Buy-side trade participation and greater access and use of trading tools, coupled with the fragmentation of liquidity pools (driven by RegNMS Order Protection Rule, a. k. a. trade through proposal), have resulted in more disintermediation of the sell-side trader. Specifically, the buy-side traders are using more and varied dark pools and crossing networks that are removing direct sell-side participation.
Further, traditional buy-side clients (versus pure quant and HFT firms) are greatly improving their in-house capabilities to develop trade algorithms (black boxes). The abundance of applications, use of more trading tools (EMS/OMS), and consumption of data are allowing buy-side clients to press harder on their brokers to demonstrate value. The sell-side traders have data coming in from all sources at a very fast pace. It’s high-speed data that is coming too fast to handle, digest, and make sense of.
Pre-trade analytics, at a basic level, is the examination of historical and current price and volume data. It attempts to help firms determine where to send orders and when, whether to use algorithms or manually trade an order, and what the opportunity cost of not acting on information might be. Most of the major investment firms have developed their own analytic models and might not necessarily rely on canned tools.
Even as the industry struggles to deliver effective pre-trade analytics, the demand for accurate predictive information isn't going away. Rather, in the high-frequency and smarter trading environment the demand for pre-trade analytics continues to grow. Though tighter technological integration is bound to come in time, if pre-trade analysis is to come to full fruition, firms may also need to find a way to share more information without putting their trading strategies at risk. For example, look at temporal technology that is needed to provide accurate time-slicing and the ability to map that activity and maintain the benchmark.
Analytical platforms are changing to mesh data at rest and data in motion, along with new/different data types and deeper/broader histories of these data sets -- all from varied sources. Underlying architectural solutions can increasingly meet these market demands for pre-trade analytics. The goal is then to find the high value opportunities to test and develop advanced pre-trade analytics.
As part of this ongoing dialogue the below list contains a few examples aimed at identifying ways to develop advanced pre-trade analytics. Some examples supplement current analytics, others can help automate manual activities, and some look at how to apply analytics to better understand client trading activity. Tools are now available to enhance the broker’s ability to observe trading behavior, characteristics, and shortcomings to educate and provide value to buy-side clients and achieve best execution.
Client execution and order fills: Analyze execution venues and routing business.
o Looking at orders that don't get filled and determining where these orders go
o Trying to understand relationships between the order and the path it takes: crossing, dark pool, electronic market makers, agency, and last is lot markets, for example
o Applying graph and path analytics to examine network efforts, decision nodes, and predicated outcomes
Client algorithms and trade strategy: Trying to understand why a buyside algorithm might not be hitting its objectives. Trading and service desks can use alerts to help understand how client trades and algorithms are benefiting them. Example: Say a client’s VWAP strategy is falling behind. How can you help traders be more proactive and help clients. Others examples where pre- and post-trade analytics can be used:
o Do orders sit on top of the queue?
o Can you use post-trade analysis to understand if strategies are slipping?
o Can you see a disconnect on a venue and dark pool that is underperforming (algorithm side) or trader decision on execution timing?
News and social sentiment: Sales traders are being asked to help clients understand sentiment. Can the trader desk have the ability to understand velocity of sentiment in addition to sentiment itself?
o Network measure of social participants: Understand who high reference individuals are and the clout that they have with recommendations that are valuable and useful. Building short-term factor models to move towards trading.
Broker segmentation: Understanding the generational segmentation of your broker base:
o How are younger, middle-aged, and older brokers utilizing systems?
o Are they having challenges/success with their trading desk tools and effectiveness for clients?
o For OMS applications, trading needs to understand the different types of users.
The march for the buy-side to understand and own more of the trading process more actively will continue. That said, with all the reams of data that people are collecting there is still a lack of good context. That is part of the opportunity for the sell-side. The move should be to apply analytics to pre- and post-trading more actively, not only for execution, but also to better understand the customer’s activity and where to be consultative using data-driven advice. This is the inevitable advancement, the electrification of the high-touch broker business model, which can help give that stronger market context for clients.
This will begin to include, not just trade and execution data, but also the other new data streams (social velocity, news, machine data, graph network -- traders and systems -- analytics, etc.) that, when combined will remove time and guesswork out of what is influencing a trade and offer clients a more concrete, granular viewpoint. These data-driven types of broker services and solutions need to become more bundled and complete. When more relevance is restored, that advice can help bring the trader closer to investor.Sean O'Dowd leads the Global Capital Markets program at Teradata for Industry and Marketing Solutions. In this role Sean focuses on industry strategy, marketing and field enablement. Areas of focus span financial market structure, regulations and technologies that impact the ... View Full Bio