWR.

Go have a chat with some retirees who were working in the 1970s.  Annual raises were double digits and home mortgages/rent/staples similarly so.  Compounding double digit inflation, especially for an ER, would be no joke.  You should run a few scenarios on cFiresim where you switch from historic inflation to having 'extra spending' with constant inflation at 5% and above.  Even 3% starting SWR has failures.  Interestingly, since the 1980's markets have been as buoyant as they were during the high inflation 70's but without the high inflation this time around.  It's no wonder retirees are feeling good these days.

[steveo]

One look and some thought was enough to get me completely over "is 4% enough?" and leaning more and more towards "5% is probably plenty".

Yes! I'm excitedly waiting for my investments to reach 5%WR. 4%WR seems sane, but cautious. Going any lower and for me personally the trade off between opportunity cost for extra time worked plus the negative health impacts don't compute.

I'm far more worried about my health failing or death than I am running out of money once I get into the 4%WR-5%WR range.

I can understand going lower but it's not really going lower from my perspective. It's more stating I'm going to retire and spend a lot more money. So prior to retiring I'll save up enough to go on some fancy overseas holidays or something like that. It's more a play on expenses.

Truthfully though I don't really care about fancy holidays. I'm happy just reading books and going for walks every day. 5% has always been my uncle point but I may choose to go to 4%. I really don't think though this is required. I think I'm contemplating this now just to get to a million dollars in investable assets. I wonder if for a lot of people it's more about keeping score than it is about a safety net.

[Retire-Canada]

I wonder if for a lot of people it's more about keeping score than it is about a safety net.

I suspect a lot of people are so conditioned to work that facing retirement is very challenging and the fear around that is a great motivator to look for reasons to keep working. Shooting for a low withdrawal rate accomplishes that goal.

[EscapeVelocity2020]

I wonder if for a lot of people it's more about keeping score than it is about a safety net.

I suspect a lot of people are so conditioned to work that facing retirement is very challenging and the fear around that is a great motivator to look for reasons to keep working. Shooting for a low withdrawal rate accomplishes that goal.

I think it'll be interesting in a generation or two, when FIRE community folks prove that early retirement at 4% (for 30 - 40 yo's) is a great lifestyle, if this kicks off a 'new normal'.  There's already a lot of upheaval toward a 'gig economy' / informal-employment like Uber, Etsy, AirBnB, blogging...  I wholly believe that the vast majority of the American workforce feels overworked, under-compensated, and unfulfilled.  I wish there were a quick solution, but not everyone can get the traditional employment opportunity I got just like not everyone will get a FIRE like Pete's.  But there seems to be more folks on the cusp of ER / ERE, just not yet a critical mass / tipping point.  Maybe in 5 - 10 years?  These are interesting times, whether we are creating the wave or just riding it...

[bob22]

I think you will see FIRE become even bigger. By phasing out "traditional" defined benefit pensions, companies are helping this transition. I actually think that I'm one of the few at my company that likes the fact that they've phased it out in favor of higher levels of defined contribution to 401(k) type plans. It really means that my financial ability to retire is (even more) de-coupled from my age.

[dividendman]

If you have saved up 25 years of spending, in my opinion, it's not going to help to save more because you're just incapable of figuring out what the world will look like 15 years from now, never mind 25.

Also, the pace of change is accelerating, so you're even less likely to figure anything out.

The world has changed, but it doesn't seem like the need for retirement savings has hardly changed at all.  We still invest in stocks and mutual funds and now ETFs.  A lot more investment opportunities have opened up during this time.  Perhaps the biggest change is that I can go on my computer and buy a basket of stocks in Egypt if I want, effortlessly and at little or nor cost, in only a minute. 

There is also a lot of research on investing, and twenty more years of investing ups and downs on the record books, but I don't see change in the world as a big problem for the individual investor.

Yes, there will probably (and maybe not even that high of a probability) be retirement savings needed far down in the future (assuming tech gains, etc. don't allow us to have a UBI).

That's not my point. My point is, since we can't predict these massive changes, why do we think we can predict market returns? Any static plan assuming any variables is necessarily inadequate over long time frames because we just don't know what is going to happen.

As has been said many times, the likelihood of any retirement plan succeeding depends more on the agility of the retiree to adapt to changing conditions rather than the size of the portfolio (after some sizable amount) precisely because conditions change so frequently.

