Are Ray's Buckets of Money fatally flawed?

[ziggy29]

In other words, the "perfect storm" of 2008 was nver back-tested because there was no real precedent to measure..........

In terms of the magnitude of the loss I'd say 1929-1932 was more than a solid precedent, and arguably so was 1973-74 (except that rates on cash didn't tank as hard then). But there's only one, maybe two data points here, which isn't a sufficient sample size to provide much comfort that the markets will rally back "just in time" to save our portfolios yet again.

 

[W2R]

In terms of the magnitude of the loss I'd say 1929-1932 was more than a solid precedent, and arguably so was 1973-74 (except that rates on cash didn't tank as hard then). But there's only one, maybe two data points here, which isn't a sufficient sample size to provide much comfort that the markets will rally back "just in time" to save our portfolios yet again.

It'll be easier to identify comparable precedents once we are completely through this. Things always seem so much clearer in hindsight, to me, anyway.

I am not at all confident that this won't be followed by runaway inflation that could be worse than anything that has happened yet. But then, who knows. Just thought I'd add that cheery observation.

 

[FinanceDude]

I am not at all confident that this won't be followed by runaway inflation that could be worse than anything that has happened yet. But then, who knows. Just thought I'd add that cheery observation.

Runaway inflation would be better than STAGFLATION.......

 

[FIREd]

I am not a specialist of the "bucket" strategy, but if it were me I would use dividends from the income and equity buckets to replenish (at least partially) the cash bucket on a regular basis and if the cash bucket needed further replenishing sell some assets either from the income or equity bucket on a yearly basis based on performance (I'd be selling treasuries from the income bucket right now). Would it be " bucket blasphemy"? I call it the "total return Norwegian bucket strategy" and that's how I (kinda) intend to do it in retirement...

 

[Independent]

I don't think Ray or almost anyone else thought we would have a 40% loss in equities, along with bonds getting a haircut of 15-20% on a lot of them, plus a fed funds rate of .25, leading to MM funds getting yields of less than 1%.

In other words, the "perfect storm" of 2008 was nver back-tested because there was no real precedent to measure..........

You are correct that this drop is historically unique. But anyone recommending a strategy like this ought to test it against the low returns in the late 60's and 70's. The S&P 500 lost 30% of its (real) value in the 10 years starting in December 1968, and was even for the 15 years with the same starting date. (Those numbers include reinvesting dividends.)

Maybe his strategy worked very well for that period. I don't know, but I think it would be interesting to see.

 

[audreyh1]

YES! I've always used Armstrong's approach (plus a cash bucket - my own personal twist before I even heard of Ray Lucia). IMO, it only makes sense to keep rebalancing out of equities in the good years. This keeps replenishing your "cash" and low-volatility buckets (or bonds) and lowers the volatility of the overall portfolio. I think it's perfectly logical to think that after a run of several good years equities are more vulnerable to a correction or crash. So if you aren't harvesting some of those gains in the good years - then what the heck are you doing?!?! Surely not trying to maximize long-term return! LOL!

Does Lucia's approach REALLY have you not taking at least some of your withdrawals out of equities during the good years? When are you supposed to replenish the other buckets if not after the good years? That's the only thing that makes sense to me!

Audrey

 

[ziggy29]

Does Lucia's approach REALLY have you not taking at least some of your withdrawals out of equities during the good years? When are you supposed to replenish the other buckets if not after the good years? That's the only thing that makes sense to me!

When I've listened to his show, he does advocate occasionally replenishing the near-term buckets when stocks have been doing well. But there really never seems to be any mechanical strategy to do it, and that's always concerned me. It seems like without it you're left to "gut feel" and "emotion" which have torpedoed a lot of portfolios. I'd rather follow a fixed, well-defined process to know HOW much to take off and WHEN.

A variant I've toyed with is something like: "when the return on your stock bucket exceeds the average assumed return for stocks (say 9%), take a defined percentage of the excess return and use it to replenish the income buckets." I wonder if there are any backtests developed on a model like this. On one hand it seems intuitive to take some off the table when equities are doing well, but on the other -- how much?

 

[ERD50]

When I've listened to his show, he does advocate occasionally replenishing the near-term buckets when stocks have been doing well. But there really never seems to be any mechanical strategy to do it, and that's always concerned me.

If it is not mechanical, there really is no way to back test it. And that makes it impossible to determine if it would do better/worse than any strategy, using historical data.

This is starting to sound more like voodoo to me. It's a mild version of the people who show up here and say " Well, I told everyone to sell when the DOW hit X". How do you evaluate that?

