How relevant is the relevance criterion on economic releases?
The importance of macroeconomic data is highlighted by the particular attention made by the financial press, especially to scheduled economic news. Every day, Bloomberg and other financial and economic data agencies publish news to be released in each country during the next few days. In addition, they collect forecasts of macroeconomic indicators from some qualified economists, such as banks or financial analysts, and publish their estimates of the news value. They also give us an idea about the importance of each indicator across the relevance criterion.
Relevance ratings appear as level icons in Bloomberg, which are graphic representations, or as an icon with different colors in other websites. Anyway, the relevance criterion is made for all scheduled economic news all over the world and is always calculated as the number of alerts that are set for the corresponding economic event relative to all alerts set for all events in the selected country. In Bloomberg, Each supplementary bar within the icon indicates a higher percentile range. It is possible to set multiple alert types (audio, message, pop ups…), and Bloomberg takes this into account by normalizing results.
The relevance criterion is a good indicator of the weight of each scheduled economic news release and it one of the sole indicator available in the industry to measure the ex-ante relative popularity of the event. But one must take into consideration three main aspects when using it to detect market movers:
1) Static measure: It is calculated using the number of alerts at some point in time.
2) No international comparison: as the relevance criterion is based on the total alerts per country, it is by construction a local rather than an international indicator, which allows for comparisons of the importance of macroeconomic data within a country, but not between different countries. It can be of interest to classify scheduled economic news in one specific country, especially when many announcements are made at the same time and coincide with some movements or jumps in markets.
3) Because the criterion is based on the number of alerts and not on the importance of the indicator, there should be a relationship between alerts and importance. A high relevance rate implies that many users add alerts on this schedule event and thus, they consider it as an important source of information when valuing prices. But the reverse is not always verified, which means that a small relevance rate does not necessarily show that the economic event is irrelevant, especially when it is a new indicator. To illustrate this, table (1) reports all the tickers about the FOMC meetings. Though this kind of information is crucial for markets as shown above, the US FOMC treasury Purchase Program is only with 3.53% of relevance. This may be due to the fact that it is a new added ticker and that Bloomberg users are not necessarily informed about this indicator.