How hard is it to RE at the market top (2014)?

[Animorph]

I like the glide path, but I think it works in the generalized case only because if you are in danger of draining the portfolio it will raise the equity percent and that might dig you out of the hole. Plus the hole won't be that that big with all the bonds. You could also do that dynamically.

With a standard 4% initial WR plus inflation, I think what you want to do is get to the point where the withdrawal amount has become (and stays) less than about 3% of your portfolio. At that point FIRECalc now says you have a 100% success rate, and by that standard you have avoided any early sequence of returns failures.

The fastest way to such a reduced WR is an aggressive portfolio and fast growth in the first few years of retirement (or saving more initially). That requires some luck. You can go conservative for the first few years, but then it will take longer to grow enough to reach the 3% WR. And instead of being vulnerable for the first few years, you're vulnerable for a few extra years. So some luck will still be required. I still like the conservative approach, but I have not seen data that proves it is any better. Psychologically it should be easier to take an early market dip with a conservative portfolio.

Anyway, if I'm taking out less than 3% of my portfolio (a higher percentage in later years), I'm comfortable with whatever the market is doing.

[Totoro]

Quote:

Originally Posted by Animorph You can go conservative for the first few years, but then it will take longer to grow enough to reach the 3% WR. And instead of being vulnerable for the first few years, you're vulnerable for a few extra years.

True. What it does is spread out a concentrated risk in one year over a few years.

So you'll be less at risk in year one, and a bit more in year four.

For those aiming to reduce maximum risk exposure at any time, that makes sense.

[photoguy]

Totoro/Audreyh1

It's good to know that you both like the work. I'm generally very skeptical of all academic papers because I've seen how errors can come up. Replication and independent verification is always a good thing and especially important for withdrawal studies where you don't get second chances once you pull the plug. As this work is newer, I don't believe the withdrawal method has been tested as much as other approaches.

I do think the general idea (of rising equity allocation) is novel and it definitely runs counter to what most are recommending. And I'm not categorically against the idea, but just think for studies like this, the details are extremely important as changes in assumptions can result in very different outcomes/conclusions.

It does look like the author's have done a good job with both testing various assumptions and looking at more than just the single binary metric of success/fail (papers that look at just success/failure is pet peeve of mine). However, in their own tables, the outcomes are mixed depending on the scenario. In some cases, a rising glidepath improves success rates by a few percent (e.g. 71 to 74% - not sure if this is significant in a practical sense) but in others the glidepath hurts (scenario b -- lower future returns). Thus my recommendation to be cautious and really understand the method (and study) in detail.

[Totoro]

Photoguy: I've been in academia myself for a few years, I know how messy the kitchen can get

[Cat-tirement]

Quote:

Originally Posted by W2R No matter what the market conditions, IME eventually you get to the point where you are so determined to retire, and have your finances so well in order, that a supreme confidence develops that you can, and WILL, retire on schedule come h*ll or high water and nothing can stop you. Once you feel that way, then nothing can stop you as you charge towards your goal like an unstoppable locomotive.

I can relate to that. Once I realized ER was feasible, it went from "can I do it" to "how soon can I do it". With that kind of thinking (and inspiration from these fora) I moved the target date up a few times, in order to retire ahead of the original ER target. (Only 42 working days to go...)

[Animorph]

OK, here's a FIRECalc answer that I like.

You retire at whatever SWR you feel comfortable with. FIRECalc says if you have at least 20% equities, initial WR's of 3% or less, adjusted for inflation, have a 100% success rate (all default FIRECalc settings, including a 30 year length). Then you monitor your yearly withdrawal percentage.

If you find your WR increasing to 3.5%, you need to adjust equities to 30% to 80% to maintain a 100% success rate.
At 4% WR equities should be close to 65% to optimize the success rate at about 96%.
At 4.5% WR equities should be close to 85% for a peak success rate of about 83%.
At 4.75% WR equities should be close to 95% for a peak success rate of about 78%.
At 5% WR and above, 100% equities provides the best success rate, starting at about 74% and decreasing as WR increases.

The maximum improvement in success rate between 50% equities and 100% equities is about 20 percentage points. This is a significant but not gigantic improvement.

You can tweak the numbers for a different time horizon, your fear-reduced WR, or whatever, but this gives a nice plan for handling a big dip in the market by adjusting your portfolio equity percentage for maximum survivability at all times. Raising equity percentage only if required by a poor sequence of returns.

Now, who's going to actually follow this plan and go to 100% equities is a severe bear market?

