Intelligent Design

Stephane  Loiseau , deputy global head of execution services, Societe Generale, explores how market microstructure knowledge and dynamic monitoring throughout the trade lifecycle can be key to best execution.  

Today’s is a complex trading environment. Buy-side firms have never faced such a vast array of choices, tools and data. But at the same time the market is a lot more challenging than in the past. Liquidity is increasingly fragmented; it is also scarce, due to macro- economic factors.And trading data, including time and sales, is not delivered in the most ideal way.

This complexity puts the onus on the buy-side trader to consider his/her decisions carefully, and to ask: how do I achieve best execution now? It is not enough to simply decide which algorithm to use and hope that less-than perfect transaction cost analysis (TCA) will reward you with a good result afterwards.

Instead, asset managers need trading strategies that provide a cradle-to-grave approach, following a series of checkpoints throughout the trading lifecycle and continuously making adjustments as needed. A smarter trading model, based on active use of market signals and trading data, is emerging.

At the outset, the trading desk needs to understand the investment objective of the portfolio manager that initiated the order. Based on this partnership with the fund manager on the investment decision, the trader can then decide speed of execution, prioritise between size and price, and select the specific channels that will be used to execute the order. These channels may be used in combination, i.e. a trade can be initiated via a broker on risk, continued using high-touch sales trader input, then expanded via DMA, all this, while a sub-portion of the order (a ‘split’) can be rested in a dark pool in order to capture further liquidity opportunities.

As the above example illustrates, execution strategies have become a lot more complex than just using a volume-weighted average price (VWAP) algorithm from 08.00 to 10.00 and benchmarking the outcome against a simple VWAP price over the period. Using ‘in-trade’ analysis of the market microstructure is critical to making proactive trading choices. It’s important to consider questions such as whether to use dark order books, which trading venues to select for a particular strategy, how to deter- mine the optimal order size (and price) for each venue selected and how to increase or decrease participation in each venue based on pre-agreed outcomes. For example, the trade-off between maximising the liquidity capture while minimising the observed price impact would influence whether to place the majority of an order in a dark pool and whether to trade via other channels simultaneously while waiting for a fill in order to minimise opportunity cost.

No more autopilot 

Monitoring the trade as it runs is vital. Ask yourself, are you waiting too long in the dark pools and therefore missing out on liquidity or price opportunities elsewhere? If you are trading on multiple platforms at once, being very active could help detect liquidity opportunities in a fragmented marketplace but could also increase signalling risk, resulting in unwanted market impact. Equally, dynamically monitoring the impact of individual trading choices as they are implemented intra- day allows quick adjustments to the strategy which will significantly improve the execution outcome.

Such is the range of price and liquidity conditions in the lifetime of a trade in today’s market that traders need to monitor trade data continuously. As trading parameters move, it is important to recognise when intervention is needed. Market microstructure analysis can and should be used to inform how to respond, once that point has been reached.

Liquidity distribution and prices available through each execution venue will vary through the trading day depending on external market factors but are increasingly also based on which market participants are interacting in the venue and with what type of order flow. Some venues will have more high-frequency flows than others in general, and some at certain times of the day. Refresh rates, price volatility or order book behaviour can help indicate the nature of participants in a venue.

As mentioned above, there is no shortage of data flowing from the multiple trading venues in Europe. The challenge is to exploit the force of the numbers, not be overwhelmed by them. SG’s quality of venue measure (QVM) tracks information leakage by comparing any movement in the mid-point of a stock experienced immediately after a trade in a dark or lit pool to the stock’s normal levels of price volatility. We use this to inform our smart order router and algorithms intra-day as well as post- trade but it is just one example of how dynamic micro-structure analysis can aide execution performance.

Once the execution is done, you can use the TCA to look at the outcome and assess how successful your trading strategy was in achieving best execution for the portfolio manager. But remember, TCA is never an end in itself – it is simply the end-point of a process that should involve active buy- side participation and thorough implementation of a comprehensive trading model from start to finish.