-ERD50

 

[ziggy29]

This is starting to sound more like voodoo to me. It's a mild version of the people who show up here and say " Well, I told everyone to sell when the DOW hit X". How do you evaluate that?

The only reason I think it sounds anything like "voodoo" is because a lot of this relies on gut feel and thus (as you say) can't be backtested.

But intuitively, I see nothing "voodoo" about suggesting that you shouldn't be as heavy in stocks when valuations (according to most conventional measures) are at nosebleed levels as when the markets are going on sale. This seems obvious to me (and did before October '07). What's never been clear to me is how to create a solid, backtested and mechanical AA decision based on it. I suspect it can be done, but I'm probably not bright enough (or good enough with databases ans spreadsheets) to figure it out.

I mean, if NOT buying stocks when valuation is high and loading up on them when valuation is low is voodoo, then Warren Buffett has been one helluva witch doctor.

 

[FinanceDude]

If it is not mechanical, there really is no way to back test it. And that makes it impossible to determine if it would do better/worse than any strategy, using historical data.

This is starting to sound more like voodoo to me. It's a mild version of the people who show up here and say " Well, I told everyone to sell when the DOW hit X". How do you evaluate that?

-ERD50

You don't.........remember, EVERYONE is an EXPERT on the Internet............

 

[audreyh1]

Well - this is where Armstrong does have a specific approach.

- Withdraw your annual amount and rebalance when equities have had a good year.
- Withdraw from cash/bonds only when equities have had a bad year.

Basically he doesn't want you withdrawing from equities "when they are down", but to use your cash/high-quality bonds while you wait for the market (equities) to recover. He recommends starting out with at least 7 years in cash and very high-quality short-term bonds in your cash/bonds allocation as this should allow sufficient time for the market to recover after a bad crash.

I don't remember what he says about rebalancing by buying equities after a bad year. But I have interpreted his approach (or made my own) that you can rebalance by buying equities after a bad year AS LONG AS you maintain the minimum number of years in cash/bonds.

In my case, I have held 10 years after-tax expenses as the minimum for my cash+bonds allocation. I confess that in my recent rebalancing last year after the crash and this year, I found I couldn't bring myself to go below 12 years cash/bonds which forced my equity AA down to 55% from where it had originally been 58%. But I still did manage to buy some downtrodden equities both late last year and this year.

Audrey

 

[ziggy29]

But I still did manage to buy some downtrodden equities both late last year and this year.

Sort of like taking cod liver oil. It doesn't go down easily, it leaves a bad taste in your mouth, but in the long run it seems like the healthy thing to do...

 

[ERD50]

The only reason I think it sounds anything like "voodoo" is because a lot of this relies on gut feel and thus (as you say) can't be backtested.

I agree, and that is all I really meant. It probably came across stronger than I intended. Not "voodoo" in the extreme sense of the word, but when something is untestable, because it has not been defined, it's moving closer to "voodoo".

-ERD50

 

[ziggy29]

I agree, and that is all I really meant. It probably came across stronger than I intended. Not "voodoo" in the extreme sense of the word, but when something is untestable, because it has not been defined, it's moving closer to "voodoo".

I don't think the concept is purely untestable, but there are a lot of permutations and slight variations one could make, and choosing the "best" of them would feel like excessive data mining.

Having said that, it would be interesting to see Monte Carlo simulations on various AAs with "valuation-based rebalancing" over (say) 30-50 year periods, much as we do with FIREcalc. A good candidate would do one of four things compared to a traditional fixed AA with periodic rebalancing:

(1) Increase the expected long-term return without an unacceptable increase in risk;

(2) Significantly reduce the risk without an unacceptably large drop in expected return;

(3) Keep the return about the same with less volatility;

(4) Keep the volatility about the same with a higher return.

 

[mathjak107]

I don't think Ray or almost anyone else thought we would have a 40% loss in equities, along with bonds getting a haircut of 15-20% on a lot of them, plus a fed funds rate of .25, leading to MM funds getting yields of less than 1%.

In other words, the "perfect storm" of 2008 was nver back-tested because there was no real precedent to measure..........

i certainly have to agree with you on this point

 

[ERD50]

I don't think the concept is purely untestable, but there are a lot of permutations and slight variations one could make, and choosing the "best" of them would feel like excessive data mining.

Having said that, it would be interesting to see...

Fair statement. It's really the detail levels that are not defined, they may not be all that significant.

Anyway, if I understand this right, that the goal of 'buckets' is to keep you from having to draw down equities over a 14 year down period, it seems this could be modeled in FireCalc.

