Unravelling MiFID II one year on: The great transaction cost confusion
Andrew Robbens, Investment Specialist, discusses the treatment of transactions costs and how these cost calculations can be misleading when compared across different asset managers.
The second Markets in Financial Instruments Directive (MiFID II) came into force a year ago. Investors are well aware of most of the regulatory refinements made under MiFID II, but some areas are still causing confusion. In particular, it appears that many investors do not fully understand the treatment of transactions costs and how these cost calculations can be misleading when compared across different asset managers.
Calculating total costs
MiFID II requires fund providers to provide details on the total transaction costs, both explicit and implicit, incurred by their investment funds.
Most of the costs included in the MiFID II calculations have been around for some time and are well known to investors—for example, one-off charges covering initial charges, redemption charges and distribution fees; ongoing charges covering annual management fees and operating expenses; and explicit transaction costs covering broker commissions, taxes and levies, and stock lending. MiFID II has, however, made these costs much more clearly visible.
On the other hand, it’s apparent from the many interactions that I have had with clients that there is a certain degree of confusion regarding what is covered by “implicit transaction costs”. It’s important therefore for investors to fully understand how these implicit costs, and the methodologies used, can impact the total cost figures for different fund providers—making comparisons difficult, if not impossible.
What are implicit transaction costs?
First of all, I think it is important to state that implicit transaction costs are a result of operating in the financial markets and are not an actual physical charge made by an asset manager. These implicit transaction costs are not a charge where an asset manager benefits financially. Before we venture further into the transaction cost calculation, it’s worth pointing out that a high transaction cost number should not be considered in isolation. It is also worth repeating that it’s difficult, if not impossible, to compare the transaction cost figure of one asset manager with another given the variety of methodologies being used, and the fact that some firms may use estimates in their calculations, which may have to be revised as they begin to incorporate actual data.
To explain the complexities of implicit transaction costs in more detail, let’s look at a couple of examples that will hopefully help to illustrate where implicit transaction costs actually come from and what they mean for investors.
Example 1: Buying a stock following positive newsflow
The first example is of a manager who trades quickly after the release of some positive news regarding a company, which results in a rally in the share price. The asset manager sends the trade to the executing broker when the share price is £1.00 (the arrival price). However, due to the fact that there is a high demand for this stock, it may take some time to actually execute. By which time the share price may be £1.01 (the execution price).
The implicit transaction cost in this example would be the difference in the arrival and execution prices, which in this example would be 1%. If the share price reached £1.03 on day 12, the performance for the asset manager would be 1.98%.
Quick trading following release of positive newsflow
Now let’s assume that another manager reacts more slowly to the information. This may be due to the type of strategy they manage, or they may have been unavailable when the news was released, or they simply took more time to analyse the impact of the news.
Whatever the case, by the time the manager executes the sale, the rise in the company’s share price has slowed after the initial burst of trading activity, and stands at £1.02. The trade eventually gets executed at £1.0225, resulting in a much lower implicit transaction cost of 0.25%. But the performance of this holding for the manager at day 12 is only 0.73%.
Slower trading following release of positive newsflow
If we compare these two examples side by side, we see that the manager who has acted quickly may have incurred a higher implicit transaction cost but would have benefited much more in terms of performance.
Implicit trading costs comparison: Buying after positive newsflow
Example 2: Selling out of a stock following negative newsflow
The relationship between implicit transaction costs and performance is similar when several managers decide to sell out of the same stock due to the announcement of negative newsflow, but execute their sale orders at different times.
Manager 3 in the chart below is slower to sell than Manager 1 and Manager 2. At the time when Manager 3 executes their sell order, the share price has fallen to a level that is tempting for value investors waiting for an attractive entry point.
As these value investors begin to buy the stock, the share price recovers from its low. The result is a negative transaction cost for Manager 3 over the time it takes to achieve the sale, which looks very cost efficient for the portfolio but may actually deliver the worst performance outcome for a client.
Selling a stock following release of negative newsflow
Again if we look at these three managers and the theoretical outcomes, the lowest transaction cost does not necessarily mean the best result in terms of performance.
The arrival price methodology
Calculating implicit transaction costs in this way, based on the “arrival price”, is the methodology expected under the PRIIPS (Packaged Retail and Insurance-based Investment Products) regulation. However, the arrival price methodology does, as illustrated, run the risk of classifying shorter-term alpha as a cost.
A manager whose process allows them to make investment decisions while price discovery is still occurring following an event is likely to show higher-than-average transaction costs. It does not necessarily follow that the manager is less likely to outperform their index once all costs are taken into consideration.
At J.P. Morgan Asset Management we use the arrival price methodology. We use real data, from the previous three years, for our ex ante implicit transaction cost calculations. It is apparent, however, that asset managers are using a wide range of methodologies to measure transaction costs, and that the quality of data used can also vary. It is therefore important for investors to be aware that it is not always appropriate to compare the transaction costs figures across funds managed by different asset management firms.
It is also worth highlighting the trading resources at J.P.Morgan Asset Management which ensure that we are able to achieve best execution in accordance with applicable laws and regulations for our clients.The Global Equity trading desk operates in 5 key locations; London, New York, Hong Kong, Tokyo and Taiwan. In 2018, JPMAM traded $1.2T or about $4.8B per day. ($748bn of that was cash equities or about $2.9bn per day which equated to 0.8% of the Equity value traded globally). Global JPMAM executes 6500 orders per day in 70 markets across the globe. Thanks to our significant scale and investments in technology and analytics, we’ve been able to lower our trading costs resulting in $1B in client savings over the last 5 years; this is down to a mix of highly-skilled traders, well-deployed technologists and a well-resourced analytics team. There are currently more quants and technologists co-located on the desk than traders, a quite profound evolution of the trading function and how it is transforming from an art to a science.
Our trading desk was voted best buy-side desk for each of the last 3 years; the Equities team rang the opening bell at the Nasdaq this past summer to celebrate the achievement. In November 2018 our trading team won three awards at the annual Leaders in Trading Awards in London, hosted by “The Trade” magazine: Trader of the Year (Long-Only): Neil Joseph; Rising Stars: Alex Hunter and Alexandra Tidy.
Conclusion: Look beyond the headline figures
In summary, investors should be aware that implicit transactions costs:
Are not an explicit cost;
Are calculated using different methodologies by different firms;
Do not necessarily result in the best outcome when they are lower;
J.P. Morgan Asset Management are calculating transaction costs in line with guidelines and on a conservative basis
Investors should therefore proceed with caution when comparing transaction costs from different managers. This also illustrates why performance comparisons need to be made net of fees. Net performance includes all costs, explicit and implicit, and so remains a valid comparison approach between fund managers and products whichever transaction cost methodology has been chosen.
Hopefully this has clarified some of the confusion around quite a complex subject that has taken me some time to really get my head around too.
Andrew Robbens is an Investment Specialist within the J.P. Morgan Asset Management UK equity team.