As detailed in this section of the Wikipedia page on the Yield Curve, we can construct the yield curve from the money market as follows:
- The LIBOR rates give us the short end of the curve (t < 3m)
- IR futures give us the midsection of the curve (3m <= t <= 15m)
- IR swaps give us the long end of the curve (1y <= t <= 60y)
On the right-hand side of the page there is a table showing typical inputs to the money market curve:
|Type||Settlement date||Rate (%)|
|Cash||Tomorrow next rate||5.59375|
There is a footnote stating the data is for lending in US dollar, taken from October 6, 1997.
Concentrating just on the midsection of the curve (the futures), I see in the contract specification for CME's Eurodollar futures that the underlying is the 3 month LIBOR rate.
So, at expiration, the Dec-97 future settles at the current 3 month LIBOR rate.
Does this, in effect, mean that the Dec-97 future gives us the Mar-98 point on the curve? (3 months after Dec-97)
In other words, if we look at the previous row in the table, which is the 3 month LIBOR rate (5.71875%), and given the data was taken on the 6th October 1997, this is the rate at 6th October + 3 months = 6th December 1997, whilst the Dec-97 future rate on the 6th October 1997 was 5.76%, which is the rate at settlement of the Dec-97 future (13th December 97), forward 3 months, because it's underlying is 3 month LIBOR, so ~13th March 1998.
Is that correct?