Commodity markets for the 21st century April 24, 2009
Posted by James Webster in : finance, gadgets, development , trackbackRecently there were a few news stories relating to potential shortages of NAND flash RAM due to Apple ordering a large volume of chips from Samsung (rumoured to be heading to a 32Gb iPhone).
From the DigiTimes article:
Downstream memory suppliers are striving to grab more NAND flash chips to meet substantial volumes of short lead-time orders from device makers
Clearly there is extensive risk for flash memory suppliers and OEMs in this market. I was wondering if an active futures market for buyers and sellers of memory would be feasible. A bit of further research uncovered DRAMeXchange which provides market data for various memory products (similar to Platts, one of the major market data providers for the energy market). Could they go a step further and set up a liquid market for buying and selling risk around memory supply? Obviously not all chips are created equal; the contract you use to hedge your risk may be for delivery of a particular type of flash memory that is slightly different to the one that you will be purchasing in the physical market. In financial/commodities markets this is known as basis risk and is understood to be something that needs to be monitored when hedging a risk exposure.
I also came across Zerobeta’s blog recently (via Park Paradigm). An interesting insight into Ztail (TechCrunch):
Whats interesting is if Ztail is selling naked puts on spot ipods, they have a synthetic position that is short the call option and long the futures. While I understand that they don’t really have that position, but it is an interesting position nonetheless and their guaranteed price should be a good (albeit hopefully low) gauge on what a 1 year out ipod futures are trading at.
Another interesting article over at Ars Technica, Why high-performance computing needs financial engineering:
If (Richard Bookstaber) is correct with his recently floated hypothesis that “the days for high frequency trading are numbered,” then this would be pretty bad news for Intel, AMD/ATI, and NVIDIA.
As biotech moves further from research to commercialisation it is the obvious candidate to pick up the slack in HPC demand if high-frequency trading does in fact fall off (a forecast I disagree with). Could a GPGPU-powered compute grid help a team win the Archon Genomics XPrize?
Finally, the source code for the JPMorgan CDS model (which I discussed here) was released and is available at cdsmodel.com.
Comments»
no comments yet - be the first?