Let's say someone saves up 20x their yearly earnings and then says "fuckit, i'm putting this into TIPS and not working". It'll probably last at least 20 years. If it turns out in year 10 that this person realizes they aren't likely to die in 10 more years, do we expect they'll just continue to draw down the remaining 10 years of savings until they're eating cat food? I don't think anyone on these forums would do that. They would adapt.

Finally, I read upthread about 3% and 2% WR. This starts to get into the really silly territory. If you're 40 and have 50x savings - you've won. Junk that 50x into TIPS and you're fine. You're not living to 90, don't worry (and if you do you still have social security not to mention everything could be free by then or some other gigantic change like WW3 makes savings irrelevant).

[tooqk4u22]

Finally, I read upthread about 3% and 2% WR. This starts to get into the really silly territory. If you're 40 and have 50x savings - you've won. Junk that 50x into TIPS and you're fine. You're not living to 90, don't worry (and if you do you still have social security not to mention everything could be free by then or some other gigantic change like WW3 makes savings irrelevant).

As conservative as I am even I know that anything about 2% is silly talk.  4% to 3%...ok...and given the point in the cycle, really understandable 3% only takes 33% more money and depending on your earning status and SR this might not be a big deal or take that long.  But 2% - the reason why that is ridiculous is because it takes 100% more money than 4% (and 67% more money than 3%) to accomplish, that's why it is so safe, and that doesn't factor in that even in this low rate environment divis & interest (for intermediate term or longer) are 2% or better and PE or CAPE are also higher than 2%.  S

[sol]

It's not going to become the new normal anytime soon, with as little as people save already.  A 19% portfolio survival with a 50-yr timeframe doesn't appeal to me either.

You keep quoting these ridiculous scenarios without any supporting evidence or citations.  If you had actually read this thread, which we all know you have not and will not, you would know that the 60, 50, and 40 year times frame are barely different from the 30 year time frame in terms of success rates, for the asset allocations this community uses. 

You can of course find higher failure rates if you choose stupid asset allocations, which seems to be what you're doing here. 

How many times do I have to repost this image before you go back and read this thread explaining it?



edited to add:  look at that top left corner.  An inflation adjusted 3% SWR has never failed, in the history of the US stock market.  Never. 

[MDM]

edited to add:  look at that top left corner.  An inflation adjusted 3% SWR has never failed, in the history of the US stock market.  Never.

The withdrawal rate in question was 4%.  cFiresim is telling me 19% failure rate - what am I doing wrong. Don't you just have to change one number.

If your time period is 60 years, 19% = 1 - ~81% seems consistent with the chart, does it not?

[onewayfamily]

But there seems to be more folks on the cusp of ER / ERE, just not yet a critical mass / tipping point.  Maybe in 5 - 10 years?  These are interesting times, whether we are creating the wave or just riding it...

I'm definitely seeing more and more of those 3-minute, simplified (usually too simplified) snippets on different news and morning shows profiling someone who has or is planning on retiring early...if that's any measure.

[dougules]

I think it will become less likely.  If more and more people need to direct some of their paycheck to a 401k and fail to do so sufficiently, the acronym will stand for
Financially Inadequate, Retirement Eliminated.

I think the main thing that will stop FIRE becoming significantly more common, is the simple and age-old fact of life that most people aren't able/willing to delay gratification to such a degree as to enable FIRE or anything close to it. This is the same thing that keeps our economy chugging along at its current rate though I suppose, so I probably shouldn't complain too much.

I think delaying gratification is hard enough for those of us that can.

The people that impulse purchase are bad for the economy in the long run.  They go crazy in the good times, then have nothing left in the bad times, so they make the economy more unstable. 

[TomTX]

Micro analysing cape or P/E ratios or historical data to try and figure out the specific reason why portfolios fail or trying to get to a ridiculously low WR (unless you are working because you love it) all miss the point that we have imperfect information because we are to a degree trying to predict the future.

Is the 4% rule immune from this criticism?  People have argued that the US market has also changed over time.  Therefore isn't trying to predict the adequacy of future withdrawal rates merely another flawed attempt to predict the future usinh imperfect information?

If you are waiting for perfect information, you will work unt you die.

Congratulations. You didn't have a portfolio failure.

But you guaranteed a life failure.

Look at the big picture. Every year you keep working, you have sacrificed a significant fraction of healthy, active retirement.

[Retire-Canada]

Look at the big picture. Every year you keep working, you have sacrificed a significant fraction of healthy, active retirement.

+1 - Years at the prime of your life. You can make more money, but you can't get more time.

