Buckets & Buffers

[Mysto]

As I continue my quest to build the "perfect" retirement portfolio - I am moving toward the theory of a buffer or bucket approach.

I've just finished reading Asset Dedication by Huxley and Burns. They make a case for having 5 - 10 years of required income in a fixed income ladder. (In other words a self annunity) The rest of the funds are 100% in equity index. The idea is that you have eliminated much of the volatility risk with the income ladder so the rest should be invested for maximum growth. This basic idea is very appealing to me and is the way that I am currently leaning.

I have also seen other methods suggested that use a three bucket approach with a short - medium - and long term portfolios.

I would like to hear from people in the group that use this buffer or bucket approach and how you go about it. (Grumpy - I know you do and I'd like to more about your structure and rules). I'd also like to hear from those that don't agree with this approach and why.

I strive to learn.

 

[unclemick]

Heh heh heh

Theory, practice and hindsight at year 13 of ER:

Theorywise I'm in the Norwegian widow/YMOYL(Your Money or Your Life book) - make your steady income stream cover expenses via interest/divs/pension/etc.

With hindsight a strong case for buckets could be made - ala stretching and spreading out lumps - a mish mash of severence/unemployment, selling and consuming RE, reinvesting DRIP divs until spin off, buyout/merger, plain sell gave a lump to smear out spending. Note - small pension kicked in at age 6 of ER.

Right now I'm working on sweeping stuff into VG Prime MM and establishing a steady take out budget - ie the income model. Have some more pesky DRIPS to kill off - put the stocks into VG Broker and sweep divs into MM.

Theorywise I'm a mess - income with pension/early SS/trad IRA, have a small bucket of Roth conversion set aside for my 80's just in case.

Interesting debate - I'm listening.

heh heh heh

 

[Ol_Rancher]

I use the 'buffer' approach. Right now I have the buffer in 5% CDs at Corus bank.

I do not own any bonds or bond funds since my defined benefit teacher retirement (valued at 25 X  yearly annuity payment) provides balance. Early S.S. in two more years will add back the losses to inflation due to the unCOLAed pension.

I adjusted the equity portfolio by moving out of any emerging market funds, growth funds, or small cap funds. I bought equity/income funds and growth/income funds that have a long history of dividend returns and  less volatility.

I am totally debt free.

I feel I am well positioned to weather any market decline while a market run up will provide good returns.

When the CDs mature I will evaluate the CD rate vs short term bond funs, intermediate bond funds and  tips.

 

[grumpy]

Mysto,

I am retired with a cola'd pension that covers all basic living costs. I have 5 years of the "extras" in the Vanguard MM fund. I created that fund by selling shares I had held for many years in a taxable account in the Lord Abbett Affiliated Fund. I still own quite a few of those shares and take the dividends and capital gains (>$8K last year) in cash. I also have a substantial portfolio of dividend paying stocks which generate more than $8K in dividends each year. As a result, in almost two years of FIRE the MM account has remained almost untouched (even though we have been furnishing our new house). I guess my planning was overly conservative. I will be looking to move some of the MM funds to short term bond funds or laddered CD's sometime soon. If and when I need to replenish those funds I will sell out of my taxable accounts. It looks like our tax deferred accounts can just sit and grow until age 70 and a half. Those accounts are nearly 100% in equities (mostly index funds). I don't hold any bonds or bond funds at the moment. I do have a small percentage in the TSP G - Fund just to provide some ballast.

Grumpy

 

[Lusitan]

I'm still striving to lean too, and at 31 I'm probably about 10 years away from ER. I suppose I'm using somewhat of a short-medium-long term bucket strategy myself.

Long-term: I'm fully funding my 401(k) and IRAs for my wife and myself, and using an asset allocation appropriate for retirement starting in 30 years. This bucket gets funded as top priority, as it is my worst-case "normal retirement" fund that will provide the income from age 60 onward.

Mid-term: This is where I invest funds to provide income from age 40 to 60. Basically, this is my "ER fund". The asset allocation in this bucket is appropriate for planning to use it starting in 10 years.

