Modelling the price history of individual (monthly) COMEX gold futures contracts over extended periods of time doesn't work very well, largely because liquid trading life of each individual contract is very short. Because of the very high correlation between contracts, most hedgers and traders concentrate (or stack) their activity in a near month - currently June15. This is obvious in the volume at the bottom of Skol's charts.
So how do we build a long term history of gold futures prices?
There are two solutions:
1) If you are a chartist, it is simpler not to bother with futures prices at all. Most chartists use spot price histories instead. The correlation between the two so close to 1 as doesn't matter, and an analyst can build a history back 40 years.
2) However, many banks have OTC gold price risk for as much as 10 years into the future, and may have open positions in untraded COMEX gold contracts as far out as 2019 to hedge these positions. To calculate their regulatory capital requirement, they need dayly price on all the COMEX contracts that they have positions in. Under the Basel market risk capital accords, the histories must have a minimum of two years of prices and be updated every trading day.
At this point I start to have nasty flashbacks!
There is a very clever solution to this problem which is interesting to less than 100 people in the country, but I am not one of them anymore! Do I bother explaining....?
