Using HP-Filtered House Prices to Identify Bubble Components

Summary

The document proposes separating house prices into a fundamental component and a bubble component using the Hodrick–Prescott filter. It describes a threshold based on the standard deviation of the filtered cyclical component: price deviations beyond that threshold would be flagged as bubbles. The approach is linked to a cited empirical measure of asset-price bubbles, but the document does not provide an implementation or results for housing data.

The author asks whether this rule is reasonable and seeks other methods to use as benchmarks. No replies or comparisons are included, so the proposal’s validity, sensitivity to filter settings, and performance against alternatives remain unresolved. The text is best read as a research question about operationalizing bubble detection rather than evidence that this threshold reliably identifies mispricing.

Key ideas

• The proposal applies the Hodrick–Prescott filter to split house prices into trend and cyclical components.
• It flags price deviations beyond a standard-deviation threshold as a possible bubble.
• The document cites an existing empirical approach but gives no housing-market results.
• It asks for alternative bubble-measurement methods to use as benchmarks.

Tags

# Calculating the fundamental value of house price to separate bubble component from the price


# Calculating the fundamental value of house price to separate bubble component from the price












The bubble in asset price is defined as the deviation of the asset value from its fundamentals, empirically Mendoza and Terrones (2008) measure the bubble as the deviation of an asset price from the standard deviation of its cyclical component. The Hodrick-Prescott is used to derive the cyclical component. Based on this procedure, I want to divide the house price to fundamental and bubble components in which the portion of the prices that is above the standard deviation of the cyclical component is flagged as a bubble. does this seem reasonable?

In addition, I would like to do the same with other methods so I can use it as a benchmark, but I don't know if there is one. Could anyone please suggest other methods used for this purpose?

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.