Smart guys question 4% SWR strategy

[AdventuresAddict]

I share this view. I think investing solely in a balanced fund and withdrawing at the 4% SWR is both inefficient and riskier. I had proposed my personal view that a better plan would be to combine an annuities portfolio with a balanced fund. One major reason why an annuity should be included is it protects you against the uncertainties in the financial markets. 4% might be sustainable historically, but history may not repeat, especiall now that the financial market is flooded with derivatives that add to the instability. The recent subprime crisis highlighted the depth of the problem. But I am afraid that still more crises may occur with more people going online, more speculators, more pressure on the fund managers to deliver returns that are no longer sustainable. An annuity gives us more certainty. A few annuities spread across different highly-rated insurers and denominated in different currencies smooth out the carrier risks and currency risks. This annuities portfolo should provide a survival income. And, they should provide more income than the 4% SWR for most who are older, because annuities take into consideration the mortality factors when determining the pay-out. The balanced portfolio, on the other hand, should be used to serve the needs of liquidity, potential capital growth and legacy to heir. The 4% from this balanced fund could be withdrawn to pay for discretionary expenses above the survival costs, which is paid for with annuity pay-outs.

 

[Rich_by_the_Bay]

Did your study include just the U.S. stock market or all of the world stock markets?

Because not everyone knows Bob and because of his respect for our board rules against self-promotion, let me jump in on his behalf to mention that he has authored one of the more highly respected books on early (semi-)retirement. For those seeking further explanation of his suggestions, I recommend that you read Bob Clyatt's Work Less, Live More: The Way to Semi-Retirement.

That said, I now return you to your regularly scheduled programming.

 

[rs0460a]

I share this view. I think investing solely in a balanced fund and withdrawing at the 4% SWR is both inefficient and riskier. I had proposed my personal view that a better plan would be to combine an annuities portfolio with a balanced fund. One major reason why an annuity should be included is it protects you against the uncertainties in the financial markets. 4% might be sustainable historically, but history may not repeat, especiall now that the financial market is flooded with derivatives that add to the instability. The recent subprime crisis highlighted the depth of the problem. But I am afraid that still more crises may occur with more people going online, more speculators, more pressure on the fund managers to deliver returns that are no longer sustainable. An annuity gives us more certainty. A few annuities spread across different highly-rated insurers and denominated in different currencies smooth out the carrier risks and currency risks. This annuities portfolo should provide a survival income. And, they should provide more income than the 4% SWR for most who are older, because annuities take into consideration the mortality factors when determining the pay-out. The balanced portfolio, on the other hand, should be used to serve the needs of liquidity, potential capital growth and legacy to heir. The 4% from this balanced fund could be withdrawn to pay for discretionary expenses above the survival costs, which is paid for with annuity pay-outs.

While I totally agree with you (assuming you are speaking of an SPIA), be aware that you are "treading water" with this group. Few (if any) have an annuity (again, an Single Payment Immediate Annuity) in their forecast (or as in my case, actual) retirement income plan.

I won't argue the point. All I know is for me (and my DW) it works. As far as your indivudial plan goes, well, it's your plan (and your life). No need to "convert/ convince" you.

However, if you want further info how it works for me, just ask ...

- Ron

 

[cute fuzzy bunny]

Few (if any) have an annuity (again, an Single Payment Immediate Annuity) in their forecast (or as in my case, actual) retirement income plan.

Another fallacy. Many members here have SPIA's as part of their financial plan and as far as I know, there are only one or two people who dont like annuities of any kind.

What doesnt fly here are annuity salesmen with funny math and questionable approaches, or annuities with excessive costs and weak payouts.

I look at them from the same perspective I look at all investments: whats the risk, whats the benefit, is there enough benefit to offset the risk.

As ESRBob's research has shown, even a small reliable income stream can drastically improve portfolio survivability, especially if you can squeak by on that income alone. The feed from an SPIA can provide that stream. Or a pension. Or a dozen other things.

Diversification is also a benefit with annuities...I'd put them as the sixth or seventh asset class I'd put money into. Thing is I'm only down to #4.

 

[FinanceDude]

The general rule is that you can take 4% for that year. That's the whole point of the 4% rule. Over time, the 4% compensates for up and down markets, based on historical worst-of-time scenarios.

Also, the general rule is that taxes are included in the 4%.

Well, that is the broad-based assumption that we are attempting to decry,no?? So, if your portfolio is down 10% the FIRST year your retire, you're still ok to take 4%? Moshe Milevsky has research that suggests otherwise.........

