Has anyone tried/used the Guyton/Klinger method for a VPW

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
Hexdump
Posts: 1626
Joined: Fri Oct 10, 2008 7:28 am
Location: Houston, Texas

Has anyone tried/used the Guyton/Klinger method for a VPW

Post by Hexdump »

With thanks to siamond, original thread here.
http://www.bogleheads.org/forum/viewtop ... 0&t=120430
I guess you have to think about your personal objectives. Are you so sure that you'd like to have less money to begin with, and more money as you get older? Including when you get real old and rather frail?

Personally, I know I'd like to have a tad more money in my early days of retirement, as I guess I will be more active by then, but then my level of activity might decrease. That being said, my medical expenses will probably increase, so this is a bit of a tough call.

I have looking at the Guyton/Klinger variable method, and although this sounds a tad complicated (albeit easy to automate with a spreadsheet) and may require a bit of customization (e.g. define floor/ceiling on my spend to avoid silly situations), this makes a lot of sense to me.
http://schulmerichandassoc.homestead.co ... ofiles.pdf
This makes a lot of sense to me and I was wondering if anyone had put it into practice and how did it work out.
User avatar
siamond
Posts: 6008
Joined: Mon May 28, 2012 5:50 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by siamond »

I am going to do it, using a slight variation on the G-K rules, starting next year (2015). I got rid of the MWR rule (no real value-added), and added a conditional floor/ceiling. Quite easy to automate the math with an Excel sheet.

At a high-level, it shares some key principles with VPW, with withdrawals essentially being a function of the current size of your portfolio, and adjusting up and down as the market gyrates. Hence benefiting from rosy scenarios, while properly reacting to dreary scenarios. Similarly to VPW, one could say "it cannot fail" (in the sense that your portfolio never goes to zero). But of course, your portfolio (and withdrawal) can go pretty low if things get very sour.

I did a ton of modelling & backtesting with it, comparing with multiple other methods (including a customized form of VPW), and for my own situation, this seems the best match, allowing variable withdrawals while aiming at preserving my capital for an arbitrarily long life, including possible LTC steep costs at the end, and preserving a proper bequest for my kids. Oh, and also the withdrawals adjustments year over year are significantly smoother than VPW or a simple % of your portfolio. And side income (recurrent or lump sum) is easy to process, just add to the principal (or subtract to the withdrawal for small amounts), and everything adjusts itself.

Overall, I think that VPW (or a variation of it) is brilliant for some personal circumstances (e.g. desire to spend it all), but in my case, G-K seems a better match.

If you want to play around, this calculator (http://www.cfiresim.com) allows you to model both CVPW and Guyton-Klinger (simplified, without MWR). Playing with the fees, you can also see the effect of grimmer future returns than what happened in the past, which I would strongly advise.
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

Can anyone explain the guyton Klinger method in terms I can explain it to my mom with?
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by steve_14 »

Unfortunately such precise rules, while they may back test perfectly, don't have much value when applied to an uncertain future. The standard advice - pick a SWR and stay flexible based on your health, investment returns, and spending needs, cannot be improved upon IMO.
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing things a certain way because "that's how we always did it before".
Last edited by coachz on Sat Aug 06, 2016 9:12 am, edited 1 time in total.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by richard »

coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.

If you do enough testing of historical data, you will find a strategy that works brilliantly, for essentially the same reasons if you have enough people toss a coin 10 times you will find people who can get 10 heads in a row. Betting that these people can repeat their performances would not be a wise move.

There's no assurance that SWR that worked in the past will work in the future, but that does not mean they will not work in the future.
Leeraar
Posts: 4109
Joined: Tue Dec 10, 2013 7:41 pm
Location: Nowhere

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Leeraar »

coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Because only one scenario applies to you, and you can optimize it as it unfolds. You don't care about all the scenarios that are not happening where your strategy will also work.

L.
You can get what you want, or you can just get old. (Billy Joel, "Vienna")
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

But that seems to me the same reasons not to trust SWR.
richard wrote:
coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.

If you do enough testing of historical data, you will find a strategy that works brilliantly, for essentially the same reasons if you have enough people toss a coin 10 times you will find people who can get 10 heads in a row. Betting that these people can repeat their performances would not be a wise move.

