In focus

The transaction costs manual: our "how to" guide to a complex topic

Discussing transaction costs is seldom an easy endeavour and never a short one. There are too many aspects to cover, starting from what they are and how they arise to the more difficult questions of how to measure them and how to use reported figures.

There are myriad factors affecting the cost of trading and the transaction cost figures disclosed to investors just about scratch the surface of what goes into their making.  

Now that transaction cost figures are becoming more widely available to investors through regulation such as MiFID II, we have put together a transaction costs “manual” to help all users of these figures navigate this complex area.

- Download and read the full Transaction costs manual here (25 pages).

Our manual covers what people need to know about transaction costs, their connection to best execution, their measurement, and their relationship to returns. It also outlines how to use reported figures and, equally important, how not to use them. We outline the key points in seven Q&As below.


1. Why do transaction costs exist?

Transaction costs are a necessary part of investing. In order to achieve a return one must trade, that is, buy or sell a security, at some point, even if it is a buy and hold investment. Trading is costly. It includes payments to the government in the form of a transaction tax and to those who facilitate the transaction (commission). These payments are “explicit costs”. There are other costs, which do not reflect explicit payments to third parties but are “implicit”, and which reflect a value loss due to market friction. These include the difference in the price one pays when buying and the price one gets when selling the same security (spread), the impact that any given transaction has on the market due to the way each order affects supply and demand (market impact) and the market risk involved in the time it takes for a transaction to complete (delay cost).

At a fund level, transactions can be triggered for many reasons. The portfolio manager may make an active investment decision to buy or sell a holding. There could be market movements that affect the relative value of a portfolio’s holdings and a need to revert to a specific balance. Investors may be buying or selling units of a fund so that the fund itself has to replicate the purchase or sale of all the underlying holdings to reflect this activity.

The reported transaction cost figures do not convey the reason behind the transactions nor whether the explicit or implicit component played the bigger role. However, the reported return figures always reflect all the transaction costs.


2. What is the best way to measure transaction costs?

There is no single “best way” to measure total transaction costs. Measuring explicit transaction costs is straightforward as these involve specific payments to third parties for each transaction. Estimating (which is more apt than “measuring”) implicit transaction costs is a science in itself. The value lost to the market during a trade is a function of many interconnected factors such as: the price one is willing to trade at, the size of that trade, the number of participants in the market and the price at which each one is willing to trade, the total volume that is being traded, the expected price movement, the trading decisions of everyone else independently of one’s own trade, whether others in the market get a hint of one’s intention to trade in a certain way before the trade is completed, etc.

There is no single methodology which can summarise all these factors in one meaningful number. There are multiple ways to capture them and each one has its advantages and disadvantages. Regulation has largely opted for a methodology that tries to capture information leakage and market impact but the mechanics of it also capture the impact of the trades that others place in the market during the time the trade is moving from order to completion. The resulting figures include this market “noise” and are not necessarily reflective of the investment experience.

3. Does best execution mean minimising transaction costs?

No, or at least, not only. Best execution means executing a trade in a way that achieves the best possible result. Cost is one component of many, including the likelihood of the trade being completed, how fast it is completed, the price at which it was completed, given the prevailing conditions in the market such as liquidity. The trader has an incentive to minimise cost but not at the expense of the overall outcome, for example, by using a broker who charges very low commission but has a poor record on settling trades or not placing a large order to try and avoid market impact but then to lose out by not investing in a security that has great potential to increase in value.

Much of what contributes to best execution goes beyond the individual trade and its cost. It involves the ability to consider all possible outcomes and the opportunity costs of choosing one specific course, having the necessary infrastructure and access to trading venues and counterparties, and having effective governance in place to review regularly policies and outcomes.

In short, cost is the tip of the best execution iceberg.

4. Are higher transaction costs bad?

Not necessarily. Transaction costs are not like the management fee and other ongoing charges where one could say that higher charges mean a more expensive fund. A fund incurring high transaction costs can be the result of the fund trading in securities with traditionally high commission or trading during volatile conditions and/or in illiquid markets. A high transaction cost number may be the outcome of a one-off shift in investment strategy or it may reflect a large number of client requests to buy or sell fund units. It may even be the outcome of one expensive transaction which was crucial for achieving a high return, such as divesting a large position in a company before the value of that company drops significantly.

The context in which transaction costs are incurred is crucial in understanding the figure itself.

5. Do transaction costs reduce returns?

Not necessarily. It is always true that for two funds with similar holdings and thus, similar gross returns, the one with the higher ongoing charges will have the lower net return. If a fund decides to lower the ongoing charges then automatically (that is, all else being equal) the net return increases. That is not true for transaction costs. If a fund decides to lower the transaction costs by transacting less or transacting in more securities with low trading costs then the fund will end up with different holdings and thus, a different return. It cannot be predicted in advance whether this return will be higher or lower.

Funds with high transaction costs can achieve both high and low returns so going for a fund with a lower transaction cost does not guarantee a higher return.


6. Can one compare transaction costs across funds?

No. This is not just due to firms using different methods to estimate implicit transaction costs. Even if there were one perfect way to estimate total transaction costs and everyone was using it, it would still be hard to draw any meaningful conclusion by comparing just the transaction costs.

Every trade is unique. It can be triggered by different events, concern a different security, be subject to different liquidity levels etc. The specific environment in which each trade is executed will never repeat itself. This goes back to the point that high is not necessarily bad. Hence, looking at transaction costs should always be done in the context of the prevailing environment, the broader fund characteristics (asset type, risk, etc.) and the investment outcome.

This remains relevant when people look at “all-in” cost figures, which aggregate ongoing charges (that can be compared and have a linear relationship with returns) with transaction costs (that cannot be compared in isolation and have no linear or indeed any type of relationship with returns).

7. Does this mean one can’t use transaction costs?

Transaction cost figures can be used but not in isolation. People need to think of the broader context and consider other fund attributes such as:

  • In what securities does the fund usually trade and what were the market conditions in the period reported?
  • Is the fund actively or passively managed?
  • What is the net return?
  • What are the ongoing charges?
  • What is the risk?
  • How are transaction costs estimated?
  • Are there any anti-dilution mechanisms in place?

This is by no means an exclusive list and there is no specific meaning in the order in which the above attributes are listed. It simply serves to show how much of a context is needed to form an opinion about the level of transaction costs.

Read the full report

The transaction costs manual 25 pages | 1,637 kb