Born lucky? Sequence of returns risk

[Focus]

Wade Pfau's posts are always interesting and thought provoking. Here's the latest, on sequence of returns risk.

These are very different outcomes for people who otherwise behaved exactly the same. ... The actual maximum sustainable withdrawal rates experienced over 30-years ranged anywhere from 1.9% to 10.9% for reasons beyond one's control reflected simply by the luck of when they retired.

This isn't news to most forum members, but nevertheless the implications are profound. Many of us hope that doing all the "right" things (e.g., diversifying in every way possible) will dampen the effects of volatility and allow us to exert some control over our financial future. But ultimately timing appears to override everything for those of us who are invested in the market. I think this is why ER often feels like a leap of faith.

At the end, Pfau offers these suggestions on how to de-risk: 1) Strengthen Social Security (and avoid turning it into defined contribution), 2) Avoid constant inflation-adjusted spending, 3) Consider holding fixed income assets to maturity or using annuities, 4) Consider reducing downside risk through other approaches, including financial derivatives, 5) Consider rising-equity glidepath, 6) Consider "safe savings rate" approach.

 

[ERD50]

There sure is a big delta in his Monte Carlo numbers and the historical data in FIRECalc.

FIRECalc shows the SWR for 100% success over 30 years is 3.59%, that's a long way (almost 2x) from the 1.9% he uses. The upper bound is pretty close, @ 10.3%.

The point still holds, though I'm not sure I agree with his suggestions.

-ERD50

 

[MDJO]

There may be other strategies that listers are using to mitigate risk in the early years of retirement. Would love to hear about them. My own personal strategy is holding five years of living expenses in assets that will not decline in value if held to maturity. Assumption is that over a five year period, it is rare to lose significantly (more than 2 - 3%) in a diversified portfolio....any thoughts out there?

 

[GrayHare]

To my knowledge, Pfau's 1.9% occurs while blindly increasing withdrawals to pace inflation, regardless of how precipitously a portfolio declines during the inevitable, periodic market downturns. My plan is to cut spending during such downturns, if it appears necessary, and give time for the portfolio to recover. Such spending adjustment should boost the SWR above 1.9%.

 

[ERD50]

To my knowledge, Pfau's 1.9% occurs while blindly increasing withdrawals to pace inflation, regardless of how precipitously a portfolio declines during the inevitable, periodic market downturns.

But FIRECalc also does that.

My plan is to cut spending during such downturns, if it appears necessary, and give time for the portfolio to recover. Such spending adjustment should boost the SWR above 1.9%.

I believe people way over-estimate the effect of spending cuts. Think about it - those drops that you see in the first few years of a bad cycle are around 40% (even w/o any withdraws). So how much effect can cutting spending from 4% to 3% have? A single percentage point is swamped by that 40% dip.

You can model this in FIRECalc, just add some income for a few years in the 'other' tab, with an amount equal to your spending cuts. I've found it takes huge cuts for a long time to have much of an effect. Plus, in the recent downturn, we recovered quickly. I would regret giving up two years of discretionary spending in my prime years. You may see that differently, but I'd still suggest you try some numbers to get a feel for what it takes.

-ERD50

-ERD50

 

[GrayHare]

OK, some numbers, first without a downturn. $1m portfolio, year 1 withdraw $40k, leaving $960k. Year 2 add 3% inflation and withdraw $41.2k, leaving $918.8k in portfolio.

What if there is 40% downturn? $1m portfolio, year 1 withdraw $40k, leaving $960k, then a 40% downturn reduces it to $576k. Year 2 blindly increase withdraw to $41.2k, so portfolio is $543.8k. Recovery happens, boosting portfolio up to $891.3k, which is $27.5k less than if downturn had not happened.

What if due to downturn the annual withdrawal is reduced to 3%? $1m portfolio, year 1 withdraw $40k, leaving $960k, then a 40% downturn reduces it to $576k. Year 2 withdraw $30k, so portfolio is $546k. Recovery happens boosting portfolio up to $910k, which is only $8.8k less than if downturn had not happened.

Cutting spending by $11.2k helps portfolio preserve $18.7k more than not cutting... so, 68% of the loss is covered by one year of cutting. To survive longer downturns, use a cash bucket so you don't need to sell portfolio holdings before they recover.

