Theoretical Fair Value for Index Futures

[Carroll Slemaker <cslemaker1@xxxxxxxx>]

I'm impressed by the sophistication and depth of technical knowledge
of some of the contributors to this list!  Perhaps one of you could
shed some light on something I find puzzling.

I continuously calculate a theoretical fair value ("TFV") for the front
S&P contract and display an indicator representing the relationship
between this theoretical value and the current actual value.  When I
first began doing this, years ago, the actual fluctuated back and forth
around the TFV, rarely moving very far away from it.  Whenever the
actual moved above the TFV by an amount greater than the buy threshhold,
I would see an immediate response in the NYSE TICK index as buy programs
kicked in, and likewise when the actual dropped below the sell
threshhold I would see the sell programs begin.

Recently (last year or so???), however, the behavior has changed.
I now never see this kind of instant response to any particular value
of my indicator.  And stranger is that the premium of the contract now
behaves option-like, as if it had time-value.  When a contract first
becomes the front contract, with 3 months to run, it has a very high
premium, well above the buy threshhold, and this premium gradually
decays over the life of the contract.  Today, for example, in spite
of the sharp drop in the market (SPX down 30), the March contract was
well above TFV most of the time, was above the buy threshhold
frequently, and dipped slightly below TFV on only a few brief
occasions.

To ensure that some of my parameters used in calculating the TFV have
not gotten out of whack, I periodically compare my calculated value
with that reported by Bloomberg and they are generally very close.
Any ideas???

Thanks.

Carroll Slemaker