Non-Financial Risks, Regulation and Innovations

Transaction Cost Analysis in Europe: Current and Best Practices

This report, which was commissioned by HSBC Investment Bank, reviews the conditions in which buy-side firms (traditional and alternative) are currently monitoring transaction costs and investigates the various issues related to transaction cost analysis in the context of the Markets in Financial Instruments Directive (MiFID) due to be enforced in November 2007. This directive contains an important provision related to Best Execution.

The MiFID is establishing a very open and competitive market for execution and other investment services. The new requirement is for regulators to ensure that the development of new execution venues, systematic internalisation and the end of the concentration obligation (ie the obligation in certain Member States to route all negotiations through a central regulated market) will not result in less efficient or transparent markets that could harm the end investor.

Transaction cost Analysis (TCA) is obviously very likely to form the cornerstone of any future academic and industry development related to this new Best Execution obligation. In this respect EDHEC Risk Advisory, sponsored by HSBC, have conducted a pan-European survey aiming at better understanding the current state of the industry when it comes to assessing transaction costs and the level of readiness in terms of complying with Article 21.

EDHEC Risk Advisory questioned 127 buy-side firms with the support of a structured questionnaire on a pan-European basis. 26 responses were provided by European hedge fund managers while 101 responses were sent from traditional asset management firms giving us the opportunity to analyse the difference between these two fairly specific populations in terms of trading and execution. Pre-trade analysis forms an essential part of ensuring that best execution can be achieved. With only half of respondents confirming the use of pre-trade analytics out of 83 responses, it is striking to see that pre-trade analytics may not have yet reached the desk of the majority of buy-side firms. This situation is rendered even more problematic in light of the nature of the forthcoming best execution obligation, which encompasses a significant obligation of means.

If pre-trade analysis allows the intermediary/trader to optimally design its execution strategy, it is again striking to note that situations where a trading desk decides to return the order unfilled when execution conditions are not favourable seems to be only possible with one traditional asset manager out of seven, and one hedge fund management company out of four. For the majority of our respondents, once decided, the order will have to be executed at all costs; the conditions usually attached to the order may therefore be considered as null. This surprising result confirms that changes are desirable not only for the trader but more globally in the entire relationship between the manager and the trader.

One very important pre-requisite for performing post-trade transaction cost analysis and implementing an automated processing of order flow is the use of technology from the initiation of the order and its handling in a format that is acceptable by most vendors and intermediaries. FIX (The Financial Information eXchange protocol) is obviously the natural response to this need and has seen a tremendous development over the last five years. With 57 respondents out of 79 (72 per cent) confirming the use of FIX to encode transactions and process them electronically, the industry has clearly embraced the technology revolution whole-heartedly.

This very rosy picture, however, has to be dealt with carefully as only 60 per cent of hedge fund respondents claim they are using FIX, which demonstrates that smaller firms are slower to adopt the protocol, and probably the digitalisation of their entire execution flow. More importantly FIX Protocol today does not allow full encoding of all forms of parameters that may be attached to an order; the survey respondents indicated that less than 20 per cent transmitted instructions to the intermediary with a specific price, volume or benchmark objective, thereby making transaction cost measurement potentially more difficult.

Finally, transaction cost analysis cannot be performed without intraday timestamps being made available during the entire flow of the order from the buy-side to the intermediary and back to the issuer. It is therefore a concern to see that two thirds of respondents may not be in a position to document the time the order was responded to.

When it comes to assessing transaction costs post-trade, a logical approach consists in estimating transaction costs ex-ante and comparing the actual result of the intermediary with the initial estimates. The fact that only one third of respondents have taken that route confirms the difficulties inherent in estimating transaction costs ex-ante, probably because of a clear lack of consensus on models.

One of the consequences of this difficulty is the reliance on the most common, but also the least reliable, benchmarks. With 90 per cent of our respondents declaring use of VWAP, it is clear that the absence of consensual methodology for assessing transaction costs ex-post is resulting in the total failure of all efforts made to assess the costs incurred by the end investor.

The prominence of VWAP as an indicator may be justified by the simplicity of its implementation and its apparent ease of interpretation for the end user. It is nevertheless probably not the most effective indicator of execution quality, especially for large size transactions. The full report published by EDHEC Risk Advisory analyses in detail the most common TCA benchmarks and reviews the many associated issues and pitfalls. To support further development, an innovative framework (EBEX) for an absolute measure of transaction cost is also suggested within the report. This framework has been designed to attempt putting a stop to the endless debate of what the appropriate benchmark may be depending on the specifics of the situation, and make execution quality (with regards to price) comparable from one order to another, from one intermediary to another.

In line with this current relatively poor approach to measuring transaction cost, it is interesting to note that 40 per cent of respondents would not take into consideration their ex-post transaction cost analysis when reviewing the allocation made to their brokers. If the selection of an intermediary obviously cannot solely rely on transaction costs, it is striking to see that such a large number of firms do not review the performances of their brokers according to such an important quality factor. Once again, it is interesting to note that a piece of regulation may induce adverse effect by creating a sense of confidence best execution has been achieved while the reality of the facts can not be demonstrated.

One year from the implementation deadline, it is also astonishing to see that 26 per cent of traditional asset managers still believe the MiFID is not related to their business or are not even aware of the Directive. Even worse is the situation in hedge fund management firms, with half of respondents being mistaken as to the importance of the MiFID for their firm.

Defining a clear execution policy, documenting it and informing the client of all details of this execution policy forms a significant part of the duty of best execution as set out by the MiFID. Once again, with one third of hedge funds having failed to clear the path towards a documented execution policy, and with one fund manager out of ten having to look at the question, there seems to be room for greater detailed awareness within certain business functions within many buy side firms.

URL for this document:

Hyperlinks in this document: