markturansky · November 24, 2016 17:33
diff --git a/gistfile1.txt b/gistfile1.txt
 Pension Math
 March 2, 2011 at 10:13am
 Several people have linked to the following blog entry on Fortune:
 
 http://blogs.forbes.com/rickungar/2011/02/25/the-wisconsin-lie-exposed-taxpayers-actually-contribute-nothing-to-public-employee-pensions/
 
 I think the “update” section is extremely important, particularly the last paragraph:
 
 “As a result, the taxpayers do not contribute to the public employee pension programs so much as serve as insurers. If their elected officials have been sloppy, the taxpayers must stand behind it. But if the market continues to perform as it has been performing this past year, don’t be surprised if the funding crisis begins to recede. If it does, what will you say then?”
 
 One of the most important and costly items in any pension is the insurance guarantee alluded to in this paragraph and I wanted to spend a bit of time trying to explain how expensive this guarantee actually is.
 
 There are many ways to value pension plans – none of them particularly easy to follow or explain in a short note.  I’ll try to cut through a lot of the complexity with a fairly straightforward example that shows exactly how valuable the state guarantees actually are.
 
 The Fortune blog article links to a study done by the Pew Center on States:
 
 http://downloads.pewcenteronthestates.org/The_Trillion_Dollar_Gap_final.pdf
 
 In the “Key Findings” section of this study on page 3, is this paragraph:
 
 “In fiscal year 208, which for most states ended on June 30, 2008, states’ pension plans had $2.8 trillion in long-term liabilities, with more than $2.3 trillion socked away to cover those costs.”
 
 So, a roughly $500 billion shortfall, which represents a funding level of just over 80%.  Not bad, right?
 
 Well . . . not so fast. 
 
 Another study by Josh Rauh at Northwestern – see the first link here:
 
 http://www.kellogg.northwestern.edu/faculty/rauh/
 
 Calculates a shortfall of about $2.5 billion ($1.9 billion of assets and $4.4 billion of liabilities – so only 43% funded.
 
 How could the results be so different?
 
 Here’s the math.
 
 Pension plans are allowed to do a very strange thing with their liabilities.  They can discount these liabilities at a rate equal to their assumed rate of return.  A quick example and then a slightly more detailed one demonstrate the point.
 
 Suppose that you are required by law to pay someone $100 in a year.  How would you value that liability today?  One approach would be to say that one year interest rates are close to zero (0.23% as I type this), so I need to put roughly $99.77 in a bank account in order to have $100 in a year, so the value of the liability is $99.77.
 
 With pension math, though, you can play a different game.  You can assume, as must funds do, that you will somehow earn 8% this year and, thus only have to put about $92.60 in the account and value the liability accordingly.    Of course, no matter what the account earns you have to pay the $100 in a year, so if the account has less you’ll have to come up with the money some way or other.
 
 The $7 difference in my example may not see like that big of a deal, so let’s look at an example that is more in line with liabilities in pension plans.
 
 For that example, I’m going to go with a slightly modified version of the one that Arnold Schwarzenegger used in his Wall Street Journal article:
 
 http://online.wsj.com/article/SB10001424052748703447004575449813071709510.html
 
 “Few Californians in the private sector have $1 million in savings, but that's effectively the retirement account they guarantee to public employees who opt to retire at age 55 and are entitled to a monthly, inflation-protected check of $3,000 for the rest of their lives.”
 
 The inflation-protected math makes the example more complicated than necessary, so I’m going to look at the slightly simpler problem of valuing a pension that pays $3000 per month (so, $36,000 per year) starting at age 55. 
 
 So, for my example, let’s pretend that starting in 10 years we are contractually required to pay someone $36,000 per year (for simplicity assume we make the payment on the first day of each year) and that person is going to live to be 84.
 
 We’ll look at that present value of that obligation assuming a 4% discount rate as well as assuming an 8% discount rate.  For fun, try to guess how big the difference will be before you read on.
 
 Using a 4% discount rate, the present value of this liability is about $430,000 and with an 8% rate you get a value of about  $200,000.  What appears to be a small change – moving the discount rate from 4% from 8% - more than doubles the present value of the liability!!
 
 What are the implications?  Well, the “our assets will earn 8% forever” assumption means that the pension fund managers need to invest in risky assets.  As we all should know quite well by now, the values of risky assets can go up and they can go down.   With the 10 year treasury rate hovering around 3.5% and the 30 year treasury rate hovering around 4.5% it is hard to imagine how you are going to earn 8% in perpetuity (particularly on $2.3 trillion of assets).
 
 Pension fund managers (and others) will retort – but our average rate for the last 30 years has been 8%, so it is not even remotely aggressive to assume we’ll keep doing that.    The fact that rates on 30 year government securities in 1981 were 12% and rates on 10 year government securities in 1981 were also about 12% should scare the hell out of you.
 
