No abstract

No abstract

The econometric literature on measuring returns on intangible capital is vast, but we still know little about the effects on productivity of different types of intellectual capital (R&D and patents) and customer capital (trademarks and advertising). The aim of this paper is to estimate the marginal productivity of different types of intangibles by relying on the theoretical framework of the production function, which we apply to a large panel of Italian companies. To this end, the European accounting system makes it possible to compare the impact on productivity of intangibles measured from expenditures (as usual in Anglo-American studies) with the impact of intangible assets reported by companies in their balance sheets (a measure which is available in the Italian context, for example, but less common in the literature). Our results contribute two main findings to the literature. First, among the intangible components, the highest marginal productivity is that of intellectual capital, customer capital and intangible assets. Second, the use of accounting information on intangible investments is crucial to find high effects of intangible assets on productivity, while intangibles measured from expenses seem to play a more limited role. Preliminary results obtained from sub-samples mimicking the presence of spillovers deliver higher effects of intellectual capital on productivity, suggesting that intangibles' social value is larger than the part we can estimate with individual firm data.

This paper proposes an empirical model for the modi®ed pecking order theory (MPO) in which both trade-o¨(TO) and pecking order (PO) models are nested. The MPO model is speci®ed as an error-correction mechanism and applied to a vast panel data-set. Unlike previously estimated ®nancial models, it avoids a number of problems: the mis-speci®cation of dynamics, the approximation of the target leverage using the historical mean, the constrained estimation of the free cash¯ow components in a unique parameter. The MPO model is particularly good at explaining``hybrid'' systems (neither marketbased nor bank-based) such as the Italian one, in which companies are a mixture of two types: TO-type ®rms with a long-term optimal debt ratio towards which they converge; PO-type ®rms for whom the short-term availability of internal funds for investment may interfere with the process of adjustment towards the target leverage. Finally, the MPO model enables us to separately test the individual relevance of each of the restricted (``pure'') TO and PO models: results con®rm their mis-speci®cation and clearly point towards the excellent empirical performance of the MPO model.

The problem of instrument proliferation and its consequencesoverfitting of the endogenous explanatory variables, biased instrumental-variables and generalized method of moments estimators, and weakening of the power of the overidentification tests-are well known. This article introduces a statistical method to reduce the instrument count. Principal component analysis is applied on the instrument matrix, and the principal-component analysis scores are used as instruments for the panel generalized method of moments estimation. This strategy is implemented through the new command pca2. 1995, Staiger andStock 1997) and by the correlation between the sample moments and the estimated optimal weighting matrix: sampling errors are magnified in the weighting matrix (Altonji and Segal 1996). Poor estimates of the variance-covariance matrix of the moments lower the power of the specification tests such as the Sargan/Hansen test for overidentifying restrictions, which suffers from a severe underrejection problem (Sargan 1958;Anderson and Sørenson 1996;Bowsher 2002). A strategy to reduce the instrument count in panel GMMOverall, such evidence supports the importance of properly addressing instrument proliferation, although this problem is often overlooked in empirical analyses; indeed, strategies to reduce the instrument count such as lag truncation and collapse (Roodman 2009a) are used only seldom in empirical applications. In addition to these two operational strategies, already implemented in Stata, the selection of correct or optimal instruments from a large set of potential candidates has received attention in a broader, more theoretical perspective. This latter stream of literature has developed statistically grounded methods for consistently selecting the GMM conditions and has investigated the statistical properties of the estimators exploiting the resulting sets of moments. Relevant contributions in this area include the information criteria methods and downward and upward testing procedures of Andrews (1999) and Andrews and Lu (2001), the Lasso-type instrument selection of Caner (2009) and Belloni et al. (2012), and the GMM shrinkage methods of Liao (2013). A recent contribution of Caner, Maasoumi, and Riquelme (2014) provides an extensive overview and a simulation-based comparison of moment-selection approaches.Our aim in this article is to tackle the issue of instrument proliferation by providing a statistically grounded and directly implementable procedure that reduces the instrument count. We advocate the use of principal components analysis (PCA) of the instrument matrix as a way to shrink the available instruments into a set of linear combinations of the original variables (the scores of the PCA). The weights used in such orthogonal combinations follow from the main features of the data and reflect the contribution of each variable to the total observed variability. We label this strategy "principal components instrumental variables reduction" (PCIVR). 1 PCIVR exploits the same tool as that found in Doran and ...