 

[dbr]

I did some simple calculations using the PMT function- yes, at a 2% real return, in 30 years, you'd be left with nothing at the 4.46% withdrawal. That would equate to worst case for the 4% SWR using Bengen's calculations.

But, this is an ER forum so you need to take more than 30 years into account. I calculated for 40 years - 3.655%, for 50 years - 3.182%. And then there are the practical issues as others have pointed out.

Through his example, I understand his points that the 4% SWR method causes you to leave money on the table (he refers to it as paying a surplus), can be replicated for less (how?), and if the preferences of the retiree are known, can probably be done for even less (again how?).

But he lost me after that. What do we do about it? Also his example of a single 30 year period left me cold - that's our most important and least controllable variable - our lifespan!

Are there some mathematicians here who could simplify his math for us?

Here is a data point one can consider in any way one wants: A 65 year old investor can go to the SPIA quote screen at Vanguard and learn that an inflation indexed SPIA funded with $100,000 has an annual payout of $6780 initial increased by inflation, guaranteed for the life of the individual. That is a lifetime withdrawal rate of COLA'd 6.8% with an endpoint of zero wealth at death. The reason this tool is not sensitive to individual lifetime is the insurance principle of pooling longevity risk over a large population.

 

[W2R]

Many members here have SPIA's as part of their financial plan and as far as I know, there are only one or two people who dont like annuities of any kind.

As ESRBob's research has shown, even a small reliable income stream can drastically improve portfolio survivability, especially if you can squeak by on that income alone. The feed from an SPIA can provide that stream. Or a pension. Or a dozen other things.

Diversification is also a benefit with annuities...I'd put them as the sixth or seventh asset class I'd put money into. Thing is I'm only down to #4.

That's an interesting way to look at annuities! I had never thought of them as an asset class, so much as essentially a reduction in living expenses. They do provide an income stream from an entirely different source, so they also provide diversification in that sense.

Although I don't really have anything against annuities in certain specific circumstances, and would consider one, the ever-lower interest rates upon which the lifetime immediate fixed annuity payments are calculated seem to have made that decision for me. If we were living in the higher interest rate environment of 20 years ago, I would buy one in a heartbeat and my monthly income from it could be half again as much.

 

[audreyh1]

I like the 'Modified 4% Rule' -- take 4% of your portfolio value every year and live on that. That way you'll be making annual adjustments -- both up and down. Almost like saying "Every Year is Year 1". If you operate that way, you're guaranteed to never run out of money (As some wags have noted, 4% of your last dollar still leaves you with 96 cents...)

Seriously, the data on this method suggest that compared to the original 4% Rule, you'll take out as much or more withdrawals over the long run, with a far higher portfolio survival record -- survival being defined as keeping up the real value of the portfolio over time, which also means your inflation-adjusted level of income stays up, and you'll still have your money. The only hard part is during down years when you need to tighten your belt. But since everyone else will be also, it might not feel so hard.

That's what I follow with the additional tweak of having a "short-term" - i.e. 2 to 3 year cash account for living expenses. This latter just helps smooth out the year-to-year variations as well as being there for the occasional large-item purchase/expense. It also helps keep the focus on market fluctuations over a longer (2 year) time frame rather than what is happening in the current year.

I never could accept the idea of some fixed initial percentage plus annual inflation adjustment. I felt the need to respond to fluctuations in my portfolio - down or up.

Audrey

 

[Independent]

I didn't read the whole thing, just the excerpts provided above, but I interpret this to mean....

that at the end of 30 years of a 'guranteed' 4.46% SWR, you are also 'guaranteed' to have zero dollars left.

Is that correct? That is how I interpret 'never has a surplus'. If you assume a fixed % return (I don't know how he predicts future returns, or does he just buy 30 year bonds?), you can simply self-annuitize that over 30 years. Is that what the article is saying?

Fine approach - at the same time I can call and have my gravestone carved at today's wages with the year '2038' on it.

edit/add:

OK, I can go plug that into a SS, but 4.46% sounds right if you self-annuitize and assume a 2% real return.

-ERD50

Right. This is another article that purports to be very "sophisticated" in terms of investment risk, but completely ignores the mortality risk.

It's like someone living in Mississippi who will tell you all about his hurricane insurance, but is entirely blind to the possibility of a tornado.

