Calculate SWR if CPI is inaccurate?

[ExHermit]

The difference between the reported CPI and our personal cost of living has increased recently. For example, our rates for electricity, gas, and water/sewer utilities have had double digit increases each year for the last 3 years. Food prices are rising rapidly. Foreign travel costs have jumped with the drop in the US dollar against foreign currencies. These things are significant parts of our budget. We spend very little on things which have become cheaper lately such as clothes or electronics.

I have been retired for 11 years and DW for 7 years. There are, and always will be, plenty of things that we are not doing and do not need to do but would enjoy if we could safely afford the extra spending.

I have my own calculations in the style of Firecalc since I need to model some things, like our small pensions with partial CPI adjustment, that Firecalc cannot handle. Both my calculations and Firecalc indicate that our spending is more conservative than necessary. However, both of these calculations use the CPI to adjust future spending, which seems like a head in the sand assumption.

Adding a yearly spending increase above the CPI to my calculations drastically effects portfolio survivability. For example increasing each year's spending by (CPI + 2.5%) has the same effect on success probability as cutting the initial portfolio value by 40%.

I hope that this is too pessimistic, but maybe not. In any case, the 2.5% figure was pulled out of thin air.

Does anybody have suggestions for determining what is a "safe" withdrawal rate when the CPI does not accurately reflect our cost of living?

ExHermit

 

[twaddle]

If you want to increase your withdrawal inflation factor, then the only "safe" answer is to grow a bigger nest egg. Probably not what you wanted to hear.

Or, do as I do, and add specific hedges to your portfolio for your specific spending pattern. For example, if you're exposed to foreign currency fluctuations, doesn't it make sense to have a larger exposure to foreign investments in your portfolio?

 

[Cb]

Does anybody have suggestions for determining what is a "safe" withdrawal rate when the CPI does not accurately reflect our cost of living?

ExHermit

If you'd like to model a CPI + .xx% inflation rate just add as many .xx% to your portfolio's annual expense ratio in FIREcalc.

Cb

 

[ExHermit]

Thanks for responding.

twaddle:
Clearly, larger portfolios support larger withdrawals no matter what is happening with inflation. I'm not so sure about choosing investments based on types of expenses. I eat a lot of yogurt, but I don't see that as a particular motivation to invest in yogurt producers or dairy farms. The relevant question for any potential investment is "how will this investment, in combination with my other investments, effect my expected return and risk?" Surely I should invest where I anticipate the highest return, consistent with the level of risk that I am willing to accept, regardless of how I anticipate spending the money.

Cb:
I don't think Firecalc's expense ratio will work for this purpose. The expense ratio deducts a fixed percentage of the portfolio balance each year, so the added expense grows or shrinks with the portfolio balance. Increasing the inflation rate for spending purposes, on the other hand, results in a compounding increase in spending each year. A simple example: if the inflation rate is a constant 3%, in 30 years it will cost $2.43 to buy what one dollar will buy today. If the true inflation rate is 2.5% higher, 5.5% total, you will need $4.98 to buy that dollar's worth of goods in 30 years.

As I said, I pretty much pulled the 2.5% figure out of thin air. So I have little confidence in this particular adjustment for an understated CPI. Also, some have argued that the CPI was more accurate until recently (until 1992 IIRC). If that is true, then the correction should only be added to historical inflation values after 1992. There may be some totally different approach that makes more sense.

Our decision to retire early (each retired at age 46) was made years ago. Things have gone well so far and I don't anticipate any problem maintaining our basic standard of living. The question is how much purely discretionary additional spending we can afford: Do we spend 6 weeks in France next spring or only one month? How often do we go out to eat? etc. The answer depends A LOT on the true future inflation rate.

While we could always cut back later if necessary, it is not nearly as much fun to cut back on things you have become used to as it is to add new luxuries. I would rather keep our spending to a sustainable level, and increase it slowly if all goes well, than over spend and have no choice but to cut back later. On the other hand, I have no shortage of ideas about fun things to do with money, and no desire to leave money unspent when we are both dead.