[dividendman]

Micro analysing cape or P/E ratios or historical data to try and figure out the specific reason why portfolios fail or trying to get to a ridiculously low WR (unless you are working because you love it) all miss the point that we have imperfect information because we are to a degree trying to predict the future.

I wasn't trying to figure out a reason why portfolios fail or trying to get to a ridiculously low WR.

My whole point was that I think that the composition of one's portfolio should shift in response to high P/E ratios (or high P/B ratios). 

If you're retired on a 4% withdrawal rate and the market drops 30%, you probably adjust your strategy - you consider living on less for awhile, or possible pick up some side income. 

I think you should look at the composition of your portfolio similarly.  You look at the 21.5 P/E and 3.0 P/B ratios on the US stocks and the 16.6 P/E and 1.6 P/B ratios on the non-US stocks and maybe think Hmmm historical data tells me non-US should perform better going forward until this imbalance equalizes, let me sell of some of my VOO and pick up a bit more of this VEU just to be safe.

Um... isn't that just portfolio balancing? I have 50% VTI/VOO, 30% VEU, 20% BND.... so if US stocks tank i'll necessarily be buying more of them when I re-balance (same with x-US and bonds).

I'll be doing what you're saying automatically.

If you're saying i should change the weightage based on market conditions then I think you can fool yourself into thinking you can time the market which is dangerous and why investors don't achieve average market returns over the long run.

[Tyson]

Runewell, you like the idea that things like CAPE and other measures can predict the future.  They cannot.  But as long as you feel like they can, you'll continue to hold on to it as a means to have some level of control over the future. 

If you're going to index, at some point you'll have to accept that the market is gonna do what it's gonna do, and the only rational response is to just let it ride (ie, buy and hold).  Then once a year, re-balance your portfolio so you maintain your 80/20 (or 60/40 or whatever) stocks/bonds ratio.

[AdrianC]

Um... isn't that just portfolio balancing? I have 50% VTI/VOO, 30% VEU, 20% BND.... so if US stocks tank i'll necessarily be buying more of them when I re-balance (same with x-US and bonds).

I'll be doing what you're saying automatically.

If you're saying i should change the weightage based on market conditions then I think you can fool yourself into thinking you can time the market which is dangerous and why investors don't achieve average market returns over the long run.

Quibble: you're not going to receive average market returns over the long run because you don't own the market - you're biased towards US stocks. Is that a market-timing call?

[dividendman]

Um... isn't that just portfolio balancing? I have 50% VTI/VOO, 30% VEU, 20% BND.... so if US stocks tank i'll necessarily be buying more of them when I re-balance (same with x-US and bonds).

I'll be doing what you're saying automatically.

If you're saying i should change the weightage based on market conditions then I think you can fool yourself into thinking you can time the market which is dangerous and why investors don't achieve average market returns over the long run.

Quibble: you're not going to receive average market returns over the long run because you don't own the market - you're biased towards US stocks. Is that a market-timing call?

Yeah, i'm also not going to achieve it due to bonds :(

I don't really have a good reason for doing the 5:3 ratio between US and international. I just started off with that since I wanted to get started and read about it on bogleheads and don't want to change now. There is a good argument to be made that it should be half-half since the US represents about half of the global market.

It's not timing because i'm not changing it based on anything that happens... i set it and that's it. Timing implies I will change the ratio based on the events of the day... I will not.

[Tyson]

Runewell, you like the idea that things like CAPE and other measures can predict the future.  They cannot.  But as long as you feel like they can, you'll continue to hold on to it as a means to have some level of control over the future. 

If you're going to index, at some point you'll have to accept that the market is gonna do what it's gonna do, and the only rational response is to just let it ride (ie, buy and hold).  Then o8nce a year, re-balance your portfolio so you maintain your 80/20 (or 60/40 or whatever) stocks/bonds ratio.

Vanguard disagrees with you, saying on page 2 of https://personal.vanguard.com/pdf/s338.pdf

We confirm that valuation metrics such as price/earnings ratios, or P/Es, have had an inverse or mean-reverting relationship with future stock market returns, although it
has only been meaningful at long horizons and, even then, P/E ratios have “explained” only about 40% of the time variation in net-of-inflation returns. Our results are similar whether or not trailing earnings are smoothed or cyclically adjusted (as is done in Robert Shiller’s popular P/E10 ratio).

They say that: historically higher P/E's (and CAPEs) have result in lower returns going forward. 

I also have no reason to believe that the selection of a portfolio at one point in time precludes making adjustments later. 