 

[lazygood4nothinbum]

so much of this i don't understand. but at least i have the most important half the problem solved. i only worry about running out of money. if "the price paid for funding... unspent surpluses" is the cost of a good night's sleep, that is a luxury that having solved the important half of the problem allows me to afford.

if i wanted to spend more money i'd still be working for it. i'm reminded of my cousin telling me about selling his father's house years ago on star island in miami. "if i'd kept the house, i'd be rich today," he said with a smile on our way from the restaurant to his privately owned new york city office tower.

 

[heyyou]

Please consider the historical perspective of Bengen's studies. Investors had endured 1966-82 with only dividends as gains, and severe inflation coupled with 10-15% CDs for those who did have cash. So in the early 1990's, no one had answered the question "How much $ do I need to retire?" Bengen answered that question with his non-engineering-background, computer analysis. Is anyone suprised that ERs were not included, or that 15 years later, there is room for improvement? It was a first, and a big step forward, not perfection, and there will be other steps, some of them forward.

In theory, no one requires a base level of steady retirement income, but most of us do. So we have cash positions or steady income sources to supplement the 4% WD. That is income diversification, as others have mentioned.

There is too much money riding on the CPI for it not to be influenced by those few who can. I will continue to take my chances in the volitile stock market before I commit more than a fraction of my funds to inflation based securities. There is always some investment that I missed early, that in retrospect, was a good investment back then. Such is life.

As mentioned earlier, you can't hammer the risk out of 40-50 years of living off of your portfolio. Get used to it.

 

[ERD50]

I played with the '95% rule' in FIRECALC, and the result didn't make sense to me.

Didn't want to hijack this thread, so I started a new one in the FIRECALC section, if anyone is interested:

http://www.early-retirement.org/forums/showpost.php?p=645431&postcount=1

-ERD50

 

[ESRBob]

Did your study include just the U.S. stock market or all of the world stock markets?

Yes -- in the first round of studies we just used US data, but in the second round I had access to the Dimson Marsh Staunton series of historical international equities data. The result was that the portfolios with the international data fared a little worse, but not materially. It could be because of your point that the 4 horsemen have shown up in a lot of places during the 20th century, trashing markets in their wake, and that perhaps they haven't shown up in as much force or frequency in the US keeping our market performance on the upper end of international norms. As you say, it could be our turn again, at which point we'll be glad to be globally diversified!

As for YouBet's question -- a little cushion in this plan is always welcome. If you are cut right to the bone, then yes, in a down year you could find your withdrawal puts you below your minimums. Two things might help -- my data actually supports a slightly higher withdrawal rate (4.5% or so) since the widely diversified portfolio and the way of withdrawing from it reduce volatility and support portfolio survival compared to the benchmark portfolios that got the 4% benchmark started.
Second, I am a big fan of using all your free time in ER to find something you like to do that can make you a little bit of cash if you needed it, but that in no way feels like w&#k. With that in hand, you can generally find a way to rustle up a few thousand dollars a year to fill the gap in your budget if you need to.

Audrey -- I like the cash cushion idea. You can generally get to it using the published Work Less Live More portfolios if you simply hold a slug of your fixed income in short term bond funds or even a medium term CD. It has the safety or nearly the same safety as cash while giving you a rate of return close to the medium term bonds the portfolio is based around. But it's a common tweak many people like to make.

I always loved the idea of some annuity income in the portfolio, however you classify it, but I haven't ever found inflation adjusted annuities with a suitable rate of return for middle aged people, and from a company who you know you can trust to be around for 50+ years. If there were such a thing (inflation-adjusted bonds come close, but the real return now is so lame) then I'd want to lay off some of that market risk for sure. There was a short time that inflation-adjusted bonds paid over 4% real, and I realize too late that I should have been a buyer then..

 

[brewer12345]

Bob, IIRC the trough of the 1966-1982 retirement killer period had the retiree drawing roughly 75% of the inflation-adjusted starting withdrawal using the 4%/95% rule, right?

 

[youbet]

As for YouBet's question -- a little cushion in this plan is always welcome. If you are cut right to the bone, then yes, in a down year you could find your withdrawal puts you below your minimums. Two things might help -- my data actually supports a slightly higher withdrawal rate (4.5% or so) since the widely diversified portfolio and the way of withdrawing from it reduce volatility and support portfolio survival compared to the benchmark portfolios that got the 4% benchmark started.
..

OK. I understand and agree. The "withdraw 4% of the portfolio balance plan" would call for a larger starting portfolio and, all other things being equal, probably more time in the harness. The more aggressive "withdraw 4% plus inflation plan" would call for a smaller starting portfolio but would entail more risk.

