02.19.2014

Sell Side Sizes Up Risk

02.19.2014
Terry Flanagan

Sell side financial institutions face stiff fines and potential reputation risk for failing to adhere to prudent risk controls, as the Knight Capital episode clearly illustrated. The market maker was left close to bankruptcy in 2012 following an algorithmic error that caused erratic trading activity and left the firm with billions of dollars of unwanted securities. The company estimated that it lost $440 million because of the software glitch.

“Sell side firms need time-varying risk limits,” said the head of trading at a large buy-side institution. “Knight Capital probably hit its risk limits of within 15 minutes of their total trading on day, but if they had risk limits that increment throughout the day, they would have lost a lot less money.”

The SEC’s rule 15c3-5, known as the Market Access Rule, requires broker-dealers to have systematic risk management controls to prevent trading errors and the entry of orders that exceed preset credit limits, already adequately addresses risks from customer orders.

“Most of the sell-side firms that I’ve dealt with are adequately prepared,” said Walter Ferstand, regulatory and compliance subject matter expert at compliance systems provider NICE Actimize. “Those firms that didn’t show an appreciation when 15c3-5 came out are getting fined by Finra.”

Financial institutions are subject to increased scrutiny and heavy fines by regulatory bodies and courts. In addition to the need to comply with external regulations, firms’ own internal policies plus the need to supervise activities across the organization, identify issues and record actions taken in a timely fashion. Firms need compliance capabilities that detect, prevent and deter non-compliant activities today, and offer the flexibility to meet future regulatory compliance needs.

“Among the more sophisticated buy side firms, most have thought about what’s going to jeopardize their own capital and have been methodical and through in what they’ve implemented, but the brokers have this mentality that it’s not their capital,” said the head of trading. “As long as they have the credit with the clearinghouses to settle their trades, most brokers have unlimited trading accounts because they have so much capital.”

The trader’s firm has installed different types of limits and redundant risk checks throughout its system to protect clients and itself. “If there’s any type of irregularity, it needs to be stopped before it hurts our capital,” the trader said. “The vast majority of our net worth’s are in the firm, so we are the clients as well, because we own significant portions of the capital. We have redundant internal risk checks and independent third parties, which a lot of companies don’t have.”

As a compliance subject matter expert, Ferstand works closely with clients and prospects to evaluate, educate, and advise on trading compliance requirements, ensuring each firm has the solutions to meet their specific business needs. He specializes in cross-asset electronic and DMA trading requirements, including Reg NMS 605 and 606, OATS, TRACE, TRF, and MSRB Compliance Reporting.

“You have to look at both the spirit and the letter of the rule,” he said. “You need to be able to differentiate between what actives are permitted and what are not, such as between legitimate market making and proprietary trading. You need to have this take place in real time; you can’t do that in T+1.”

Related articles

  1. Algos, Post-Trade Top FCM Concerns

    TT Splice provides industry-first functionality for synthetic multi-leg spread trading.

  2. Algorithmic Trading Broadens Appea
    Daily Email Feature

    Trading Smarter With Algo Wheels

    Modern wheels can incorporate many different data points.

  3. Asset managers leave money on the table when using VWAP algos for low-urgency orders.

  4. The firm is leveraging its newly acquired quantitative trading expertise to generate new client algorithms.

  5. Congress Unlikely to Act on HFT

    The algo provides an alternative to VWAP for minimizing implementation shortfall.