ExHermit

 

[twaddle]

I'm not so sure about choosing investments based on types of expenses. I eat a lot of yogurt, but I don't see that as a particular motivation to invest in yogurt producers or dairy farms. The relevant question for any potential investment is "how will this investment, in combination with my other investments, effect my expected return and risk?" Surely I should invest where I anticipate the highest return, consistent with the level of risk that I am willing to accept, regardless of how I anticipate spending the money.

So, an investment's return is only a function of some risk premium?

Earnings growth has nothing to do with it?

Currency exchange rates have nothing to do with it?

If 4% works as a SWR for US consumers owning US stocks, let's imagine it also works for foreign stocks owned by foreign consumers.

So, if you plan to spend $10,000 each year in France, you'll want about $250,000 worth of French investments, n'est pas?

Why wouldn't this work to inoculate you from the "inflation" effects you're seeing due to the falling US dollar?

Likewise, if you spend $10,000 per year on dairy products, why wouldn't a $250,000 investment in the dairy economy ensure that you can safely pay those dairy bills? It probably wouldn't boost your overall returns, but it would probably track your dairy inflation pretty well.

I'll grant you that I've never seen a study to support this idea, but it might be fun to test my hypothesis.

In any case, I know of no way to predict future inflation -- even personal inflation, so I'm not sure how else you would even approach the problem.

 

[Cb]

Cb:
I don't think Firecalc's expense ratio will work for this purpose. The expense ratio deducts a fixed percentage of the portfolio balance each year, so the added expense grows or shrinks with the portfolio balance. Increasing the inflation rate for spending purposes, on the other hand, results in a compounding increase in spending each year. A simple example: if the inflation rate is a constant 3%, in 30 years it will cost $2.43 to buy what one dollar will buy today. If the true inflation rate is 2.5% higher, 5.5% total, you will need $4.98 to buy that dollar's worth of goods in 30 years.

Seems to me the extra money is gone forever either way, and the effect compounds.

Cb

 

[SecondCor521]

Seems to me the extra money is gone forever either way, and the effect compounds.

Cb

The expense ratio is a percentage applied to the portfolio balance. The CPI is a percentage that is applied to the portfolio withdrawal. If one follows the standard FIREcalc model of withdrawing 4% initially then adjusting the amount by the CPI every subsequent year, any relationship between the portfolio balance and the withdrawal amount can diverge rather quickly.

I thought there was an option in advanced FIREcalc to model inflation based on CPI plus x% or at least a constant inflation fact of x%. But I don't think that is really the OP's question.

My best response to the OP's question is to attempt to model your own personal inflation rate based on the past 10 years of increases, with adjustments made up or down based on any changes (like if you were paying college expenses in the past and you're not going to in the future). Other than that, it's the age old question of trying to have the check to the funeral home bounce, and there's no perfect answer to that.

2Cor521

 

[Cb]

The expense ratio is a percentage applied to the portfolio balance. The CPI is a percentage that is applied to the portfolio withdrawal.

2Cor521

Oh...

Cb

 

[kcowan]

We had the same dilemma and have decided to increase CPI by 1.5% in our projections. Then the challenge we have is to live to that budget each year. As long as we do not exceed the budget we will remain whole.

We also use 9% for portfolio returns. So far we have been exceeding the returns and underspending the budget. The budget is because we have gotten creative in our spending in the first 5 years of retirement. And we expect the portfolio returns we achieved in that period to be exceptionally high.

Our current SWR is down to 3.5% not that that matters as it is the original that counts.