There's no assurance that SWR that worked in the past will work in the future, but that does not mean they will not work in the future.
rmark1
Posts: 359
Joined: Sat Jun 16, 2007 7:43 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by rmark1 »

'Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced?'

US data 1926-2013 is commonly used, not 'all' the historical data. Nor historical investing costs.

A lot of financial articles should start with the disclaimer 'my best guess'.
richard
Posts: 7961
Joined: Tue Feb 20, 2007 2:38 pm
Contact:

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by richard »

So?

See the last sentence of my post.
coachz wrote:But that seems to me the same reasons not to trust SWR.
richard wrote:
coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.

If you do enough testing of historical data, you will find a strategy that works brilliantly, for essentially the same reasons if you have enough people toss a coin 10 times you will find people who can get 10 heads in a row. Betting that these people can repeat their performances would not be a wise move.

There's no assurance that SWR that worked in the past will work in the future, but that does not mean they will not work in the future.
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by longinvest »

The Guyton/Klinger method is quite complex.

Have you looked at the Bogleheads Wiki VPW Page? It is a much simpler VPW developed on this forum (http://www.bogleheads.org/forum/viewtop ... 0&t=120430).
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
User avatar
midareff
Posts: 7711
Joined: Mon Nov 29, 2010 9:43 am
Location: Biscayne Bay, South Florida

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by midareff »

steve_14 wrote:Unfortunately such precise rules, while they may back test perfectly, don't have much value when applied to an uncertain future. The standard advice - pick a SWR and stay flexible based on your health, investment returns, and spending needs, cannot be improved upon IMO.

I think your idea is much better Steve.

But to get to the paper itself; When I got to the returns they modeled their portfolio's with, as with Bengen's, I find a problem. S&P returns at 9.62% :annoyed , long Treasuries @ 4.99% :annoyed , and 3 month T bills at 3.74% :oops: . Really? Inflation at 3.13% ... ?? The inflation rate for the last 53 years has been 3.92%, and is that low based on abnormally low CPI-U for the last dozen years. IMHO beginning a retirement withdrawal schedule based on the cited returns to start a 30 year retirement with a WR of 5.5% in today's environment is asking for disaster, as is 5.3% for a 40 year retirement. Sequence of returns risk looks significant (to me) from moment one.

I do agree that cuts and bonuses will help the longevity of any portfolio but we have to make several assumptions here as well...... does the retiree have sufficient starting fluff in their budget to sustain cuts? To be able to sustain a couple of years of cuts (and avoid the cat food years) that assumes significant fluff was in the budget to start with. We all have a floor below which we would not be comfortable by a reduction of lifestyle, and an idea of what we would do with extra money in great return years.... a ceiling.

Applying rules to these floor and ceiling variations in lieu of thoughtful flexibility seems to imply we didn't have the brains to get here (retired) in the first place. I've also looked at the VWR spreadsheets posted earlier by longinvest. With due respect to longinvest's work the problem with every pre-established finite withdrawal plan I've seen is one size doesn't fit all. There are retirees that have more income than they spend and want a nice cushion in their late years, and those that are retired and not getting by with money leftover. If you will; the Mercedes vs. the Kia retirees. Add to that the unknown of investment returns, the CPI-U, the Fed's manipulation of interest rates, Swans Black and White, your healthcare costs, your longevity, etc., .. and any finite plan, even one with programmatic cuts and bonuses starting at year one will cover all of this?

Good luck with that..................... with a high Shiller indicative of lower equity expectations coupled with the Fed's pressure on interest rates moderating from record or near record lows, I think any strategy starting with a 4 or 5 percent number could be asking for trouble.
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

"but that does not mean they will not work in the future." So doesn't that also apply to Guyton/Klinger ? SWR and backtesting are both simply past data with no guarantee of future performance so I should weigh them both equally suspiciously ?
richard wrote:So?

See the last sentence of my post.
coachz wrote:But that seems to me the same reasons not to trust SWR.
richard wrote:
coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.

If you do enough testing of historical data, you will find a strategy that works brilliantly, for essentially the same reasons if you have enough people toss a coin 10 times you will find people who can get 10 heads in a row. Betting that these people can repeat their performances would not be a wise move.

