Econometrics

Group2 BMAN70211

Abstract

  This research aims to examine determinants of debt maturity, investigate the trend of using a short-term debt in US firms, estimate the speed of adjustments (SOA) towards target debt maturity. We find that Volatility, Abnormal earnings, and Market to Book have a positive relationship with short-maturity debt, while R&D expenses, natural logarithm of size, asset maturity, term structure and leverage are inversely related to short-term debt. Additionally, US firms tend to use more short-term debt after 2008, and the SOA is possibly  from 0.4221 to 0.7496.  

1. Introduction

Choosing short-or long-maturity debt is an important financial decision for companies (Custodio et al., 2012), it is because using different maturity debt may lead to distinct consequences of firms financial operations, so it draws our interest to provide an empirical analysis concerning the corporate debt maturity.

The objectives of this paper, firstly, is to find out determinants of debt maturity; secondly, is to investigate whether there is an upward trend of using a short-term debt in US firms or not; thirdly, is to estimate the speed of adjustments (SOA) towards target debt maturity conducted by firms.

In order to achieve the aforementioned objectives, we study the evolution of debt maturity in US firms from 2000 to 2014, and we use several econometric models and methods to analyze statistically.

We choose short-term debt maturity less than three years as the dependent variable, and use market-to-book ration, abnormal earnings, research and development (R&D) expenses, leverage, natural logarithm of size, the natural logarithm of size squared, asset maturity, asset volatility and term structure as control variables.

By estimating pooled OLS model (POLS), we find that all the control variables given are significant at 5 % level, while the model including the industry dummies is more accurate. In order to eliminate the fixed effect, we use first differencing (FD) model and Fixed-effects (within group) estimation, both of the results indicate that the original POLS is biased and inconsistent. Additionally, we use the Random effect model to estimate the effects of the time invariant variables, however, the results aren’t exactly similar as that of FE. After performing the Hausman test, we can conclude that the Fixed effect is at least more efficient and unbiased than the Random effect model. Besides, when we add time dummies to the baseline model, the result shows that year 2008 is the turning point with regard to firms using more short-term debt than the year before, and the continued upward trend can be mostly explained by the significant impact of financial crisis happened in 2008.

It is worth noting that based on the literature, research often argues that debt maturity and leverage are jointly determined, so to address the endogeneity problem, we choose PPE as an instrument variable (which is tested to be valid), and we re-estimate the debt maturity equation by using 2SLS regression with IV estimator. Hence, the endogeneity of leverage is controlled and we could estimate firm’s debt maturity with exogenous variables. Inaddition, the Hausman test will be introduced to illustrate the validity of instrumental variable.        Moreover, we examine the dynamics of debt maturity using the partial adjustment model, we find that POLS and FE either underestimates or overestimates the SOA, so we apply more advanced methods which are AHIV1, AHIV2, FDGMM, SYSGMM respectively, however the diagnostic tests suggest all of these estimators are invalid, then we argue that DPF may be more suitable for models.

The structure of this paper is as follows. Section 2 illustrates predictions regarding factors that may affect debt maturity which are based on the study of prior researches. Section 3 provides summary statistics of the variables considered in our research. Section 4 elaborates the econometric models, relevant diagnostic tests as well as the regression results. Section 5 presents the conclusion of our findings.

2.Literature review and predictions

R&D

The debt maturity is predicted to be correlated with research and development (R&D) expenses. The firm with high level of information asymmetry who is R&D intensive is more likely to choose short-term debt (Custodio et al., 2012). In addition, according to Myers’s model, the growth firm who invests substantial amount in R&D is more likely to use more short-term debt (Berlin, 2006). Therefore a positive relationship between short-term debt and R&D expense is expected.

Maturity matching

According to the rule of thumb of Myers’s model, the matching maturity of asset and liability is important, and many empirical studies suggested that firms with long asset maturity choose the long-term debt (Berlin, 2006). Therefore, there exists a negative relationship between asset maturity and short-term maturity.

Volatility

Asset volatility measures the credit risk of a firm (Datta et al., 2005). The firm with higher asset volatility is more likely to default, and the long-term debt market is more difficult to enter, so the firm tends to choose short-term debt instead (Custodio et al., 2012). Therefore, it is expected that the asset volatility is negatively related to the debt maturity.

Firm size

According to prior studies, large companies tend to have less asymmetric information, agency problems and better credit quality (Scherr and Hulburt, 2011). Therefore, it is easier to invest in long-term debt market. Small firms, on the other hand, are more likely exposed to agency issues related to debt (Petit and Singer, 1985). Therefore, Antoniou(2006) argues that there is a positive relation between firm size and long-term debt. Moreover, Titman and Wessels (1988) point out that short-term debt and firm size are negatively related. Following these empirical researches, we expect a negative relation between firm size and short-maturity debt. Firm size squared is also used to capture non-linear relation between credit quality and short-term debt (Diamond, 1991).

Term structure

Previous research illustrates that firms prefer long-term debt when term structure increases (Newberry and Novack, 2000). Brick and Ravid(1985)argue that firm could accelerate interest deduction and increase corporation value by issuing long-term debt with upward term structure. Based on these studies, we assume that term structure has a negative influence on short-term debt.

Abnormal earnings

According to Custodio et al (2013), firms with better-quality projects, as expressed by abnormal earnings, are more likely to issue short-term debt. Therefore, it is predicted that there is a negative relationship between abnormal earning and debt maturity.