 

[ExHermit]

To clarify my question a bit: It is not so much the rate of inflation that I am concerned with, it is the difference between the CPI and personal inflation rate. If we assume that the past is a guide to the future (a much discussed topic), then Firecalc-like computations provide a reasonable estimate of the probability of being able to maintain a given level of inflation adjusted spending for a given amount of time. The Firecalc methodology provides for adjusting spending to match the change in the CPI. If the CPI does not accurately reflect personal inflation this methodology results in an ever-decreasing standard of living, something I hope to avoid.

One of the main attractions of Firecalc is that it does not require made up or estimated numbers for investment return or inflation. The underlying model is based on actual historical facts. The only estimated number is how long the money must last. (For my own purposes, I don't pick a particular length of time. I use life expectancy tables from the Social Security website which are based on census, death certificate, and Medicare data. I run the calculations out until we would be 120 years old, but discount portfolio failures based on the probability that one of us is still alive in the year of failure. No arbitrary estimates required. The gory details are not relevant here.)

Inserting an arbitrary number (2.5% above CPI) greatly reduces my confidence in the Firecalc model. I don't know that 2.5% is really the correct number. Even if I could accurately compute my personal inflation rate for the last ten years I would still have my doubts. I don't think that the difference between CPI inflation and personal inflation is really a constant. Historically, the December to December CPI inflation rate has ranged -10.3 to +20.4. Even if 2.5% is the correct number for the last ten years, when CPI inflation was confined to the range 1.6% to 3.4%, will it remain the correct number in the event of deflation or double digit CPI inflation?

I may be looking for a warm fuzzy feeling where the reality is that great uncertainty is unavoidable. Quantum mechanics makes me uneasy too, but it seems to describe the way the world works.

ExHermit

 

[twaddle]

It's still not clear to me what you're looking for. The CPI is based on a set of weights applied to individual components. Everybody's individual set of weights differs depending on their own spending pattern. And, of course, nobody's spending pattern is constant. In my experience, lifestyle changes completely swamp any inflationary effects on my spending.

So, trying to predict your future cost of living increases seems impossible to me.

The market only cares about the CPI. It doesn't care about your personal rate of inflation. So, if you know you are dependent on some oddball expense that isn't in sync with the CPI, then hedge against the particular inflation factor that worries you.

Otherwise, try to live within your budget, and if things aren't working out, try to adapt.

Edit: the BLS has a lot of component-specific data here: Consumer Price Index Home Page

If you really want a fun project, create your own inflation index based on your personal set of weights, then run that inflation factor against the historical data, and see how it changes survivability.

 

[audreyh1]

If you are worried about your standard of living dropping after many years in retirement, then you have no choice but to pad your retirement fund considerably. This means working longer and saving more, obviously sacrificing today for tomorrow.

It's a tough tradeoff.

Audrey

 

[cute fuzzy bunny]

This is one of the topics that usually draws out the flamebaiters attempting to undermine any reasonable discourse on the topic. Lets see how it goes this time.

The CPI is a fine measure of inflation for urban workers who rent an apartment and dont pay for health care. If thats your profile, then the CPI should track your costs pretty well.

Where I think some people are going to get themselves in a deep sticky mess is if they plan for the CPI in their investment planning and they're not urban workers who rent and dont pay their own health care. There seems to be at least a portion of folks that feel that they can plan in this manner, invest in items returning a "real" return vs the CPI, and safely live their lifestyle going forward free of the effects of inflation.

As you've noted, and the BLS themselves explains when they say in detail that the CPI is not THE measure of inflation but A measure of inflation, the devil is in the details and everyones budget will be affected differently.

One way to look at this is to try to get a rough idea of your spending and inflationary pressures that apply to your intended lifestyle. The painful part is that I suspect most people would see that their budget adjusted for their real rate of inflation is incompatible with their portfolio size. If your actual costs inflate significantly due to health care, energy and housing costs at a rate of 4-5%, you'll probably come to the conclusion pretty quickly that you're not going to make 4%+4-5%+taxes as a regular rate of return year in and year out. If the CPI is running at 2% and your PRI is 4%, CPI indexed securities arent going to be much help either.