There's no assurance that SWR that worked in the past will work in the future, but that does not mean they will not work in the future.
User avatar
archbish99
Posts: 1649
Joined: Fri Jun 10, 2011 6:02 pm

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by archbish99 »

I've read most of the Guyton papers I could find, and while I basically understand his method, it's not terribly well defined. There are elements of it that I've kept, and elements I've decided serve no purpose. From memory, here's where I landed:
  • Separate buckets for near-term withdrawals and long-term assets; replenish near-term when market is good: This ultimately boils down to being more in cash when the market is "high," and putting more into stocks and bonds when it's not, which is market timing. Bad plan. Pick a sufficiently conservative AA and stick to it.
  • Draw first from long-term assets that have gone up: Also known as rebalancing. Move on.
  • Put guard-rails around the withdrawal amount: This is the element I find actually useful. Instead of blindly taking your inflation-adjusted SWR amount each year, or cutting back drastically out of panic, Guyton proposes and tests an additional withdrawal rule that defines when to cut back, and how much. And conversely, when your portfolio has grown enough that you're not taking enough.
The short form is this: Say you decide on a 4% SWR.... Each year, you don't just inflation-adjust the amount, you also sanity check the amount as a percentage of the current portfolio. If it's more than 20% above your initial SWR target (so for 4%, if it's over 4.8% of your portfolio), cut it by 10%. If it's more than 20% below your initial SWR target (so for 4%, 3.2%), increase it by 10%. (There are some hand-wavy statements about being able to eliminate the upper guard-rail if you're getting close to the end of your life expectancy.)

His backtesting and Monte Carlo stuff suggests that if you do this, even a conservative AA will last with very high confidence with a SWR of 5%. It's still possible to fail -- a multi-year nosedive that doesn't recover for years more can still deplete your portfolio following these rules. And others have argued that any rational investor/retiree would cut back on their spending if they see the market crashing around them.

But to me, the value is in a functional, pragmatic definition of what it means for your portfolio to have fallen "too far," and what to do about it. Take a bad year in which your portfolio drops by 15%. That means your initial 4% is now 4.7% of the remaining portfolio. Do you need to cut your withdrawal by 15%? Do you sit tight and assume a recovery will come along? Having a guideline that says "this is bad enough to cut back, but this isn't" is useful. If you have your own guideline, use what works for you.
I'm not a financial advisor, I just play one on the Internet.
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

Thank you for the explanation and input. I realize I am a total noob at this but the logic is compelling to me:

https://www.youtube.com/watch?v=tH_Shw9RXB4
User avatar
LadyGeek
Site Admin
Posts: 95686
Joined: Sat Dec 20, 2008 4:34 pm
Location: Philadelphia
Contact:

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by LadyGeek »

coachz wrote:Thank you for the explanation and input. I realize I am a total noob at this but the logic is compelling to me:

https://www.youtube.com/watch?v=tH_Shw9RXB4
The video is by Wade Pfau, titled: Sequence-of-Return Risk: Gorilla or Boogeyman?

He mentions Guyton, focused on sequence-of-returns risk, i.e. the market will drop when you need the money the most.
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

Thank you, next time I will post more info about the link. Thanks!
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by longinvest »

LadyGeek wrote: The video is by Wade Pfau, titled: Sequence-of-Return Risk: Gorilla or Boogeyman?

He mentions Guyton, focused on sequence-of-returns risk, i.e. the market will drop when you need the money the most.
Great video blog!

Just for comparison, I've simulated using the much simpler Bogleheads Wiki VPW for a year 2000 retiree, using a 3-fund portfolio allocated as 25% U.S. Stocks / 25% International Stocks / 50% U.S. Bonds.

Here are the results, so far.

Image

Image

Image
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
User avatar
archbish99
Posts: 1649
Joined: Fri Jun 10, 2011 6:02 pm

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by archbish99 »

My issue with the VPW method is that it responds too quickly. Good years shoot up, while bad years drop. The advantage of the Guyton approach is that things stay relatively steady year-to-year; only after a severe shock, or several years of good/bad returns do you experience a big change. (Of course, the counter-argument to that is that running the same portfolio through the Guyton rules, you cut back in both 2003 and 2009 following major portfolio drops the year before, but never regained the cut in real terms. Of course, neither did the portfolio, which is kind of the point.)
I'm not a financial advisor, I just play one on the Internet.
longinvest
Posts: 5682
Joined: Sat Aug 11, 2012 8:44 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by longinvest »

archbish99 wrote:My issue with the VPW method is that it responds too quickly. Good years shoot up, while bad years drop. The advantage of the Guyton approach is that things stay relatively steady year-to-year; only after a severe shock, or several years of good/bad returns do you experience a big change. (Of course, the counter-argument to that is that running the same portfolio through the Guyton rules, you cut back in both 2003 and 2009 following major portfolio drops the year before, but never regained the cut in real terms. Of course, neither did the portfolio, which is kind of the point.)
I see what you mean.

