We examine the relationship between the monetary policy uncertainty (MPU) and the price bubbles in U.S. oil futures, including WTI crude oil future, heating oil future, and gasoline future. The Log Periodic Power Law Singularity (LPPLS) model is firstly used to analyze and validate the price bubbles of U.S. oil futures. We find that (1) the timing of local peaks of MPU closely aligns with the occurrence of price bubbles in U.S. oil futures, (2) MPU is significantly correlated with crude oil future price bubble but not with heating oil future (or gasoline future) price bubble, and MPU significantly amplify the magnitude of price bubble risk for crude oil future, (3) MPU is an effective set of factors for machine-learning based identification of all three U.S. oil future price bubbles in the non-linear model, even though the impact of MPU on heating oil (or gasoline) future price bubble is not significant in logit regression as the linear model. Collectively, our findings highlight the significance of MPU in relation to price bubbles in U.S. oil futures, offering new insights for investors and policymakers to explore the (non)-linear relationship between MPU and price bubbles in energy markets.

中文翻译：

---

货币政策的不确定性和能源市场的价格泡沫

我们研究了货币政策不确定性（MPU）与美国石油期货（包括 WTI 原油期货、取暖油期货和汽油期货）价格泡沫之间的关系。对数周期幂律奇点（LPPLS）模型首先用于分析和验证美国石油期货价格泡沫。我们发现（1）MPU 局部峰值的时间与美国石油期货价格泡沫的发生密切相关，（2）MPU 与原油期货价格泡沫显​​着相关，但与取暖油期货（或汽油期货）不显着相关。 (3) MPU 是非线性模型中基于机器学习识别所有三种美国石油期货价格泡沫的有效因素集，即使尽管 MPU 对取暖油（或汽油）未来价格泡沫的影响在 Logit 回归中并不像线性模型那样显着。总的来说，我们的研究结果强调了 MPU 对于美国石油期货价格泡沫的重要性，为投资者和政策制定者探索 MPU 与能源市场价格泡沫之间的（非线性）关系提供了新的见解。