Ahead of the Curve | Prattle Data Anticipates Rate Expectations

  • Wednesday, September 28, 2016

Plotted alongside Bloomberg’s World Interest Rate Probability function* (WIRP), Prattle’s Fed speaker data** quantitatively illustrates the effect the Fed’s tone has on market expectations.

Figure 1: Fed Speaker Trend and the WIRP


Figure 1 demonstrates that Fed speaker sentiment correlates with–and often leads–market expectations. In late April, weeks before market expectations rise, sentiment begins to climb. The sentiment then drops off sharply in early May–a trend that once again takes weeks to materialize in the WIRP. This dynamic continues through the rest of the time frame, with Fed sentiment tracking with or leading the market.

Prattle data can be broken down by communication type, allowing users to discover how the different channels of Fed communication influence various market segments. The trend illustrated in Figure 1 was constructed using only Fed policymaker speech scores. When the scope is expanded to include individual speakers, minutes, and FOMC statements, the relationship between the two trends*** is even more obvious:

Figure 2: Fed Residual Trend and the WIRP


The fact that Fed sentiment not only correlates with but leads market expectations is especially indicative of the value of Prattle’s central bank data as a trade signal. Traders with access to Prattle’s data can effectively forecast market expectations, allowing them to ideally position themselves for upcoming price movements.

Interested in Prattle data? You can try it for free.

Click here to Request a Trial

* The WIRP is a popular measure of the market’s interest rate expectation.

** Prattle’s models are based on the historical relationship between central bank language and market reaction, which is used as basis of evaluation for future communications. The scores are normalized around zero and range between -2 and 2, negative numbers indicating dovishness and positive numbers indicating hawkishness.

*** It should also be noted that the sentiment trend in Figure 2 was generated using residual scores. Residual scores represent the tone of a communication compared to the rolling, 12-month average for that individual communication type or speaker. Raw scores represent the tone of a communication compared to the average of all communications.

Disclaimer: the forecasts provided herein are based upon sources believed by Prattle Analytics, LLC D/B/A Prattle, to be reliable and to be developed from models which are generally accepted as methods for producing economic forecasts.

Prattle cannot guarantee the accuracy or completeness of the information upon which this Report and such forecasts are based. This Report does not purport to disclose any risks or benefits of entering into particular transactions and should not be construed as advice with regard to any specific investment or instance. The opinions and judgments expressed within this Report made as of this date are subject to change without notice.

Copyright © 2016 Prattle Analytics, LLC.