Is your broker keeping up?
Over the past year, the world’s financial markets have been turned on their heads. Responding to this turmoil, buy-side traders have dramatically increased their use of electronic trading including algorithms in an attempt to reduce costs and information leakage, while increasing control of their orders.
The increase in volatility in the last quarter of 2009 was reflected in the average daily move of MSCI Asia- Pacific Indices spiking to 5.1%, and spreads blowing out from 16bps to almost 22bps. Liquidity dried up dramatically between Q3 2008 and Q1 2009, with the average daily traded value for the region’s largest markets dropping by more than 35% and daily volume curves becoming erratic.
These unpredictable trading conditions across Asia-Pacific mean the region’s traders are now catching up with Europe and US in their adoption of electronic trading.
Electronic trading tools have developed at breathtaking speed to handle demand for efficiency and productivity. Many of Asia-Pacific’s buyside firms have moved swiftly from direct market access to first generation algorithms, largely simple time slicing tools, and beyond to second generation algos which often referenced static historical data, including volume curves per stock, and would divide the orders into even smaller slices to avoid being gamed.
Brokers such as Société Générale have also developed third-generation algorithms, adapted from proprietary traders’ statistical arbitrage techniques to execute clients’ VWAP, implementation shortfall or with volume orders. These algos utilise a combination of historic static data per stock, such as volume curves, average size on the touch, historic volatility and market beta, as well as realtime stock and index futures data. The algorithms then calculate momentum, correlation and finally a statistical relative value indicator (SRVI).
Simply put, the next generation of algorithms uses a combination of historic and real-time data to take advantage of intraday price or volume anomalies to improve execution performance and capture liquidity by trading more actively when favourable. As the diagram shows, when the SRVI suggests the price will rise, the algorithm will buy more aggressively than the historical volume curve would otherwise suggest.
Recent market chaos has amplified the differences between sell-side offerings and exposed the shortfalls of trading technologies that only rely on historic data to predict daily trading patterns.
Adapting to market realities
Great variations in market structure and regulation across Asia-Pacific - such as rapidly changing short-selling rules or different trading patterns due to unique opening/closing auctions or up-tick rules - also present a significant challenge to traders.
These differences can be illustrated by observing the average spreads in each of the Asia-Pacific markets. Considering the Asia-Pacific markets trading more than USD$500 million per day, the average market spreads start at 5bps in India, one of the lowest spreads globally, up to 37bps in Singapore which is almost twice the region’s average of 19bps.
Algorithms which utilize real-time data to improve trading strategies, and have been optimized by local teams according to local market idiosyncrasies, can greatly improve a trader’s performance, cost efficiency and consistency (i.e. lower standard deviation of results). For example, Société Générale’s Indian algorithms use real-time data to identify iceberg orders and then send larger orders to capture the hidden liquidity.
Efficiency has also been improved via developments such as the Alpha trading engine designed for 12 regional markets including India by Société Générale’s dedicated Asia-Pacific quantitative research, development and trading team, In Q1 2009, more than 75% of Société Générale client orders in the region were traded using the trading engine, allowing dealers and clients to focus on the remaining 25% of orders.
The pace of change in the Asia-Pacific trading environment requires continuous development of electronic trading tools. Although mainly limited to Japan, fragmentation of liquidity caused by the emergence of alternative trading venues is making best execution increasingly hard to achieve without algorithms.
And just as in Europe and the US, algorithms should now sweep systematically through internal crossing engine (like Société Générale’s recently launched Alpha x Europe), dark pools and lit pools prior to trading in regular markets.
The efficiency and control offered by algorithms is beyond doubt. But as performance differentials widen, due diligence by the buy-side trader to ensure that the sell-side broker has invested in the latest technology has never been more important.