SBI Liquidity Market in Japan provides online services for foreign currency exchange (FX) trading. The company has used Microsoft SQL Server software for several years, during which time the company has increased throughput from 35,000 to 200,000
transactions per second. To handle a volume of transactions that continues to grow and to support expansion worldwide, the company needed a faster, more responsive trading system. To fulfill that need, it decided to become an early adopter of Microsoft SQL
Server 2014 In-Memory OLTP. SBI Liquidity Market is now achieving better scalability and easier management with the built-in feature and expects to strengthen its competitive advantage with a trading platform that is ready to take on the global marketplace.
||We decided it was necessary to use SQL Server 2014 in-memory OLTP to manage sales aggregation as a plan with future vision, especially if you consider the exponential growth in the number of transactions that will accompany future
| Katsuyoshi Miwaki
Part of SBI Holdings, a financial services conglomerate in Japan, SBI Liquidity Market Company provides online services for commodity traders such as banks and securities firms. Based in Tokyo, SBI Liquidity Market is a global leader in the foreign exchange
(FX) market. With monthly transaction reaching approximately ¥101 trillion (US$1.04 trillion)—more than the Japanese government’s annual—the company needed to improve the performance, availability, and reliability of its transaction system.
Since its founding in 2008, SBI Liquidity Market has based its mission-critical FX trading system on a Microsoft platform. Its first system included Microsoft SQL Server 2005 software, the Windows Server 2003 operating system, and the Microsoft .NET Framework.
At the time, the solution’s processing capability was sufficient to handle throughput of 35,000 transactions per second. But as the trade in foreign currencies continued to expand throughout the region, SBI Liquidity Market realized it needed to upgrade.
The company wanted to be ready for a future in global FX trading and needed a platform that could bring together traders from multiple countries. To succeed, the trading system would need to handle an increasing volume of transactions and provide near-real-time
“FX trading is expected to grow significantly in the future in Asia,” says Tatsuo Shigemitsu, CEO of SBI Liquidity Market. “When that happens, the number of transactions will increase by up to 30 to 40 times as well. We needed to immediately develop faster
and more available transaction systems that could provide strong performance and the transactional accuracy needed for that scale.”
The mission-critical FX system handles operations called cover trades, which connect customer orders to the interbank market. Dealers cover trades on the basis of aggregated results of sales data using this system, and if that aggregation process takes time
due to transaction bottlenecks, for example, then latencies can arise. As a result, SBI Liquidity Market needed to significantly speed the aggregation of transaction data, minimize latencies, and improve the precision of the aggregation results.
To improve transaction processing and customer service, SBI Liquidity Market decided to become an early adopter of an in-memory OLTP solution available with SQL Server 2014.
SBI Liquidity Market has worked extensively with CSK WinTechnology Corporation, which provided support for building and running the trading system. In June 2012, Yukio Kumazawa, Executive Technical Fellow at CSK WinTechnology, attended a conference at Microsoft
headquarters in Redmond, Washington. When Mr. Kumazawa learned about SQL Server 2014, he realized that SBI Liquidity Market could benefit from OLTP in-memory technology. “SBI Liquidity Market immediately came to mind when I heard about the in-memory OLTP solution
in SQL Server 2014.” Mr. Kumazawa says. “One of the challenges faced was achieving good performance for tables with the largest transaction processing load, the ones that enabled seeing the status of foreign exchange trades. I consulted with one of our managers
and decided to recommend the in-memory OLTP solution from Microsoft.”
Cover trading systems have to be able to process inserts and updates very quickly. Conventional database implementations have long suffered from locking and latching problems. SBI Liquidity Market and CSK WinTechnology believed that the powerful updating
function of the in-memory OLTP functionality would virtually eliminate those problems.
Yuta Yoshikawa, Manager, Systems Development Section, IT Division at SBI Liquidity Market, notes, “It wasn't the first time someone had suggested the in-memory OLTP function in connection with increasing the processing speed for large amounts of transactions.
However, introducing another company's in-memory database into our system, which we had based on SQL Server, presented a lot of problems, so the idea was put to one side. But when we heard that it would be offered with SQL Server 2014, we thought it would
be worth a shot.”
The company wanted to continue using existing Microsoft product assets and expertise. In addition, because the OLTP solution was built in to SQL Server 2014, SBI Liquidity knew it would be easy to manage.
Testing began in October 2012. By the end of January 2013, the verification process was complete, and the new trading system was deemed a success. The team used operational data, and the results were in line with expectations. According to Kazuo Tsutsumi,
Applications Specialist at SBI Liquidity Market, who took part in the verification, “Conventional databases tend to have longer and longer processing times as the number of items to process grows. With the in-memory OLTP solution, however, that tendency was
nowhere to be seen. The processing time remained the same, no matter how much had to be processed, and the proof is clear as day in the numbers. The results of the verification were extremely heartening in terms of what you can expect in a real-life situation.”
Yuta Yoshikawa, Manager of IT Division at SBI Liquidity Market, says that after transitioning to a production environment, the company expects to achieve similar results. He says, “During the introduction phase of the trading system, our thinking was that
it had to be able to process 5,000 orders per second, which was the initial target during testing. That said, looking at the verification results, we hoped that it would be able to process that many orders in milliseconds or even microseconds, and not one
second. Naturally the actual production system had to accommodate not just OLTP transactions, but other peripheral application processes. We want to work it out so we can maximize the potential of using SQL Server 2014 in-memory OLTP in a real-life environment,
while still maintaining the overall balance of the system.”
Currently, SBI Liquidity Market can process a maximum of 200,000 transactions per second. The company plans to start using the new in-memory OLTP solution in production by the end of 2013 when it has finished determining optimal server and storage configurations.
SBI Liquidity Market will be the first company to use SQL Server 2014 in-memory OLTP in Japan. The company is also looking at additional functionality, including the SQL Server 2014 AlwaysOn high-availability feature, and In-memory columnstore for data warehouse
which is expected to improve performance as well.
||By working with Microsoft to provide a trading system based on SQL Server 2014 that can be used easily by anyone anywhere in the world ... we can expand the FX market in a healthy and ever-growing fashion.
| Tatsuo Shigemitsu
SBI Liquidity Market
With a trading system based on SQL Server 2014 in-memory OLTP, SBI Liquidity is gaining faster performance, better scalability, and improved support for global expansion.
Speeds Performance by Tenfold
The enhanced trading system is expected to improve service to end users. “SQL Server 2014 in-memory OLTP is the perfect tool for a developer looking to speed up in-memory database processes,” says Kazuo Tsutsumi, Appliacation Specialist at SBI Liquidity
Market. “More than anything else, we have to proactively respond to the wishes of customers who trade using the system. We expect to increase overall performance of the system tenfold in the future, which means that at the same time we have the opportunity
to incorporate the wishes of our customers tenfold as well. The future looks bright!”
SBI Liquidity Market can now process huge volumes of information quickly and accurately, and keep pace with growth as well. “The important question was how to put together a system that can process massive amounts of transactions quickly and accurately,”
says Katsuyoshi Miwaki, Manager, CSK WinTechnology. “It's not that we couldn't performance tune a conventional SQL Server database, but we decided it was necessary to use SQL Server 2014 in-memory OLTP to manage sales aggregation as a plan with future vision,
especially if you consider the exponential growth in the number of transactions that will accompany future growth.”
In addition to gaining better performance and scalability, SBI Liquidity Market expects to reduce operational costs by taking advantage of the built-in OLTP feature.
Strengthens Global Competitive Advantage
By taking advantage of next-generation database technologies, the company is gaining better support for worldwide expansion. “We believe that entering the global market, with a focus on Asia, is a very important future challenge for us,” says Tatsuo Shigemitsu,
CEO of SBI Liquidity Market. “If FX demand continues to grow from the domestic market to countries in Asia, and more people in more countries start trading, there will eventually be a demand for an environment that allows trading any time and in every country.
By working with Microsoft to provide a trading system based on SQL Server 2014 that can be used easily by anyone anywhere in the world and that offers trading opportunities that are fair for professional dealers and private investors alike, we can expand the
FX market in a healthy and ever-growing fashion.”
For More Information
For more information about Microsoft products and services, call the Microsoft Sales Information Center at
In Canada, call the Microsoft Canada Information Centre at
Customers in the United States and Canada who are deaf or hard-of-hearing can reach Microsoft text telephone (TTY/TDD) services at
Outside the 50 United States and Canada, please contact your local Microsoft subsidiary. To access information using the World Wide Web, go to:
For more information about SBI Liquidity Market products and services, call or visit the website at: www.sbilm.co.jp
For more information about CSK WinTechnology products and services, call (81) (3)33432503 or visit the website at:
This case study is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.