Short-term: This is basically just an "emergency fund" where we stash a minimum 13 months of living expenses plus savings for any large purchases (car, downpayment on new house, etc.). The 13 months living expenses are in a CD ladder spread out over 13 months, with one that renews every month. The rest is in Vanguard VMMX money market fund.

When you sum it all up, it's just a shorthand way for me to create an overall asset allocation that is right for my "big picture" plan. But making these conceptual buckets allows me to prioritize and tweak my plans a little bit which I find helpful as I learn more about the best asset allocation etc.

 

[audreyh1]

Some of us learned the "Seven year rule" - always have 7 years of needs in cash and short-term cash equivalents (high quality short-term bond funds qualify) as part of your portfolio.  That way you can "wait out" down equity years, and replenish your cash/short-term bonds when equities have a good year.  This rule is actually used for withdrawal - withdraw from your cash/short-term bond bucket in equity down years, and withdraw from your equities bucket in equity up years.

I have found that using straight asset allocation with a large enough cash/short-term bond allocation covers this case just fine.

Say for example you hold 55% equities, and 45% cash/short-term bonds.  (cash plus short-term bonds is similar to a fixed income ladder of short duration).  As long as your cash/short-term bonds covers at least 7 years of required income, it can function in a similar way.

With the rule that your portfolio be 25x your required income (i.e. 4% SWR), 45% cash/short-term bonds represents 11.25 years of required income.  If you used the 7 year rule, you could still rebalance your portfolio if equities declined precipitously (thus take advantage of a good buying opportunity as asset allocation seeks to do), as long as you never let your cash/bond allocation drop below 7 years of needs.  So you might not rebalance as aggressively perhaps without this constraint, but you might be better able to sleep at night yet still get some advantage of asset allocation.

Being able to sleep at night is a key important characteristic of any investment discipline.

OK - I hope that wasn't ridiculously confusing!

Audrey

 

[TromboneAl]

Well,

Here's a theoretically foolproof way to win at the roulette wheel: Bet $1.  If you lose, bet $2.  Continue on, such that whenever you lose, you bet an amount that will win back all your losses if you win. 

This will work, except for one problem: at some point you may not have enough money to place your bet.  At that point, the betting system fails in a big way.

In the same way, the buffer system isn't reliable, because at some point, you may have to replenish your fixed income buffer when the market has been down for a while.   At that point you withdraw money from the stock portion when stocks are down.  You are then subject to the problems that come from withdrawing from a volatile asset class.

In fact, you're probably going to have to continually add some to the fixed portion, since it isn't enough by itself for the rest of your life.  If you say "I'll only do that when the market is up," then you're trying to time the market. What does the book say about replenishing your fixed income ladder?

 

[MasterBlaster]

Audreyh:

Just so you know...

a 4% SWR represents 25x your annual income (not 20x as you indicated).

 

[audreyh1]

In the same way, the buffer system isn't reliable, because at some point, you may have to replenish your fixed income buffer when the market has been down for a while.   At that point you withdraw money from the stock portion when stocks are down.  You are then subject to the problems that come from withdrawing from a volatile asset class.

The idea is that you pick a time duration in which chances are very good that the market will have recovered by the time you need to withdraw from the equity portion.

At ten years, odds are extremely high that the market will have recovered.

Audrey

 

[audreyh1]

Audreyh:

Just so you know...

a 4% SWR represents 25x your annual income (not 20x as you indicated).

Ooops! - Thanks! fixed my post.

Audrey

 

[WilliamG]

Hi Mysto - I retired in 2000 at 55 and have been using the "buckets/buffer" approach. First exposed to it by Frank Armstrong articles. I also agree with Audrey that you can look at your portfolio's total asset allocation with this "reserve" as a sub-set.

I turn 62 next year and will begin taking SS. Have done a 6 year income/expense spreadsheet based on retirement experience so far and that guides what portfolio needs to produce, including special expenses such as new cars. Have been using Quicken all this time and this really helps in understanding your spending.