Seems to me that most people break into one of two categories: one has a huge portfolio and slogs on without much concern and the other "discovers" that their spending is outstripping their investment returns. The latter effect has been somewhat muted due to a 14 year bull market that was only briefly interrupted by a 2 year bear.

The latter group also changes their spending/shopping patterns, stops buying some items/services or changes their lifestyle dramatically to compensate. Unfortunately a lot of these aids are one or two iteration changes. You go from shopping at the local fancy market to costco to buying dented cans at the grocery outlet emporium. Not much room to go down from there unless you buy a packet of seeds and a shovel.

Perhaps we've also scratched into why older retirees spend less as they age. It might have less to do with not wanting to spend money anymore and more to do with things becoming more expensive than their social security adjustments are compensating for, and watching their nest egg shrink too fast. Anecdotally, my dad feels that after 12 years on SS that the buying power on his money has shrunk significantly. Increases in medicare seem to be overwhelming the SS CPI adjustments, and the rest of his costs are heavily in the food and energy sectors.

I'd suspect that ER's who reside in a low cost of living area and plan to live in an LBYM mode will be just fine. Those in high cost areas that want a full featured health plan, to travel, and indulge in other expensive hobbies that are thinking that 4%+cpi will do them fine...are in for a rude awakening when the current up-nearly-every-day market slows down for 5-10 years.

The short answer is to roughly figure your PRI's delta from the CPI. If your PRI is more than a percent or two higher than the CPI and you're squeaking by on a 4% SWR, you're not gonna make it long term. That level of spending is simply going to be higher than a reasonable portfolios returns. You need to work longer or take a highlighter to the budget.

"Hedging against your inflation items" can be done for a few items, but its pretty risky and not a strategy I'd recommend. You can buy commodities or energy funds or health care funds. You'll feel a bit displeased with those investments when someone comes up with an oil alternative energy product that doubles or triples production while costing half as much, or when the health care system gets nationalized and all the profit gets squeezed out of it. In other cases I just cant see any benefit. Milk costs have doubled around here but I dont see the stocks of the large conglomerates that own most of the dairy production going up very much.

 

[dbr]

In Advanced Firecalc under the Options tab you can type in any amount you want for inflation. Did I miss anyone in the postings above pointing this out?

 

[DRiP Guy]

Terrific post, CFB. Very nice synopsis of the issue and how people are dealing with it, and the possible result of that. Me, I'll stay adaptable - as a single guy with LBYM tendencies, I think I can make the needed adjustments to spending and therefore WR, year-to-year as I eyeball my nestegg. Perhaps the bull will continue or reappear after a short absence, and I'll be in tall cotton. But it is also possible that things will chill out in an extended down market, and I will be a bit more frugal than I had hoped. I am looking at it in the perspective that a bit of austerity sure did not harm my parent's generation or the one before them, so I am sure will get along just fine. If it comes down to 'selecting bargain brands instead of designer labels', or else continuing to stay harnessed to the yoke until a traditional retirement age, then I sure know where I come down, personally! Where is unclemick, my hero!

 

[maddythebeagle]

"Hedging against your inflation items" can be done for a few items, but its pretty risky and not a strategy I'd recommend. Milk costs have doubled around here but I dont see the stocks of the large conglomerates that own most of the dairy production going up very much.

Buy a cow?

 

[cute fuzzy bunny]

Yeah but then you'd have to get into that whole milk discussion about washing and homogenizing and hormones and imports and why George Bush is somehow involved. Oh yes, and Hitler.

 

[audreyh1]

Yeah but then you'd have to get into that whole milk discussion about washing and homogenizing and hormones and imports and why George Bush is somehow involved. Oh yes, and Hitler.

LOL!

And don't forget farm subsidies!

Seriously, if you are going to go that route you might as well do the "self-sufficiency" thing - growing your own food, living in a "green" energy efficient house with lots of solar power, water collection system, etc.

