Actuarial Outpost
 
Go Back   Actuarial Outpost > Actuarial Discussion Forum > Pension - Social Security
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions


Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 09-18-2010, 03:24 AM
Mojo Mojo is offline
Member
 
Join Date: Oct 2001
Posts: 1,092
Default The Illusion of Pension Savings - NY Times

The Illusion of Pension Savings
http://www.nytimes.com/2010/09/18/bu...sion.html?_r=1


"Cuts for workers not yet hired do not save much money in the present — but that’s where actuaries can work their magic. They capture the future savings for use today by assuming, in essence, that 100 percent of today’s work force is already earning tomorrow’s skimpier benefits. When used in actuarial calculations, that assumption has a powerful effect"

can someone elaborate this part? thanks
Reply With Quote
  #2  
Old 09-18-2010, 12:16 PM
Jeremy Gold Jeremy Gold is offline
Member
SOA AAA
 
Join Date: Jul 2003
Location: New York City
College: Wharton PhD 2000
Posts: 448
Default

Quote:
Originally Posted by Mojo View Post
The Illusion of Pension Savings
http://www.nytimes.com/2010/09/18/bu...sion.html?_r=1


"Cuts for workers not yet hired do not save much money in the present — but that’s where actuaries can work their magic. They capture the future savings for use today by assuming, in essence, that 100 percent of today’s work force is already earning tomorrow’s skimpier benefits. When used in actuarial calculations, that assumption has a powerful effect"

can someone elaborate this part? thanks
Although the article doesn't spell it out -- too techy and too long for the Times -- there are two different methods being referred to.

The first method, and the core of the paragraph you cited, is called Ultimate Entry Age Normal Cost (UEANC). Most of us learned individual EAN where, for each current employee, we go back to their entry age (uniwinding their salaries by backing off the salary scale) and compute the PV of their total benefit and the PV future pay (PVFP). Since, at entry, all costs are normal cost (there can be no past service cost), the normal cost rate is PVB/PVFP. That rate is applied to this year's pay to get NC and the same rate is applied to the updated PVFP (from attained age to exit) to get the PVFNC. The PVFNC is subtracted from the current PVB to get the accrued actuarial liability (AAL). The assets are subtracted from the AAL to get the UAAL which is then amortized. In public plans this amortization is often done as a level percentage of pay over thirty years of open group payroll resulting in negative amortization for perhaps the next 17 years. The rate of funding of the UAAL starts at about 5% of the UAAL with interest at about 8%, so that after one year a UAAL of 100 has become 103.

The average speed of funding the PVFNC is, say, 10% of the PVFNC and the speed of funding the UAAL is about 5% of the UAAL (in the first of thirty years, but forever when the 30 years is rolling, i.e., every year is the first year). This means that switching $1 from PVFNC (by lowering the NC rate) to the UAAL slows funding from 10 cents to 5 cents.

Under the UEANC, developed in days when computers were slow and mainframe time expensive, the NC rate is determined by using one sample life to represent the average employee at entry. If the representation is realistic, this method produces an acceptable approximation to the more elaborate indivdual EAN method.

But public plans today have begun to cut back on benefits for new hires. How shall an actuary handle the two tiers? Under UEANC and EANC, the PVB is the same. The difference is in the PVFNC vs. UAAL split of the PVB. If, under UEANC, the actuary uses two sample lives, one for the newest hire, one for people covered by the older higher tier one benefits, it can still be a good approximation. If the actuary uses a blend of the two, it can still be pretty good. But suppose the actuary says that ultimate normal cost means the rate of normal cost that will ultimately arise when all active employees are in tier two with lower benefits. Then the PVFNC can drop sharply, e.g., by 20% and the decrease will cause the UAAL to increase by the same amount. But each dollar transferred from PVFNC to UAAL reduces corrent contributions by 5 cents. So even though only 2% of the total population (and only a few basis points of the total PVB) consists of new hires, the plans actuarially required contributions can drop by as much as 10% immediately.

That is what the article is talkiing about. It says that the actuaries assume that everybody is earning the new lower rate and (although the article doesn't mention it) that the excess of the old over the new for all tier one employees is now in the UAL where it is slow funded through negative amortization.

Later today I will describe what is going on in Illinois.
Reply With Quote
  #3  
Old 09-18-2010, 12:17 PM
Jeremy Gold Jeremy Gold is offline
Member
SOA AAA
 
Join Date: Jul 2003
Location: New York City
College: Wharton PhD 2000
Posts: 448
Default

Mary Pat,

You might want to cross-reference this thread in the PPinT 2 thread.

Jeremy
Reply With Quote
  #4  
Old 09-18-2010, 03:11 PM
limabeanactuary's Avatar
limabeanactuary limabeanactuary is offline
Mary Pat Campbell
 
Join Date: Jan 2010
Studying for Anglo-Saxon
Favorite beer: Bass Ale
Posts: 14,142
Default

Thanks Jeremy, I will.
Reply With Quote
  #5  
Old 09-20-2010, 03:09 AM
Mojo Mojo is offline
Member
 
Join Date: Oct 2001
Posts: 1,092
Default

From the man himself!
Thanks

Quote:
Originally Posted by Jeremy Gold View Post
Although the article doesn't spell it out -- too techy and too long for the Times -- there are two different methods being referred to.

