Search

ABSTRACT: This study investigates the relation between the corporate bond market and the main debt financing alternatives available to firms in Brazil from 2004 to 2014. In this period, three main sources of debt financing were available: loans from commercial banks, loans from the National Development Bank, and the issuance of international bonds. We analyze the characteristics of each alternative, as well as their advantages and disadvantages in relation to corporate bond issuance, considering three main aspects: i) total cost, which includes the interest rate, taxes and fixed costs; ii) maturity; and iii) contracting speed. The results corroborate the initial hypothesis that, although the alternative debt sources competed with corporate bonds, there was also a complementary relation among them.

This paper aims to create a bondholders’ protection index (BPI) and to investigate what the influence of this index would be on multimarket funds’ allocation in corporate bonds. Understanding this relation is relevant because only about 1.36% of multimarket funds’ portfolios correspond to debentures. This study advances the literature by covering a topic little discussed in a Brazilian context, proposing the creation of a BPI, which would be related to the number of automatic maturity clauses, which guarantee immediate payment to bondholders in cases of the rupture of a contract. This research comprised 926 debentures series issued in Brazil from 2009 to 2017, and 1,753 multimarket funds, which allocated some portfolios’ percentage in these securities. In creating the BPI, we contemplated 15 restrictive clauses, which the most common correspond to negligent business performance, liquidation, dissolution and bankruptcy, and restrictions related to company structure. Moreover, we examined less common restrictive clauses as well, including indebtedness policy, shares issuance and amortization, and ratings downgrading. Regarding data analysis, we employed multiple linear regression models, with pooled estimators, applying the standard error correction by White’s robust matrix (1980). The main results suggest that BPI positively effects multimarket funds’ allocation in debentures. Furthermore, this influence is more intense in indentures with higher number of clauses with automatic maturity. Thus, this study contributes to literature about restrictive clauses, since it demonstrates that debentures’ flexible and adaptable structure seems to be interesting for the main bondholders in Brazil.

Neste trabalho, teve-se como objetivo identificar o impacto do risco de liquidez nos retornos excedentes esperados das debêntures no mercado secundário brasileiro. Foram realizadas análises de regressão em painel desbalanceado com dados semestrais de 101 debêntures ao longo de oito semestres (primeiro semestre de 2006 ao segundo semestre de 2009), totalizando 382 observações. Sete proxies (spread de compra e venda, %zero returns, idade, volume de emissão, valor nominal de emissão, quantidade emitida e %tempo) foram utilizadas para testar o impacto do risco de liquidez nos yield spreads das debêntures. O yield spread foi controlado por até dez outras variáveis determinantes que não a liquidez (fator de juros, fator de crédito, taxa livre de risco, rating, duration, quatro variáveis contábeis e volatilidade de equity). A hipótese nula de que não há prêmio de liquidez para o mercado secundário de debêntures no Brasil foi rejeitada apenas para três das sete proxies (spread de compra e venda, valor nominal de emissão e quantidade emitida). Os prêmios encontrados são bastante baixos (1,9 basis point para cada 100 basis point de incremento no spread de compra e venda, 0,5 basis point para um aumento de 1% no valor do valor nominal de emissão e 0,17 basis point para cada menos 1.000 debêntures emitidas). De qualquer forma, houve perda na eficiência das proxies de liquidez após correção das autocorrelações e potenciais endogeneidades, seja por meio da inclusão de efeitos fixos, da análise de primeiras diferenças ou da utilização de um sistema de três equações. Esses resultados apontam para a suspeita de que o risco de liquidez não é um fator importante na composição das expectativas dos investidores no mercado secundário de debêntures.En este estudio se tuvo como objetivo identificar el impacto del riesgo de liquidez en los rendimientos en exceso esperados de las obligaciones en el mercado secundario brasileño. Se realizaron análisis de regresión con datos de panel desbalanceado de 101 obligaciones durante ocho semestres (del primero semestre de 2006 al segundo semestre de 2009), con un total de 382 observaciones. Se utilizaron siete proxies (spread de compra y venta, %cero returns, edad, volumen de emisión, valor nominal de emisión, cantidad emitida y %tiempo) para evaluar el impacto del riesgo de liquidez en los rendimientos de las obligaciones. El diferencial de rendimiento fue controlado por hasta diez otras variables determinantes a excepción de la liquidez (interés, riesgo crédito, tasa libre de riesgo, rating, duration, cuatro variables contables y equity volatility). La hipótesis nula de que no existe una prima de liquidez para el mercado secundario de obligaciones en Brasil fue rechazada sólo para tres de los siete proxies (spread de compra y venta, valor nominal de emisión y cantidad emitida). Los premios encontrados son muy bajos (1,9 basis point para cada 100 basis point de aumento del spread de compra y venta, 0,5 basis point a un aumento del 1% en el valor nominal de emisión y 0,17 basis point para al menos 1000 debentures emitidos). De todos modos hubo una pérdida en la eficiencia de los proxies de liquidez después de la corrección de las autocorrelaciones y potenciales endogeneidades, ya sea por medio de la inclusión de efectos fijos, del análisis de las primeras diferencias o del uso de un sistema de tres ecuaciones. Estos resultados apuntan hacia la sospecha de que el riesgo de liquidez no es un factor importante en la composición de las expectativas de los inversores en el mercado secundario de bonos.This study aims to identify the impact of liquidity risk on expected excess returns of Brazilian corporate bonds in the secondary market. We performed a battery of regression analysis with semiannual unbalanced panel data for 101 securities over eight semesters (first semester of 2006 to the second half of 2009), totaling 382 observations. Seven proxies (bid/ask spread, %zeroreturns, age, amount outstanding, face value, number of bonds, and %time) were used to test the impact of liquidity risk in the yield spreads. Ten other variables (Slope Factor, Credit Risk Factor, risk-free rate, rating, duration, four accounting variables, and equity volatility) were used as yield spread determinants. The null hypothesis that there is no liquidity premium for bonds in the Brazilian secondary market was rejected for only three proxies (bid/ask spread, face value, and number of bonds). The premiums observed were quite small (1.9 basis points for each 100-basis-point increase in the bid/ask spread, 0.5 basis point to 1% face value increase, and 0.17 basis point for each issue of at least 1000 bonds). In both cases we observed a loss of efficiency of liquidity proxies following correction for autocorrelations and potential endogeneity, either through the inclusion of fixed effects, first differences analysis, or simultaneous equation analysis. These results point to the fact that liquidity risks may not be of great importance to the expectations of the Brazilian corporate bond secondary market.

O objetivo nesse trabalho é a construção da Estrutura a Termo de Taxa de Juros (ETTJ) das debêntures que se encontram disponíveis no mercado brasileiro, usando o modelo Nelson-Siegel (1987). As curvas de juros são divididas de acordo com o tipo de indexador e com a classificação de risco atribuída às debêntures. Por fim, foi possível mensurar o spread que existe entre o mercado de títulos privados (debêntures) e os títulos públicos.El objetivo en este trabajo es determinar la estructura temporal de tipos de interés (ETTI) de los bonos que se encuentran disponibles en el mercado brasileño, por medio del uso del modelo de Nelson-Siegel (1987). Las curvas de rendimientos fueron divididas de acuerdo con el tipo de índice y con la clasificación de riesgo atribuida a los bonos. Fue posible medir el spread entre el mercado de bonos corporativos (debentures) y el mercado de los títulos del Estado.The aim of this work is to establish a term structure of interest rates for corporate bonds in the Brazilian market, using the Nelson-Siegel model (1987). The yield curves were divided by the type of index and the rating level. It was possible to measure the spread between the corporate bonds market and the government bonds market.