Have settled on a 50/50 equity/non-equity allocation. Use 20% for Reserve which covers 5-10 years of expenses (if you sweep interest and dividends) and is also there for big ticket items when needed. We use a 5 year CD ladder in IRA with each rung in ladder currently being more than enuf to cover what I project to withdraw each year over next 6 if market is down. In taxable accounts use IBonds and treasuries for remainder of reserve. This allows for 30% in IRA bond funds; less volatile than stock but higher return for ongoing income.
Currently have 10% each in Vanguard High Yield Corp, TIPS and GNMA.

 

[TromboneAl]

The idea is that you pick a time duration in which chances are very good that the market will have recovered by the time you need to withdraw from the equity portion.

At ten years, odds are extremely high that the market will have recovered.

How do you decide, during those 10 years, when to replenish the cash?

In a sense, this system isn't so different from a standard rebalancing system. For example, let's say you've got $1M bucks, and spend 40K per year. So you have 400K (10 years) in your ladder, and 600K in stocks. That's the same as a 60/40 allocation.

 

[Mysto]

Good Stuff!  Keep it coming please.

A few comments from me.  First, I am not defending the book - it just got me thinking more about an area I was already exploring.

Some possible thoughts.

Unclemic2 - I've enjoyed reading your posts on the widow but I'm not convinced that investing in div stocks is best at this current time (I have been wrong and could be again)

audreyh1 - I like your approach - the only prossible problem I see it that because your ladder is actually in bond "funds" that the draw down is variable (I'm not sure that is the correct term here but) unlike external bond or cd ladders that have a known rate of return - but I like the ability to better rebalance and it does allow for upside.

TromboneAl - Remember I am not trying to defend this one book but rather explore a method of trying to limit variability risk - having said that a 10 year period would allow for average equity returns in all but "I think" one period in past history (1973- decade was about 1.9%) so even in a down year you should have a substantial increase of your portfolio- then if the market is up, build a new ladder - if it is down only build a year at a time until an up market. The book has several approaches to the problem.

A big piece of this is the sleep factor.  I have a low risk tolerance but if I know that I'm covered for 10 years I could be a little more confortable in a much larger equity share than I otherwise would have.  As the major growth (historically) has been in equities that would/should be a good thing.

Having said that, in trying my scenarios in firecalc - I have found that 100% equities in the portfolio is probably a bad idea anyway.  By using at least 15-20% I can improve chances of success at maximun withdrawal with only a slight decrease in average portfolio growth.  (I suspect that that is due to periods of extremely slow equity growth as I stated above or very high interest rates) so if I continue with this plan I would probably use 10% short or intermediate treasuries and 10% TIPS to help in case of low equity returns and some inflation smoothing.

More ideas please.  I'm learning with every post - and any comments I've made do not necessarily reflect how I will feel tomorrow.



:

 

[nun]

How would you design a 10 year ladder investment? If you had $1M, need $40k income and put $600k in stocks where would you put the $400k for the 10 year ladder. CDs of various length?
Next how would you rebalance?

 

[youbet]

 I have a low risk tolerance

That's the important point to note Mysto!  And there is nothing wrong with that.  You just have to structure your allocations to accomodate your own investing personality.  Your allocation structure should emphasize stability of principal.

No doubt, generous cash reserves can help keep you from selling equity positions in a down market.  The Huxley and Burns book illustrates one example of managing cash reserves.  Keep reading and pick a method, or invent a hybrid method of your own, and do it!

 

[Mysto]

Hi nun

If I do this, I would build my ladder with CDs 1-10 years (currently paying about 5.2 or so - easy and insured) It could also be done with treasuries or corporate bonds but it is more difficult (strangely the book authors have a web site where for $99 you can get a bond allotcation worked out <g>)

In the book method there is no rebalancing because everything else is in 100% equities and you can't rebalance 100%.

youbet

The problem with risk tolerance (and they do a good job of exploring this in the book) - is that the measure of it is usually bunk. But a reason for exploring buffers is that if I was to select a portfolio based on "questionaires" ( I done at least 20 of them) I would probably have a 30/70 equity/bond allocation. With a buffer approach I think I can confortably handle something more like 60/40 or better and in my head I know that should reap great benfits.