Take your expenses, enter any pensions, SS, and dividends from your equity portfolio as a faux-pension. Now, *exclude* equities from your portfolio, and include only your fixed holdings. Adjust your fixed until you survive 14 years.

Now, that would not match up those 14 years with 14 years of a down market, but it would give you some ballpark ideas.

-ERD50

 

[Rich_by_the_Bay]

I don't remember what he says about rebalancing by buying equities after a bad year. But I have interpreted his approach (or made my own) that you can rebalance by buying equities after a bad year AS LONG AS you maintain the minimum number of years in cash/bonds.

I believe the only time he has you buying equities directly is if a) stocks are down, bonds are up and b) still have an imbalance favoring bonds after you've taken your yearly allowance from bonds.

In other words, you rebalance the two buckets after taking your distribution for the year, so under some situations you are buying stocks. In the current recession, that could get interesting and you'd be buying a lot of equities this year. Of course you then have to rebalance each portfolio internally as well.

 

[audreyh1]

How much equities you buy depends on how much you can draw down your cash/bonds allocation after withdrawal and maintain your minimum number of years expenses (minus 1 year I suppose). If your bonds have been hurt too (as pretty much everything but treasuries has), there might not be as much available as one would think to buy equities!

Audrey

 

[Gpond]

Do you hear the 3rd hour today? Good ole Ray ranted on and on about he had no problem with the all expense paid for retreats/vacations for companies that received TARP money. (AIG Spa Retreat is an example).

Seems good like looking out for us Ray was giving his corporate sponsors a nice reach around.

 

[mathjak107]

although i havent been motivated to do it recently ,a few years back i pulled 15 year blocks of time out going back decades and averaged them out...lo and behold thru crashes, recessions, prosperity etc i was hard pressed to find a period of 15 years where the long term average wasnt within 1% of each other with a 50/50 mix

now i was thinking about the above and pulling the money first out of bucket 3 in good times ... im not so sure you wouldnt severly diminish bucket 3's long term return by constantly siphoning more money out of it and not letting it just run up maximizing the return on bucket 3 because your pulling the money out before it really really had a chance to grow.

as down as we are bucket 3 was at 4,000 15 years ago and still hovering around 8,000 today....while in this case it would have worked out selling some of bucket 3 earlier think about selling some of bucket 3 off only 2 years into a 10 year bull market.... that would kill bucket 3's growth

 

[Rich_by_the_Bay]

although i havent been motivated to do it recently ,a few years back i pulled 15 year blocks of time out going back decades and averaged them out...lo and behold thru crashes, recessions, prosperity etc i was hard pressed to find a period of 15 years where the long term average wasnt within 1% of each other with a 50/50 mix

now i was thinking about the above and pulling the money first out of bucket 3 in good times ... im not so sure you wouldnt severly diminish bucket 3's long term return by constantly siphoning more money out of it and not letting it just run up maximizing the return on bucket 3 because your pulling the money out before it really really had a chance to grow.

as down as we are bucket 3 was at 4,000 15 years ago and still hovering around 8,000 today....while in this case it would have worked out selling some of bucket 3 earlier think about selling some of bucket 3 off only 2 years into a 10 year bull market.... that would kill bucket 3's growth

I can't vouch for the numbers in my case, but I really do think that you get maximal returns by leaving B3 alone for as long as you can -- not just from Ray but from other academic studies. "Bonds first" give the best returns overall. My concern is the volatility issue, which narrows down as your 15 year distributions come to the last few years. If the market is brutal then like it is now, it could be painful. I might be willing to trade some returns for peace of mind by rebalancing a bit more aggressively, though not overly so..

 

[audreyh1]

I don't understand trying to maximize long term returns when you are in withdrawal mode. To me maximizing long term returns also guarantees maximum (short-term) volatility, which means you are going to go through some pretty scary periods on your way to that large portfolio when you are too old to enjoy it. Syphoning off some equities during good years is a way of lowering the volatility of the overall portfolio after it gets riskier due to appreciated equities. So you are trading off your maximum long term return but gaining short term lower volatility.

Since long-term performance against inflation and portfolio survival of certain AA ratios is well researched and documented, a retiree can choose an AA that best meets their personal tradeoff of long-term return/short-term volatility. Theoretically speaking, of course.

Well, anyway, that's why I rebalance (ala Armstrong) instead of using Lucia's version of buckets.

Audrey

 

[Nords]

That black swan--she's an ugly bird.