When designing the VPW spreadsheet, the various bogleheads involved played with ceilings and floors to try to smooth things; it didn't help when it was most needed. Of course, it reduced withdrawal volatility during normal, boring years. But, when really bad or really good returns happened, they had a significantly bigger impact on withdrawals. It's one of those "pay me now or pay me (more) later" kind of thing.

One trick that can really dampen volatility is increasing the bond allocation of the portfolio, but this will reduce both withdrawals and returns. Mainly, it's the volatility of the portfolio that determines the volatility of withdrawals with portfolio-return-driven withdrawal methods such as VPW and Guyton. It's a risk/return trade-off, as usual.
Variable Percentage Withdrawal (bogleheads.org/wiki/VPW) | One-Fund Portfolio (bogleheads.org/forum/viewtopic.php?t=287967)
User avatar
siamond
Posts: 6008
Joined: Mon May 28, 2012 5:50 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by siamond »

For sure, there is no size fit all when it comes to withdrawal methods, or at least, I doubt there is. As usual, one has to ponder about your goals in selecting what works for you. The G-K method, when stripped to its fundamentals (essentially stay centered on a fixed % of your current portfolio, with those guardrails to make you smoothly adjust your withdrawals - as nicely explained by archbish99), will work very well for some folks, or not for others. And it's really quite trivial to program in an Excel sheet or equivalent.

If you wish to spend all your money, or if you can't deal with some variability in your spend, or if you have all the money in the world, then forget it, there are better ways. But if you aim at keeping your portfolio value somewhat stable (on average) for an indefinite time, can deal with some spending variations but not too much, and would like to take advantage of the ups of the markets while dealing with the downs of it, then this is a very sensible method. And to be clear, it DOES adjust to grimmer future returns than what occurred in the past - or better ones, actually. Again, this is centered on a fixed % of your current portfolio, not your initial portfolio, so if things go bad, well, withdrawals will shrink accordingly.

To make your own assessment, there is at least one retirement calculator that implemented a basic form of the Guyton-Klinger "guardrails": http://www.cfiresim.com/input.php. One can play on the initial spending to indirectly affect the % of the portfolio that the guardrails are centered on. And one can play on the fees (try 1% or 2%) to model much grimmer returns than actually occurred in the past.

If you play with it long enough, you'll see that the smoothing mechanism is really quite good, not only for the immediate effect of avoiding jittery effects on your spend, but also avoiding to overreact to market gyrations and be there for the long run. Which I found crucial to do, actually. This is where the flat 4% of the initial portfolio method does convey a bit of wisdom, it can take 20 years for things to return to 'normal', and going through crazy jitters in the meantime will do nothing good for your blood pressure, and little to help your portfolio. I don't agree that this doesn't properly address big crisis (or big bull markets), by the way. Just run simulations with sensible inputs, and you'll see.
Last edited by siamond on Mon Jun 30, 2014 10:55 am, edited 1 time in total.
User avatar
siamond
Posts: 6008
Joined: Mon May 28, 2012 5:50 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by siamond »

Now something that keeps boggling my mind... All those assertions that 'manually' adjusting your withdrawals based on current circumstances is the way to go... I don't know, but if somebody would assert that point decisions every year (hence likely driven by emotions, recency bias, etc) was the way to manage your investments in the accumulation phase, I suspect that the poster would find little support... Here, this is the exactly the same, in a distribution phase, better NOT risk to be driven by such emotions... I would even argue that it is MORE important to avoid emotional decision-making during your retirement. Hence the beauty of using a solid formulaic method like VPW or G-K (pick the one matching your goals). And spreadsheets make it trivial to do the little math required.