<--->

Later today I will describe what is going on in Illinois.
Reply With Quote
  #6  
Old 09-20-2010, 07:11 AM
campbell's Avatar
campbell campbell is offline
Mary Pat Campbell
SOA AAA
 
Join Date: Nov 2003
Location: NY
Studying for duolingo and coursera
Favorite beer: Murphy's Irish Stout
Posts: 84,980
Blog Entries: 6
Default

I would like to hear Jeremy's thoughts on the Illinois death spiral.

http://www.bloomberg.com/news/2010-0...erfunding.html

http://www.chicagotribune.com/news/o...,5498288.story
__________________
It's STUMP

LinkedIn Profile
Reply With Quote
  #7  
Old 09-20-2010, 11:06 AM
Dan Moore's Avatar
Dan Moore Dan Moore is offline
Member
SOA AAA
 
Join Date: Jan 2008
College: University of Dallas
Posts: 2,135
Blog Entries: 1
Default

Section B of GASB 27 lists the permitted valuation methods for the calculation of ARC (annual required contribution), which is analogous to FAS 87 NPPC. The section on EAN reads:

Quote:
B-2 Entry Age Actuarial Cost Method of Entry Age Normal Actuarial Cost Method
A method under which the Actuarial Present Value of the Projected Benefits of each individual included in an Actuarial Valuation is allocated on a level basis over the earnings or service of the individual between entry age and assumed exit age(s). The portion of this Actuarial Present Value allocated to a valuation year is called the Normal Cost. The portion of this Actuarial Present Value not provided for at a valuation date by the Actuarial Present Value of future Normal Costs is called the Actuarial Accrued Liability...
The silence of ASOP 27 on calculating a normal cost allows the actuary to define the normal cost as they wish. I support the closing of this loophole.

If I were an auditor (which I'm not), I would not accept the use of the UEAN for calculating a GASB 27 ARC, because GASB 27's definition of EAN doesn't include the UEAN in its scope.
__________________
The best time to plant an oak tree is twenty years ago. The second best time is right now.

Last edited by Dan Moore; 02-10-2012 at 04:50 PM.. Reason: wrong asop
Reply With Quote
  #8  
Old 09-20-2010, 11:30 AM
Jeremy Gold Jeremy Gold is offline
Member
SOA AAA
 
Join Date: Jul 2003
Location: New York City
College: Wharton PhD 2000
Posts: 448
Default Illinois

The draft actuarial report for the State Employees' Retirement System of Illinois includes the following:

"In making these determinations, the required contribution shall be calculated each year as a level-percentage of payroll over the years remaining to and including fiscal year 2045 and shall be determined under the projected unit credit actuarial cost method."

It may not be entirely clear from this one sentence but the method used is an open group projection where plan assets and future contributions through 2045 are used to:

1) Pay all benefits paid through 2045
2) Leave the plan 90% funded at the end of FY 2045.

Contributions are determined as a level percentage of open group rising payroll and are not "determined under the projected unit credit actuarial cost method." The only place PUC comes in is in defining the AAL as of 2045.

That explains the following comment in the NY Times article:

"Illinois issued public documents this year naming its cost method as one that did not permit the cost of future employees’ benefits to be factored into the current year’s contributions. The apparent contradiction caught actuaries’ attention."

When funding is developed as a level percentage of rising group open payroll, effectively the projected payroll each year appears in the denominator of a ratio that determines the percentage of pay to be contributed each year. The benefits paid and the AAL in 2045 appear in the numerator. New entrants are included in the numerator (benefits) and denominator (payroll).

This method is used in five Illinois state plans.

Under this method funding ratios, already very low, drop for more than a decade before bottoming out, subsequently rising sharply to 90% of 2045 PUC in 2045. I have attached an excel graph of the funded ratio pattern for the total of the five plans.

Within the last year, the Illinois legislature adopted a cheaper plan for new hires beginning in 2011 (Senate Bill 1946).

Contributions for FY 2011 dropped by about 25% as a result. The closed group of current employees continues to earn benefits under the old formula. The contribution decrease begins immediately and is as large as it is because, by 2045, virtually every active employee will be under the new plan. So the effect is similar to the UEANC described above.

The new benefit formula caps pay to be recognized in the numerator at approximately 106,000 which affects few of today's employees. But this cap increases at 1/2 of the CPI while employee pay is likely to increase more rapidly and remains in the denominator. This explains the very backloaded funding graph. How realistic is it to assume that the cap will not be raised at the first sign of decent funding?

In the largest plan (Teachers), the 2009 valuation shows an $11 billion asset loss. Effective with the 2009 valuation, $9 billion is deferred and only $2 billion is recognized in the valuation. They switched from the market value of assets to an actuarial 5-year averaging method after they saw the asset losses that occurred in fiscal 2009. Does this meet the letter of ASOP 44? Maybe. But the spirit of ASOP 44 is that asset valuation methods should be unbiased. An unbiased switch to 5-year averaging would be adopted before the start of the fiscal year to which it first applies.

