In response to my noting the collective $1 trillion hole for the state pension funds a reader asked "Why does a return of 6-8% seem OK for a pension portfolio, while 4% is the expected return for a personal portfolio?"
There are a few things here to address and perhaps clear up. First thing is that the reader might be apples and oranging a couple of numbers. Generically speaking a portfolio might average 8% per year over some long period of time which is different than a safe withdrawal rate. The issue with the state pensions, or any pensions really, is that they have liabilities that must be paid every year in the form of pensioner benefits. If a worker is entitled to $1773 per month then the fund has to pay him that amount every month, period.
In any 10 or 20 year period 8% might be the average annual number like maybe in the 1990s but in another time period like decade just ended it might less, even negative. The problem created by a negative decade is obvious, the fund pays out the benefits as the value of the fund shrinks (or maybe not depending on any funding but you get the idea).
Maybe not, but it seems like the last two decades were very extreme; one very good and one very bad but whether that is true or not one thing that is true is that no matter what the average per year it will rarely hit that average number in a given year. One look at a Stock Trader's Almanac will tell you that.
If a newly retired individual has a $1 million portfolio and plans to take out $55,000 he will be just fine that first year if the stock market goes up 10% and his portfolio goes up 8%. In a simplified world after that type of year he will have $1,025,000. If in his second year the market is up 2%, the investor matches it and takes the same $55,000 he will finish the second year with $990,500. So two up years in the market to start but he has already below where he started.
Lets say that the third year turns out to be 2008 but he does a great job avoiding the full brunt and he only goes down 15% in a down 38% world but thinks he can take out the same $55,000. He would be ending that third year with $786,925 and his $55,000 is now a 7% withdrawal rate Being down 15% is a very generous assumption.
If the third year was not 2008 and the market went up 8% and our investor was up 10% the portfolio would be $1,034,550. Whether the third year is 2008 or not is a matter of luck. Whether an investor ever encounters a 2008 in their retirement is a matter of luck. Clearly this investor starting with a 5.5% withdrawal rate is subject to the vagaries of the market but he is just one person and there are not that many moving parts.
Running a pension has far more moving parts and less flexibility on payouts and investment policies but faces the same varies of the market. Additionally there are pensioners coming and going all the time making it more complicated. While I'm not going to crunch the numbers for this post it has been noted in many places that often a disproportionate amount of the total appreciation in a bull cycle comes from just one year (think about 2003's contribution to the bull ended in October 2007). This means that the typical year might have returns that are less than the average but the state pensions still have the same obligations no matter what.
It is with this sort of thought process as a backdrop why although a portfolio might, I say might, average 8% over some period of time it makes sense to make the withdrawal rate as small as possible keeping in mind a reasonable income need and survivability of the portfolio. I believe most studies find an optimal amount being 4.2% which many people tend to round down to 4.0%.
From there it gets cloudier; 4% and then adjust for inflation every year some would say. I don't get this one. What if after ten years $1 million starting point is still $1 million but inflation works out such that the adjustments for same mean you take out $50,000?
The denial that surrounds this concept will cause an awful lot of misery I am afraid. As I mentioned during the week (albeit with different numbers) it is unlikely that someone who has accumulated $1 million has a lifestyle, even if it is modest, that only requires $3333 per month.
If you think about it there are all sorts of variables at play starting with when you retire that can either make it very easy or very difficult. This is why I believe in living modestly relative to your income and working longer or in other words creating a large margin of safety in your numbers.
One thing not mentioned above but that we have talked about many times before is expensive one off events like expensive home issues or medical events. If you can live comfortably on 4% that is quite commendable but what happens if your roof has to be replaced or you have some sort of foundation issue? Someone with a margin of safety can better navigate these sorts of things than the person who uses his margin of safety to buy a boat. Nothing against boat owners but I heard a guy (over my iPod) bitching about his boat expenses at the gym yesterday--he is underwater and needs to sell it fast.
The easiest way, IMO, to have enough money is to work on getting the overhead down. It is a lot easier to cover your nut if your nut only consists of utilities, insurances and food.
I watched a good chunk of the Czech Republic/Latvia hockey game last night. Nice to see that Latvia was able to put a team together and send them to the tournament (not a sarcastic comment).
Proud Member Of
Random Roger
© 2009 | Gadget Theme Blogger Template
Original WPtheme By Blogger Template | Converted By Gadget Lover
| Distributed by eFree Blogger Help
Original WPtheme By Blogger Template | Converted By Gadget Lover
| Distributed by eFree Blogger Help
15 comments:
A boat owner who is underwater? Very good!
i should have acknowledged the double entendre
Roger,
I really don't understand your explanation. The required return of pension plans has indeed been unrealistic. Just look at how many have been dumped on Pension Benefit Guaranty Corporation (i.e. federal government). Here is a notable example, United Airlines, http://bit.ly/bpt0WQ. A few years ago the financial press was filled with articles regarding the attempt to change accounting rules to smooth over poor assumptions about portfolio growth. As I recall, the subject was part of Berkshire Hathaway's annual report.
Structurally, pension plans should have constant cash inflows from which to pay benefits while contributions to an individual's plan should stop at retirement. This feature should allow government pensions to have a portfolio with higher risk/reward ratio. Although 8% is too optimistic (IMO) considering the S&P composite long run return is on the order of 9% on a nominal basis.
Also, although the PBGC doesn't backstop government plans, each locality probably has at least an implicit taxpayer guarantee which could lead to some risk-seeking behavior on the part of the investment managers.
Speaking of the PBGC, isn't it logical that all these public pension funds that are underfunded get dumped on the PBGC? Yes, it is still taxpayers money but the PBGC has payout limits that are significantly less than some of these pensions have promised to pay. It seems the only logical solution to this problem.
The article link in anon 8:39's post says that some of the pilot's took a 50% haircut, maximum befefit at the time being $45,000 per year.
I would venture to say that a big problem with some pension plans is that the current workforce (inflow money) is substantially smaller than the legacy workforce collecting pension benefits. This would be especially true in the manufacturing sector.
If you think about it, the 8% return is likely a reverse engineered figure that allows a company to meet expected financial performance. Many companies may need more like 10% or more, but that would shift the income statement from gain to loss. It is this type of financial shenanigans that conservative investors like Buffett warn us about.
I used to have a defined benefit plan for my employees. Too risky. I terminated it and now have a profit sharing plan. I am still responsible for portfolio performance, but I am no longer at risk of coming up with funds if the investment pool falls short according to an actuary's calculations. I also don't stray too far from a 60/40 allocation with low cost index funds due to fiduciary responsibilities.
I think the point of rogers post was to discuss the thought process, not defend pension fund forecasts.
That is all old-school thinking anyway. Hedge funds target volatility -- and an expected sharpe ratio based on target volatility.
Saying 'the long run return of the market is 7% [or 4% or 8% or whatever]' is not all that useful given rogers points --- lumpy returns and large drawdowns causing major problems in your own personal situation.
There is a statistical term for this problem: leptokurtosis. When average returns are low relative to standard deviation, you have a lot of risk for relatively low return. Some of the better hedge funds understand this, when will the rest of the world catch-up?
The right way to think is in terms of target volatility and then executing a good well-thought out plan with this in mind. If you can't tolerate volatility given your situation, then there are strategies for that. If you choose to ignore volatility, then expect large drawdowns, easy enough concept.
I created an image for this post here:
http://www.etfreplay.com/blog/post/2010/02/20/ETF-Statistical-Distributions.aspx
The virtue of a defined benefit is that you can define targets based on duration of cash flows; eg, if you knew a cash requirement of $5000 (or any increment) cash was needed in March of 2020 you could buy treasury strips in that amount maturing on that date; is there was a COLA requirement you could do the same with TIPS or combine the two. That reduces variance a lot but there is reinvestment risk; e.g., you may have to buy more bonds than expected when rates are low.
Flipping fchris's point (thanks for the charts, good illustration), many equity or equity/bond mix curves tend to platykurtosis -- the tails are rather fat compared to the peak -- and variance usually sufficiently large that (following Roger's point) on any given date a person withdrawing money could be facing a loss in principal which withdrawing a fixed amount would compound into an effectively greater loss.
Shorter version of the Black Swan model: Most markets are better described by some power function rather than a normal curve which means variance could be much larger than expected on any given date; e.g., in a normal curve an event four standard deviations (SD) away from the mean is very rare but in a power function even larger sigma events can happen multiple times within a relatively short period*
*Nearly a year ago, last March, I tracked equity and commodity market moves of 16 SD or more, two or three times in a week. That might happen in a normally distributed system maybe once or twice between the big bang and the heat-death of the universe (little statistical hyperbole there) so, given that the vast majority of financial models make normal assumptions, there might be a bit of a problem there.