Almost everyone that participates on this forum seems to have some source of RE income (pension, working spouse, part time work, likely future inheritance, manage real estate, etc.) beyond their FIRE portfolio and future SS, and that includes me. Still, I find the cases where people are planning very early retirements on only portfolio generated incomes and who want to retire as soon as possible to be the most interesting to hear about and analyze!

 

[ESRBob]

Bob, IIRC the trough of the 1966-1982 retirement killer period had the retiree drawing roughly 75% of the inflation-adjusted starting withdrawal using the 4%/95% rule, right?

Brewer, I'd never actually done those calcs as a %, so I dusted off the spreadsheet and have an answer.

My method ended up giving a real withdrawal 80% or less of the original withdrawal in 5 of the 30 years, twice going to 69% of the original real withdrawal. In 9 of the years, it was 90% or less of the original withdrawal. These were all during the dark days of the 70s until 1982.

It was a case of making a series of mostly small, annual adjustments to the portfolio during hard times (low market returns and high inflation), but with the outcome that the real value of the portfolio 30 years later was about triple that of the Traditional 4% Rule portfolio, 1.4x the original value in 1966 vs about 45% of the original real value for the Traditional 4% withdrawal.

Worth noting that in 17 of the 30 periods, the real withdrawal was above the original withdrawal. In the old method, you never get above -- you only match the original real withdrawal.

Still living at the bottom of the trough on 75% or so of the original wouldn't have been easy, but in seems clear that it would have been worth making these tradeoffs to have survived all that economic malaise without having to go back to full time career work because of a decimated portfolio.

 

[ESRBob]

OK. I understand and agree. The "withdraw 4% of the portfolio balance plan" would call for a larger starting portfolio and, all other things being equal, probably more time in the harness. The more aggressive "withdraw 4% plus inflation plan" would call for a smaller starting portfolio but would entail more risk.

Almost everyone that participates on this forum seems to have some source of RE income (pension, working spouse, part time work, likely future inheritance, manage real estate, etc.) beyond their FIRE portfolio and future SS, and that includes me. Still, I find the cases where people are planning very early retirements on only portfolio generated incomes and who want to retire as soon as possible to be the most interesting to hear about and analyze!

Youbet,
Not necessarily a larger starting portfolio, as long as you're prepared either to be lucky or to do a little part time work if you need to.

 

[audreyh1]

Audrey -- I like the cash cushion idea. You can generally get to it using the published Work Less Live More portfolios if you simply hold a slug of your fixed income in short term bond funds or even a medium term CD. It has the safety or nearly the same safety as cash while giving you a rate of return close to the medium term bonds the portfolio is based around. But it's a common tweak many people like to make.

I do have most of my fixed-income in short to medium term bond funds as well as some cash in my portfolio. But that's because Armstrong convinced me that such FI correlated less with equities. And it's really there for rebalancing against equities. Having a cash cushion in addition takes some of the angst out of rebalancing when equities are down - i.e. seeing my cash/FI pot shrink while I buy stock funds on sale (and wondering if I'll be doing it yet again in a year). I just find it mentally useful to have them separated.

Of course, having a 2 to 3 years cash in addition to the portfolio means I actually had 27x to 28x of my annual needs. I started out this way because we padded a generous "travel budget" for use right after retiring, and I liked psychological isolation of the cash buffer so much that I kept it. I'd never even heard of buckets! But now I see it also as a way of averaging out over fat years/lean years at least in the short term. And if things really go south, I can start making my short term cash fund stretch farther.

Audrey

 

[unclemick]

Hmmm - even with the number of ER's here - I doubt anybody matches up - one to one.

Plus starting in 1993 - real estate income, dividend stocks, a small starting cash pile from severance, temp work, small pension in 98, SS in 2005, and Mr Market in my historical period so far - have allowed me to make changes as I went along. 2008 ER is not 1993.

The 4% rule is 'the one' which survived all prior history math wise. ESRBob's work gives me a good feel for tweaking/making adjustments in the stretch since math is not real life - I may not croak at precisely 84.6.

Sort of general principles keeping me on the right playing field.

heh heh heh - Rude and crude 25X expenses on up to the more elegant problem solving.

 

[Kronk]

The premise in the article is something I've been mulling over for a while. It seems like most retirees are cutting spending back with the downturn in the stock market. And no one seems to think the CPI accurately reflects inflation. So a rigid CPI-adjusted percentage just doesn't seem to make sense.

