Earlier this year Kepler Cheuvreux won the accolade of best Sell-Side Equities Trading Desk in Markets Media’s first European Markets Choice Awards. We invited Patricia Shin, Managing Director and Global Head of Execution Services at Kepler Cheuvreux to reflect upon the pathway to their success.
As I sit here 37 years into trading, it doesn’t seem that long since we were writing ‘buy’ on a blue ticket and writing ‘sell’ on a pink ticket and the ‘market’ reference was the stock exchange. All the OTC stocks needed three sets of markets from three brokers, to then trade at the best price. That was the ‘best execution’. The current generation of traders may laugh at the fact that settlement was handled physically by the clearing houses. The physical stock delivery was done by two people to make the cut-off window at noon. Why two, you may ask? Well, in case one got hit by a car, you can still make delivery. How efficient was that?
So, what was wrong with the way things were? Well stock exchanges charged too much, monopolising the markets, which in turn brought competition and fragmentation to the marketplace. As a broker, we are constantly developing to maximise the value of the only true constant in the world that is change. The goal as always is to efficiently maximise liquidity for our clients, whilst minimising market impact and avoid information leakage. Now we’re in a world of information and data overload, where the added value is not finding the information but filtering out the noise to focus on the information that truly matters.
Here at Kepler Cheuvreux, we operate in 12 local regions across all the major European cities and New York, where we continue to grow and evolve. Contrary to the experience of many firms in the financial services sector, Brexit was not an issue. Our multi-local setup meant that we did not have to moderate our structure. We continued to focus our efforts on delivering innovations for our clients, which included an overhaul of our entire execution platform. Our new modular architecture can be easily customised for client’s needs. Working with best-in-class providers, we have implemented a new OMS, new SOR, new market gateways, new market data provider coupled with our new algorithmic platform, all while the market was reeling from a pandemic. Put simply, we now have far more flexibility, are incredibly stable, with ultra-low latency.
To compete with the bulge bracket banks, it was imperative to provide a global reach in execution. On top, we needed to be a ‘go to broker’ for clients for natural liquidity. We are no longer known as just a European broker, our third largest market in execution is in US equities and we have our own US customised algos using our in-house logic. We supply agency only liquidity, interacting with 99.5% of all addressable liquidity in the market. We are client centric; venue agnostic and our strategies are 100% performance driven. We are also one of the very few who can trade listed bonds electronically on the primary market.
As I have touched on, the role of a sell-side broker has changed immensely. We now more closely resemble a tech company than a traditional broker, a requirement for success in today’s world. Our clients similarly continue to evolve, none more so than in the setting of benchmarks, quite often unique to a portfolio manager and naturally traders want to outperform their benchmarks. As execution experts, we spend a large portion of our time in the execution consultancy process, predominantly for algo wheels, discussing strategy customisation with our partner clients, pushing the boundaries of what is best execution.
Our algorithms are built on a philosophy of avoiding complexity whenever possible and concentrating on innovations that can truly make a difference. Case in point; a lot is made about machine learning, which is a very broad topic. In order to achieve best execution it is our view that quantitatively-derived models, such as those exploiting machine learning, are best suited to provide the required adaptability and flexibility, along with the depth and breadth of available liquidity, that today’s constantly changing market conditions demand.
QAI, short for Quantitative Algorithmic Infrastructure, has been built by our algo-quant team following the successful overhaul of our electronic platform to provide the building blocks to assemble our next generation of algorithms and will leverage machine learning to derive the data used by the algorithms to make valuable trading decisions.
The first innovation we are rolling out builds upon a home-grown volume forecast model and will find its direct application into our Target Close algorithm. Essentially, Target Close becomes adaptive, re-allocating quantities intraday as the day unfolds to better target the closing price while containing market impact.
The second wave of innovations will leverage the work we put into deriving optimal schedules from the trade-off between market impact and price objectives.
We will first apply this new methodology to Dark Scheduling whereby the allocation to the Dark will be reassessed throughout the day to capture liquidity better, while avoiding adverse selection.
Our work will culminate this year with the release of our new Implementation Shortfall algorithm, using the same method for both lit and dark adaptive scheduling, centred around market impact estimates, updated in real-time for feedback on price.
©Markets Media Europe 2021