[tuixiu]

Quote:

Originally Posted by Cat-tirement (Only 42 working days to go...)

But who's counting

[urn2bfree]

I am confused on this SWR scenario: I have a million dollars, I have it invested in a 70/30 stock/bond AA- I plan to withdraw $40,000 a year and adjust up by the inflation rate - classic 4% SWR...BUT my great uncle Fester dies and leaves me exactly $40,000, so I spend that instead of touching my portfolio....during my year of spending Uncle Fester bequest the market tanks and is only worth $600,000 by the time I need to make my first withdrawal....so am I still going to be okay taking $40,000 x 1.inflation rate ? Which is now starting my withdrawals off a smaller start and a larger percentage?

[photoguy]

I think you would be crazy to pull 40k out of a 600k portfolio.

The thing to keep in mind is that your initial firecalc run showing that you could pull 40k per year out is now a year old. In that year you've gained a lot of additional information (the market tanks) and you should update your withdrawals ( I.e. Lower them) to reflect this change.

[urn2bfree]

Quote:

Originally Posted by photoguy I think you would be crazy to pull 40k out of a 600k portfolio. The thing to keep in mind is that your initial firecalc run showing that you could pull 40k per year out is now a year old. In that year you've gained a lot of additional information (the market tanks) and you should update your withdrawals ( I.e. Lower them) to reflect this change.

But SWR theory would have me pulling out $40,000 from an even smaller portfolio and still claim I will be ok- without Uncle Fester's bequest I would have withdrawn down to $960,000 and the market would have dragged that down to $576,000-- and classic SWR theory is that even with the market tanking in that first year I should be able to take out MORE than $40,000 the next year (adjusted up for inflation- which barring a depression there would be some inflation)...so it seems given market history that the recovery from big dips has been so good that starting with a 6.67% SWR should be ok...in such scenarios in the past... It seems hard to believe but that is what the theory claims...so it would seem to me that your "max safe amount" is 4% of your max portfolio value that it ever reaches, adjusted up for inflation from that peak time.

So if my portfolio is now $5,000,000 but I am not actually touching it for the next 10 years...in 10 years my SWR is still going to be $200,000, (adjusted up for 10 years of inflation) even if my portfolio has sunk some during those 10 years..

Again I am not sure this is right, but I cannot see how it could work to spend from a portfolio at 4% of its peak and adjust up no matter what for 30 years and you survive, but if you -for what ever reason -do not touch your portfolio at its max value for any length of time--until after it drops -and then you start spending the exact same amount you would have spent had you started at peak --it has to work out at least as well....

Show me how this does not work...based on the same data that supports the Trinity SWR..not some predicted future of horrible returns that so far has not happened...

[springnr]

Quote:

Originally Posted by urn2bfree I have it invested in a 70/30 stock/bond AA

You would dispense your 40k annual funds from the 30% bond portion, allowing the 70% stock allocation time to recover.

The stock is down xx amount, you still own the same number of stocks.

[REWahoo]

Quote:

Originally Posted by springnr You would dispense your 40k annual funds from the 30% bond portion, allowing the 70% stock allocation time to recover. The stock is down xx amount, you still own the same number of stocks.

+1

Not only would you want to withdraw the $40k from your bond allocation, you'd also want to rebalance back to 70/30 by selling bonds and buying equities. (In your $576k portfolio example your AA is in the neighborhood of 50/50 after the decline.) That takes real guts but history (FIRECalc) tells us that strategy is the path that leads to success.*

*Note that a big downturn in the market during the initial few years of the withdrawal phase can lead to portfolio failure prior to 30 years. See the Start Here intro page to FIRECalc for an example illustrating this.

[MichaelB]

Quote:

Originally Posted by REWahoo Not only would you want to withdraw the $40k from your bond allocation, you'd also want to rebalance back to 70/30 by selling bonds and buying equities. (In your $576k portfolio example your AA is in the neighborhood of 50/50 after the decline.) That takes real guts but history (FIRECalc) tells us that strategy is the path that leads to success.

+1

[CaliforniaMan]

Quote:

Originally Posted by urn2bfree I am confused on this SWR scenario: I have a million dollars, I have it invested in a 70/30 stock/bond AA- I plan to withdraw $40,000 a year and adjust up by the inflation rate - classic 4% SWR...BUT my great uncle Fester dies and leaves me exactly $40,000, so I spend that instead of touching my portfolio....during my year of spending Uncle Fester bequest the market tanks and is only worth $600,000 by the time I need to make my first withdrawal....so am I still going to be okay taking $40,000 x 1.inflation rate ? Which is now starting my withdrawals off a smaller start and a larger percentage?

