A New Model For Fixed Income
With Enrico Melchioni, Director, LIST Group
LIST’s experience of best execution began in 2007 when the first MiFID directive came into force. In Italy, one of List’s operational countries, the directive was applied immediately to both equity and fixed income.
Back in 2007, the basic approach to best execution was to implement a “static best execution” logic: given the list of venues included in the Investment Firm Execution Policy, the best trading venue for each security (or security class) was defined on the basis of the statistical analysis of historical data performed off-line. Incoming client orders were then routed to the “best venue” simply taking into account the security parameter of the orders.
We decided to push the best execution process one step further, introducing the “dynamic best execution”: we use the same statistics to prepare a basic ranking of possible execution venues, but just before executing the trade, live market data collected from the different trading venues are analysed, computing an estimated total consideration that includes broker fees and exchange taxes. By bringing a wider variety of broker and venue real time data into the metrics, we can integrate more holistic pricing and liquidity data, so that when an order is executed, it is done on the basis of a comprehensive and updated set of information, achieving a deal-by-deal, pre-trade best execution.
This model, initially devised for the equity space, has been adapted to also incorporate the specificity of fixed income trading.
Move to the buy-side
In the past few years the fixed income market has evolved and is now quite a different landscape from the one we were used to 10 years ago. Trends on the buy-side show that there is broad interest in speaking with multiple dealers or to join one of the many new initiatives for fixed income trading which promise peer-to-peer trading or pre-trading facilities.
The reasons for this are two-fold: we are experiencing very low rates in the Euro area (that drives investments toward riskier, less liquid assets in search of higher yields), and the increased regulation forces a higher cost of capital onto the banks. The combined effect of these two factors is that it becomes economically difficult for the sell-side to keep a position on a bond and so the buy-side cannot find liquidity when it wants it.
In this new scenario, the buy-side needs tools to connect to more than one execution venue, be it a sell-side, a buy-side network or a crossing venue. Technology can greatly help in this process, supporting the trader in retrieving the relevant information from a variety of sources, pulling all the data together in a single place and evaluating the best execution based on the updated and aggregated information.
We have already witnessed similar trends in the equity space: when MiFID I went into force, the concentration of trading on national exchanges was removed, so that all of a sudden a number of different exchanges flourished and the same equity was traded in different places. What firms really needed was some re-aggregation that allowed them to connect to the different trading venues and reconstruct the trail.
MiFID II is now creating a similar scenario for fixed income. Firms need to perform best execution across multiple execution venues, but in this case, the trading models are different: the prevailing market model is not the limit order book; rather liquidity is fragmented across different platforms mainly based on request for quote (RFQ) models.
Our clients ask us to be able to connect simultaneously to multiple platforms and to aggregate data from different sources (regardless of the trading model) so that they can have a comprehensive view of how the market is behaving.
When it comes to illiquid products, some logic that works perfectly well with liquid security does not represent the optimal solution. A process that takes into account OTC circuits and RFQ based venues is needed.
Last but not least, we are witnessing the birth of several initiatives that propose new business models based on peer-to-peer cooperation and pre-trade information sharing. The trading system must be capable of easily integrating with new trading models to take advantage of new opportunities.
All these specificities of the fixed income market may suggest that the buy-side industry needs to start thinking like the sell-side in terms of technology; being able to connect with different venues and to analyse data. This points to many questions including whether to outsource the technology or to develop it, and how best to do so given that the final shape of the regulation is still unknown.
MiFID II best execution implementation
Any execution desk has best execution as its goal. Most desks already have best execution logic in place. The problem is that MiFID II requires a specific set of obligations, with varying degree of compliance in each firm. Therefore, the actual impact of implementing MiFID II will vary from firm to firm. MiFID II also requires a standardised approach, which means that even if a firm already has a best execution policy in place, the firm will have to modify the parts of it that are no longer compliant.
However, firms that have started work on this will have an advantage when the final regulations are announced.
On the other hand, it could be argued that other aspects of MiFID II, such as the post-trade reporting, are very complex and that they are not adding any value to trading activity. So this element is likely to be one that people will try to implement at the last possible moment.
Post-trade to pre-trade
Trading platforms are generating vast amounts of data, which is only going to increase in quantity over time. These large data lakes contain a wealth of information that can be reaped with the proper technology. Firms need to start thinking in terms of data mining and Big Data technology in order to analyse the information that is already available within their systems.
By analysing trading data it is possible to create a link between pre and post-trade worlds: information retrieved from post-trade data can be fed back into the pre-trade activity, enhancing the execution process.
Data analysis can help the buy-side in understanding which brokers were efficient in serving a specific bond or bond class. For the less liquid fixed income products, the buy-side and sell-side are both focussed on keeping track of who is willing to trade a particular securities. This requires considerably more data and relationship management than for an automated asset class like equities.
Data mining technology can help answering the above needs, and also provide new perspectives. We have been conducting some research in applying advance clustering algorithms to trading data in order to understand what clients are doing and to find temporal correlation among the trading activities. For instance, it can be quite easy to identify those firms that lead and those that follow after a given stimulus from the market.
The industry is transitioning from an era where the execution tool was the only item firms needed, into an era where firms also need to manage the information alongside the ability to execute because the market is becoming more complex and fragmented. It is a period of transition where firms should try to act on all possible information that is available within the system.
A continuation of this transition is the growing importance of industry initiatives that involve the buy-side. Decisions are being driven by those firms who are actively involved with industry initiatives as they try to find ways to solve the problems the market is facing.
When firms make decisions about their technological infrastructure, they should envisage an infrastructure that is capable of managing all the exiting trading models and is also open to new initiatives, so that it is easy to integrate with new market models that may arise. The largest buy-and sell-side firms will probably already have internal development teams that can work toward this goal, but smaller companies will be looking to outsource this ability.
We’d love to hear your feedback on this article. Please click here