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

[NW-Bound]

is your 3% calculated only on your equity / fixed income portoflio, or do you include your cash in the portfolio before calculating the 3%.

I include the 25% cash in the denominator. That cash is earning a bit above 2% now. It is really I-bonds, stable value fund, etc..., and not 100% true cash.

 

[robnplunder]

I am in the same boat and plan to leave at the end of this year or mid next year. Thanks for starting the thread because I'm looking for other ideas. I share your concerns and here is where I landed (so far):

1) Start with what ERD mentioned, 100% success in FireCalc and ******** with constant spending power. I too believe that I'll spend less as I get older, but I just can't get myself to plan for that.
2) Add a 5% cushion to that 100% portfolio number (if I can stand work that long)
3) Reduce my equity allocation from 65%-70% to 50% (I'm at 100% success with both allocations). This will allow me to better absorb an early sequence of returns problem. I'll increase that over 5 years or so to about 60% and stay there.
4) My plan is to take SS at 70. I'll take it earlier if the above still aren't enough.

I didn't list cutting expenses because I really don't want to. Of course I will if I need to, but at this point I'd rather work and extra year or two. I'm usually have a pretty decent risk tolerance but, as has been mentioned before, you only get one shot at this.

I am with you on 2 (in the form of house equity), 3 & 4. It's 1 that keeps me doing OMY. I get 95% from FIRECalc with Bernicke spending model. But I admit my expense is bloated (large "misc" section).

 

[NW-Bound]

BTW - I get a 93% success rate in FIRECalc with my proposed 20% / 10% hit. Its not horrible but the first zero value is at 26 years which makes me age 76 ... way too young to be broke !

I don't know if we can really use FIRECalc in the deterministic sense that we would like to. It can only tell us the rough WR that we can expect. Too much tweaking, and we are actually expecting the future to duplicate the past, which I am sure it will not (however, it may rhyme).

Just for kicks, instead of using the default portfolio of Total Market/Treasury in FIRECalc, I played with the more detailed "Mixed Portfolio", and was able to get a higher WR without going negative. Try the following portfolio: "30% US Small Value, 30% US Large Value, 40% 1-month Treasury". That last one is almost like cash. FIRECalc will say you can get up near 4.4%WR without going broke in 30 years.

Along the same vein, I posted some other results in this past thread: http://www.early-retirement.org/forums/f28/optimal-firecalc-portfolio-69523.html. Take all these with a grain of salt, per the earlier caveat.

 

[Fermion]

Why not just go a bit more cash for awhile until you get comfortable. Next time PenFed has 5 year 3% CDs, grab a few years of those to cover living expenses.

The USA can't survive another major downturn so it will not happen. A 50% drop in the market would wipe out the fragile gains of most of the public pension funds and collapse the government. There are not a lot of ways to bail out this time. Conclusion: The people that are able to keep the market running along will do so.

There may be a 10% or 20% correction, but nothing much more than that.

 

[NW-Bound]

And how about this FIRECalc portfolio: "30% US Small Value, 25% US Large Value, 45% 1-month Treasury"? You can go up to 4.5% WR on that for 30 years.

Do you believe that? See how a bit of tweak can change things quite a bit? We should not take it too far either with Monte Carlo or historical simulations like FIRECalc. A guideline is all we can get.

 

[Options]

If you are really worried about high equity prices, consider starting with 35% equity exposure, and then allowing it to rise very gradually over the decades in retirement. Kitces and Pfau have shown that this "rising equity glide path" approach improves retirement fund survivability over maintaining a straight AA over the same period, mainly because it handles bad early years better.

Should Equity Exposure Decrease In Retirement, Or Is A Rising Equity Glidepath Actually Better? | Kitces.com

If they had published this approach in the late 90s, I might have considered it.

+1
I intend to do just that. Reduce equities to 40% and increase perhaps another 5% 5 years into retirement. What is hardest about retiring into a market top, should it turn out to be that later, is sequence of returns risk. First 5, 10 years into retirement are crucial for PF returns. That said, a low withdrawal rate cures a multitude of portfolio ills...

 

[ejman]

I don't know if we can really use FIRECalc in the deterministic sense that we would like to. It can only tell us the rough WR that we can expect. Too much tweaking, and we are actually expecting the future to duplicate the past, which I am sure it will not (however, it may rhyme).

Just for kicks, instead of using the default portfolio of Total Market/Treasury in FIRECalc, I played with the more detailed "Mixed Portfolio", and was able to get a higher WR without going negative. Try the following portfolio: "30% US Small Value, 30% US Large Value, 40% 1-month Treasury". That last one is almost like cash. FIRECalc will say you can get up near 4.4%WR without going broke in 30 years.

Along the same vein, I posted some other results in this past thread: http://www.early-retirement.org/forums/f28/optimal-firecalc-portfolio-69523.html. Take all these with a grain of salt, per the earlier caveat.

I like this approach NW-Bound. Let's reconvene in another 10,000 years - by then we might have enough data points to reach some meaningful conclusions

 

[fh2000]

And how about this FIRECalc portfolio: "30% US Small Value, 25% US Large Value, 45% 1-month Treasury"? You can go up to 4.5% WR on that for 30 years.

Do you believe that? See how a bit of tweak can change things quite a bit? We should not take it too far either with Monte Carlo or historical simulations like FIRECalc. A guideline is all we can get.

I noticed using 'Mixed portfolio', FireCalc calculates 49 cycles instead of 114 cycles with 'Total Market'. Maybe the smaller sample size favors Mixed Portfolio option.

 

[daylatedollarshort]

Here are some articles on stocks in retirement that have influenced our thinking:

How to avoid sequence-of-return risk - Robert Powell's Retirement Portfolio - MarketWatch

The Biggest Urban Legend in Finance
https://www.researchaffiliates.com/Our%20Ideas/Insights/Fundamentals/Pages/F_2011_March_The_Biggest_Urban_Legend.aspx

We're more in the won the game why keep playing camp. The Fidelity RIP shows us chugging along at ages 100+ with no financial issues, yet the Fido planner recommends a riskier AA mix. I don't remember the exact worst year right now, but I think the worst year loss for the portfolio they recommended was something like 50%. I can't see risking losing half our life savings early on in retirement when there is no need to do so.

(Note: I use an inflation adjusted annual expense amount based on my own retirement budget estimates and looking at the Consumer Expenditure Survey and not their health cost projections.)

 

[NW-Bound]

I like this approach NW-Bound. Let's reconvene in another 10,000 years - by then we might have enough data points to reach some meaningful conclusions

As I said, a 3 to 4%WR is all we can get from these simulations. To my 3% WR, our SS will be the reserve, and then our home values. If it comes down to that, I will be hitting the road in my small motorhome. A vagabond life is adventurous and fun, though my wife does not agree to it yet. So, me worry?

I noticed using 'Mixed portfolio', FireCalc calculates 49 cycles instead of 114 cycles with 'Total Market'. Maybe the smaller sample size favors Mixed Portfolio option.

The sectored market performance data in FIRECalc is more limited than the total market history, hence the smaller number of cycles. However, there have been studies showing that conservative value stocks historically outperformed growth stocks. And as the total market encompassed growth stocks, it also trailed value stocks. It's something to keep in mind.

Of course, if one can pick just the right growth stock, he will do very well for a while. However, no stock can grow to the sky like Jack's bean stalk, so timing is important.

 

[CaliforniaMan]

Why not just go a bit more cash for awhile until you get comfortable. Next time PenFed has 5 year 3% CDs, grab a few years of those to cover living expenses.

The USA can't survive another major downturn so it will not happen. A 50% drop in the market would wipe out the fragile gains of most of the public pension funds and collapse the government. There are not a lot of ways to bail out this time. Conclusion: The people that are able to keep the market running along will do so.

There may be a 10% or 20% correction, but nothing much more than that.

Wow, your confidence is reassuring. However not quite as reassuring as when economist Irving Fisher said "Stock prices have reached what looks like a permanently high plateau"

 

[Fermion]

Wow, your confidence is reassuring. However not quite as reassuring as when economist Irving Fisher said "Stock prices have reached what looks like a permanently high plateau"

Curious, when did he say that? At the 2012 highs, 2013 highs, or just now at the 2014 highs?

Edit. Oh, Irving Fisher. Isn't he quite dead?

 

[Bestwifeever]

I'm hoping we're not at market top - I gave notice today and retirement starts in 2 weeks....

Someone is going to have a great summer! Congrats!

 

[CaliforniaMan]

Curious, when did he say that? At the 2012 highs, 2013 highs, or just now at the 2014 highs?

Edit. Oh, Irving Fisher. Isn't he quite dead?

Yes, he said it in 1929 just before the crash.

 

[redduck]

NW-Bound
Give me a museum and I'll fill it. (Picasso)
Give me a forum ...

Join Date: Jul 2008
Posts: 10,000


"With 25% in cash and a 3% WR, I should be able to survive several years of famine. Even a bad stretch of Biblical proportion was only 7 years, right?...."

+++++

Congratulations on post # 10,000.

 

[NgineER]

I am considering a glide path approach starting with a $3MM portfolio.
What do you think of filling up with 80% bonds AA and work on depleting that over the next 25 years. When hitting 75, out of bonds, the $600,000 would hopefully have compounded nicely oven 25 years even with a bear market to provide sufficient funds for another 25 years.
Is this a realistic approach? Still 14 years out for me, so just thinking about various options.

 

[REWahoo]

What do you think of filling up with 80% bonds AA and work on depleting that over the next 25 years.

I think with a 25 year time horizon you'd be giving up substantial gains and risk taking a painful loss in purchasing power from inflation. History indicates an AA with less than ~40% in stocks fares poorly over a 30 year time frame. From FIRECalc:

 

[REWahoo]

One more thought: If you followed your proposed 25 year plan you'd gradually increase your equity allocation from 20/80 to 100/0. How concerned will you be near the end of those 25 years - when you are 75 - about a big downturn in the stock market?

 

[pb4uski]

Yes, he said it in 1929 just before the crash.

Well, in the long run, he was right.

 

[NW-Bound]

Congratulations on post # 10,000.

Thank you for the reminder! I did not notice at all this milestone. Time flies while you are having fun.

 

[photoguy]

I am considering a glide path approach starting with a $3MM portfolio.
What do you think of filling up with 80% bonds AA and work on depleting that over the next 25 years. When hitting 75, out of bonds, the $600,000 would hopefully have compounded nicely oven 25 years even with a bear market to provide sufficient funds for another 25 years.
Is this a realistic approach? Still 14 years out for me, so just thinking about various options.

I'd be very skeptical and cautious about using a rising glide path. The idea comes from a paper by Pfau and Kitces:

Reducing Retirement Risk with a Rising Equity Glide-Path by Wade D. Pfau, Michael E. Kitces :: SSRN

AFAIK they are the only folks advocating this approach. I certainly wouldn't take this approach without fully understanding their work including their assumptions, methodology, data, and outcomes (not just failure probability but the entire distribution of returns). This approach is especially concerning given the relatively low expected returns for bonds.

I have not read this paper in depth, but some things you may want to consider:

- Most papers on w.r. are written using higher withdrawal rates (4-5%) as failures are more common then. But many (at least on this board) are choosing to cut risk by going with lower w.r (say 3% or less). If you're using a lower withdrawal rate, the conclusions in the paper may not apply.

- I wouldn't trust the results until it's been replicated using both historical simulation (e.g. Firecalc) and with MC methods. Academic papers are often filled with errors which escapes the notice of the reviewers. There's also a good chance that they have bugs in their experiments (this is widely true for academic code) which may impact their results

- Pfau is very fond of MC methods and I think used it in the paper. I'd double check his assumptions as he has done things like somewhat arbitrarily reducing US equity returns by 2% in his simulations.

 

[Totoro]

I'd be very skeptical and cautious about using a rising glide path. The idea comes from a paper by Pfau and Kitces:

What a gliding path offers fundamentally in my opinion is, in some cases, a better balance throughout time between increasing risk (stock crash) vs. reducing uncertainty (longevity, spend levels). Especially when you want to leave an inheritance.

It can make a lot of sense on an emotional and rational level.

Case in point: my mother is retired since 3 years or so. She will start receiving a small pension in three years, and is married. Her mother is still alive. We have the following uncertainties:

• Pension. It will be certain in three years time. We have a rough idea now but the government keeps changing the rules.
• Grandma may or may not start running up healthcare costs. As she gets older we get more certainty around that. For example, just last year her husband (my grandfather) died suddenly. He might also have ended up in the ER or nurse care for several years.
• My sister may or may not get into money issues
• Mother may or may not continue her current spending levels which are quite high. We'll know better in a few years as she ages how that evolves.
• Her husband has a rather high pension, with survivor benefits. Since this is her (and his) second marriage they have separate finances right now.
• Tolerance for risk. She never saw stocks drop 30% or more in her life since she was fully invested in her own business. We don't know how she will react when it does. Having a few positive years first OR a small amount in stocks eases the ability to "stay the course".
• In all likelihood, not all grandchildren have been born yet. If there is a child born with special needs, she'll want to provide extra.

Which each passing year all these uncertainties become more certain, and therefore we can increase risk exposure if things go a certain way.

In these first retirement years for her, having a reduced downside risk is worth more than later on.

 

[Fermion]

It would be interesting to model a portfolio success rate based on a variable SWR factoring one's lifestyle.

How would you calculate success rates based on:

3% SWR for the first 10 years, 4% SWR for the following 10 years, then 3% SWR after that.

This might come about for someone retiring at age 47 and living a low budget lifestyle for 10 years, maybe working some part time job they find fun for a little extra cash, then at age 57 doing a bit more lavish lifestyle with a higher SWR, then having SS kick in at 67 and drop back to 3% SWR.

 

[Willers]

I get many more failures with ******** because of the treatment of bonds

I've noticed that too. ***** results give me less fun money than FireCalc...

 

[daylatedollarshort]

I am considering a glide path approach starting with a $3MM portfolio.
What do you think of filling up with 80% bonds AA and work on depleting that over the next 25 years. When hitting 75, out of bonds, the $600,000 would hopefully have compounded nicely oven 25 years even with a bear market to provide sufficient funds for another 25 years.
Is this a realistic approach? Still 14 years out for me, so just thinking about various options.

It depends on your expenses and other income, like Social Security, pensions or part time work. Many here have retirement lifestyles and annual expenses that require higher withdrawal rates than a very conservative portfolio could support and/or they are okay with a riskier portfolio in order to increase their net worth.

If your nest egg is large enough in relation to your spending and other retirement income, you may be able to have zero or a low allocation in stocks throughout retirement and still show a 100% success rate, plus leave an estate, in the retirement planners to a very old age.

In the Fidelity RIP, it shows what the worst year loss could have been for various types of AAs. You might play around with the various AAs, check out the worst year and think about your personal tolerance if that worst year happened in year one of your retirement.