The book also makes an argument as to the purpose of bond allocations and how to design it that I won't get into here but it is interesting (however there are things in the book that I heavily disagree with)

Still listening to other people that are using buffer approaches or are agin' it. Allocation - the risk factor - and withdrawal strategies are what it is all about IMHO

 

[audreyh1]

How do you decide, during those 10 years, when to replenish the cash?

Gosh I wish I could find the Frank Armstrong article where he describes very specifically when to replenish (anyone have the link?).  I think it goes something like this.  Say after your initial allocation is set up, you have couple of bad equity years.  You don't sell any equities during this time, but you draw down two years expenses out of your cash/bond allocation.  In year 3, the stock market recovers and then some.  In this year, you may be able withdraw some of your expenses from your equities, plus replenish your cash/bonds if enough is left over.   This is probably the more typical scenario.

So the ten year case is just to cover the unlikely worst case scenario where equities go down and take ten years to recover - you would withdraw from your cash/bonds during this entire time.  It's all about risk tolerance.  You match your cushion to your risk tolerance.

In a sense, this system isn't so different from a standard rebalancing system.  For example, let's say you've got $1M bucks, and spend 40K per year.  So you have 400K (10 years) in your ladder, and 600K in stocks.  That's the same as a 60/40 allocation.

That's right!  It's not that different - just an added refinement.  It provides a cushion and it tells you which bucket to withdraw from each year based on market conditions and the performance of your portfolio.  It provides a limit so that you don't rebalance so much during a major market downturn that you lower your "cushion".   It's also a way to make sure you have a sufficient amount allocated to cash/short-term bonds to survive a major market downturn without losing your mind.

I generally keep a larger allocation to cash/bonds than the 7 year rule simply so that I can use some of cash/bonds portion to buy some equities during bad market years, but I just won't let that allocation drop below the "7 year rule".  Peace of mind.

Every good stock market year - portfolio gets rebalanced to bring up the cash/bonds allocation.

Audrey

 

[audreyh1]

audreyh1 - I like your approach - the only prossible problem I see it that because your ladder is actually in bond "funds" that the draw down is variable (I'm not sure that is the correct term here but) unlike external bond or cd ladders that have a known rate of return - but I like the ability to better rebalance and it does allow for upside.

I don't see my drawdown as being variable because I hold bond funds. First of all - you have to match the duration to your goals. For the 7 year case, say you had 2 years in cash, and 3 years in ultra-short term bond fund (duration 2 years or less) and 2 years in short-term bond funds (duration 5 years or less). This is kind of like a fixed income ladder. All dividends are being reinvested in back into each bond fund in this scenario. The duration of the bond fund means it has sufficient time to recover from changes in interest rates before you need to withdraw the money.

I just use bond funds because I find them easier to manage (i.e. more convenient) than CDs or holding individual bonds. Only reason.

I also want to make clear that "cash and high quality short-term cash equivalents" is key here. No room for junk bonds, emerging market or riskier bond funds here. You don't want to go out too far on credit risk with this approach because riskier debt tends to suffer when equities do. Only government and the very top rated corporate short-term debt are the ones appropriate for this bucket.

Audrey

 

[youbet]

I must be missing something on this "buffering" concept.

If I have a 60/40 allocation. The 40% is a diversified mix of MM, short, medium and long instruments, all high quality (treasuries, CD's, funds, etc.) If the equity side is up, I rebalance by spending from there or selling and transfering or both. If the equity side is down, I rebalance by spending from the fixed side and, if equities seem to be a bargain, selling fixed and buying equities. The fixed portion generally contains enough cash or short term money to cover a few years expenses.

What's the difference between this and "buffering" other than apparently "buffering" uses some rules regarding how much cash or short term money to carry?

"Buffering" sounds like a gimic name for allocating and periodic rebalancing.

 

[brewer12345]

"Buffering" sounds like a gimic name for allocating and periodic rebalancing.

