|
GSCS related: What benchmarks do you think pension funds should use? General questions: Are transaction costs really important enough for me
to spend valuable time analysing? What benchmarks do you think pension funds should use? Pension funds should consider one benchmark that measures trader performance against the market and one that measures implementation shortfall (i.e. the overall impact on investment performance of executing trades. If information concerning “release time” is available then the fund can use Available Volume Weighted Average Price (avVWAP) for the former and Release Time Price (RTPx) as the latter. For many funds this level of precision in analysis is not available because of limitations of the data available. In these situations use of Daily VWAP and Trade Date Open or T-1 Close as the relevant two measures is appropriate. While the latter measures are not as accurate when looking at specific transactions, they are statistically reliable in terms of overall average performance assessment. What kind of custom reporting is available? GSCS is willing to create and regularly produce any kind of specific client reporting that may be desired. This can range from summary “senior management” reporting through to detailed reconciliation and identification of specific trades. Rolling summaries to highlight trends are also available. GSCS provides reports for each manager used by a pension fund. GSCS is the first company to offer reports that comply with the disclosure recommendations contained in the FSA policy 05/5 covering the breakdown of execution and research commissions. Do you provide qualitative interpretation of the data? Yes. GSCS will provide interpretive commentary in either written or verbal form to its clients. Helping clients better understand why the trading results they have achieved have occurred, and how they can be improved in the future will be the key element of all transaction cost analysis, far more important than simple data collection and presentation. Helping pension funds work with external managers to improve performance without being unduly confrontational or critical is an important element for any long term relationship with pension funds. What does the service typically cost? Some smaller clients pay very low fees, reflecting the straightforward nature of the analysis they receive, the standard nature of reporting and the low number of transactions involved compared with an investment management company. However where clients have specific interest in transaction costs and demand higher levels of sophisticated reporting and analysis fees are somewhat higher. Only the very largest and most sophisticated of funds would pay more than EUR 10,000 p.a.. How frequently should we review information with our managers? Twice a year on average. Reports should be reviewed internally each quarter. From this review managers whose performance is relatively weak should be identified and discussions held with them as part of their regular investment review process. If a manager needs to be contacted for this reason more than twice a year then it would suggest that there are some serious inadequacies in their approach to and management of execution and trading. What should we expect from our managers in terms of transaction cost analysis? As with investment, managers should be able to explain how they approach the trading process, including explaining why particular types of brokers may be used for specific types of trade or security. They should also be able to explain how and why they measure their execution performance and illustrate the process by showing results for any specific client on request. At a minimum they should offer as part of their regular reporting an analysis of their overall trading performance. In addition with implementation of FSA policy 05/5 it will be appropriate to ask managers for a breakdown of total commissions between “execution commissions” (paid to facilitate individual trades themselves) and “research commissions” (paid for provision of specific research e.g. company analysis, economic analysis etc.). Managers should also be able to provide a statement of what actual research was purchased, what it cost and be in a position to justify the cost based on value derived by the pension fund from the research. This will make for a much more effective dialogue between pension funds and managers about what the pension funds’ money is actually being spent on and why. To the extent that their own analysis, or that obtained directly by a pension fund, raises questions about performance, managers should be able to respond to the specific questions concerning how and why a particular trade or trades was handled. The level of detail should be comparable with that which they would provide in respect of an investment idea that is not performing as well as expected. Will you present findings to our trustees/investment committee? Yes. As noted in the question on interpretation, GSCS firmly believes in helping clients understand the nature of transaction costs and how they can and should be managed. Trustees have an important role to play within the overall process. Can you help us determine how we should act on the information provided? Yes. However GSCS recognises that investment managers are selected and retained for reasons that are more important than their ability to manage transaction costs. It is important that any advice in relation to possible actions to be taken should recognise that reality and be sensitive to it. In the final analysis the client has to manage the overall relationship and therefore may feel unable to act in a way that GSCS might recommend, even though they may agree with the GSCS recommendation.
