Where are Compliance teams left in a new era of Trade Surveillance automation?
The Rising Cost of Financial Crime
The cost of financial crime of firms is as much seen by the figures spent on Compliance and Trade Surveillance technology as it is by the plethora of fines levied on firms by regulatory bodies. Having the necessary technology in place to avoid the reputational damage associated with investigations and fines is a necessity, and the increased spending on compliance related technology is justifiable.
To put into context, recent figures published from BAE systems, shown below, indicate that nearly half of all respondents predicted increased compliance spending over the next 12 months.
The Trend towards ‘Big Data’
As Trade Surveillance spending is increasing, the range of software available is becoming progressively diversified. This diversification can be partially linked to the trend of ‘Big Data’. The problematic amount of data available to compliance teams is recognised by NICE, a global supplier of trade surveillance technology to a number of Tier 1 & 2 Banks. NICE frame their work with clients within the concept of manipulating “relevant” data. But, what is determined to be relevant data, and what qualifies as an exception to raise?
To explore this, we have to look at how alert thresholds are determined today.
There is an initial testing period which involves the trial of different thresholds; and the subsequent correspondence between analyst and data program provider. This process can be characterised as follows:
Places the duty of responsibility for the use of the program on the in-house team.
Requires relevant experience of setting suitable thresholds to identify instances of suspicious trading.
Reliance on user initiative to recognise any drawbacks to current limits, and update accordingly.
Trade Surveillance cannot evolve with market data alone.
The future of Trade Surveillance lies in creating a holistic model for which, there needs to be an insight into the intent of rogue traders. This places increasing importance on the surveillance of Audio and Electronic Communication (AComms and EComms).
Looking at how the concept of ‘big data’ will be filtered down to ‘relevant data’ within AComm/ EComm surveillance, the importance is placed on the Natural Language Processing (NLP) software, which aims to automate part of the process. NLP will aim to provide context to the data surveyed, via automated interpretation of language to construct both a market and original trader profile. With the application of NLP to AComm/ EComm surveillance, there is an increased potential for a reduction in the number of false positives that are generated- as anything escalated will contain key words or phrases that can be used as relevant evidence to explore any suspicions of abusive trading practices.
The integration of this kind of technology into the realm of Trade Surveillance returns us to the original dilemma.
Even with Automation, there are still clear benefits to human intervention.
What must be remembered is that NLP technology will need constant refinement. Simplistic terms and phrases can be highlighted, but name detection and hints at changing a communication venue present a different challenge, which may not be wholly perfected. This means that, in part, NLP technology will only be as diligent as the employees who operate it. Whilst NLP software can be used to produce relevant samples that need to be reviewed, it will only be used as part of the evidence for escalation of alerts. This again cements the status of the analyst as they will have to combine this data with the market movements that have been observed.
There is strong evidence in favour of the ongoing engagement of compliance professionals in the realm of trade surveillance. At FinTrU, we provide analysts who have experience of Trade Surveillance drawn from utilisation of sources which are both on and off-market. This means that, with the potential move into a holistic approach to trade surveillance, we are well placed to provide a service complete with professionals who are extremely capable of integrating new data sources into their trade surveillance process, as well as implementing steps to improve existing processes. This can be offered across an increasing number of markets, including (but not restricted to) foreign exchange; commodities; cash bonds and equities.
Conor joined FinTrU in 2014. Conor has since worked as part of the Non Market Risk Team within the Fixed Income Division of a Tier-1 Investment Bank.
Conor is a BA Hons History graduate from The University of Manchester. Keen to develop a career in financial services, Conor subsequently come on board with FinTrU through our original graduate academy intake.
The FinTrU Financial Services academy is a bespoke programme, comprising intensive training in financial markets, and financial regulation with a focus on data analysis and management.
Within his current role as part of the Non Market Risk Team, Conor operates as part of the Fixed Income Trade Surveillance team. More specifically, he focuses on the Foreign Exchange market. Conor is responsible for the oversight of EMEA FX trading alerts generated by firm traders, and the escalation of valid alerts to supervisors. This is done across multiple, industry-standard platforms.
Alongside this, Conor’s involvement within the wider Trade Surveillance team at FinTrU has increased his exposure to the management of FX Front Running alerts and the Futures market, relating to alerts caused by potential instances of washed trades; fictitious orders and ramping globally.