Reading in the local Singapore paper, "Singapore Exchange plans to offer a futures contract based on dynamic random access memory (DRAM) computer chips." The idea is "to help manufacturers and users manage their risks effectively." Little is said about the specific contract definition aside from, "[contract] to comprise 10,000 pieces of DRAM valued at about $70,000."
Here is yet another dead-contract-walking introduced by the SGX.
The problem screaming loudest about this contract is the contract specification itself. Whatever the specification of the DRAM the contract mentions today will be old in six months, out-of-date in a year, and an antique in two years. So what sort of future curve would that ever generate? One with enormous backwardation and no backend liquidity. State-of-the-art today will not be in a year, thus a futures contract for that DRAM will slowly slide into worthlessness.
Other problems? I've been told that there isn't good two-sided customer demand for this instrument. Lots of manufacturers would love to sell forward their DRAM production as a hedge, but there are no large stakeholders who want to purchase forward. Either the consumer just passes-on the cost of the DRAM into whatever product he's building, or for the end user, the price of the memory is a nominal percentage of the device they're buying. Regardless, there is no incentive for hedge buying forward at anywhere near the same scale as the DRAM manufacturers to sell.
I think the basic reason this idea is doomed is that the underlying commodity is technology and thus so dynamic you can't define a decent benchmark for it. There is a continual deprecation process at work plus the problem of technology step-changes that on short notice might permanently disrupt the entire market. DRAM manufacturers would probably be better served trying to find proxy-hedges in the equities markets or somewhere else.
Posted by Nils Blutig at September 17, 2002 10:42 PM | TrackBack