Park that IPO, let’s have an NSE initial penance offer
It’s hard to believe that the absence of a simple random-number algorithm has thrown a question mark around almost one-third of NSE’s revenue
Singapore: Try this at home. Anywhere on a Microsoft Excel spreadsheet, type =RAND(). You’ll see a decimal number between 0 and 1. Close the worksheet and reopen it a second later, and you’ll see a different figure.
It’s hard to believe that the absence of a simple random-number algorithm has thrown a question mark around almost one-third of National Stock Exchange of India Ltd’s revenue.
The chairman of India’s No. 1 bourse is confident of taking it public before the end of this year. Yes, the NSE needs to give existing investors a chance to exit. But the awful gaps in systems, processes—and most importantly, ethical behaviour—revealed by a forensic study by Deloitte Touche Tohmatsu, mean that offering penance (and penalties) would be more helpful before the exchange starts its $1.5 billion IPO.
Late last year, just as the NSE was getting ready to file its draft prospectus, Gadfly advised the exchange to drop the hubris it carried with its lineage. As a government-sponsored alternative to the broker-owned casinos that dotted the capital-markets landscape until the 1980s, the NSE has been a tremendous success, consistently trading four to six times as much equity as its rival, the much older Bombay Stock Exchange, or BSE Ltd.
In recent years, though, the combination of proximity to government and market leadership has led to swollen heads and bad behaviour. Both were on display in co-location, which allows traders to place their computers close to the exchange’s so they can pare microseconds when executing trades. NSE offered the service to high-frequency traders. But not fairly.
The exchange didn’t use something like RAND() to grant fair access to its tick-by-tick system. Those who connected to servers first saw securities prices sooner. It took the Deloitte analysis to discover that the development team was using a “randomizer” for servers other than those handling real-time price data. In other words, the exchange knew it was giving an unfair advantage to some traders, and profited from it.
Then there was “Server 27”. It was a backup, which worked parallel to the primary server. Clients were told not to connect to it except in an emergency. But not everyone was asked to lay off Server 27, and not all the time.
Co-location services account for 30% of the exchange’s revenue, which for now is going into an escrow account set up at the regulator’s behest. NSE chief executive officer Chitra Ramkrishna resigned last December. Vice chairman Ravi Narain, a co-founder, left this month.
The old method of disseminating price data has been junked. Still, it’ll take a more decisive round of spring cleaning by incoming CEO Vikram Limaye to fix the culture.
As for cozy relationships with the authorities, the Securities and Exchange Board of India, the market regulator, has a new chairman. Ajay Tyagi has no incentive to go easy on the NSE. Any preferential treatment would be rightly seen as repeating the behaviour that got the exchange into this mess.
Looking at the BSE, which listed in February and has given IPO investors 36% returns to date, it’s clear that the NSE has caused a serious opportunity loss to stakeholders including Morgan Stanley, Goldman Sachs Group Inc., Citigroup Inc., Temasek Holdings Pte, and Tiger Global Management LLC.
Still, the share sale by investors can wait. First, the NSE should make haste with a penance offering. Bloomberg Gadfly