Ramin please tell us a little about what your day to day job and responsibilities at DWS involve.
I am a Senior Trader at DWS where my responsibilities include trading FX and Rates products along with their derivatives. Furthermore, I focus on the automation of trading workflows and the data driven optimization of existing processes. A key part of my role hence revolves around monitoring and analysis of our trading activities leveraging both internal data and insights from external TCA providers. My work aims for continuous improvements in execution quality and a state of operational excellence which allows us to provide the best possible trading service for our investors.
How would you describe the key objectives and guiding principles of your trading desk and the dealing activities it undertakes?
The overarching principle of our trading desk is to fulfill the best execution principles. We are a traditional buy-side execution desk, hence operating in an agency model which makes our primary focus ensuring that trades are executed in the most efficient manner and at the best possible price. We strongly invest in automating workflows wherever human involvement does not provide a clear value-add which reduces operational risks, streamlines processes, improves resilience and decreases time to market. To enhance our decision making and therefore our trading outcomes for both manual and automated workflows, we utilize internal and external data sources. This ensures that our clients benefit from consistently improving trading platform.
Top class trading practitioners like yourself are now expected to have a thorough understanding of macroeconomics, market structures and securities pricing coupled with knowledge of data analytics, statistics and machine learning. Do you think the more traditional skillsets required of the job such as technical trading acumen and an ability to manage liquidity relationships are likely to remain equally important for the foreseeable future?
I think that traditional skill sets, especially knowledge about market structure will remain highly relevant and will most probably become even more critical in the foreseeable future. An in-depth understanding of how markets operate is a crucial skill to have to be able to navigate markets. Managing relationships to trading counterparties also continues to be essential, as especially during periods of high market volatility or other stress scenarios, personal interactions may help in securing better outcomes for our investors. While traditional voice sales relationships are still important, they are now complemented by electronic sales coverage and increased interactions with trading platforms and external TCA providers.
Let’s talk a little about FX algorithmic trading. How important are execution algos becoming in your day to day trading operations and what are the advantages of using these toolsets?
Execution algos are an integral part of trading operations and have become essential to our execution strategy. These algorithms among other things allow us to minimize the spread costs by warehousing the market risk and potentially reducing market impact by breaking up large orders and executing them over time. Therefore, FX algos enable to us to reduce transactions costs and ultimately delivering better outcomes for our investors.
What are your main objectives when undertaking algorithmic FX trading and what types of orders are usually a good fit for them?
Our primary objective when undertaking algorithmic FX trading is to lower the transactions costs, directly benefiting our investors. Discretionary orders which are above a certain size threshold are normally a good fit. The thresholds are continuously calibrated on a currency pair level based on the transactions costs that we are facing on our RFQ trades. Even though an order size might be above the calculated threshold, our traders could decide to execute in a different way as they may want to avoid certain market events and hence regard the direct risk transfer as favorable for the investor.
How do your source your FX algorithms and what influences that?
We are conducting a yearly review of our FX algorithms to be up to date on the latest developments in the market and new product launches of our algo providers. This review includes a qualitative part which queries points such as conflicts of interest, segregations policies, safety features or technical abilities, and a quantitative, TCA driven part. All algos passing the qualitative review form our FX algo universe and are then used based on their historical performances.
How much real-time visibility are you looking for on how an algo is performing during the execution process?
Real-time visibility paired with inflight analytics is a key feature for an FX algo suite as it allows for monitoring of execution performance and other metrics like fill rates or aggressiveness during the execution. Live data enables the trader to make better informed decision on whether there is a need to adjust the algo’s parameters to meet the desired execution outcome. As the traders are ultimately responsible for the performance of each algo trade, they should be able to see all the relevant key metrics during the execution process.
How do you analyse the results of your algorithmic FX trading to see how effective it is and whether it is meeting your objectives?
We make use of an external TCA provider which helps us conducting an in-depth analysis of FX algo executions. This allows us to compare performance across algos and currency pairs and therefore enables us to assess each algo’s effectiveness in meeting our objectives.
Key metrics we focus on include several reversion statistics, which help us getting a better understanding of each child order’s market impact and Sharpe ratio-like measures that indicate risk-adjusted performance. Looking at your key metrics post-execution on a sufficient large set of outcomes then allows you to for example draw conclusions about the under-/outperformance of certain algos in certain currency pairs or whether you should be more cautious about fills in specific ECNs.
In what ways are you leveraging data and analytics to help you make more effective use of FX algos?
The use of data and analytics starts from the moment an order arrives in our blotter. Firstly, we assess whether order parameters and current market conditions make it a good candidate for an algo execution.
For the subsequently algo selection process, we are looking at our analysis of the historical performance within the specific currency pair, ensuring we select the best fitting algo. During execution, we oversee the algo’s progress and performance using the real-time analytics.
Post-execution, we review the key metrics by both our algo provider’s and our external TCA provider’s reports. We also evaluate post-execution statistics on an aggregate level, as this helps smoothing out the individual trade variances and therefore provides a clearer picture about the actual algo performance.
What advice would you give to other firms who may be looking to exploit the power of algorithmic FX trading?
I think it is essential to have a robust framework for continuous evaluation of your FX trading outcomes. The insights from this framework then enable you to make informed decisions on when to use FX algos versus other execution protocols. Partnering with an external TCA provider helps creating independent and comparable statistics about each algo execution, allowing firms to conduct objective performance reviews.
Apart from the analytical points, it is also crucial to consider your overall trading workflows, especially how you desire to submit your FX orders into the algos, since more manual processes could lead to less efficient workflows, increased operational risks and increased time to market.
