PERS Calculator - Uses 1978 and 2001 Actuarial Equivalency Factors (NO LOOKBACK YET)
The PERS calculator link below will be updated quite frequently as PERS and I work out the details of their calculations. My program will NEVER be as accurate as the programs PERS uses, but my goal is to produce realistic estimates that enable you to plan responsibly as PERS undergoes this transition from the 1978 to the 2001 actuarial tables.
Please note that the estimates my program produces might well be dead wrong -- and I will accept no responsibility if that is the case. Use this program at your own risk! By all means, you should get estimates directly from PERS. Since they cannot provide estimates for post-2003 retirement dates yet, my program should give you some clues about what may befall you if you retire afterwards.
PERS does not endorse anything I do, but they have given me relevant information whenever I've asked. I am grateful to Phil, Paul, and Scott at PERS who've taken time to answer my (often) persistent questions about the gnarly details of benefit calculations. I still have much to learn. I'm also grateful to Dale Orr and Jim Voytko, who have also taken the time to answer my questions, some of which relate to implementating the new tables. Representative Greg Macpherson took time from his schedule to explain some of the thornier details of HB 2004-A.
CAVEATS:
1) The program does not calculate benefits under the option of purchasing "waiting time" or other prior service credits. This requires information that few of us have handy. The calculations aren't complicated, but do YOU remember exactly how much you earned during the first 6 months of your career with a PERS employer?
2) The program is not currently programmed to calculate either the 12/31/03 "lookback" method adopted at the 8/12/02 PERS Board meeting OR the 6/30/03 "lookback" embodied in HB 2004-A, soon to be considered by the Oregon Senate after having been approved 51-8 in the Oregon House. I intend to program the Legislative "lookback" calculation in the next iteration of the program (probably early March). I understand the calculation but it adds at least one order of magnitude of difficulty to the error checking in the program.
3) While I calculate the benefit increment adjustment mandated by the Oregon Supreme Court as codified in ORS 238.380, the results here are approximations based on reverse-engineering my own benefit estimates, attending PERS counselling sessions and trying to decode the hidden meaning of ORS 238.380 (I tried playing the Beatles' "White Album" backwards, but it didn't help me at all :-). There are so many exceptions to ORS 238.380 that I have no way of anticipating them. If I had more benefit estimates from PERS, I could more easily check the calculations.
4) Please keep in mind the distinction between "separating from employment" and "retiring". You *cannot* retire unless you are "separated". The latter must precede the former. PERS only allows retirements on the first of a calendar month.
5) The program calculates OPTION 1 benefits (no refund, no beneficiary), as well as Option 2 (100% Survivor Benefit), 2A (100% Survivor Benefit with Option 1 reversion), 3 (50% Survivor Benefit), 3A (50% Survivor Benefit with Option 1 reversion) and 4 (15 Year Annuity Certain) for both the 1978 and 2001 Tables. It also provides the Option 12 (refund of member balance) estimate, as well as Option 13 (refund of member and employer balance), which took effect for retirements on or after 1/1/2003.
6) The "lookback", once programmed, is still only hypothetical. There is NO LOOKBACK yet approved anywhere except in the Oregon House. The version in HB 2004-A, which is what I'm currently programming, still has to be approved in the Oregon Senate and signed by the Governor before it becomes law.
7) There is NO way to "finesse" an estimate of your "lookback" benefit, although you could estimate the "lookback" balance. If you are interested in estimating your balance on 7/1/03, simply put in a separation date of 6/30/03 and a retirement date of 7/1/03. The Option 12 balance will be the approximate balance "frozen" under the HB 2004 version of the "lookback" UNLESS you have a significant stake in the Variable. If so, the 6/30/03 balance will only be as good as your ability to predict the stock-market performance over the next 3 months or so. (If you have that gift, please share will all, including the Oregon Investment Council).
