Interpreting the Latent Dynamic Factors By Threshold FAVAR Model

Approximate factor models with restrictions on the loadings may be interesting both for structural analysis (simpler structures are easier to interpret) and forecasting (parsimonious models typically deliver superior forecasting performances). However, the issue is largely unexplored. In particular, no currently available test is entirely suitable for the empirically important case of non-stationary data. Building on the intuition that defactoring the data under a correct set of restrictions will lower the number of factors, we propose a procedure based on the comparison of the number of factors selected for the raw and de-factored data. To control and reduce the risk of rejecting valid constraints we develop a bootstrap procedure, shown analytically to be asymptotically valid and by simulation to have good small sample properties.JEL: C12, C33, C55

show abstract

“…so that: (17) which shows that the resampled common component (λ i2 F 2t ) * also depends upon u e ε only.…”

Section: Appendix Proof Of Lemmamentioning

confidence: 87%

Evaluating restricted common factor models for non-stationary data

Iorio

Fachin

2021

Econometrics and Statistics

View full text Add to dashboard Cite

show abstract

“…One possibility is to use the Bai-Ng information criterion to determine the number of factors present in the macroeconomic data vector X t . However, as shown by Tuzcuoglu and Hoke (2016), diff erent time spans might lead to diff erent numbers of factors. Also, the Bai-Ng criterion does not solve the issue of how many factors we should include in the VAR model itself.…”

Section: Appendixmentioning

confidence: 99%

Placing the Czech Shadow Banking Sector under the Light

Hodula¹,

Macháček²,

Melecky³

2020

Prague Economic Papers

View full text Add to dashboard Cite

The size of the shadow banking sector (SBS) has more than doubled in the Czech Republic over the last decade. This places a potential burden on policy makers. On the one hand, the SBS complements regular banking by expanding access to credit and investments, enabling better risk sharing and maturity transformation, and supporting market liquidity. On the other hand, SBS activities can put the stability of the financial system at risk and amplify its procyclicality by exacerbating the buildup of leverage and asset price bubbles. We implement a FAVAR model of the Czech economy to determine the impact of macroeconomic factors on the SBS. We find that the SBS: (i) is sensitive to changes in market interest rates and term spread; (ii) exhibits great procyclicality; (iii) can act as a complement to regular banking and satisfy some additional demand for credit. We also define some potential risks of continued growth of the SBS, linked to our empirical evidence.

show abstract

Predictive Regressions Under Asymmetric Loss: Factor Augmentation and Model Selection

Demetrescu

Hoke

2018

SSRN Journal

View full text Add to dashboard Cite

Interpreting the Latent Dynamic Factors By Threshold FAVAR Model

Cited by 3 publications

References 0 publications

Evaluating restricted common factor models for non-stationary data

Evaluating restricted common factor models for non-stationary data

Placing the Czech Shadow Banking Sector under the Light

Predictive Regressions Under Asymmetric Loss: Factor Augmentation and Model Selection

Contact Info

Product

Resources

About