With broker-dealers continually seeking new methods of enhancing efficiency and reducing costs, Evgeny Sorokin, chief product officer at Devexperts delves into the role that automation can play and how best to approach emerging technologies.
With broker-dealers seeking ways to enhance efficiency and reduce costs, many are turning to automation. Unfortunately, broker-dealers can’t trust artificial intelligence (AI) yet with crucial aspects like compliance and risk management. Recent Large Language Models (LLM) achievements show remarkable results in summarising existing information and recognising patterns in mixed content like recorded sound, images, and video. However, decision-making processes that are a crucial part of compliance and risk management are far from being fully automated by the usage of the LLM-like AI. In particular, so-called AI hallucinations are a noticeable side effect of the LLM application in different spheres of IT and automation.
Fortunately, there are multiple ways to automate these business areas with good old automation software crafted specifically for its purpose, time-tested, and configurable to suit each brokerage’s needs.
Limiting risk exposure
Risks emerge when clients sell options, short stocks, and take advantage of a leverage provided by a broker. Different broker-dealers have their own criteria for measuring an acceptable level of risk the client takes intraday and whether to limit it or not. Say, if a client exceeds a safe number of orders daily, their orders should be alerted or sent to a risk desk for review before being further routed to execution. Other limiting parameters can be the total used margin, daily unrealised loss, or the total size of derivative products’ position in the client’s portfolio. Regulators oblige broker-dealers to monitor client risk if they offer derivatives trading. However, broker-dealers themselves may put additional controls in place to avoid situations when their clients can’t cover losses due to a lack of funds.
In this scenario, risk exposure can be automated to detect such clients and their risky operations. Broker-dealers can fine-tune risk management to assign risk profiles or bespoke risk parameters on the individual account or account group level. At the same time, the software will not distract the risk desk department with clients’ orders that do not increase the risk exposure or are below the accepted risk level.
Post-trade operations
When it comes to operations, involving corporate actions can be entirely automated without the need for interruptions from a broker-dealer’s staff. In the case of a stock split and reverse stock split, merger, or acquisition, clients’ portfolios can be automatically updated with the new number of equities they’re eligible for. Dividends can also be automatically represented as cash transactions or as an equity increase.
As regards reconciliation, we’ve learned through our work with brokers that they spend too much time on manual reconciliation due to legacy systems and missing integrations between them. To keep up with growing trading volumes and the number of end users, as well as the recent implementation of the T+1 settlement rule, brokers should automate reconciliation with the latest technology.
The perfectly automated process looks the following way: a broker’s trading platform is integrated with their venue of choice for custody, clearing, and settlement. The tool syncs transaction data in the client’s trading platform with the venue’s “Start of Day” (SOD) files. If there’s a mismatch between the broker’s data and SOD files, the broker’s trading officers use a specially developed web UI to approve adjustments in the trading platform database. The process might be further automated by automatically adjusting the trading platform records and intrade syncs. The tool supports individual and bulk corrections. Thus, a broker’s staff doesn’t have to search for mismatches and update them manually in something as outdated as an Excel file.
In addition, brokers might get a substantial competitive advantage if they implement a tool for clients that notifies them about all prospective commissions and fees they’ll have to pay before sending orders for execution.
Order routing is another key factor. Smart and thoroughly thought-through order routing poses no risk and only brings gains to broker-dealers. While being aligned with the best execution rule, a broker-dealer has the right to route client orders not only to public exchanges but also to dark pools and alternative trading systems (ATSs).
What’s different about sending orders to dark pools and ATSs is that broker-dealers can save on execution fees and get paid for order flow.
Why automation is required here — dark pools and ATSs prefer orders manually issued by traders and not algorithmically generated orders, let alone pay for them. Intelligent order routing allows us to define execution destinations per order/account according to this aspect.
Besides this basic filtering, there are four automation scripts for intelligent order routing according to: traded volume, percentage of orders sent to a particular venue, user groups, primary and secondary destinations, routing time, depending on trading hours.
The bottom line
Automation undoubtedly improves brokers’ operations. It makes their processes more efficient, reduces risks, and improves overall performance. By using automation tools in important areas like risk management, post-trade operations, reconciliation, corporate actions, commissions, and order routing, broker-dealers can make their work easier and gain an advantage in the industry.
As automation changes, broker-dealers should carefully consider each automation solution’s benefits and possible impacts. This careful approach helps broker-dealers make good decisions that meet their goals and the rules. Ultimately, using automation well can help broker-dealers succeed in a changing and competitive market.