In Detroit, were Orr’s pension estimates in ‘good faith’?

August 28, 2013

On June 14th, when Detroit emergency manager Kevyn Orr released his creditor proposal detailing how he wanted to treat creditors of the city, including retirees, I was aghast. Orr’s report said that the city’s pensions, which have been listed as exceptionally well-funded for years, were suddenly in desperate shape. What had been the one strong piece of Detroit’s balance sheet was suddenly reported to face medium-term collapse. Was there securities fraud in the financial reporting or had the investments in the pension funds blown up?

I’ve written numerous times that Orr likely over-inflated the pension liabilities.

We now have analysis from Morningstar that shows Orr probably did overstate Detroit pension liabilities by at least $1.5 billion. The Morningstar analysis is based on the recently released report that Orr commissioned from the actuarial firm Milliman.

Here is how Detroit reported the condition of its two pension funds in the official 2012 financial statements of the city (page 145):

Note that the two funds are showing actuarial value of assets at $6.8 billion as of 2011, with funded ratios of 82.8 percent and 99.9 percent. Here is what Orr originally claimed the pensions assets were as of 2013, via Morningstar, using a 7 percent investment return assumption and a reduced amortization period of 15-18 years:

Orr uses a “market value of assets” of $4.9 billion, versus the actuarial value of $6.8 billion recorded by the pensions in the city’s official financial reports. This is the essence of pension voodoo. Orr is allowed to calculate his assumptions in whatever way he wants, while the pension funds did their calculations following industry conventions.

But the real story is how Orr, seeming to disregard Milliman’s analysis, overstated the weakness of the pension’s condition. Compare Orr’s analysis of the pensions in the June 14 creditor proposal above with the Milliman report he commissioned (dated June 4) below, using the same assumptions:

Milliman’s use of a market value asset valuation method increases the pension assets by $1.5 billion and the funded levels by 18 percent. The unfunded value of the pensions, according to Milliman, drops from the widely touted $3.5 billion to $1.9 billion. Milliman’s estimate is much closer to what the pensions had reported in official financial statements.

Federal bankruptcy judge Christopher Rhodes has ordered that access to bankruptcy information do not require confidentiality agreements. Orr’s previous strict enforcement of secrecy requirements are curious given how wobbly his data is.

The important question is what Orr knew about how overstated his initial pension analysis was. Does this episode prove that Orr did not “negotiate in good faith” with creditors, which is a requirement for eligibility for Chapter 9 bankruptcy protection? Or are Orr and his team just not qualified to lead a complex bankruptcy? Judge Rhodes and all parties (including the media) need to look more critically at Orr’s numbers and his assertions. Taking Orr’s information at face value could be perilous.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Defending calculations based on “actuarial standards is nuts. John Bury is right. By this article Ms. Long has reduced her credibility to zero.

Posted by areopagetica | Report as abusive

1. The discount rate only affects the present value of accrued liability. It does not affect the actuarial value of assets. Indeed, you can see this in the very tables you cite. Note how the actuarial value does not change, regardless of discount rate used? Pretending it affects the actuarial value of assets is deeply misleading to your readers. I hope you correct this.

2. Using the market value of assets is not “voodoo.” It’s the actual value of assets that the plan has on hand. If the plan needed to sell its assets, today, it would not have the smoothed actuarial value. It would have the market value. Smoothing losses over a period of 7 years is simply a way to dampen asset volatility. But pretending that it’s the more accurate reflecting of the plan assets’ actual value is pretty ridiculous.

3. According to the data you cite, the use of a market value asset valuation method DECREASES the pension assets, contrary to your claim otherwise. You see how the “actuarial value” is higher than the “market value,” in the tables posted?

I sincerely hope there are meaningful corrections made to this post, as it contains a number of elementary errors.

Posted by ingramlaw | Report as abusive