 

[clifp]

Cutting spending by $11.2k helps portfolio preserve $18.7k more than not cutting... so, 68% of the loss is covered by one year of cutting. To survive longer downturns, use a cash bucket so you don't need to sell portfolio holdings before they recover.

Well 11.2K of the preservation came from spending less, so the benefits are actually only $7.5k or less than 1% of the portfolio. It is also worth noting that bear markets generally are much shorter and the declines steeper than recoveries. So realistically you would have cut spending dramatically in 2009 and 2010 and probably 2011 before getting back to near 2007 levels in 2012.

But even if we use your one year recovery scenario. 1 year is 3.3% of a 30 year retirement, going from 40K spending to 30K is pretty dramatic, cutting 27% out of my budget in one year would make me a pretty unhappy guy.

I think a far more realistic solution is to delay the purchase of big ticket items capital improvements during a bear market. Examples include car purchases, home improvements, and the fabulous once or twice in a lifetime vacation. For instance my mom and sister have been urging me to remodel the kitchen for many years. When the bugged me to remodel in 2009 (cause the contractors will be cheaper) I said hell no, I don't want take the money out of the market. Now 4 years later with my stock investment having almost doubled, I have much less of problem spending more than a year expenses on home improvement projects (roof, kitchen, solar).

The good news is I think this is human nature to cut big ticket discretionary spending during bad times. I think anything more drastic than Clyatt 95% rule is unrealistic.

 

[ERD50]

OK, some numbers, first without a downturn. $1m portfolio, year 1 withdraw $40k, leaving $960k. Year 2 add 3% inflation and withdraw $41.2k, leaving $918.8k in portfolio.

What if there is 40% downturn? $1m portfolio, year 1 withdraw $40k, leaving $960k, then a 40% downturn reduces it to $576k. Year 2 blindly increase withdraw to $41.2k, so portfolio is $543.8k. Recovery happens, boosting portfolio up to $891.3k, which is $27.5k less than if downturn had not happened.

What if due to downturn the annual withdrawal is reduced to 3%? $1m portfolio, year 1 withdraw $40k, leaving $960k, then a 40% downturn reduces it to $576k. Year 2 withdraw $30k, so portfolio is $546k. Recovery happens boosting portfolio up to $910k, which is only $8.8k less than if downturn had not happened.

Cutting spending by $11.2k helps portfolio preserve $18.7k more than not cutting... so, 68% of the loss is covered by one year of cutting. To survive longer downturns, use a cash bucket so you don't need to sell portfolio holdings before they recover.

OK, thanks for the numbers, but $18.7K is under 2% of the portfolio. Not a big number over the long run. Obviously, amplified over several more years, but it's still a very small effect, and I would think most people would have to make some pretty big sacrifices in lifestyle to accommodate a cut of that magnitude. How would you pull $11.2K out of a $41.2K budget? I wouldn't think $41.2K includes all that much 'what the heck' kind of items that one could do without for years on end.

If your budget can handle that much change, then maybe it works for you. But I've seen a lot of posts where people just say they'll buy hamburger instead of rib-eye, have fewer dinners out, etc.

Alright, so here's a FC run:

A) I took a 35 year portfolio length, since we are generally talking ER here. A $40,000 infl adj spend requires a $1,043,168 portfolio to provide a 95.4% success rate.

B) To increase that success to just get to 100%, I started adding infl adjusted income of $10,000 in 2015 (two years into a downturn), to simulate the cut in spending.

C) I found you had to extend that cut out to 2026, a full 12 years. So 12 of the first 15 years of your retirement, probably your best years for travel and other discretionary spending, have a full 25% of your spending cut, and that must be an awful lot of discretionary spending for most people.

D) A straight $36,027, with no cuts provides 100% success. And you would likely be able to raise that as time goes on if no downturn occurred. IMO, a 10% cut long term is much more manageable than a 25% cut for 12 of the first 15 years. That would be going really deep into discretionary for most.

-ERD50

 

[Huston55]

Wade Pfau's posts are always interesting and thought provoking. Here's the latest, on sequence of returns risk.

