Exclusive | NSE co-location audit: 3 brokers misused secondary server and ignored exchange warning

In its submission on May 28, 2018, OPG said that it was facing various issues while connecting with the server. The Delhi-based company claimed that there were around 35,000 disconnections to the primary server over 365 days between 2012 and May 2014 (approximately 98 disconnections per day). They connected to the secondary server as a contingency measure to minimize the risks faced due to unforeseen disruptions in the Tick by Tick (TBT).

Also read: What is the NSE co-location case?

OPG used the algorithm provided by “Omnesys” and “Greeksoft” to perform trading through their NSE co-location. However, former Chief Technology Officer Ravi Apte, said in a statement, “[On May 2, 2018] I was on the board of Omnesys for a brief period representing NSE. There could be an advantage of a few microseconds in logging in first/early to the TBT server.”

In July 2010, OPG entered into a contract with “Algotech” to perform arbitrage trading using OPG’s membership and capital Algotech in 2010 was using the platform developed by Chanakya. The wife of Suprabhat Lala (senior NSE official) and sister in law of Ajay Shah and Krishna Dagil are directors in Chanakya.

Moneycontrol has a copy of Deloitte’s forensic report on the three brokers.

When contacted over phone, OPG said that company has not anything wrong.

"We were not aware of any benefit of using the secondary server which we have stated in the statement also. Since the report is very technical we will that we will try to convince that regulator that we did not gain anything out of shifting to the secondary server. We have not done anything wrong in this case".

SMC Global

In its forensic audit, Deloitte found that SMC global had connected to the secondary server even after several reminders from the exchange to avoid doing so. As per the NSE circular, the secondary server was to be used only when a primary server is not available. In the audit, Deloitte found that between 2010 and 2015, SMC had connected to the secondary server on 389 days in the Future and Options segment.

The report said, “Out of the above 389 days that SMC connected to the secondary server, we observed that it had made complaints to COLO (co-location) support on 210 days. However, for the remaining 179 days, on which SMC was connected to the secondary server we did not observe any complaint made to the COLO support or the exchange. For 210 days that SMC had connected to the secondary server, it made complaints only on 64 days to COLO support about connectivity or network issues.”

As per the forensic audit, the NSE COLO Support team stated that SMC had connected to the secondary server, and despite several reminders, SMC continued to use this server instead of the prescribed one. SMC failed to provide any reason behind connecting to the secondary server without a corresponding complaint regarding connectivity issues.

"We used secondary server when the primary server was not working properly. And, whenever we were asked to shift to the primary server we immediately shifted. We had made complaints to NSE colo support system many times that the primary server was not working properly," said an SMC Global senior official.

"We have told our version to auditors, now we will try to convince the regulator also," he said.

He also clarified the NSE disciplinary committee has already cleared this matter. After the Deloitte report, the regulator has not pointed out anything adverse to us and we are not part of the show cause notices issued by the regulator.

Barclays Securities

Barclays installed its infrastructure and went on the NSE COLO in October 2010. However, Barclays said it never received any guidelines from NSE that restricted it for connecting to the secondary server, except for an email dated December 2, 2011, from NSE Co-location Support which stated that the members are not supposed to connect to the secondary server until the exchange asked them to do so.

Barclays used the TBT segment and used a secondary server for Future and option and the cash market. Deloitte’s forensic audit of the UK-based broking house revealed that Barclays connected for 35 days on the server in the F&O segment and 36 days in the cash market segment.

Barclays in its response dated May 10, 2018, said that their IT team was responsible for switching between the primary and the secondary server. It was not aware of the reason why its IT team did not use the primary server despite multiple follow-ups from NSE. The IT team was transferred to an entity named Squarepoint Capital around December 2015.

On September 8, 2017, SEBI had ordered a forensic audit of the brokers that used the co-location facility.

“This is not material that Barclays can comment on. With regard to any regulatory matters, it is our policy to co-operate fully with our regulators, and render such assistance as we are able to provide,” Barclays Securities said in an email response to queries sent by Moneycontrol.