That would definitely help counteract the ravages of inflation, and it could be challenging and fun and provide some folks with a very enjoyable retirement "project".

If you have other priorities, the above sure sounds like w-o-r-k.

Audrey

 

[twaddle]

Seriously, if you are going to go that route you might as well do the "self-sufficiency" thing - growing your own food, living in a "green" energy efficient house with lots of solar power, water collection system, etc.

That would definitely help counteract the ravages of inflation, and it could be challenging and fun and provide some folks with a very enjoyable retirement "project".

Yup, the ultimate hedging strategy is to buy everything today that you'll need in the future.

So, get the farm, fill a LARGE tank of gas, go to med school, and you can mostly stop worrying about inflation.

 

[kcowan]

Welcome back CFB. I missed you. How is the house sale and move progressing?

PS Our fallback plan is to move to Mexico. We can cut our costs by 40%: 3.5x.6=2.1% SWR and still maintain our travel plans. We had some great results from a convertible debenture: paying 6.5% since 2004 and currently valued at a 51% premium over face value.

 

[ExHermit]

I was hoping CFB would respond to this thread, as he has previously posted thoughtfully on the topic of personal vs. official inflation.

Seems to me that most people break into one of two categories: one has a huge portfolio and slogs on without much concern and the other "discovers" that their spending is outstripping their investment returns. The latter effect has been somewhat muted due to a 14 year bull market that was only briefly interrupted by a 2 year bear.

I think I represent a third (maybe rare?) category. I don't have a huge portfolio but I see our spending rate as a variable largely under our control. "ExHermit" is literally true. I know from experience that all you really need is water, food, and clothing and shelter sufficient for the local climate. If you are warm, dry, fed, and not in pain, then you are doing OK. Everything else is gravy. Don't get me wrong, I like gravy. In fact my questions are all about maximizing my gravy supply. The catch is that more gravy is good up to the point that the cost depletes your resources and leaves you feeling deprived later on. Affordable gravy purchases are great fun. Excessive gravy purchases are not good value for money. I am trying to determine, as closely as possible, the break point between affordable and excessive.

Since our small pensions started in 2005, our cash withdrawal rate has been less than 2% of each year's Jan. 1st portfolio balance. It looks like 2007 will come in at about 1.3%. Our total budget, including set asides for amortization of irregular expenses and totally discretionary expenses, is 3.2% of our current portfolio. Pensions currently cover 37.7% of that budget. Social Security, if I start drawing it in 4 years, should cover about 25% of our current budget. For the almost 19 years that I have complete records, our compounded annual investment return has exceeded the compounded annual CPI inflation rate by 6.48%. (I know, I know, way too nerdy about the whole deal.)

I think we are in good shape. In fact I hope we can increase our discretionary spending. The absolute level of future inflation is obviously a factor, but is largely accounted for in Firecalc type calculations. A personal inflation level above the "official" rate is not accounted for in the Firecalc model and has a huge impact on the real value of our (partially) CPI adjusted pensions and on the real value of the SS "pension".

So far I can identify three sources of personal vs. official divergence. 1) Personal expenditure weightings do not match the weightings used in the CPI. 2) The inflation rate in our locality may very from the national urban average. 3) The "improvements" made to the CPI calculations over the years reduce the reported CPI. The CPI seems to have morphed into something other than tracking the cost to maintain a constant standard of living. See Consumer Price Index research series using current methods, 1978-98 Monthly Labor Review - Find Articles
for a discussion of the effect of changes made in 1978-1998.

Items 1 and 2 do not necessarily increase your personal inflation. Over the very long run they might have neutral effect. Item 3 only does one thing, reduce the reported CPI.

Lots to play with here. Probably measuring with a micrometer and cutting with an ax, but a hermit in the north woods gets real good with an ax.

ExHermit

 

[twaddle]

Probably measuring with a micrometer and cutting with an ax....