Haven't read the book yet, but it's on my list. Spouse says that right now our defensive strategy could best be defined as "gobble gobble"...

I still like the overall approach, and will stick with it but I think I will rebalance more often (maybe every couple of years). Or maybe rebalance to maintain a 7 year cushion, not 2 years. Sure, my returns may be a little lower overalll but I'll maintain a larger, longer cushion for times like this.
Gotta change with the times. If Ray or anyone else comes up with a better-documented revision, I'm all ears. Maybe even look at a SPIA when I'm in my 70s if interest rates are favorable.

I don't think "when" or "how much" is as important to rebalancing as much as the discipline of having a system. "1 January" or "110 minus age" or "after a triple top with a head & shoulders formation", or whatever system is most likely to be followed by its disciple. One of the main benefits of DCA & value averaging is that investors are more likely to stick with the system through all markets.

Here's a FPA article on mechanistic rebalancing recommended by Walkinwood:
http://www.early-retirement.org/forums/showpost.php?p=605754&postcount=9
So now we have our asset allocations set to 23% +/- 20%-- in other words, rebalance when one of them drops below 18% or rises above 28%. Oddly enough it hasn't been a problem in this "no place to hide" black swan.

As for SPIAs: Rich, have you read Milevsky's "Are You A Stock Or A Bond?" ? You might be in for a pleasant surprise about SPIAs:
Raddr's Early Retirement and Financial Strategy Board :: View topic - New book: Milevsky's "Are You a Stock Or a Bond?"

A variant I've toyed with is something like: "when the return on your stock bucket exceeds the average assumed return for stocks (say 9%), take a defined percentage of the excess return and use it to replenish the income buckets." I wonder if there are any backtests developed on a model like this. On one hand it seems intuitive to take some off the table when equities are doing well, but on the other -- how much?

Spouse and I have been through the Oct 1987 plunge and our ER portfolio dropped over 40% during both 2000-2002 and 2008. As life-affirming as our portfolio's survivability may be, the novelty is wearing thin.

From a historical & Monte-Carlo perspective we're doing fine, but from a "sleep at night" perspective she's getting restless. The discussion of "taking some off the table" has come up, although I doubt we'll ever end up with a Buffettesque $40B cash stash waiting for a market such as this.

Best idea we've come up with was to continue reinvesting dividends as long as our asset's share prices are below their long-term moving averages. When they rise above their MAs then we'll start taking dividends in cash and putting them in money markets or long-term CDs. At some point some of that cash may be permanently set aside, but I suspect that it'll go back to work when the share prices drop below their long-term MAs. I suppose part of the discipline of the strategy would be waiting for rock bottom or waiting for a CD to mature before dumping it into an underpriced asset. But instead I think we'd just keep trickling cash into whatever asset drops below 18%.

In the meantime we've refinanced a mortgage and we're either re-renting or selling our rental home. It's always good to be able to play defense.

 

[haha]

90 % of this thread is the bear market talking. 18 to 24 months into the next bull we'll be hearing a different song.

Ha

 

[mathjak107]

I don't understand trying to maximize long term returns when you are in withdrawal mode. To me maximizing long term returns also guarantees maximum (short-term) volatility, which means you are going to go through some pretty scary periods on your way to that large portfolio when you are too old to enjoy it. Syphoning off some equities during good years is a way of lowering the volatility of the overall portfolio after it gets riskier due to appreciated equities. So you are trading off your maximum long term return but gaining short term lower volatility.

Since long-term performance against inflation and portfolio survival of certain AA ratios is well researched and documented, a retiree can choose an AA that best meets their personal tradeoff of long-term return/short-term volatility. Theoretically speaking, of course.

Well, anyway, that's why I rebalance (ala Armstrong) instead of using Lucia's version of buckets.

Audrey

although you maximize your bucket 3's returns by leaving it alone and also get highest volatility; you also get the smoothest performance long term also.... while the 15 year ride can be bumpy its also pretty consistant... its that bucket 3 consistancy that actually smooths out the bumps and to some extent makes the system work at peak efficiancy ....

its like daily , yearly and even in a decades time the market returns are everywhere but when you get to that 15 year mark its almost scarey how the return normals out to about the same percentage no matter what the events of any 15 year period. .

soooo you actually have a choice i see after putting our collective heads together here which is why this is my favorite financial forum...

you can have a smaller withdrawl rate and less bumpy ride in the short term by taking money out of 3 and refilling or living on that money in the good years

or you can leave 3 alone and get a higher return and withdrawl rate but a volatile short term ride