Quote:

Originally Posted by photoguy I think you would be crazy to pull 40k out of a 600k portfolio. The thing to keep in mind is that your initial firecalc run showing that you could pull 40k per year out is now a year old. In that year you've gained a lot of additional information (the market tanks) and you should update your withdrawals ( I.e. Lower them) to reflect this change.

While the FIRECalc runs may say that in the past you would have been successful following the WR plus inflation, it says nothing about the future. Photoguy is correct, of course the next year you have more information.

Since we don't know about the future in essence we are always in year 1 of the rest of our retirement, and you have what you have. I doubt if many people follow the inflation adjusted withdrawals blindly, without looking at the real balance. I know I won't. I think the correct use for FIRECalc and similar calculators is to get a baseline, what is reasonable based on past experience. It tells us what might be reasonable, but is not a precise recipe for future withdrawals.

[nun]

Quote:

Originally Posted by CaliforniaMan While the FIRECalc runs may say that in the past you would have been successful following the WR plus inflation, it says nothing about the future. Photoguy is correct, of course the next year you have more information. Since we don't know about the future in essence we are always in year 1 of the rest of our retirement, and you have what you have. I doubt if many people follow the inflation adjusted withdrawals blindly, without looking at the real balance. I know I won't. I think the correct use for FIRECalc and similar calculators is to get a baseline, what is reasonable based on past experience. It tells us what might be reasonable, but is not a precise recipe for future withdrawals.

I agree. There should be feedback between your principal balance, rate of return and withdrawal. I have always thought of FireCalc results as a maximum sensible withdrawal rate and plan on withdrawing 2% max and probably less.

[ERD50]

Quote:

Originally Posted by photoguy I think you would be crazy to pull 40k out of a 600k portfolio. The thing to keep in mind is that your initial firecalc run showing that you could pull 40k per year out is now a year old. In that year you've gained a lot of additional information (the market tanks) and you should update your withdrawals ( I.e. Lower them) to reflect this change.

But that has nothing to do with the historical information in FIRECalc, and how one interprets that output.

Historically, a 4% WR provides ~ 95% success rate, and that, by definition includes retiring at market peaks, market troughs, and everything in between, maintaining that inflation adjusted WR throughout.

Which is why I prefer to target a 100% success rate, that says that historically, I could have retired at any time, and maintain that inflation adjusted WR regardless of the roller coaster ride.

As long as the future isn't worse than the worst scenarios of the past (not just worse than average), that WR will succeed.

If you want to protect against future record bad scenarios, you need to add to the cushion. I'll add that I think many people overestimate the effect of cutting spending for a few years. Going from say 3.5% to 2% for a few years is a drop in the bucket compared to a portfolio drop of 50%.

-ERD50

[NW-Bound]

Quote:

Originally Posted by urn2bfree But SWR theory would have me pulling out $40,000 from an even smaller portfolio and still claim I will be ok- without Uncle Fester's bequest I would have withdrawn down to $960,000 and the market would have dragged that down to $576,000-- and classic SWR theory is that even with the market tanking in that first year I should be able to take out MORE than $40,000 the next year (adjusted up for inflation- which barring a depression there would be some inflation)...so it seems given market history that the recovery from big dips has been so good that starting with a 6.67% SWR should be ok...in such scenarios in the past... It seems hard to believe but that is what the theory claims...so it would seem to me that your "max safe amount" is 4% of your max portfolio value that it ever reaches, adjusted up for inflation from that peak time. So if my portfolio is now $5,000,000 but I am not actually touching it for the next 10 years...in 10 years my SWR is still going to be $200,000, (adjusted up for 10 years of inflation) even if my portfolio has sunk some during those 10 years.. Again I am not sure this is right, but I cannot see how it could work to spend from a portfolio at 4% of its peak and adjust up no matter what for 30 years and you survive, but if you -for what ever reason -do not touch your portfolio at its max value for any length of time--until after it drops -and then you start spending the exact same amount you would have spent had you started at peak --it has to work out at least as well.... Show me how this does not work...based on the same data that supports the Trinity SWR..not some predicted future of horrible returns that so far has not happened...

Yes, one can draw 4% WR from portfolio at market top and survive, but that survival means the balance is barely above 0 after 30 years.

On the other hand, drawing 4% WR from portfolio at market bottom and then adjusting only for inflation when the market recovers means one will die a decamillionaire after 30 years.

Oh, one must have the guts to buy low sell high rebalance too.