I don't care who says it - it's wrong to listen to it and it's even more wrong to act on it.  Vanguard doesn't know.  Buffet doesn't know, runewell doesn't know.  No one knows, that's the point.  That's why you create your asset allocation and your investment plan and you stick with it. 

The approach your taking (removing $$ from high CAPE investments and moving them to a low CAPE investment) involves both trying to time the market AND pick winners.  That's going to end badly.  That type of thinking is exactly why "active" investors don't beat passive investors over time. 

You might wish it weren't true.  You might wish that your magic CAPE numbers give you an edge and insight over all the dumb indexing schlubs.  But you're wrong.  And it's gonna cost you money. 

But I can see that what I'm saying isn't even making a dent with you, so I'll just stop. 

[runewell]

So let's get this straight: 
You don't own the market, stock portion of your portfolio is overweighted US and you have 20% bonds.

Do you rebalance?  Isn't that market timing too?  If having a certain allocation is important, why don't you reallocate more often then you do?  If you only rebalance as often as you do, do you think you have some special information on how often to rebalance?   

[dividendman]

So let's get this straight: 
You don't own the market, stock portion of your portfolio is overweighted US and you have 20% bonds.

Do you rebalance?  Isn't that market timing too?  If having a certain allocation is important, why don't you reallocate more often then you do?  If you only rebalance as often as you do, do you think you have some special information on how often to rebalance?   

I think the whole point is the rebalancing occurs on an arbitrary timeline that is independent of what's going on in the world, so it's not timing the market.

Let's just define timing the market since there appears to be confusion.

"timing the market" means using trends in GDP, stock valuations, geo-political events or other data to buy or sell positions in the market in an attempt to beat the market average.

Saying on Feb 28th I will re-balance is precisely NOT timing the market because you're going to rebalance regardless of what happens from now until then.

[steveo]

I don't care who says it - it's wrong to listen to it and it's even more wrong to act on it.  Vanguard doesn't know.  Buffet doesn't know, runewell doesn't know.  No one knows, that's the point.  That's why you create your asset allocation and your investment plan and you stick with it. 

This is exactly the approach to take. My asset allocation isn't perfect. I also recognise that it will probably turn out just fine so I stupidly don't look at any indicators and invest whenever I save up 10 grand.

[steveo]

So let's get this straight: 
You don't own the market, stock portion of your portfolio is overweighted US and you have 20% bonds.

Do you rebalance?  Isn't that market timing too?  If having a certain allocation is important, why don't you reallocate more often then you do?  If you only rebalance as often as you do, do you think you have some special information on how often to rebalance?   

No it's not. Timing the market like you are suggesting is utilising leading indicators or opinions or whatever. Re-balancing is simply adjusting to market conditions post some market event.

[Classical_Liberal]

I think you should look at the composition of your portfolio similarly.  You look at the 21.5 P/E and 3.0 P/B ratios on the US stocks and the 16.6 P/E and 1.6 P/B ratios on the non-US stocks and maybe think Hmmm historical data tells me non-US should perform better going forward until this imbalance equalizes, let me sell of some of my VOO and pick up a bit more of this VEU just to be safe.

This is one of the few things I can "sort-of" agree with runewell. BUT (big but),  not based on market valuations (too unpredictable in the short term); rather based on what someone's personal financial goals are presently and in the medium term (next five years or so).  Sometimes it makes sense to adjust portfolio composition to meet specific goals.  Maybe to decrease volatility for a period, increase liquidity, cash flow, or to hedge against certain risks at particular points in life. In this sense, I think a "set it and forget it" mindset can be counterproductive.

[AdrianC]

Yeah, i'm also not going to achieve it due to bonds :(

I don't really have a good reason for doing the 5:3 ratio between US and international. I just started off with that since I wanted to get started and read about it on bogleheads and don't want to change now. There is a good argument to be made that it should be half-half since the US represents about half of the global market.

It's not timing because i'm not changing it based on anything that happens... i set it and that's it. Timing implies I will change the ratio based on the events of the day... I will not.

Yeah. I was just ribbing a little.

This year I increased our international and decreased US. Why? Well, I've been thinking for a while that I should be closer to the world market. The valuation disparity certainly helped me to make the decision. So far I'm a market-timing genius...

But we're getting off topic. I do have an on topic point/question - next post.

[CanuckExpat]

If it brings you joy to save 20x your expenses and retire, do it.

It is good to have at least a 3/4 chance.