There are numerous other gimmicks in the Illinois valuations.

An actuary who advises the group (CoGFA) that advises the legislature wrote to the CoGFA:

"I agree that the language in these reports is somewhat misleading in stating that the projected unit credit actuarial cost method was used to determine contributions. In fact, the projected unit credit actuarial cost method is used only to determine the total actuarial liability in the year 2045 and then the state contribution is determined as the level percent of payroll needed to attain a 90% funded ratio in the year 2045. This is what is required under state law. Other portions of the report do point out that this is what is being done. But, I do agree that it would help to clarify that the projected unit credit method is used only to determine the total actuarial liability in the year 2045."

and:

"The reason that the State is able to make lower contributions under Senate Bill 1946 is that by the year 2045, total actuarial liabilities are projected to be substantially lower as most employees by that time will be covered under the new, lower benefits. Therefore, the level percent of payroll that the State needs to contribute to reach this lower total actuarial liability is also lower, even though there has been no reduction in benefit for most employees. This is why I have pointed out that the current funding plan may not be an appropriate one for the benefit changes under Senate Bill 1946."
Attached Files
File Type: xls Illinois funding.xls (36.5 KB, 284 views)
Reply With Quote
  #9  
Old 09-20-2010, 11:41 AM
Jeremy Gold Jeremy Gold is offline
Member
SOA AAA
 
Join Date: Jul 2003
Location: New York City
College: Wharton PhD 2000
Posts: 448
Default

Quote:
Originally Posted by Dan Moore View Post
Section B of GASB 27 lists the permitted valuation methods for the calculation of ARC (annual required contribution), which is analogous to FAS 87 NPPC. The section on EAN reads:



The silence of ASOP 4 on calculating a normal cost allows the actuary to define the normal cost as they wish. I support the closing of this loophole.

If I were an auditor (which I'm not), I would not accept the use of the UEAN for calculating a GASB 27 ARC, because GASB 27's definition of EAN doesn't include the UEAN in its scope.
The attached pdf was extracted from a GRS audit of a Buck valuation for Texas.

Quote:
Originally Posted by GRS Audit of Buck
Buck uses a variation of the EANC that bases the normal cost on the benefits that are available to
new hires (the ultimate entry age normal cost or UEANC method). This is a very common
variation in the public sector. However, Buck uses a different variation for purposes of
developing the GASB disclosures. All of the other public sector plans that we are aware of that
use the UEANC have taken the position that the UEANC method is acceptable under GASB.
GASB specifically requires that the method used for GASB reporting must be the same as the
funding method, if the funding method is acceptable under the GASB parameters. GRS believes
that the UEANC satisfies the GASB parameters and therefore should be used for GASB
reporting. We can see an argument that leads to the conclusion that the UEANC method does
not satisfy the GASB parameters, based on a very literal interpretation of the definition of EANC
included in the statements. Based on discussions with GASB staff, we do not believe that such a
literal interpretation is intended. For example, when we discussed the open group aggregate
method with GASB staff for another client (a method we thought did not meet the GASB
parameters), the response was, “It is aggregate so it is one of the methods. If the open group
variation is acceptable for funding purposes under the actuarial standards, GASB would not wish
to second guess the use of the method for accounting purposes.” We recommend this issue be
discussed with the ERS outside auditors.
This sort of flies in the face of the GASB 27 Glossary posted by Dan which says UEANC is not GAAP kosher. If, as GRS reports, GASB staff does not follow a literal interpretation of what they wrote, one must wonder whether GASB really understands the funding methods it approves. It seems that if the name of the method matches then, per GRS, GASB succumbs. As for relying on its actuarial acceptability, parts of ASOP 4 (3.11a and 3.11b, IIRC) require the allocation of NC to track service or compensation. Allocating to past service the difference between old plan and new plan benefits that is being earned on future service by most plan members appears not to meet this part of ASOP 4.
Attached Images
File Type: pdf grs buck extract.pdf (120.9 KB, 303 views)

Last edited by Jeremy Gold; 09-24-2010 at 04:38 PM.. Reason: to make important paragraph of attachment more prominent
Reply With Quote
  #10  
Old 09-20-2010, 12:08 PM
Chilango's Avatar
Chilango Chilango is offline
Member
SOA
 
Join Date: Aug 2002
Location: Washington, DC
College: Indiana University
Posts: 231
Default

If the amortization period is allowed to reset each year under the level percent of payroll method, then the excess that shifts from the NC to the UAL from using the ultimate EAN method may never get funded because of the perpetual negative amortization.
__________________
Over 10,000 practice exam books and study guides sold.
http://www.amazon.com/Philip-Martin-...y/e/B001JS02O6
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


All times are GMT -4. The time now is 05:45 PM.


Powered by vBulletin®
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.
*PLEASE NOTE: Posts are not checked for accuracy, and do not
represent the views of the Actuarial Outpost or its sponsors.
Page generated in 0.61418 seconds with 8 queries