Stephane Loiseau, deputy global head of execution services, Societe Generale, explores how market microstructure knowledge and dynamic monitoring throughout the trade lifecycle can be key to best execution.  

Today’s is a complex trading environment. Buy-side firms have never faced such a vast array of choices, tools and data. But at the same time the market is a lot more challenging than in the past. Liquidity is increasingly fragmented; it is also scarce, due to macro- economic factors. And trading data, including time and sales, is not delivered in the most ideal way.

This complexity puts the onus on the buy-side trader to consider his/her decisions carefully, and to ask: how do I achieve best execution now? It is not enough to simply decide which algorithm to use and hope that less-than perfect transaction cost analysis (TCA) will reward you with a good result afterwards.

Instead, asset managers need trading strategies that provide a cradle-to-grave approach, following a series of checkpoints throughout the trading lifecycle and continuously making adjustments as needed. A smarter trading model, based on active use of market signals and trading data, is emerging.

At the outset, the trading desk needs to understand the investment objective of the portfolio manager that initiated the order. Based on this partnership with the fund manager on the investment decision, the trader can then decide speed of execution, prioritise between size and price, and select the specific channels that will be used to execute the order. These channels may be used in combination, i.e. a trade can be initiated via a broker on risk, continued using high-touch sales trader input, then expanded via DMA, all this, while a sub-portion of the order (a ‘split’) can be rested in a dark pool in order to capture further liquidity opportunities.

As the above example illustrates, execution strategies have become a lot more complex than just using a volume-weighted average price (VWAP) algorithm from 08.00 to 10.00 and benchmarking the outcome against a simple VWAP price over the period. Using ‘in-trade’ analysis of the market microstructure is critical to making proactive trading choices. It’s important to consider questions such as whether to use dark order books, which trading venues to select for a particular strategy, how to deter- mine the optimal order size (and price) for each venue selected and how to increase or decrease participation in each venue based on pre-agreed outcomes. For example, the trade-off between maximising the liquidity capture while minimising the observed price impact would influence whether to place the majority of an order in a dark pool and whether to trade via other channels simultaneously while waiting for a fill in order to minimise opportunity cost.

No more autopilot 

Monitoring the trade as it runs is vital. Ask yourself, are you waiting too long in the dark pools and therefore missing out on liquidity or price opportunities elsewhere? If you are trading on multiple platforms at once, being very active could help detect liquidity opportunities in a fragmented marketplace but could also increase signalling risk, resulting in unwanted market impact. Equally, dynamically monitoring the impact of individual trading choices as they are implemented intra- day allows quick adjustments to the strategy which will significantly improve the execution outcome.

Such is the range of price and liquidity conditions in the lifetime of a trade in today’s market that traders need to monitor trade data continuously. As trading parameters move, it is important to recognise when intervention is needed. Market microstructure analysis can and should be used to inform how to respond, once that point has been reached.

Liquidity distribution and prices available through each execution venue will vary through the trading day depending on external market factors but are increasingly also based on which market participants are interacting in the venue and with what type of order flow. Some venues will have more high-frequency flows than others in general, and some at certain times of the day. Refresh rates, price volatility or order book behaviour can help indicate the nature of participants in a venue.

As mentioned above, there is no shortage of data flowing from the multiple trading venues in Europe. The challenge is to exploit the force of the numbers, not be overwhelmed by them. SG’s quality of venue measure (QVM) tracks information leakage by comparing any movement in the mid-point of a stock experienced immediately after a trade in a dark or lit pool to the stock’s normal levels of price volatility. We use this to inform our smart order router and algorithms intra-day as well as post- trade but it is just one example of how dynamic micro-structure analysis can aide execution performance.

Once the execution is done, you can use the TCA to look at the outcome and assess how successful your trading strategy was in achieving best execution for the portfolio manager. But remember, TCA is never an end in itself – it is simply the end-point of a process that should involve active buy- side participation and thorough implementation of a comprehensive trading model from start to finish.

• e-mail
• print
• share