Roger, RW,
I was thinking about what Taleb was saying in the Russian Conference about hyper inflation. What should a portofio look like to take advantage of hyper inflation.
Thank in advance for your imput,
Jeff from milan, Italy
"From there it gets cloudier; 4% and then adjust for inflation every year some would say. I don't get this one. What if after ten years $1 million starting point is still $1 million but inflation works out such that the adjustments for same mean you take out $50,000?"
The assumption is that the 1,000,000 has grown with inflation as well. Its possible that it hasnt, but then indeed you are underwater.
But it really doesnt matter to get the point of the 4%.
The whole 4% came from the Trinity Study I believe. The idea is pretty simple. They asked the question "Historically speaking, what percent of a portfolio could one have spent over the last 200 or so years, and not run out of money?". Its based on historical results of stock/bond markets for the last 200 (or so) years.
They started in year 18xx and just ran the numbers for 30 year retirement. Starting in 1830 and ending in 1860. Then again in 1831 and ending in 1861 etc etc. Not sure of the dates but u see my point.
Anyway its just historical backtesting. It survived 1929 and the 1970s. Ie could you have retired in 1928, taken 4% (inflation adjusted) and not run out of money in 20 years. Not sure if it survived the Japanese 20yr bear test though.
If you google a guy called Jonathan Guyton he has come up with some decision rules that would have allowed over 5% and still survived.
Brian
The assumption is that the 1,000,000 has grown with inflation as well. Its possible that it hasnt, but then indeed you are underwater.
if that were true then i don't think my spin of whatever you got, 4% would have met as much objection when i first brought that up but that is what you are saying.
All it says is that if history is a guide, withdrawing 4% with inflation will give you approx 90% chance of survivability of your money. If you take 5% your chances of running out of money are higher etc.
What happens after year 1 is immaterial to the facts of the study, or the resulting 4% rule. Its really just a statistic :)
btw: just looked it up, the Trinity study was from 1926 not the 1800s. But I have seen similar results for the 1800s as well.
Brian
For the UK 1869 to 2005 the average year end dividend yield has been 4.9% and ranged from lows of 2.1% to highs of 11.7% (1.34% stdev).
Generally stock price gains pace inflation. So if you buy into an index then whatever dividends are paid might be considered as income, leaving the stock price to appreciate with inflation and also provide a rising income as dividends are increased over time.
On this basis then 4.9% might be considered to have been a historic safe withdrawal rate.
Much of actual income will depend upon when you 'lock-in' however. Do so when yields were 2.1% and you'll be income poorer than another who did so when yields were in double digits.
For anyone with a lump sum and some years until retirement perhaps a strategy might be to preserve the purchase power of that lump sum as much as possible until an opportunity arises to buy (lock) in at an above average yield.
Harry Browne's Permanent Portfolio or Mebane Faber's Quantitative models have historic low year on year draw-downs and might be a couple of examples of purchase power preservation styles.
Jeff,
In some ways hyperinflation is more difficult to counter than deflation because it's effects can be so variable but generally domestic cash is trash and commodities or other real assets or assets denominated in other, more stable currencies recommend themselves.
From a strategic perspective I model positions in extreme economic scenarios as insurance/hedges and prefer to buy them cheap. Right now more folks seem to be reacting to the buzz about inflation so hedging it has become more expensive but the probability of greater deflation has hardly disappeared, in fact I still consider it more likely than hyperinflation at this point because of the jobs picture and the dollar pegs, particularly to commodities such as oil.
Apologies for (another) extended comment but a comment at http://tinyurl.com/yznxakd (ht "steve from virginia") reminded me of the macroeconomic reasons strong energy reform is as critical as robust health reform right now. Set aside the debates about global climate change, freedom from foreign oil, etc; in a nutshell US dollars remain in demand because the world wants to stabilize their large, fixed costs and, for many nations, one of the largest is oil which is (mainly) priced in dollars.
But the US cannot print jobs or oil so the dollar (and to a certain degree the Euro) is piling up against the oil peg, mainly maintained by Arab gulf states, making the USD a somewhat "hardened" currency; not as hard as the "good old days" in the early 1930's when most of the worlds major currencies were pegged to gold and global economies crashed right and left before getting off but getting close enough.
Note that models such as permanent portfolio maintain strategic positions contra deflation as well as inflation.
Post a Comment