FXLax -

In economic/financial modeling one can use nominal values (not adjusted for inflation) or real values (adjusted for inflation).

While adjusting for inflation is a more precise modeling method, it's also more complicated because it's another value one must estimate/assume, and is open for debate. Inflation in the 1970s was far different than the 1980s...which was different than the 1990s...or early 2000s...which is different than today & forecast.

US monetary policy, set by the Federal Reserve, also shifts with regards to inflation and unemployment. The Fed has recently shifted their views and policies regarding inflation, and I don't believe anyone knows what inflation will be for the next 10, 20 or 30 years. So with that said, introducing inflation into the model is another variable and another risk.

Given that, one can still build a nominal model - but such a model is affected by the discount rate used. Regardless of inflation, the discount rate (i) reduces the value of a current payout N years. Net Present Value of Payout in Year N = Payout times 1/(1+i)^N. Add up all future, discounted payouts and you will obtain the overall Net Present Value of any stream of payouts.

As the discount rate increases the value of future payouts decrease....and the total Net Present Value decreases

As the discount rate decreases the value of future payouts increase...and the total Net Present Value increases

So what discount rate should a pension fund use?

That's hotly debated by not only private pensions, but public pensions as well. If you do a simple Google Search you will see many articles. You can even research the discount rate used by the Fedex Plan. 3 Years ago, I did a lot of research on the Fedex Plan using financial reports in a university library.

The weighted discount rates used by our plan have steadily fallen from the 7.9%-8.5% range (1996-2000) to 7.7%-6.3% range (2001-2010)...to the 5.7%-4.0% range (2011-2017).

I have not updated my research for 2018 & 2019....sorry, I've been off doing "other things". I have the specific data on each year, but it's too much too type

So, while our $130,000 max payout has remained constant..... the Net Present Value (unadjusted for inflation) has Increased.

To my other point, the current Fedex A plan EROA (Expected Return on Assets) has also decreased (...which, from the Fund manager's perspective, makes it harder to fund the A Plan)

Those ranges are: 9.3%-10.9% (1996-2000).....10.9%-9.1% (2001-2007)....8.5%-8.0% (2008-2013)....and 7.75%-6.5% (2014-2017). Once again, I have each year but it's too much to type.

In combination, this decrease in the Discount Rate....and the decrease in the Expected Return on Assets....has put a greater burden on the company to fund the A plan. A burden that would shift to the pilots under the Variable Benefit Plan (...No thanks! The company can keep that promise, that responsibility and that burden!)

In my research, I was also able to obtain the AROA (Actual Return on Assets) for each year of the plan. Any year the AROA was greater than the EROA was a GOOD year for the plan. Any year the AROA was less than the EROA was a BAD year for the plan.

It's interesting to me that the VBB Modelor doesn't actually use the known AROA from our A Plan....rather it uses some weighted asset allocation of two other historical indexes: The S&P 500 Index and the Barclays US Aggregate Bond Index (...previously known as the Lehman US Aggregate Bond Index).

Why not? Why assume Fedex Fund managers would have performed liked the idexes? Why not look at their actual historical performance?

How would the modeler results be different if actual Fedex AROA returns were used?

In Unity,
DLax