With millions of dollars being spent on IT infrastructure, hardware and software the world’s leading financial institutions should’ve got the whole IT paradigm right, it seems not. Japan’s largest Stock Exchange, TSE, suffered a major computer glitch earlier today after an important router was broken. The system fault meant traders were in panic for 95 minutes, as a result volumes were down 20%. The exchange has an automatic back up router, however it failed to start.
System faults have been the cause of a potential bankruptcy for one of the world’s largest equity market makers, last week Knight Group had suffered a computer failure which resulted in a massive $440 million loss.
The Tokyo Stock Exchange has suffered a second fault in its system during the last 6 months, earlier this year it had a breakdown in its data distribution system although the system underwent repairs late 2011. The fault occurred at the time when companies were reporting their third quarter earnings thus impacting trade volumes for the day.
Japan’s TOPIX index is traded on the TSE, smaller rival Osaka Stock Exchange which primarily trades derivatives took a large share of today’s trading volume. The TSE has an average daily trade volume of $19 billion.
The TOPIX is trading at 743.70.
Computer glitches have been common contributors to price discrepancies in global financial markets. In May 2010 NYSE had cancelled trades in 286 securities that had rapid movements of 60%. Another incident was recorded on the London Stock Exchange when the new Linux-based Millennium Exchange went live and affected pricing on instruments.
Liquidity in FX markets help the market to trade efficiently however an erroneous trade placed at Rabobank on 12th July 2010 caused a 1% move in GBP USD, the Rabobank trades caused sterling to drop from 1.53 to 1.518 against the dollar in a short period of time.
According to Forex Magnates Reports, Japan is the world’s largest FX market.
NASDAQ openly admitted during the FaceBook IPO software glitch that their trading system stressed at 2500 trades per second & we all know the [colocated] hft bot army is capable of hitting this rate at will..
As well as being the best funded area, the market also provide the simplest of environments. It’s a set of predefined stocks going up & down in value. Where is the complexity in that?
There are three types of Banking glitch:
1. Hardware glitch re this new one – turn it off & then on again.
2. Legacy glitch as code patched up & rescued from the millennium bug piggy backs Moore’s Law to claim efficiency. COBOL ate my mortgage. The fact that the new off-shore coder misunderstands it is not something the old western coder should be shouting about.
3. The modern top-down solution glitch. As in NASDAQ BATS trading systems. Where the simple obvious efficient unobtrusive solution is dismissed for the more professionally relevant derivative approach to buying a pint of milk. It’s like it’s all designed to fail or something.. @thepoettrap