Now, of course, if you have significantly more money than you need (which I guess is often the case with BH retirees, many of them tend to be very conservative people), then the whole point is probably moot, that I can understand.
User avatar
siamond
Posts: 6008
Joined: Mon May 28, 2012 5:50 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by siamond »

Well, it turns out that Wade Pfau posted a video of Jonathan Guyton explaining his method. Check it out here:

http://wpfau.blogspot.com/2014/06/guyto ... er+Blog%29

Personally, I don't see the point (or the logic) of the first rule (freeze spending in case of portfolio loss), and I didn't see any meaningful effect of activating it or not in my backtesting. The core mechanism is really the guardrails.
User avatar
Cut-Throat
Posts: 2011
Joined: Sun Oct 17, 2010 9:46 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Cut-Throat »

archbish99 wrote:My issue with the VPW method is that it responds too quickly. Good years shoot up, while bad years drop. The advantage of the Guyton approach is that things stay relatively steady year-to-year; only after a severe shock, or several years of good/bad returns do you experience a big change. (Of course, the counter-argument to that is that running the same portfolio through the Guyton rules, you cut back in both 2003 and 2009 following major portfolio drops the year before, but never regained the cut in real terms. Of course, neither did the portfolio, which is kind of the point.)
VPW was designed to Respond Very quickly, which is why it cannot ever fail. It does not remove assets from a portfolio into the teeth of a bear market, which is the main cause of Portfolio Failure.

VPW was also designed to spend down your portfolio, so that you are not left with a Pile of money at Plan End. So, it is only conservative when it needs to be. Rather than scrimping in advance to prepare for a bear market. So, if you can live with some flexibility in your withdrawal income, VPW would be the preferred method. (My Portfolio AA should only cause a maximum downside of about 25% based on Historical Data),
User avatar
midareff
Posts: 7711
Joined: Mon Nov 29, 2010 9:43 am
Location: Biscayne Bay, South Florida

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by midareff »

It's all lovely planning which takes into account little of the real life things; actual inflation, longevity, and the market. As I recall none of these can be planned for, only recovered from... except death.

IMHO all of the SWR discussions, floor and ceiling discussions, etc., are seriously lacking in the mechanics of drawdown. If you are drawing from equities to support your lifestyle you are doing something that makes no sense to me. If you are drawing from a ST or LT bond fund that makes more sense. If that bond fund or funds have six to ten years of drawdown it makes more sense. If EQUITIES have a floor under which they don't contribute to the bond fund and a ceiling over which they do, makes even more sense to me "as the buckets flow". Bad one or two years and cut back.... nonsense, WR stretched way too tight, asset allocation and bad withdrawal planning.

My 2 cents.
User avatar
Cut-Throat
Posts: 2011
Joined: Sun Oct 17, 2010 9:46 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Cut-Throat »

midareff wrote:It's all lovely planning which takes into account little of the real life things; actual inflation, longevity, and the market. As I recall none of these can be planned for, only recovered from... except death.

IMHO all of the SWR discussions, floor and ceiling discussions, etc., are seriously lacking in the mechanics of drawdown. If you are drawing from equities to support your lifestyle you are doing something that makes no sense to me. If you are drawing from a ST or LT bond fund that makes more sense. If that bond fund or funds have six to ten years of drawdown it makes more sense. If EQUITIES have a floor under which they don't contribute to the bond fund and a ceiling over which they do, makes even more sense to me "as the buckets flow". Bad one or two years and cut back.... nonsense, WR stretched way too tight, asset allocation and bad withdrawal planning.

My 2 cents.
Actually all of these things can be planned for and compared against past history. And drawing down equities makes perfect sense. Don't confuse 'Plans' with Predicting the future.

If you don't have any flexibility to do these things, you have to get a job. IOW, you don't have enough assets to be flexible.
User avatar
archbish99
Posts: 1649
Joined: Fri Jun 10, 2011 6:02 pm