Historically, 5% withdrawal hasn't done that bad for the US market, barring great depressions and stagflation:


from here

If it brings you joy to save even more and retire, I'd say you saved a lot, good for you, no harm done and it's your life.

I'd say there could be plenty of damage done by over saving that people either ignore or is long-term enough they don't appreciate until it's too late...health and relationship damage would be two examples. ..

I take your point, and I think it's a good one. I'm optimistic in that I think people will choose what works best for their personal situation and risk tolerance. It's like an asset allocation question. Should be selected based on your willingness and ability to take risks.

My whole point was that I think that the composition of one's portfolio should shift in response to high P/E ratios (or high P/B ratios). 

If you like that sort of thing and are wondering about safe withdrawal rates, you might be interested in some of the stuff Kitces previously looked at Resolving the Paradox – Is the Safe Withdrawal Rate Sometimes Too Safe? . He looks at adjusting withdrawal rate based on valuations.

I'm only posting the most interesting tables, but read the whole thing


His suggested "rules" (or at least the ones studied more in depth) based on those results:


And as a follow up from same author: Dynamic Asset Allocation and Safe Withdrawal Rates looks at effect of reducing/increasing equity allocation by 20% in rule based fashion based on valuations at top/bottom third of historical range. It's not as easily summarized just from the results tables.

The work is interesting, I'm curious about the dynamic asset allocation stuff. Not curious enough to change my own strategy yet, but keeping an open mind. Based on your like for valuations, thought you might find that work interesting.

[Classical_Liberal]

If you like that sort of thing and are wondering about safe withdrawal rates, you might be interested in some of the stuff Kitces previously looked at Resolving the Paradox – Is the Safe Withdrawal Rate Sometimes Too Safe? . He looks at adjusting withdrawal rate based on valuations.

I've read this before and its a good article.  While I like CAPE because it's really the only measure shown to have any predictive value at all.  It's value is only in long term average return predictive power.  Even then, there are a lot of assumptions.  The main being that CAPE mean remains unchanged, all of which was discussed up thread.

I see no major issue with slightly adjusting a WR based on CAPE, but starting at 4% is on the very conservative edge. 

As far as adjusting an entire portfolio based on a single valuation indicator whose predictive power is, at best, limited to average returns over a decade? ...  Not a great idea, IMO.  The same issues arise as in any attempts to time the market.  Even if an investor manages to accurately predict lower equity returns in the next decade, how will that sequence bear out?  How is that actionable from an allocations standpoint for an investor that is passively world indexed?  Even with the chart provided WR was only slightly optimized with higher bond (ie not 100% equities, which I would argue against in almost every case anyway) allocations in high CAPE situations and those situations didn't start with ZIRP.

[fuzzhead1506]

While I like CAPE because it's really the only measure shown to have any predictive value at all. 

That is actually false.  There are prolly a dozen or more that do a pretty predictive job, and it has been argued that CAPE is not even the best.  You could even also make up your own silly measure that has some predictive power over past results.  If you dare to go down the rabbit hole of looking further into this, I suggest you start here and read the BOTH of the author's articles!

http://www.philosophicaleconomics.com/2013/12/the-single-greatest-predictor-of-future-stock-market-returns/
http://www.philosophicaleconomics.com/2013/12/valuation-and-returns-adventures-in-curve-fitting/

runewell,

I think that there is a large cohort of people that think if you so dare to time the market, you might be harming the value of the 4% rule. I think otherwise.

I think that you prolly could time the market a little.  And even if you failed to do well, I'd actually argue you could take as much as 6% or 7% and prolly be ok in pretty much any environment - but that could just be my optimism gun talking. 

The 4% rule works - get over it.  ;)

Edited: for clarifications

[TomTX]

I liked this article, it was very much along the lines of what I was thinking.  Safe withdrawal rates and portfolio longevity can be enhanced with a system of rules that recognizes that less equity exposure is ideal during higher P/E periods.  It is exactly the sort of mindset I was thinking about all along.  My portfolio was probably 85% US stocks until I lowered it to 75% just recently, and I could possibly get it down into the 60's eventually here.  If balancing your portfolio based on P/Es is wise during retirement, it's probably wise before retirement as well.

Thanks so much for posting it.

You have still failed to properly address how "Earnings" today are defined quite differently than "Earnings" 30 years ago, let alone 50 or 100 years ago, with the trend being generally more restrictive on how you report "Earnings" - thus adding an upward forcing of "P/E" over time.

