Event Studies of Policy Rate Cuts and Long-Term Bond Yields
Summary
The document considers how to separate the effect of a policy-rate cut from the effect of a quantitative-easing announcement on long-term yields. The case is a Polish central bank announcement that combined the two actions, making an event-study estimate of the QE effect difficult to interpret. The question proposes regressing changes in the ten-year yield on changes in the policy rate and its square.
The answer points to research on the causal relationship between short- and long-term interest rates and recommends considering the time window used to calculate yield changes. It notes that estimated sensitivity can depend on the horizon, citing work that compares different frequencies. The discussion offers a literature lead rather than a fully specified alternative model or identification strategy, and leaves other regression concerns unresolved.
Key ideas
- A simultaneous policy-rate cut and QE announcement complicate efforts to identify their separate effects on long-term yields.
- A proposed model relates ten-year yield changes to policy-rate changes and their squared values.
- The estimated sensitivity of long-term yields to short rates can depend on the measurement horizon.
- Relevant research on causal relationships and frequency effects can guide model design.
Tags
Full text
# Estimating the relationship between short-term intretes rates and 10Y bond yields
# Estimating the relationship between short-term intretes rates and 10Y bond yields
On the 16th of March 2020, the Polish Central Bank announced its first-ever round of Quantitative Easing. I am conducting an event study on how this announcement impacted the term structure.
The main obstacle is the fact that in the same press announcement the central bank also lowered its policy rate by 50 bps. My goal is to get an estimate of the reduction in the 10Y bond yield that would follow ONLY from the policy rate cut, based on the historical data. I have considered estimating the following equation:
\begin{align*} \Delta10Y\_Yield_{t}=\beta_{0}+\beta_{1}\Delta Policy\_Rate_{t}+\beta_{2}\Delta Policy\_Rate_{t}^{2}+u_{t} \end{align*}
What could be an alternative approach? Are you aware of any relevant literature?
## Answer by Sharad (score 1, accepted)
https://quant.stackexchange.com/a/71500
A good discussion about the complex and still unsettled relationship between short- and long-term interest rates (along with a literature review) can be found here in Section 3 of the paper below:
The causal relationship between short- and long-term interest rates
Some of the issues that you might want to take a look at in the context of your regression model:
- Time window over which you will calculate your delta: the sensitivity of long-term yields to changes in short-term rates shows dependence on the time horizon considered. See The sensitivity of Long-Term Interest Rates: A Tale of Two Frequencies
There are many other issues to consider but this should hopefully give you a good starting point.
(Apologies for not reading your question carefully the first time.)Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.