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by archbish99 »

midareff wrote:IMHO all of the SWR discussions, floor and ceiling discussions, etc., are seriously lacking in the mechanics of drawdown. If you are drawing from equities to support your lifestyle you are doing something that makes no sense to me. If you are drawing from a ST or LT bond fund that makes more sense. If that bond fund or funds have six to ten years of drawdown it makes more sense. If EQUITIES have a floor under which they don't contribute to the bond fund and a ceiling over which they do, makes even more sense to me "as the buckets flow". Bad one or two years and cut back.... nonsense, WR stretched way too tight, asset allocation and bad withdrawal planning.
By definition, if you're drawing from bonds and not from equities, then you're increasing your overall equity allocation the longer the market remains in a slump. That isn't a plan we ordinarily advocate around here, but if it works for you, have fun!
I'm not a financial advisor, I just play one on the Internet.
Leeraar
Posts: 4109
Joined: Tue Dec 10, 2013 7:41 pm
Location: Nowhere

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Leeraar »

archbish99 wrote:
midareff wrote:IMHO all of the SWR discussions, floor and ceiling discussions, etc., are seriously lacking in the mechanics of drawdown. If you are drawing from equities to support your lifestyle you are doing something that makes no sense to me. If you are drawing from a ST or LT bond fund that makes more sense. If that bond fund or funds have six to ten years of drawdown it makes more sense. If EQUITIES have a floor under which they don't contribute to the bond fund and a ceiling over which they do, makes even more sense to me "as the buckets flow". Bad one or two years and cut back.... nonsense, WR stretched way too tight, asset allocation and bad withdrawal planning.
By definition, if you're drawing from bonds and not from equities, then you're increasing your overall equity allocation the longer the market remains in a slump. That isn't a plan we ordinarily advocate around here, but if it works for you, have fun!
Yes,

And therein lies a contradiction. Let's say you have a 25-year plan, with 5-years short term in near-cash earmarked for current expenses. The rest is in equities or some mix of "riskier" investments. Next year you have a 24-year plan, with 5-years short term in near-cash earmarked for current expenses. Guess what? If you want a constant (5-year) near-cash buffer, you are effectively spending out of your long-term investments.

L.
You can get what you want, or you can just get old. (Billy Joel, "Vienna")
User avatar
coachz
Posts: 1048
Joined: Wed Apr 04, 2007 7:10 am
Location: Charleston, SC

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by coachz »

I just reread this thread and I'm more confused than before. :annoyed
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Rodc »

coachz wrote:"but that does not mean they will not work in the future." So doesn't that also apply to Guyton/Klinger ? SWR and backtesting are both simply past data with no guarantee of future performance so I should weigh them both equally suspiciously ?
richard wrote:So?

See the last sentence of my post.
coachz wrote:But that seems to me the same reasons not to trust SWR.
richard wrote:
coachz wrote:Why is backtesting dismissed so much? Is the future really that different than what all the historical data has already experienced? I find that hard to believe. And if the future IS so different then it seems SWR will be equally screwed. What am I missing ? Backtesting seems to take into account the wide variety of possible declines and inflation periods, no ? I just don't want to be doing thinks a certain way because "that's how we always did it before".
Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.

If you do enough testing of historical data, you will find a strategy that works brilliantly, for essentially the same reasons if you have enough people toss a coin 10 times you will find people who can get 10 heads in a row. Betting that these people can repeat their performances would not be a wise move.

There's no assurance that SWR that worked in the past will work in the future, but that does not mean they will not work in the future.
In addition to Rickard's important points that (1) we don't have anywhere near enough data and (2) the background "physics" if you will likely changes with time (unlike in real physics or chemistry), and (3) if you backtest 100 schemes odds are good about 5 will test out as statistically significant at the 95th percentile confidence level, it is also important to understand that in backtesting the "investor" always has nerves of steel. So what if the portfolio crashes horribly if the market recovers in time to pull the proverbial bacon out of the fire? In real life people, even professionals and the like who should know better, often panic and do the wrong thing in such times. Backtesting never has any of the all too often documented human frailties.

I don't think backtesting entirely worthless, but too many put far too much faith in it.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Rodc
Posts: 13601
Joined: Tue Jun 26, 2007 9:46 am

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Rodc »

FWIW: in real life stuff happens. You need a new roof. You have an expensive medical thing come up (insurance covers the procedure but you need to modify the layout of your home). You need assisted living. Someone totals your car and you need a new one 5 years earlier than planned. Maybe your grand kid is a world class musician and you really want to help with world class lessons. Whatever.

Or you work part time in the first years of retirement. Or what to see the world while you still can. Whatever.

Income needs and desires are not entirely smooth and predictable.

I can't imagine how one can really set out on a 30 year journey with a precise plan for every step.

