Download presentation
Presentation is loading. Please wait.
Published byBranden Ford Modified over 8 years ago
1
Understanding Growth in Phoenix Can the FFR Help Explain Economic Activity in the MSA? Charlotte D. Smith MIAMI UNIVERSITY
2
Variable of Interest: Economic Growth Index of Phoenix, AZ PHXAGRIDX : Economic Conditions Index for Phoenix-Mesa-Scottsdale, AZ (MSA) Source: Federal Reserve Bank of St. Louis Note: first-differenced series for stationarity
3
QUESTION: Is the FFR a stationary series? Source: FRED
4
Stationarity of the FFR Dickey-fuller unit root tests imply non-stationarity Null Hypothesis: series follows a Random Walk P-value greater 0.05 for 1, 2, and 4 lags p-value > 0.05
5
Stationarity of the FFR: Debate in the literature Bec and Bassil (2009) Stationary with 2 breaks Figueiredo and Shikida (2010) Linear models inappropriate Heavy tails Existence of outliers
6
Stationarity of the FFR: First-difference to stationarize Notable improvement in ACF
7
Descriptive Statistics Source: FRED Fat tails in FFR distribution
8
Source: FRED Fat tail Left-hand side Negatively skewed Distribution of FFR
9
Source: FRED Positively correlated Significant? “Eyeball” outliers Relationship between FFR and Phoenix Growth Rate
10
Regression Results FFR positively, significantly correlated with Phoenix growth rate Adjusted R-squared quite low (0.05) Simple Regression Analysis
11
Source: FRED Concerns: 1) Nonlinearity 2) Heteroskedasticity Concentration at (0,0) Recall mean and median values of FFR ~0.0 Patterns?
12
Identifying Outliers 19 outliers Studentized residuals that are > 1.96 Source: FRED
13
Influence of the Outliers Explanatory power increases Regression Results: Full Sample Regression Results: Excluding Outliers
14
Considering Functional Forms: Log-level and Log-log Explanatory power null Regression Results: Log-level Regression Results: Log-log
15
Considering Nonlinearity: Including Quadratic Term Regression Results: PHX Growth = FFR + FFR 2
16
Using ANOVA to Determine Better Model Null Hypothesis FFR has a constant marginal effect on Phoenix growth rate Include FFR 2 Reject H 0
17
Revisiting Residual Plot: Evidence of Nonlinearity? Source: FRED
18
Before January 2000 120 observations After January 2000 188 observations
19
Groupwise Regressions Explanatory power greater for later observations Regression Results: Before January 2000 Regression Results: After January 2000
20
Conclusion of Findings Both series appear to follow Random Walk processes First-difference to stationarize Positive relationship between growth and FFR Relationship strengthens using more recent data (2000-2015) Better to use nonlinear model (quadratic)
21
References Claudio D Shikida and Erik A Figueiredo, (2010). "Is the Federal Funds rate stationary? New evidence from P-ADF" Revista de Economia e Administração Vol. 9 Iss. 1 Frédérique Bec andCharbel Bassil, (2009). “Federal Funds Rate Stationarity: New Evidence'', Economics Bulletin, Vol. 29 no.2 pp. 867-872 Board of Governors of the Federal Reserve System (US), Effective Federal Funds Rate [FEDFUNDS], retrieved from FRED, Federal Reserve Bank of St. Louis https://research.stlouisfed.org/fred2/seri es/FEDFUNDS, May 24, 2016. Charlotte D. Smith ECO 690 May 24, 2016
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.