Front to Back – new technologies arriving from emerging markets

25 october 2010

International Securities Services: SIBOS 2010


It was relatively recently that Russia with its distinctive features of a significant value became part of the world financial market. Trading shares and bonds on the MICEX Stock Exchange, the main Russian stock market, still requires 100% pre-funding for each trade.

  • Starting from this point, the method of strict pre-trade risk control has been evolving as follows:
  • The advent of external automatic brokerage systems (similar to QUIK) gave rise to the demand for integral control of risks to a trader or a client for all trading venues,
  • Once the brokerage systems had become capable of providing access to foreign markets, a requirement arose to control overall client risks with consideration for T+0 and T+n payments in different currencies,
  • Considering the development of margin trading in Russia, the requirements applicable to the functional capabilities of the pre-trade control over client operations in brokerage systems should be even higher.

As a result of that development, the Russian broker platforms include an inbuilt highly developed risk manager functionality featuring complex models of position and risk evaluation.

At the same time, the development of both exchange and OTC trading platforms, on the one hand, and algorithmic trading, on the other hand, results in a growing demand for high frequency and low latency features in operation of brokerage systems such as QUIK.

The developers of broker trading systems feel themselves between a rock and a hard place: they are expected to offer high-performance solutions and, at the same time, each transaction entails complex position and risk calculations accounting for delays in the system operation. Such a development brings quite impressive results some of which are listed below:

  • Pre-trade control time per order: 1-2 milliseconds,
  • Overall transaction execution time including pre-trade control: 7 milliseconds for MICEX Stock Exchange and 8-10 milliseconds for FORTS,
  • Number of orders per day processed by a QUIK server for a major Russian broker: circa 2.5 million (number of trades: 300 – 350 thousand).

International trends of introducing standard algorithms, such as VWOP, TWOP or Iceberg, into brokerage systems are apparent in the development of Russian software platforms as well.


Another part of the present-day requirements to technologies used to carry out proprietary and client operations in financial markets is to ensure the end-to-end automation of all processes involved. Quite often, high frequency and low latency are used to describe trading operations proper, whereas the issues of their accounting and affecting position or integral risks, in the online mode and for each transaction, are tackled to a lesser extent.

An intensive development effort in recent years has been focused on risk assessment and prevention software and technologies. It is no seldom that exchanges, e.g. Shenzhen Stock Exchange or FORTS, act as integrators in providing services based on the above solutions.

Program complex midQORT offered by ARQA Technologies is one of such solutions. Middle-office software packages do not have any ‘national’ specificity, so they are generalized from this viewpoint. At the same time, there is no generality in implementation because each implementation suggests a substantial customization. Therefore, it would be impossible to get a "fully packaged" solution as far as middle-office software is concerned. Quite often, a customization adds some unconventional functionality to the system. For example, a module of company/bank liquidity management has been implemented in midQORT.

In the development of such software a new model of customer relations emerges: it is not the vendor that starts a negotiation speaking about how fine is the solution offered, but the client tells instead about problems that need a solution. The vendor listens attentively and thinks which combination of available modules and solutions should be use to reach the client's objectives with the maximum degree of accuracy. Opposite to ‘packaged’ solutions, this approach suggests fundamentally different techniques of software promotion and a new system of ‘client-vendor’ relationships in which the vendor becomes actually a part of the client's technology unit. Such a model supposes that the vendor should have sufficient resources for development and support of heavily customized products.


Solutions intended for the back office of financial institutions, like banks and investment companies, are somewhat far from the intensively discussed issues of software package operation speed. At the same time, technologies traditionally used in that area become inefficient because of the growing number of accounts and transactions, complex schedules of accounting commission fees and other charges, and gradually tightened requirements of regulatory authorities to the periodicity and content of statutory reports.

A sign of imminent changes is that back offices face inability to process the results of daily operations in a reasonable time. New generation program complexes for back office operations that enable accounting numerous transactions for a considerable number of end users in reasonable time should become a main topic of exhibitions, conferences and publications dedicated to financial accounting software. Such complexes are already available in the market. And still, the short supply of such systems keenly felt in Russia, Ukraine and other East-European countries hinders the development of DMA services for the market participants.

The back office software starts dealing with the tasks traditionally not attributable to it in the online mode, not at the end of a business day. Online tasks appear with the development of retail business, when clients want to receive from the brokerage system the reports of operations carried out for an arbitrary period of time as well as preliminary estimates of commission and tax deductions for the operations being planned.

Полная версия статьи...

Back to list
To top