Thus I don't really think there is much value in over thinking these things.
We live a world with knowledge of the future markets has less than one significant figure. And people will still and always demand answers to three significant digits.
Leeraar
Posts: 4109
Joined: Tue Dec 10, 2013 7:41 pm
Location: Nowhere

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Leeraar »

Rodc wrote:FWIW: in real life stuff happens. You need a new roof. You have an expensive medical thing come up (insurance covers the procedure but you need to modify the layout of your home). You need assisted living. Someone totals your car and you need a new one 5 years earlier than planned. Maybe your grand kid is a world class musician and you really want to help with world class lessons. Whatever.

Or you work part time in the first years of retirement. Or what to see the world while you still can. Whatever.

Income needs and desires are not entirely smooth and predictable.

I can't imagine how one can really set out on a 30 year journey with a precise plan for every step.

Thus I don't really think there is much value in over thinking these things.
Maybe, after you and spouse reach 70.

I retired at 58 with a sweetener expiring at age 62, and looked at my options. Kid in college, my and spouses' SS choices, working part time, etc.

Huge variances in our cash flow needs for the next few years.

It makes no sense for me to assure a floor now (I am 63), because at age 70 with delayed SS I will have plenty. In the meantime, our cash requirements are anything but smooth!

L.
You can get what you want, or you can just get old. (Billy Joel, "Vienna")
DFrank
Posts: 474
Joined: Thu Mar 20, 2014 10:44 pm
Location: North Idaho

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by DFrank »

richard wrote:Those starting retirement can be investing for periods of 30 or more years. Younger people for even longer. There are only about 100 years of reliable market data, so fewer than four independent data points. Four data points is not an adequate basis for reliably predicting the future. In addition, it's far from clear that conditions today are comparable to conditions 100 years ago, which would leave us with even less useful data - the world is a safer place (presumably lowering risk premiums), investment technology has improved (e.g., low cost mutual funds), market participants have learned the lessons of the past 100 years, etc., etc.
Why is it necessary that the sequences of actual data used for backtesting all be 100% independent events?

I think the roughly 100 years of data will allow backtesting of 70 or so sequences of real world market performance, and the fact that many of the years are the same in different sequences is less important than that the sequences of good and bad returns are unique for any 30 year period, and that they reflect the way markets have behaved in the real world in the past. Would it be better if we had more years of data to construct these forecasts? Yes, of course. But I think characterizing it as only 4 useful data sets may be a bit extreme.

Now, I agree that the question of whether the future will look like the past is a real concern about any of these forecast techniques. There are many good reasons why it may not.

Dave
Dave
Dandy
Posts: 6701
Joined: Sun Apr 25, 2010 7:42 pm

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Dandy »

My concern with any withdrawal strategy, especially one that is complicated, is that it will not be manageable in later years for me and my wife or helpful children. Also, anything that can drastically change the amount of withdrawal is problematic at least for me.

I think most people will manage their withdrawals well by periodic assessment of their health, expenses and portfolio balance. When times are bad most will cut back on all sorts of things to preserve their nest egg. Cars, clothes, eating out, charity, doctor visits, travel, subscriptions, etc. The "risk" they will likely face is that they could have spent a bit more and died with a bit less. I worry more about people who have too much confidence in a plan, no matter how successfully back tested, that encourages them to spend a bit more than their instincts tell them is not wise.

No matter what your initial plan in the green zone (I have enough) once you enter the yellow zone (I don't know if I have enough) you will most likely have made adjustments. Only a very unwise person will wait until the red zone (I know I don't have enough) to start making adjustments or stay the course they are on.
Leeraar
Posts: 4109
Joined: Tue Dec 10, 2013 7:41 pm
Location: Nowhere

Re: Has anyone tried/used the Guyton/Klinger method for a VP

Post by Leeraar »

Dandy wrote:My concern with any withdrawal strategy, especially one that is complicated, is that it will not be manageable in later years for me and my wife or helpful children. Also, anything that can drastically change the amount of withdrawal is problematic at least for me.
Dandy,

Consider buying an SPIA. That will give you a predictable income for life.


If the idea there is nothing left after you pass bothers you, and you have enough money, consider a bond or TIPS ladder instead.

L,
You can get what you want, or you can just get old. (Billy Joel, "Vienna")
Post Reply