So the article contends that you're really underspending in a whole lot of those scenarios in order to get a 95% success rate. That part makes sense to me.

I like ESRBob's method, and I'll probably use something like that in a dozen years or whenever it is I ER. It is great that there are so few numbers that go into figuring out your annual withdrawal: your portfolio value, last year's withdrawal, and the whatever your floor percentage is (95% or whatever).

 

[brewer12345]

Brewer, I'd never actually done those calcs as a %, so I dusted off the spreadsheet and have an answer.

My method ended up giving a real withdrawal 80% or less of the original withdrawal in 5 of the 30 years, twice going to 69% of the original real withdrawal. In 9 of the years, it was 90% or less of the original withdrawal. These were all during the dark days of the 70s until 1982.

It was a case of making a series of mostly small, annual adjustments to the portfolio during hard times (low market returns and high inflation), but with the outcome that the real value of the portfolio 30 years later was about triple that of the Traditional 4% Rule portfolio, 1.4x the original value in 1966 vs about 45% of the original real value for the Traditional 4% withdrawal.

Worth noting that in 17 of the 30 periods, the real withdrawal was above the original withdrawal. In the old method, you never get above -- you only match the original real withdrawal.

Still living at the bottom of the trough on 75% or so of the original wouldn't have been easy, but in seems clear that it would have been worth making these tradeoffs to have survived all that economic malaise without having to go back to full time career work because of a decimated portfolio.

Thanks, Bob. You had actually shared your spreadsheet with me a while back and I went and looked at what the downside was in budget etrms. Just couldn't find the sheet now and wanted to make sure my memory was at least in hand grenade territory of the truth.

For a retiree with some flexibility, I think the 4%/95% rule makes a great deal of sense.

 

[youbet]

Youbet,
Not necessarily a larger starting portfolio, as long as you're prepared either to be lucky or to do a little part time work if you need to.

Well....ahhh.....OK. It would mean needing a larger starting portfolio for me as luck isn't my strong suite and my career/trade/profession was much more compatible with working an extra year or two than going part time later. But I understand your point.

It's interesting how making the 4% + inflation plan safer always involves having more money. Lower withdrawal rate = larger portfolio (more money). Work part time = more money. Extra cash "on the side" to use in down markets = more money. Only withdraw 4% of the current portfolio balance = more money. Damn. It just keeps coming back to the fact that if you want more money to spend in retirement, or less risk of running out, you have to have more money!

The only exception I've come across are schemes that use an SPIA or delayed SS to provide a cola'd income stream and I haven't finished investigating that yet. Good article about it on the Retire Early Home Page.

 

[youbet]

I liked psychological isolation of the cash buffer so much that I kept it. I'd never even heard of buckets! But now I see it also as a way of averaging out over fat years/lean years at least in the short term.

Watch it Audrey! You're likely to be the victim of a patent/copyright infringement lawsuit! Ray Lucia has a whole parking lot full of Class A diesel pusher motor homes he's won as a result of law suits against retirees doing what you're doing without using the "buckets jargon!"

Be careful!

 

[ESRBob]

YouBet,
Yep, you heard it here first: More money is usually a good thing

I agree, counting on luck isn't really an investment strategy. Still, the thing about all these SWR studies is that they work around a worst case scenario. Funny thing is we forget the corollary: 90% or so of the possible scenarios (if past is prologue) will be better than the worst case scenarios, meaning 9 times out of 10 you'll have more money than your worst case would suggest. Thus your minimum portfolio should hold up fine in 9 out of 10 cases, and it's only that tenth case that would have you looking for part time work.

Now those are odds I wouldn't mind betting on, as long as the result didn't mean I'd be really scr&%$d if I lost the bet.

Still, more money is nice, if only to support a growing lifestyle and compensate for any unanticipated spending glitches (to say nothing of market hiccups). And if you can earn it with an extra year or two in the saddle during your peak earning years, without having a heart attack or life meltdown... probably worth it. Remember, you'll be FI at that point, which for some people can make work a little more tolerable.

 

[youbet]

Bob, I couldn't agree more.

I'm two years into retirement although due to collecting unemployment compensation, some delayed compensation from MegaCorp and DW getting some modest part time $$$, I'm really just starting the "real thing" financially now.

If I'm fortunate enough to live for 20-30 years or so and don't have any circumstances that throw all planning to the wind, I'm betting I'll wind up using several withdrawal schemes over time and will always be able to second guess myself at that............

Heck of a good time so far!