This isn't news to most forum members, but nevertheless the implications are profound. Many of us hope that doing all the "right" things (e.g., diversifying in every way possible) will dampen the effects of volatility and allow us to exert some control over our financial future. But ultimately timing appears to override everything for those of us who are invested in the market. I think this is why ER often feels like a leap of faith.

At the end, Pfau offers these suggestions on how to de-risk: 1) Strengthen Social Security (and avoid turning it into defined contribution), 2) Avoid constant inflation-adjusted spending, 3) Consider holding fixed income assets to maturity or using annuities, 4) Consider reducing downside risk through other approaches, including financial derivatives, 5) Consider rising-equity glidepath, 6) Consider "safe savings rate" approach.

I also am a Wade Pfau fan; his research makes me consider things in new ways. I plan to use Guyton-Klinger rules, supplemented by a cash buffer (~4yrs), back stopped by using an annuity hurdle (Fullmer or Otar) approach later if needed. Using G-K, the data indicates a worst case of two 10% reductions and six freezes. This could dramatically reduce annual income, by ~40%, if it all happened sequentially due to a meltdown. However, I have a Cola'd DB pension kicking in at year 2 after FIRE, which pretty much negates the worst case for me by replacing the lost annual income. So, roughly, that's my plan.

Well 11.2K of the preservation came from spending less, so the benefits are actually only $7.5k or less than 1% of the portfolio. It is also worth noting that bear markets generally are much shorter and the declines steeper than recoveries. So realistically you would have cut spending dramatically in 2009 and 2010 and probably 2011 before getting back to near 2007 levels in 2012.

The good news is I think this is human nature to cut big ticket discretionary spending during bad times. I think anything more drastic than Clyatt 95% rule is unrealistic.

I like Clyatt's approach but, plan to use G-K instead, as described above. It has the potential to be somewhat more punitive than Clyatt in an extended bear market but, not overly so IMO. Both Clyatt and G-K are "spending reduction" approaches but, G-K allows a higher initial SWR with the possibility of greater future income reductions; a trade off that works for me.

 

[photoguy]

To my knowledge, Pfau's 1.9% occurs while blindly increasing withdrawals to pace inflation, regardless of how precipitously a portfolio declines during the inevitable, periodic market downturns. My plan is to cut spending during such downturns, if it appears necessary, and give time for the portfolio to recover. Such spending adjustment should boost the SWR above 1.9%.

I think it also happens because in his MC model he reduces returns by 2% compared with historical (he notes this in at least one paper). You may or may not agree his reasons.

 

[clifp]

I think it also happens because in his MC model he reduces returns by 2% compared with historical (he notes this in at least one paper). You may or may not agree his reasons.

I didn't bother to read this paper in detail, but assuming he did the same as he did in the last paper reduce average US equity returns by an arbitrary 2%, while keeping the standard deviation the same, it results in a very distorted view of the world IMO.

A monto carlo simulation is of dubious value for the stock market as it is, but making these adjustment render his analysis worse than useless IMO.

He is a smart guy and his research is worth reading, but I pity the average Joe trying to accumulate a 50x nest egg for the 1.9% withdrawal when reality is most people have trouble saving 10x their salary.

 

[gerntz]

The fallacy is that when one retires isn't luck. It's a conscious choice.

 

[yakers]

The fallacy is that when one retires isn't luck. It's a conscious choice.

Or a medical necessity

 

[haha]

Most of these glib explanations of how to handle big market drops are put forth by people who haven't faced much of this, at least not without any income source other than their portfolio.

Mr Bernanke rang a bell back in March 2009, and the fire companies showed up and their water pressure was good and it was off to the races once more.

Some think that all he did was set up a bigger, more difficult to escape problem for next time, which if it is like 81-82 or 73-74 could be a much bigger deal than '08-'09.

Our mental model says to us that a ~50% equity drawdown is limit down. However, it might be that the floor, if there is one, is equity valuation. So, for example if we needed to get to a PE10 of 7 (roughly 1982 or 1974), it should take more than a 50% fall from our current PE10 of ~24.

I am not tone deaf enough to make predictions on this board, but to me anyway, we have not necessarily seen the worst that fate can throw at us.

Ha

 

[REWahoo]

...we have not necessarily seen the worst that fate can throw at us.