8) Please, please, please let me know if something doesn't look right. An example would be a scenario where the 2001 benefit (for any option) is higher than the 1978 benefit for the SAME option. I don't think that should happen and I haven't seen any examples of it happening. But, please do let me know if there is something that just doesn't look right. PLEASE READ THE DISCLAIMER NOTE BELOW THE PROGRAM LINK TOO. The 2/26/03 version had a bug in the calculation of the Option 4 benefit under the 2001 tables. It wasn't a computational error; the program simply displayed the wrong value. The 3/2/03 version ADDS the Option 0 (refund annuity) calculation. This calculation is the foundation of the HB 2004-A "lookback". I haven't finished programming the HB 2004-A "lookback". The current version - 03/04/03 - is the foundation of the version to be released at the end of the week. It switches the positions of the Option 0 and Option 1 benefits (DO NOTE THIS), adds "holes" for information irrelevant now but needed in the "lookback", and uses a color-coding scheme to identify the highest benefit. Until the "lookback" is released, the 1978 tables should always "win". Finally, the program will stop working after 3/31/04. There just isn't any point in calculating the lookback when the lookback balance will be a known amount by the time the 2003 Member Statement comes out in April 2004.
9) The latest version fixes a minor bug in the way variable rates were calculated. It also fixes a bug in the way ORS 238.380 adjustments were calculated for folks who separated from PERS employers prior to the date of their last PERS Statement. Finally, at the request of many Police & Fire members, I changed the program's limit so that you can "try" to retire down to age 50. If you aren't Police/Fire, you have been warned that the results are meaningless and you can't really retire. Just because the program gives you an answer doesn't mean you're eligible to retire.
Note: THIS IS A WINDOZE PROGRAM. If you have a current version of WINE (I use the 6/5/2002 version) and a reasonably current Linux kernel (I use Red Hat 8.0 with the 2.4.18-24.8.0 kernel) the program will run under Linux (if you're geeky enough to use Linux, then you'll know how to run it under WINE). Note that under Linux, the fonts are not rendered entirely correctly, but this is cosmetic, not computational. The program WON'T run on a Mac (I can't figure out why anyone would own one anyway) except, perhaps, with a PC emulator. Beginning with the 11/24/02 Version, you can capture the screen to your printer and get hard copy of the results. It isn't exactly pretty, but it works.
PERS Calculator for Regular & Variable Accounts (5/04/03 Version) (NEW!!!)
DISCLAIMER!!!!!!!!!!
THIS PROGRAM WILL NOT WORK FOR STATEMENTS EARLIER THAN 2002, NOR LATER THAN 2003. IN FACT, THE PROGRAM IS UTTERLY USELESS FOR RETIREMENTS AFTER 6/30/03 AND SO I WOULDN'T USE IT AT ALL RIGHT NOW. While the above program calculates the benefit under both the 1978 and 2001 AEFs, be particularly careful with the results if you have a substantial investment in the variable. The 'variable' annuity is especially tricky to program because of the dynamic way the interest rate changes with time and market performance. Consequently, whatever assumed interest rate you use for the variable is almost assuredly WRONG if you're planning to retire more than about a year from the time you run the calculator. It is easy to program the 'fixed' because you have a baseline assumption you can use (8% guarantee), but this isn't true for the 'variable'. You can 'fix' the variable balance on a particular date (e.g. 12/31/02) because you have actual interest rates to use, but going forward is nearly impossible unless you program in all the intermediate return rates between the date of the last statement and the proposed retirement date. You *could* assume an 8% return, which would probably be OK if you weren't going to retire for about 5 years, but if you're within a few years of retirement, the rate forward from 12/31/02 isn't quite as optimistic. At best the program is only as good as the rate you want to assume for the variable portion of your account. I advise you to figure low. How low? As Harry Potter is wont to say, "dunno". The 2002 final rate is -21.30% (look at the rates for each month in 2002 if you want to see what kind of horror show you could have). Everyone hopes that these negative return rates will not continue too much longer, but 2003 hasn't started out demonstrably better. Only you can decide what rate to use. Play with a bunch of scenarios (positive and negative) to get a feel for the range of variation you might expect when you retire. I think you'll be quite astonished by the results. In any case, be aware that the beta (standard deviation) on returns in the variable is quite high.
Please send any comments, criticisms, complaints, or feature requests (with calculation methods) to: feldesmanm@pdx.edu.
If you want to send lawyers, guns, or money [with apologies to Warren Zevon], I'll deny I ever wrote this program. If you disassemble the source code, you won't be able to trace it to me. If you hire a pimply-faced teenager to hack the code, nothing can implicate me, but on the off chance that I've left a stray "identifier" in the comments, I know full-time software engineers who work for some of the biggest firms who could munge the code so that it would be untraceable even if you had the Rosetta Stone.
I would appreciate an email letting me know how the program compares with the estimates you've gotten from PERS (you have gotten a benefit estimate, n'est ce pas?). I can't fix what I don't know is broken.
Return to Main PERS page