 By investing in risky assets, the pension plans are taking enormous risks with other people’s retirement money.  When it is suggested that people be allowed to invest their own social security money in the stock market through private accounts, politicians go crazy.  But no one (including the pensioners themselves) seems to mind at all that this is exactly what happens with pension plans.  It is a complete mystery to me why this kind of gambling is allowed with retirement money.
 
 Well . . . except that through the magic of accounting, it makes the cost of the pension plans to be smaller than it actually is.  As the NYTs article reports today:
 
 http://www.nytimes.com/2011/03/02/business/02leonhardt.html?ref=business
 
 “The delaying of costs is obvious. Both politicians and union leaders have decided that generous future benefits offer the easiest way to hold down spending and still satisfy workers. The result is government pay that’s skewed too heavily toward pensions and health insurance.”
 
 Anyway, back to the cost of the taxpayer guarantee . . .
 
 Since the values in the pension accounts are independent of the liabilities (i.e. the pension payments need to be made whether or not the pension plan has money) one easy way to estimate the cost of the taxpayer guarantee is to compare the value of the liabilities using a risk free rate and using the reported risky rate.
 
 As we saw in my simple example, the liabilities for pension assets are about double when you use a 4% discount rate rather than an 8% rate.  According to the PEW article linked at the beginning, in 2008 the assets of the pension plans were about $2.3 trillion and the liabilities were about $2.8 trillion..   At the risk free rate, a decent estimate of these liabilities is about doubles, so say $5.6 trillion and the true shortfall is not $500 billion, but rather $3.3 trillion, and a good estimate for the value of the taxpayer guarantee is thus about $2.8 triillion.   Hardly something to brush aside.
 
 As I’ve said to my friends who posted the original article, a major contributor to the pension fights that are going on today are the techniques used to value the pensions over the years.  Many, many people are going to get hurt because these valuation techniques have vastly understated the true liability of these programs. 
 
 You have to look no farther than the contributions to the NYC pension plans over the last 10 years to see how the valuation problems manifest themselves.  Here are the contributions since 2000 (in $ millions):
 
 
 2000 692.8
 2001 1,566.8
 2002 1,762.9
 2003 2,302.4
 2004 2,997.9
 2005 4,073.1
 2006 4,378.3
 2007 5,429.2
 2008 6,511.8
 2009 7,284.2
 2010 7,684.4
 
 The contribution rates are growing at a compounded 18% per annum since 2001.  That level of growth in the contribution rates obviously cannot be sustained much longer without the pension fund contributions being larger than the entire budget. 
 
 This is going to be an extraordinarily difficult problem to deal with and the solutions are going to be incredibly painful to many, many people.  My hope is that we’ll learn how painful cheating on the valuations was and value the obligations properly in the future.  I also hope that with proper valuations, the pension plans will stop gambling with people’s retirement money and people’s retirements will actually be safer.  I doubt either will happen, though.
	Pension Math
	March 2, 2011 at 10:13am
	Several people have linked to the following blog entry on Fortune:

	http://blogs.forbes.com/rickungar/2011/02/25/the-wisconsin-lie-exposed-taxpayers-actually-contribute-nothing-to-public-employee-pensions/

	I think the “update” section is extremely important, particularly the last paragraph:

	“As a result, the taxpayers do not contribute to the public employee pension programs so much as serve as insurers. If their elected officials have been sloppy, the taxpayers must stand behind it. But if the market continues to perform as it has been performing this past year, don’t be surprised if the funding crisis begins to recede. If it does, what will you say then?”

	One of the most important and costly items in any pension is the insurance guarantee alluded to in this paragraph and I wanted to spend a bit of time trying to explain how expensive this guarantee actually is.

	There are many ways to value pension plans – none of them particularly easy to follow or explain in a short note. I’ll try to cut through a lot of the complexity with a fairly straightforward example that shows exactly how valuable the state guarantees actually are.

	The Fortune blog article links to a study done by the Pew Center on States:

	http://downloads.pewcenteronthestates.org/The_Trillion_Dollar_Gap_final.pdf

	In the “Key Findings” section of this study on page 3, is this paragraph:

	“In fiscal year 208, which for most states ended on June 30, 2008, states’ pension plans had $2.8 trillion in long-term liabilities, with more than $2.3 trillion socked away to cover those costs.”

	So, a roughly $500 billion shortfall, which represents a funding level of just over 80%. Not bad, right?

	Well . . . not so fast.

	Another study by Josh Rauh at Northwestern – see the first link here:

	http://www.kellogg.northwestern.edu/faculty/rauh/

	Calculates a shortfall of about $2.5 billion ($1.9 billion of assets and $4.4 billion of liabilities – so only 43% funded.

	How could the results be so different?

	Here’s the math.

	Pension plans are allowed to do a very strange thing with their liabilities. They can discount these liabilities at a rate equal to their assumed rate of return. A quick example and then a slightly more detailed one demonstrate the point.

	Suppose that you are required by law to pay someone $100 in a year. How would you value that liability today? One approach would be to say that one year interest rates are close to zero (0.23% as I type this), so I need to put roughly $99.77 in a bank account in order to have $100 in a year, so the value of the liability is $99.77.