Ha

That's certainly true of any of us still living. The dead may have a different perspective.

 

[foxfirev5]

While I've had a plan that I've stuck with for quite a few years it is not absolute. Changes and the need to adapt will always loom over us.

 

[Animorph]

I think it also happens because in his MC model he reduces returns by 2% compared with historical (he notes this in at least one paper). You may or may not agree his reasons.

I think that's much of it. And the yearly growth is selected arithmetically, so no chance of much reversion to the mean.

 

[dtbach]

I am not tone deaf enough to make predictions on this board, but to me anyway, we have not necessarily seen the worst that fate can throw at us.

Ha

Like maybe Poland in 1939?

 

[Ed_The_Gypsy]

What good is "strengthening Social Security" to me? I can't do it by myself.

After I commit, all I can do is reduce my expenditures.

Is he giving advice or running for congress?

 

[Koolau]

Short term (i.e., the 8 years since I FIRED) my strategy has been to "avoid" the big down turns by not playing the game (stock portion of port in the 20% to 30% range.) This has been effective for me in that I have never really suffered any significant losses. The obvious (immediate) downside of this strategy is that 4% WDR is out of the question. Longer term, big gains are not in the cards either. Neither has been a problem so far. BUT, my fear is that the very low fixed-rate returns we see now may become semi-permanent and inflation may rear its ugly rear. That would be MY personal "perfect storm".

If I'm making any point at all, it would probably be that no strategy is fool proof and we all face uncertainty, no matter how we structure our portfolio. Use "buckets" or "highly diversified AA" or any other name we wish to attach to our FIRE plans and we all still face uncertainty. Ultimately, my "real" strategy is to remain adaptable to whatever comes. I must be ready to adjust my standard of living including the possibility of a move to a lower-cost area. Not very palatable, but it beats w*rking for a living. YMMV

 

[gerntz]

Or a medical necessity

Wouldn't that be disability vs. retirement?

 

[Huston55]

Ultimately, my "real" strategy is to remain adaptable to whatever comes. I must be ready to adjust my standard of living including the possibility of a move to a lower-cost area. Not very palatable, but it beats w*rking for a living. YMMV

Says it all doesn't it!

 

[omni550]

Wayne Pfau's latest blog post on the "Lifetime Sequence of Returns Risk" shows that retirees are very vulnerable to what happens just after they retire.

Wade Pfau's Retirement Researcher Blog: Lifetime Sequence of Returns Risk

omni

 

[rayvt]

OK, some numbers, first without a downturn. $1m portfolio, year 1 withdraw $40k, leaving $960k. Year 2 add 3% inflation and withdraw $41.2k, leaving $918.8k in portfolio.

What if there is 40% downturn? $1m portfolio, year 1 withdraw $40k, leaving $960k, then a 40% downturn reduces it to $576k. Year 2 blindly increase withdraw to $41.2k, so portfolio is $543.8k. Recovery happens, boosting portfolio up to $891.3k, which is $27.5k less than if downturn had not happened.
To survive longer downturns, use a cash bucket so you don't need to sell portfolio holdings before they recover.

For a slightly different purpose, this spreadsheet was created: https://www.dropbox.com/s/cbzvg74iyeyfwt6/SPX-monthly-1950-2013-and-IUL-test.xls

You can enter the data using a 1/1/2000 start date just before the 2001 crash and see what would have happened. (Ignore the stuff about IUL.)
That was TWO large drops, no withdrawal adjustments, and the portfolio is still staggering along okay by 2013, with $600K. Even better, by using a simple market timing to switch to T-bills, the 2013 value has grown to $1.7M.

Cash buckets are mostly a mirage. It's comforting emotionally, but financially costs you money. Spreadsheet at https://www.dropbox.com/s/xf4ma5blug27aws/SPY_Withdraw_by_CashBucket_rules.xls

 

[REWahoo]

Wayne Pfau's latest blog post on the "Lifetime Sequence of Returns Risk" shows that retirees are very vulnerable to what happens just after they retire.

Wade Pfau's Retirement Researcher Blog: Lifetime Sequence of Returns Risk

omni

Once again, someone reinvents the wheel.

The perils of a big hit during the first few years of retirement was illustrated by Dory36 in 2002: FIRECalc