A "higher P/E" is oversimplified to the point of being wrong. CAPE is even worse.

You've already been given links on the topic.

[bob22]

I agree that the 4% rule will probably work throughout retirement. But let's say I'm just really fiscally conservative, and had the ability to vary my discretionary spending 3-4% to maximize the chances that my investments will last (and that I'll have something to leave to my kids). I'd rather not just spend 3% all the time. But, do I spend 4% when the market is good and 3% when it isn't? Should I be contrarian and do the opposite? Or should I do something more complicated like try to anticipate using a market indicator or several?

[MDM]

I agree that the 4% rule will probably work throughout retirement. But let's say I'm just really fiscally conservative, and had the ability to vary my discretionary spending 3-4% to maximize the chances that my investments will last (and that I'll have something to leave to my kids). I'd rather not just spend 3% all the time. But, do I spend 4% when the market is good and 3% when it isn't? Should I be contrarian and do the opposite? Or should I do something more complicated like try to anticipate using a market indicator or several?

Perhaps consider Variable Percentage Withdrawal (VPW) - Bogleheads.org.

[maizefolk]

I agree that the 4% rule will probably work throughout retirement. But let's say I'm just really fiscally conservative, and had the ability to vary my discretionary spending 3-4% to maximize the chances that my investments will last (and that I'll have something to leave to my kids). I'd rather not just spend 3% all the time. But, do I spend 4% when the market is good and 3% when it isn't? Should I be contrarian and do the opposite? Or should I do something more complicated like try to anticipate using a market indicator or several?

I did some simulations before that suggested if you spend 4% of your starting portfolio balance when your networth is above your starting point and 3.3% of your starting portfolio balance when your network is below your starting balance you wouldn't run out of money in any historical retirement window using retirement lengths from 20 years to 100 years.

https://forum.mrmoneymustache.com/investor-alley/the-4-rule-for-those-in-their-mid-20's-completely-unrealistic/msg1368531/#msg1368531

[maizefolk]

networth is above your starting point and 3.3% of your starting portfolio balance when your network is below your starting balance you wouldn't run out of money in any historical retirement window using retirement lengths from 20 years to 100 years.

https://forum.mrmoneymustache.com/investor-alley/the-4-rule-for-those-in-their-mid-20's-completely-unrealistic/msg1368531/#msg1368531

Theoretically the amount of net worth you need every year is less than the year before, since your expected life expectancy is lower.  Even though it is hard to tell what your net worth goal line should look like, you could probably squeeze out a few more tenths of withdrawal if you had some such adjustment.  I don't know exactly what this would look like though.

Yup, that's the idea behind the death and bankruptcy graphs from further upthread.



If you read up on the variable percentage withdrawal method (link), they're also using actuarial data to increase your withdrawal rate by spending more and more of your portfolio each year the expected number of years you have left in your life decreases.

[matchewed]

https://web.stanford.edu/~wfsharpe/art/active/active.htm

In short they have to under-perform because of math and logic. All things being equal on average the performance of passive and active investors is equal. The costs for the active investor is greater therefore after cost performance is lower for the active investor.

[Retire-Canada]

This is why it's so hard to beat the market: https://www.bloomberg.com/news/articles/2017-04-09/lopsided-stocks-and-the-math-explaining-active-manager-futility

Trading costs and taxes also provide additional headwinds.

[Le Barbu]

networth is above your starting point and 3.3% of your starting portfolio balance when your network is below your starting balance you wouldn't run out of money in any historical retirement window using retirement lengths from 20 years to 100 years.

https://forum.mrmoneymustache.com/investor-alley/the-4-rule-for-those-in-their-mid-20's-completely-unrealistic/msg1368531/#msg1368531

Theoretically the amount of net worth you need every year is less than the year before, since your expected life expectancy is lower.  Even though it is hard to tell what your net worth goal line should look like, you could probably squeeze out a few more tenths of withdrawal if you had some such adjustment.  I don't know exactly what this would look like though.

Yup, that's the idea behind the death and bankruptcy graphs from further upthread.



If you read up on the variable percentage withdrawal method (link), they're also using actuarial data to increase your withdrawal rate by spending more and more of your portfolio each year the expected number of years you have left in your life decreases.

Maizeman, is this interesting graph available for anyone to enter self situation or scenario? I would like to compare ER @ 45 with a 5% WR vs a later retirement @ 50-55 with a lower WR (3.5-4%) without splitting hairs. Your graph talks so much!