Are transaction costs really important enough for me to spend valuable time analysing? Yes. It is increasingly recognised that effective trading performance is an integral part of overall investment returns. In periods when average absolute returns are at historically low levels, the importance of trading becomes even clearer. Pension funds should therefore be ever more interested in transaction costs which are often equal to or greater than investment management fees. In addition regulators, lead by the FSA in the U.K. are keen to ensure that client funds, in the form of commissions, are spent appropriately and that best execution is not affected by bundling of execution and research. Regulators believe that if pension funds are informed about how money is being spent, they will ensure that it is not wasted. They are to some extent relying on pension funds to understand and be interested enough in transaction costs to act as effective "monitors" of investment managers. How sophisticated should my analysis be? The most important aspect of the analysis is that you should be able to explain it to non-specialist trustees. The primary purpose is to make clear to your managers and your trustees that you are interested in and will monitor transaction costs over time, and that your expectation is that they will decline. Two benchmarks, one to highlight implementation shortfall and one to compare market impact are the most that should be used. The benchmarks should be easy to implement and to understand. The focus of any review with managers should be on performance trends over time, not on individual transactions, whose results are naturally quite volatile. What can I do with information about manager transaction costs? Investment managers are responsible for transactions just as they are responsible for investment decisions. Pension funds do not manage portfolios but they do expect managers to be able to explain the larger investment decisions they have made. Funds demonstrate their interest through communication and having information about investment performance. Similarly, by making clear to managers that they are interested in and have information about transaction costs, pension funds will ensure that managers do as much as possible to reduce them. If transaction costs are high, this will in the long run, affect investment performance which is the ultimate measure of satisfaction with a manager. How do I compare transaction costs across my managers given they have different mandates? Different investment styles and mandates are likely to result in different transaction costs, especially against some measures such as implementation shortfall. A pension fund should be aware of these differences and recognise that they will have an impact on performance. Pension funds should not generally compare one of their managers with others. Rather, for comparative purposes they need to know how the costs incurred on their portfolio compare with costs incurred by the same manager generally across multiple clients and for other managers with similar styles and mandates. This kind of information should be available from the transaction cost measurement service that they utilise. Why are my managers not happy with volume weighted average price (VWAP) or high/low price comparisons? Comparisons with high and low prices suffer from the fact that they represent the very best that could ever be achieved. In addition the volume of business completed at these prices is very often small and less than a typical institutional manager trade size. Managers therefore consider it unfair to be evaluated against an ideal result that they can never beat. A parallel in investment would be to judge a manager against an ideal portfolio that only included the very best performing stocks in any period. By contrast comparing execution prices to VWAP is similar to comparing investment performance to an index. Managers do not like the comparison because they know they should be able to beat the benchmark but often do not. Also with access to “algorithmic trading tools” managers can if they wish guarantee to match the VWAP price but it essentially requires no skills or experience to use the tools. Again traders believe they are contributing to improved performance and establish many reasons why they should be judged against a different benchmark which reflects this. Pension funds need to operate so that they do not simply follow a single benchmark slavishly. Any individual benchmark, if pursued as a single objective too firmly, can result in trading behaviour which is detrimental to overall investment performance. Why don't my managers all give me consistent information? For the reasons noted earlier, managers’ styles and mandates will impact transaction costs in different ways. Passive managers for example often believe that they should be measured against closing prices since this is how they can best meet their objective of tracking the index, whose calculation uses closing prices. Few active managers would believe this to be an appropriate measure. Momentum driven managers tend to dislike implementation shortfall measures because they will perform less well against them than value managers will. It is therefore the case that given a free choice managers will use a transaction cost benchmark that they believe best suits them and their activity. With the availability of multiple benchmarks however managers are able, if they wish to show the pension fund their performance against both their own preferred benchmark and a single benchmark that the pension fund is using, and explain variances. How do I know that the information I get from managers is accurate? A pension fund cannot know that information provided by managers is accurate unless it is produced for the manager by an independent third party who certifies that it is correct. Any party that may have executed trades for that manager is clearly insufficiently independent to provide such an “audit”. The best solution for a pension fund is likely to be to have its own independent service provider produce results. The costs are much lower today than they were even three years ago. The cost should not represent a barrier to obtaining independent information geared to the specific needs of the pension fund. Aren't my managers already incentivised to ensure they minimise transaction costs? Yes. If managers reduce costs they can improve investment returns. However managers have many different incentives and concerns. These may conflict with the desire to minimise transaction costs. For example a single transaction may affect many individual client portfolios. By making clear that they are monitoring transaction costs, pension funds can help ensure that all these conflicts are handled in ways that provide the maximum benefit to their particular fund. What is a reasonable amount of research commissions for me to pay? There is no “reasonable” level of commissions that should be paid. With the advocacy by the FSA in the U.K. and MiFID in Europe of transparency, managers are increasingly recognising the need to explain the nature of the research they are purchasing with commissions, its price and why it is valuable. Pension funds should encourage that communication and clarity and satisfy themselves that their money is being well spent. Why do my managers pay different levels of execution commission to the same brokers? Arguably every trade execution is different. Execution commissions reflect the skills and resources needed to complete a particular trade or group of trades over time. If a manager completes a programme trade or a transition trade with a broker the work and risk involved is not the same as that for a very large order in a single stock. Also different managers have different “buying power” with brokers based on the size of the relationship, location etc. As a result of these factors, as well as simple negotiation the rates of execution commission will always vary. A pension fund needs to understand why these variations may have occurred on their trades and to make sure they are reasonable and justified and that the manager has negotiated the most favourable deal possible in the particular circumstances. Why do managers pay different levels of execution and research commission to different brokers? For the same reasons noted above. Brokers do not all provide the same research and its value is not the same to all managers. Does a commission recapture programme affect the quality of execution and can I measure it? Commission recapture programmes are usually organized by a particular broker and in some cases more than one. If the manager executes only those trades with that broker that they would have chosen to do irrespective of the programme, then it will not impact execution performance. If however, in order to meet a “target” requested by a pension fund, a manager executes trades with the commission recapture broker that they would normally have given to another broker, then almost certainly this will result in higher transaction costs. A comparison of the transaction costs incurred by managers on trades completed within the programme will highlight over time the extra costs being incurred by the pension fund. Should I allow my managers to undertake soft-dollar trading? Generally no. In the U.K. with the new disclosure recommendations managers will show pension funds what research they have bought and at what cost. The FSA has clearly defined the parameters it considers appropriate in establishing what constitutes “research” that might appropriately be paid for by a manager using client commissions. Pension funds should ensure that they monitor this expenditure and satisfy themselves that it is consistent with these recommendations. Will hiring a transition manager lower costs as and when I decide to change managers? Usually yes. A transition from one manager to another involves a complex series of transactions. In terms of investment exposure, trading and operations there is great scope for excessive costs to be incurred. Without a dedicated transition manager it is not always clear who is responsible for making sure that any costs are minimised. There are many brokers who have specialist teams that handle transitions. Their success in reducing total costs is reflected in the fact that their adoption is now widespread. |
 |
|
FAQs
