The central securities depository of Mexico is applying a resource allocation model widely used in pricing airline tickets to settle trades more quickly and efficiently.
By using a resource allocation-centered approach, a novel one for the banking industry, the Indeval central securities depository has saved Mexican banks more than US$240 million in interest fees over a period of 18 months, according to the organization.
To aid in the exchange of securities, Indeval uses Cplex Optimizers, a mathematical library for resource allocation, offered by IBM in its ILOG mathematical programming software. Indeval incorporated ILOG into an upgrade of its securities settlement system, called Dali, that was completed in 2008.
"This is a very novel usage of this technology," said Pierre Haren, IBM vice president and former CEO of ILOG, which IBM acquired in 2008.
Dali settles over $250 million in transactions daily. Every two minutes, the system evaluates all pending trades. It matches trades so that only the net amounts of securities and cash are needed to be exchanged between participants. By executing rapid trade settlement, as this process is called, the depository has reduced the amount of cash Indeval needs to cover trades by 51 percent.
"The optimization engine minimizes the encumbrance you have on securities, and that makes it less expensive to operate," said Jaime Villase
Much like the Federal Reserve System acts a central bank for the U.S., Indeval manages securities on behalf of Mexican banks and individual investors, acting as a hub for executing national and international trades.
Formerly, Indeval would execute trades on a trade-by-trade basis, or reconcile transactions at the end of the day. Dali is the first settlement system to operate on a minute-by-minute basis. "No one before us had attempted to do this in the securities industry," Villase
Resource allocation balances the supply of a given item against the demand for that item, by using a set of equations that define the exchange process. The ILOG Cplex library is widely used in supply chain management, large-scale manufacturing, electric grid optimization, airline ticket pricing and other industrial-sized resource allocation problems.
In Indeval's case, the software is "figuring out the minimum exchange of cash that satisfies everyone," Haren said.