We have a winnah!!!

That, and it helps authors wthout a genuinely new idea to sell books.

 

[Mysto]

Hi youbet

I honestly think there is a difference "but it is small"

First - because the fixed income ladder is in stone the constant draw down would have no effect on the investment portfolio (at least until the time span - in this case 10 years when it is reloaded)

Secondly, there is some evidence that it will allow a slightly larger withdrawal rate.  I just played with a few numbers for an example:  Let's say I need an additional 3800 mo. and I have 1080000 to invest. Using firecalc starting at year one with 40/60 "conservative" equity / tips (I figured 2.0 rate) would have a success rate of 97%.  Take out 370K for a fixed income ladder (I used a 9 year ladder - that would be the current rate for a 3800 - 3% bumped annunity) and start the 3800 mo. adj for inflation at year 10 ... it rises to 100% success.  

I've already mentioned the sleep effect - I think for me it might allow a better equity allocation because of the insulated effect (you don't withdraw from a potentially hurt portfolio) but that is only a "mind game" and not finance.  

Having said all of that - I'm not fixed in my mind about the best and that's why I'm enjoying hearing about the various methods that the board is using to decrease risk and withdraw safely. You and many others have made a lot of good points to consider and have offered alternatives that I am considering.  All the while, I'm moving my assets into Vanguard with some already invested in a diversafied portfolio - more are being liquidated and moved this week.

 

[Bailer]

I have been lurking on this board for about a year, and it's time to delurk! I am in my late forties and RE'd about five years ago, with the 4% withdrawal rate and a vague idea about a self-annuity in mind. A site that has helped me quite a bit with the next five (and more) years is www.bucketsofmoney.com. The idea is to divide assets into three "buckets", short-term (the self-annuity), medium and long term. In essence, you spend down the buffer or bucket money in the asset class and let the mid- and long-term money keep up with inflation.

As a caveat, because I know how people on this site feel about financial advisors (sales dudes) the website is used to garner listeners for a radio show and to line up new customers for financial planners. But the demo is a good
illustration that helped me.

 

[donheff]

Some of us learned the "Seven year rule" - always have 7 years of needs in cash and short-term cash equivalents...
If you used the 7 year rule, you could still rebalance your portfolio if equities declined precipitously (thus take advantage of a good buying opportunity as asset allocation seeks to do), as long as you never let your cash/bond allocation drop below 7 years of needs.

It seems to me that, if the purpose of the 7 years of cash is to wait out a bad equities market, you would have to let the cash fund dwindle in a prolonged bear. Thus, if the bear went 5 years and then turned up you would start replenishing your cash accounts after 5 years. That would mean that in a bad market you would fall below 7 years of cash on hand. A form of market timing to be sure, but consistent with the purpose of the cash fund.

If you replenish the cash regardless the market why bother with 7 years. It would seem two or three would lead to the same end result.

Don

 

[audreyh1]

If you replenish the cash regardless the market why bother with 7 years.  It would seem two or three would lead to the same end result.

Don

The seven years is just to hold out on a really bad equity bear market that lasts seven years. And yes during a prolonged bear market, your cash bucket will go way down - you'll have less and less years of expenses covered as you use it up. But you WON'T be drawing on your depressed equities. And you have UP TO seven years to let your equities recover before being forced to draw them down. That's the whole point.

You just have to come up with however long of a bear market you want to be able to survive. Armstrong picked 7 based on historical data (I don't remember the odds of a 7 year period being negative), 10 is even more conservative.

Audrey

 

[audreyh1]

What's the difference between this and  "buffering"  other than apparently "buffering" uses some rules regarding how much cash or short term money to carry?

That's it! It helps you establish you how much cash or short term money to carry.

Also in a prolonged bear market you will have your equity allocation down. After a few years you will also have your cash/short-term being drawn down since you are withdrawing from that bucket. But this method prevents you from selling your equities in a bear market even though you have drawn your cash/short-term down far enough to need to rebalance. This is a significant difference from straight annual rebalancing.

Audrey