In reality, people will try to reduce expenses if they see that their WR is alarmingly above 4% when market crashes. And they will up their spending during market booms if the WR now becomes 1 or 2%.

[Animorph]

Quote:

Originally Posted by urn2bfree I am confused on this SWR scenario: I have a million dollars, I have it invested in a 70/30 stock/bond AA- I plan to withdraw $40,000 a year and adjust up by the inflation rate - classic 4% SWR...BUT my great uncle Fester dies and leaves me exactly $40,000, so I spend that instead of touching my portfolio....during my year of spending Uncle Fester bequest the market tanks and is only worth $600,000 by the time I need to make my first withdrawal....so am I still going to be okay taking $40,000 x 1.inflation rate ? Which is now starting my withdrawals off a smaller start and a larger percentage?

The classic 4% is not 100% safe, even historically. So a big market drop may mean you are on one of the failure paths. Or maybe a new future failure path.

But one of the benefits of a calculator like FIRECalc that uses historical returns in their correct order is that you get the advantages of actual market "mean reversion". In 2008, I'm not thinking "oh no, my money is gone forever!". I'm thinking I wish I had more cash to buy more cheap stocks before the price goes back up. How many "V" shaped market dips have we had?

So FIRECalc is also saying, yes the market went down a lot, but it is also likely that there will be a period of above average market returns coming. So you may be able to recover from briefly spending more than 4% of your portfolio.

And one problem with FIRECalc is that you put your portfolio balance into it and FIRECalc tests scenarios that include retiring just at the start of the Great Depression. If the portfolio balance you input was already at a market bottom (unlike right now), that might be a scenario of two big market drops in a row (the current market bottom plus any number of historical market drops). That probably exceeds the severity of any actual market drop. So FIRECalc is probably best used during a market peak, not at a market bottom. Your initial 4% SWR calc before the market drop was probably more realistic than your recalculation of SWR at the market bottom.

There have been attempts at incorporating current market conditions into retirement calculators, such as accounting for P/E, but nothing that has looked strongly predictive.

[photoguy]

Quote:

Originally Posted by urn2bfree But SWR theory would have me pulling out $40,000 from an even smaller portfolio and still claim I will be ok

I don't think that's what the SWR studies are claiming or implying.

The way all of the SWR studies work is that they pick a withdrawal rate (say 4%) and run a simulation against either historical or fabricated data (monte carlo methods). They take all of the results and say something like of 6 of the 114 cycles failed (5% failure rate). (A monte carlo simulation might say something like 1123 cycles of 20,000 failed).

One can take the results on simulations and use it as an estimate of the future success rate. But this estimate is based on all of the runs -- including ones where equity returns in the first year went up, down, and stayed flat.

In your situation we have an additional piece of information, that the market tanks in the first year. If you were to look at only the runs where the first year is a bear, you might find that the numbers look more like 4 of 46 cycles failed for a 8.7% failure rate. In other words, your failure rate has nearly doubled given that the first year is a bear. You may still be ok to keep pulling out 40k but your risk level is certainly much higher than before. To reduce risk, the obvious solution is to drop the 4% withdrawal to a lower amount.

As far as I know, all of the SWR tools and studies only look at the probability of success as computed at the start of retirement. They don't let you "update" the estimates as your retirement progresses. So the initial failure rate estimate can be wildly off depending on actual market returns.

Quote:

Again I am not sure this is right, but I cannot see how it could work to spend from a portfolio at 4% of its peak and adjust up no matter what for 30 years and you survive, but if you -for what ever reason -do not touch your portfolio at its max value for any length of time--until after it drops -and then you start spending the exact same amount you would have spent had you started at peak --it has to work out at least as well....

The Trinity authors used a fixed SWR rate because it was easy to analyze. But they did not expect actual retired investor to keep their withdrawal rate constant. Here is a quote from their paper:

Quote:

The word planning is emphasized because of the great uncertainties in the stock and bond markets. Mid-course corrections likely will be required, with the actual dollar amounts withdrawn adjusted downward or upward relative to the plan. The investor needs to keep in mind that selection of a withdrawal rate is not a matter of contract but rather a matter of planning.

(emphasis mine)
http://www.aaii.com/journal/article/...m_medium=click

[springnr]

Voila, it is my pleasure to announce the Asset Allocation; perhaps the 50/50 would appeal.

May it please you to notice that over ten years of withdraws are present in the bond portion for your serenity. If the market tanks, at your leisure, please consider a variable withdraw rate. Also one could choose alternatives such as the self correcting strategy of constant stress, or possibly learn to conversar en español durante su residencia en México.