In the immortal words of CFB, ding ding ding, we have a winnah!

You mentioned that you pulled your 2.5% CPI delta out of thin air, but you also mentioned that you've been tracking your expenses for 19 years. So, you should have a pretty good handle on your CPI delta.

I'm curious. What was the mean and variance of your CPI delta?

I eyeballed my average increase in spending over the last 20 years, and I came up with 12%/year or so. Waaaay over the CPI, but very little of that was due to inflation. And I certainly don't expect to increase my spending at the same rate for the next 20 years.

As I always have, I will continue to adapt my spending to my income.

 

[SecondCor521]

In Advanced Firecalc under the Options tab you can type in any amount you want for inflation. Did I miss anyone in the postings above pointing this out?

Sorta. My post #7, second paragraph.

2Cor521

 

[cute fuzzy bunny]

Items 1 and 2 do not necessarily increase your personal inflation. Over the very long run they might have neutral effect. Item 3 only does one thing, reduce the reported CPI.

At a ~2% withdrawal rate and champagne desires with an LBYM willingness, you should be fine. Gravy IS good, and should be served in large mugs instead of in "boats". Not that I dont admire the analog.

You've aptly identified three things that an ER needs to pay attention to. As obvious as they are, supposedly expert investors still call CPI adjusted investment products returners of "real" income.

I guess it is, if you've got a pretty flexible definition of "real", with it being equated to "real" for someone who lives somewhere other than where you do, and buys stuff other than what you buy...

The last major adjustment to CPI really WAS a huge thing. It changed a lot of real financial aspects for a retiree, but surprisingly not the expectations. A few divots in your income stream over 20-25 years is a little different from 35-45 years.

It was pretty amusing to read the data on the "CPI-E", which was a CPI index for elderly retirees, which more favors health care costs and other aspects of retired life. IMO a better measure of inflation for the average retired person, and it clocks in at a full 1-2% higher than the CPI. The comments on the folks who reviewed and deep sixed it were telling 'Its a fair measure and more appropriate for retirees than the CPI-U, but we'd go broke if we paid this much out in social security and government pensions, so lets wish it away'.

This while my hero Alan Greenspan claimed that the CPI-U overstated average inflation by 1% or more, and he was raising rates and telling joe sixpack to go get an adjustable mortgage. Theres a book I dont need to buy...

I think the nice round rule of success is that if you're pulling more than 8% in returns, taking less than 4% in withdrawals, and you're willing to be flexible on spending at least during the rough years, you'll be fine. If you structure a portfolio based on CPI indexed securities that pay under 2.5%+CPI and religiously take your 4%, good luck 20-30 years from now when you figure out that the strategy sucked.

Oh and since I mentioned CPI, inflation, TIPS and social security in the same thread...

(P.S. Annuities suck too!)

 

[ExHermit]

From twaddle

You mentioned that you pulled your 2.5% CPI delta out of thin air, but you also mentioned that you've been tracking your expenses for 19 years. So, you should have a pretty good handle on your CPI delta.

I should have made it clear that it is my investments, not my expenses, that I have been tracking in detail for 19 years. I agree with your earlier statement that lifestyle changes completely swamp inflation. I have a good idea what I spent each year, but that has only a voluntary relationship to the Consumer PRICE Index or my personal rate of price inflation. As long as my spending does not approach the minimum required to survive or the maximum I would spend if money was no object, my spending pattern is determined by what I choose to spend, nothing else.

My spending has fluctuated widely, and what I spend it on has changed radically from year to year. My highest ever yearly spending was 11 years ago, the first year I retired, when I was traveling all year on a round-the-world trip. For the last 10 years the change in my total spending, after smoothing out the large year to year fluctuations, has closely matched the CPI change. That is not a coincidence. That is how much I have chosen to spend. My choices may be different in the future, but I am endeavoring to ensure that they will continue to be sustainable choices.


ExHermit