	With pension math, though, you can play a different game. You can assume, as must funds do, that you will somehow earn 8% this year and, thus only have to put about $92.60 in the account and value the liability accordingly. Of course, no matter what the account earns you have to pay the $100 in a year, so if the account has less you’ll have to come up with the money some way or other.

	The $7 difference in my example may not see like that big of a deal, so let’s look at an example that is more in line with liabilities in pension plans.

	For that example, I’m going to go with a slightly modified version of the one that Arnold Schwarzenegger used in his Wall Street Journal article:

	http://online.wsj.com/article/SB10001424052748703447004575449813071709510.html

	“Few Californians in the private sector have $1 million in savings, but that's effectively the retirement account they guarantee to public employees who opt to retire at age 55 and are entitled to a monthly, inflation-protected check of $3,000 for the rest of their lives.”

	The inflation-protected math makes the example more complicated than necessary, so I’m going to look at the slightly simpler problem of valuing a pension that pays $3000 per month (so, $36,000 per year) starting at age 55.

	So, for my example, let’s pretend that starting in 10 years we are contractually required to pay someone $36,000 per year (for simplicity assume we make the payment on the first day of each year) and that person is going to live to be 84.

	We’ll look at that present value of that obligation assuming a 4% discount rate as well as assuming an 8% discount rate. For fun, try to guess how big the difference will be before you read on.

	Using a 4% discount rate, the present value of this liability is about $430,000 and with an 8% rate you get a value of about $200,000. What appears to be a small change – moving the discount rate from 4% from 8% - more than doubles the present value of the liability!!

	What are the implications? Well, the “our assets will earn 8% forever” assumption means that the pension fund managers need to invest in risky assets. As we all should know quite well by now, the values of risky assets can go up and they can go down. With the 10 year treasury rate hovering around 3.5% and the 30 year treasury rate hovering around 4.5% it is hard to imagine how you are going to earn 8% in perpetuity (particularly on $2.3 trillion of assets).

	Pension fund managers (and others) will retort – but our average rate for the last 30 years has been 8%, so it is not even remotely aggressive to assume we’ll keep doing that. The fact that rates on 30 year government securities in 1981 were 12% and rates on 10 year government securities in 1981 were also about 12% should scare the hell out of you.

	By investing in risky assets, the pension plans are taking enormous risks with other people’s retirement money. When it is suggested that people be allowed to invest their own social security money in the stock market through private accounts, politicians go crazy. But no one (including the pensioners themselves) seems to mind at all that this is exactly what happens with pension plans. It is a complete mystery to me why this kind of gambling is allowed with retirement money.

	Well . . . except that through the magic of accounting, it makes the cost of the pension plans to be smaller than it actually is. As the NYTs article reports today:

	http://www.nytimes.com/2011/03/02/business/02leonhardt.html?ref=business

	“The delaying of costs is obvious. Both politicians and union leaders have decided that generous future benefits offer the easiest way to hold down spending and still satisfy workers. The result is government pay that’s skewed too heavily toward pensions and health insurance.”

	Anyway, back to the cost of the taxpayer guarantee . . .

	Since the values in the pension accounts are independent of the liabilities (i.e. the pension payments need to be made whether or not the pension plan has money) one easy way to estimate the cost of the taxpayer guarantee is to compare the value of the liabilities using a risk free rate and using the reported risky rate.

	As we saw in my simple example, the liabilities for pension assets are about double when you use a 4% discount rate rather than an 8% rate. According to the PEW article linked at the beginning, in 2008 the assets of the pension plans were about $2.3 trillion and the liabilities were about $2.8 trillion.. At the risk free rate, a decent estimate of these liabilities is about doubles, so say $5.6 trillion and the true shortfall is not $500 billion, but rather $3.3 trillion, and a good estimate for the value of the taxpayer guarantee is thus about $2.8 triillion. Hardly something to brush aside.

	As I’ve said to my friends who posted the original article, a major contributor to the pension fights that are going on today are the techniques used to value the pensions over the years. Many, many people are going to get hurt because these valuation techniques have vastly understated the true liability of these programs.

	You have to look no farther than the contributions to the NYC pension plans over the last 10 years to see how the valuation problems manifest themselves. Here are the contributions since 2000 (in $ millions):


	2000 692.8
	2001 1,566.8
	2002 1,762.9
	2003 2,302.4
	2004 2,997.9
	2005 4,073.1
	2006 4,378.3
	2007 5,429.2
	2008 6,511.8
	2009 7,284.2
	2010 7,684.4

	The contribution rates are growing at a compounded 18% per annum since 2001. That level of growth in the contribution rates obviously cannot be sustained much longer without the pension fund contributions being larger than the entire budget.

	This is going to be an extraordinarily difficult problem to deal with and the solutions are going to be incredibly painful to many, many people. My hope is that we’ll learn how painful cheating on the valuations was and value the obligations properly in the future. I also hope that with proper valuations, the pension plans will stop gambling with people’s retirement money and people’s retirements will actually be safer. I doubt either will happen, though.