I am working on this corporate panel data model: LEVERAGE_it = PROFITABILITY_it + NON-DEBT-TAX-SHIELD_it + SIZE_it + LIQUIDITY_it + TOBIN_Q_it + TANGIBILITY_it + u_it. Where:
leverage = long term debt/total assets
profitability = cash flows/total asset
non debt tax shield = depreciation/total asset
size = log (sales)
liquidity = current assets/current liabilities
tobin_q = mkt capitalization/total assets
tangibility = tangible fixed asset/total assets
What can I say about the exogeneity condition? Can I assume that expected value of the covariance between error term u_it and of X_i is zero? Why? A lot of papers make this assumption but do not explain why.
Thank in advance for your response.
zero covariance between between Xi and ui implies zero correlation between Xi and ui. This assumption means that there exists no linear association between ui and Xi. It assumes homoskedastcity
Hi Conrad
Please, read the following answer with details,
https://www.researchgate.net/deref/http%3A%2F%2Fwww.stata.com%2Fstatalist%2Farchive%2F2011-08%2Fmsg01063.html
The exogeneity condition implies that the coefficients obtained from the panel data model can be used for presenting a causal story. Generally, this assumption is likely to be violated in many scenarios due to simultaneity, reverse causality or omitted variables in the system. Hence, assuming that the correlation between X and the error term is zero is not always recommended for empirical papers. At least the author should convince the readers why the assumption is valid in the given context.
For scientific and practical purposes, what are you trying to measure? Panel data are mostly used to accommodate more information, more variability, and more efficiency than pure time-series data or cross-sectional data. Panel data heterogeneity helps with evaluating intervention effect. Be mindful that correlation is not the same as causality.
Thank you all for the responses.
My purpose was to analyze the determinants of capital structure of italian listed non-financial firms. In order to do so I read some others similar papers such as:
- Determinants of capital structure of Chinese-listed companies (Jean J. Chen, 2003)
- The Determinants of Capital Structure of Large Non-Financial Listed Firms in
- Nigeria (Rafiu Oyesola Salawu, Akinlolu Ayodeji Agboola, 2008)
- Determinants of capital structure, An empirical study of firms in manufacturing industry of Pakistan (Nadeem Ahmed Sheikh, Zongjun Wang, 2011)
- Determinants of capital structure: new evidence from Spanish panel data (Alberto de Miguel, Julio Pindado, 2000)
- Determinants of Capital Structure and Adjustment to Long Run Target: Evidence from UK Company Panel Data (Aydin Ozkan, 2001)
These papers consider panel data models for studying firms capital structure. I am dealing with Italian firms, and trying to fit the panel data model (which is very similar among the papers) to the Italian firms dataset. Is my model consistent and well specified? The exogeneity condition stands?
Shreya Biswas the papers' authors above do not say anything about the exogeneity condition to convince the readers. To me, for example, profitability is not likely to be exogenous. So this is my point: why all this papers make and do not discuss this assumption? If the exogeneity assumption is violated, than the model is not well specified and all the results would be inconsistent. Am I wrong?
Sunny Nyemah I explained above what I am trying to measure. Correlation and capture the linear relationship. Casuality capture relationship (linear and not) between two variables. Am I correct? Exogenity implies also there is no casuality?
Thank you all.
In general, to explore the heterogeneity in a panel data, you need to first identify the group classifications of data in line with variables of interest.
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation