Buying Actuarial Opinions

Never has the curtain of deceit in public pension funding been open so wide as in Detroit’s clumsy attempt to create an underfunding in their two pension plans within the parameters used by public plan actuaries for decades to create the appearance of quasi-solvency.   Detroit had been paying Gabriel Roeder Smith & Company (GRS) millions of dollars over the years to mask the true cost of pension liabilities accruing.  Then last May when they needed to have the plans be severely underfunded and unsustainable they paid Milliman, Inc. (Milliman) $350,000 for that opinion (proffered on June 4, 2013 with the general public getting it today) for both the Detroit General Retirement System (DGRS) and the Police & Fire Retirement System (PFRS).

Resist the impulse to wade through these jargon-laced reports.  They were not designed for anyone to understand.  They were designed to have their conclusions accepted without question.

I do not accept them and they raise four questions that are obvious once isolated.

Milliman valued liabilities using a rate they deemed reasonable (6.3% for DGRS and 6.57% for PFRS) and two rates they were asked by Detroit to use (7% and 7.5%)  This is what they came up with using market value of assets:

MILLIMAN REPORT NUMBERS @ 6/30/12 (in millions)
………………………………..DGRS………PFRS………..TOTAL…..UNFUNDED
Market Assets……………2,159……….2,974………..5,133
Milliman Liabilities……4,433………4,528 ……….8,961………..3,828
7% Liabilities…………….4,085………4,316………..8,401………..3,268
7.5% Liabilities………….3,836………4,069 ……….7,905………..2,772

That $3.5 billion number being kicked around is Milliman projecting their 7% liability numbers to June 30, 2013:

MILLIMAN REPORT NUMBERS @ 6/30/13 (in millions)
………………………………..DGRS………PFRS………..TOTAL…..UNFUNDED
Market Assets…………..2,077……….2,918………..4,995
7% Liabilities…………….4,114………4,355………..8,469………..3,474

Milliman then projected costs out through 2022 comparing what Detroit would theoretically be putting into the plans under the current structure to what the contributions would be assuming the following changes:

  • Benefit accruals frozen
  • Cost-Of-Living-Adjustments (COLAs) eliminated
  • Replace the DB plans with DC plans at 5%-of-pay for DGRS and 10%-of-pay for PFRS

With these changes total contributions would come way down and the headlines would have been about how much Detroit taxpayers would save (as opposed to how much Detroit retirees would lose) as the changes take effect.  Obviously the unions didn’t go along so Detroit filed for bankruptcy and we all await these changes….and more.

Now, as to the obvious questions that should be asked of Milliman on their reports:

1) Asset Smoothing:

Check out all those charts and you’ll notice that the Actuarial Value of Assets is always greater than the Market Value of Assets.  This isn’t smoothing.  It’s a multiplier which Milliman properly ignored when producing their official numbers – something they are loath to do in their day job.

2) The proper interest rate:

From the Milliman study of the DGRS report (pages 2 – 3):

The investment return assumption is reduced from the 7.90% rate used in the valuation.  Results are shown on the following three bases, each of which reflect an adjustment to the liability and also are used as the projected future annual return to model future asset levels:

Exhibit I – 6.30% – Milliman’s recommended assumption based on our capital market assumptions

From the Milliman study on the PFRS  report (page 3):

The investment return assumption is reduced from the 8.00% rate used in the valuation.  Results are shown on the following three bases, each of which reflect an adjustment to the liability and also are used as the projected future annual return to model future asset levels:

Exhibit I – 6.57% – Milliman’s recommended assumption based on our capital market assumptions

So Milliman has a 6.3% recommended assumption for the the DGRS and a 6.57% rate for the PFRS based on………….??????

3) Projecting Benefit Payouts:

Absolutely critical in projecting future costs and liabilities is the estimate of what future payouts are expected to be.  In the PFRS report (page 6) Milliman took the GRS projections over the period noting:

The anticipated benefit payments developed by Gabriel Roeder Smith & Company in the projection for the 2011-2012 fiscal year were $317.7 million.  The actual benefit payments for the 2011-2012 fiscal year were $326.6 million, a difference of $8.9 million.

Our simplified baseline projection uses the anticipated benefit payments for fiscal years 2012-2013 through 2020-2021 as shown above and an anticipated benefit payment of $370.0 million for 2021-2022,

Fair enough.  Milliman used GRS projected benefit amounts to develop their estimated liabilities and costs.  But turn to the Milliman DGRS report (page 7):

The anticipated benefit payments developed by Gabriel Roeder Smith & Company in the projection for the 2011-2012 fiscal year were $225.5 million.  The actual benefit payments for the 2011-2012 fiscal year were $394.2 million, a difference of $168.7 million.  In addition, the anticipated benefit payments in the projection, as shown above, increase by 32% from 2011-2012 to 2012-2013.

Our simplified baseline projection uses the anticipated benefit payments for fiscal years 2012-2013 through 2020-2021 as shown above and an anticipated benefit payment of $406.2 million for 2021-2022,

Annual payouts from the DGRS for the 2011-2012 fiscal year were $394.2 million yet the projections assume that for every year through 2020-2021 annual payouts will be less!  Was there something unusual about the 2011-2012 year for the DGRS?

4) $350,000?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: