_{1}

^{*}

Investment is an important instrument of growth and competitiveness for non financial firms. However, these firms have limited financial resources (or liquidity) at their disposal. The financial constraint is defined as a conditionality to be met in order to have access to liquidity by assuming that the information held by shareholders is perfect, and that financial markets are efficient. We have attempted in this study to analyze empirically the impact of these financial constraint on the investments of Sub-Saharan manufacturing firms. We carried out an empirical analysis of a sample of 73 firms belonging to the different manufacturing sectors listed on the stock market during the period 1998-2009, and by taking inspiration from panel data methodology. The empirical tests emphasize the fact that the manufacturing firms of Sub-Saharan countries, including the smallest ones and those with which financial institutions have no close relations, witness an environment with a strong information asymmetry between borrowers and lenders. These firms are constrained in their access to external indebtedness due to the levelling-off of indebtedness. However, taking account of uncertainty could enrich the extension of this study.

Investment is the operation which consists for a firm in increasing its capital stock (machines, equipments of all types, goods of all kinds, etc.) with the prospect of future growth and competitiveness. On the accounting level, it represents an expenditure on a movable, corporeal or incorporeal property acquired or created by the firm, and intended to remain durably in the same form in the firm. On the theoretical level the analyses carried out on the firms’ sources of growth, investment takes up a predominant place. Its high volatility contributes to the fluctuations of the firm in a non-negligible way. Its role in the short term and its role in the long term make of investment a key variable for the firm. Understanding the evolution of investment behaviour requires an identification of the main variables which guide this behaviour.

Glesne and Legris [

The financial constraint is defined as a conditionality to be met in order to have access to liquidity, by considering the fact that stockholders information is perfect and that financial markets are efficient. The asymmetry of information between stockholders, the banks and executives can modify the structure of financing. It refers to the nature of incentives to the efficiency of the firm’s executives.

Investment therefore appears to come under the multiple and complex objectives which must also take account of the influence of long-term programmes that aim at increasing the firms’ effectiveness and competitiveness. The investment of manufacturing firms has recently and very often taken the form of external growth, that is, the repurchase of existing firms rather than a physical investment. Generally speaking, these firms increasingly hold securities.

The different financial and econometric approaches to investment collide against the problems of taking account of the financial constraints that keep firms from investing as much as they would wish. Under the assumption of one financial market, Modigliani and Miller [

But during the last twenty five years, the separation between the real sphere and the financial sphere tends to become less pronounced. This situation is due to the development of an abundant theoretical and empirical literature which deals with the impact of financial constraints on the structure of the firm as far as investment is concerned. The study by Fazzari, et al. [

Generally speaking, investment is a significant instrument for the growth and competitiveness of non-financial firms. However, they have limited financial resources (liquidities) at their disposal. This problem is recurrent in the non-financial firms of North-American countries (and it increasingly occurs in Sub-Saharan countries) and in the presence of financial constraints due to information asymmetries. The assumption used to detect a financing constraint is that of Fazzari, et al. [

Minton and Schrand [

One major question emerges from all this development: What is the influence of financial constraints on the investment decisions of Sub-Saharan manufacturing firms? In other words, do financial constraints have an impact on the investments of manufacturing firms in the SubSaharan context?

We may envisage several methods for answering these questions. A possible method may consist of specifying a reduced-form model such as Vogt [

It is for this reason that one may specify an Euler equation (version Whited [

The objective of this study is therefore to analyze empirically the impact of this financial constraint on the investment decisions of Sub-Saharan manufacturing entreprises. In other words, our aim is to explain the influence of financial constraints on the investments of manufacturing entreprises in the context of Sub-Saharan countries. To test the hypothesis of the presence of a relationship between financial constraints and investment, it is useful to examine in Section 2 below, the theoretical motivations, then in Section 3 the methodological approach, and lastly in Section 4, the results of the study and their interpretation.

In the exercise of their daily activities, firms are confronted by certain situations which may paralyze their operation. This is essentially summarized in the problems they face in financing the activities of the firm.

Actually, there exists a multiplicity of choices for a firm when it faces financial problems. But the firm may be incapable of using these alternatives to solve these problems owing, for instance, to the indebtedness clause to be respected. Thus, the firm is limited in its decision making, linked to financial constraints.

In general, under the assumption of a limited control of the managers, the stockholders prefer a distribution of part of the internal gross cash flow to a total re-investment of this liquidity in the projects. However, if external financing becomes imperative when self-financing is insufficient to finance profitable projects, firms have recourse to indebtedness and sometimes to the issuing of new stocks.

The literature on financial constraints has been an area of research interest for several authors, and each of them has chosen a firm’s characteristic which seems to explain this situation better.Fazzari, et al. [

Fazzari, et al. [

These studies open a new debate on the question of the validity of using the sensitivity of investment to cash flows as a measure of the degree of the firms’ constraints. This debate takes shape when Kaplan and Zingales ([8,13]) wonder about the effectiveness of using the sensitivity of investment to cash flows to explain the firms’ financial constraints. They suggest, for firms that disburse small amounts of dividends, that the sensitivity differences which appear are partially attributable to differences in the firms’ rates of growth. Thus, the sensitivity of investment to cash flows may be due to the preferences of managers who prefer financing their investments with internal liquidity (cash flow).

It emerges from previous works that the firms’ investment policy is sensitive to the fluctuations of cash flows. In addition, the most financially-constrained firms have higher investment sensitivity to cash flows than those facing weaker financing constraints. Other authors use the indebtedness level to explain the firms’ financial constraints.

According to the theory developed byModigliani and Miller [

In the context of financial theory, Modigliani and Miller [

According to Modigliani and Miller [

Several studies, particularly those of Nivoix, et al. [

According to Tobin [

Empirical studies on this argument were many and they gave some support to the relationship between investment and Tobin’s q. They traditionally use the average q of Tobin which is evaluated with financial market data instead of marginal q. The average q may be measured directly from the value of the firm on the market. The marginal q is the only relevant ratio in theory, which provides the signal in the investment decision of the firm’s manager. In effect, one does not observe the marginal q which is the implicit value of a new unit of capital installed relative to the cost of this unit, or which is the same thing at the optimum of the firm’s programme, the present value of profit flows which will be reported in the future by a dollar of new investment. The marginal q only coincides with the average q under certain conditions: a financial environment which verifies the theorem of Modigliani and Miller [

Vogt [

Indeed, the manager tends to over invest (to invest in projects with a negative net present value (NPV) to capture the pecuniary and non-pecuniary advantages (allocations) of the firm’s growth Jensen and Meckling [

Vogt [

The study by Vogt [

The main assumption in the theoretical work on the firm’s financial constraints is that these constraints entirely manifest themselves in the costs of external financing which are generally very high. This assumption poses two types of research problems (Heitor and Campello [

In a second approach, Heitor and Campello [

The role of financial variables in the explanation of the firms’ investment takes up an important place in the financial theory of the firm. It depends on the value of the firm, hence the use of financial ratios as explanatory variables in order to test the effect of financial constraints on investment.

The approach through Euler’s investment equation, integrated by Abel [26,27], empirically tested afterwards by Whited [

Whited [

According to Azofra and Lopez [

Under the assumption of the existence of imperfect financial markets, this model thus shows the influence of the financial situation of firms on the conditions of access to external financing (cost and availability). Therefore, the expression of our equation will be simply established. It will take into account the impact of financial ratios on the investment decision such as the effect of the debt, Tobin’s q, and the cash flows of the firm.

We take inspiration in our empirical approach from the model similar to those of Bond and Meghir [

For this reason, is therefore equal to the expectation of the present value of the sum of present and future returns which take account of taxation and the dilution property rights due to the issue of new stocks. This maximization of the value of firm is subject to the following which must be respected:

The constraint linked to the dividends of the firm, which is given by:

where and are respectively the labour rand capital variables at time t, = informamation asymmetry costs represented here by the agency function, which means, a function of the level of the debt and of capital. = new stocks issue; = dividends; = endowments to amortizations; = share of labour; = reimbursement of interests; function de production which depends on the variables labour and capital; = adjustment cost function, assumed to be convex to investment; = the level of output prices; = price of capital, and the rate of the tax on companies.

The classical constraint of capital accumulation K

The constraints linked to the issue of stocks, dividends, and debts must be positive:

And finally, the constraint linked to setting a ceiling on the level of indebtedness:

The special feature of this model is the presence a new additional constraint in the maximization program which here imposes a ceiling on the firm’s level of indebtedness, as well as an agency cost function in the definition of the dividend. This is a special feature of the model that distinguishes the classical approach which does not take the financing constraints into consideration.

For this reason, this agency function depends on two arguments, namely the level of the debts (B), and the capital stock (K). Similarly, we make the assumption that the higher (B) is, the more incentive problems are severe while the capital stock (K) reflects the guarantee level which the firm gives to its financial institution, in order to promote greater trust and lower agency costs. The specification of functions helps us to obtain the following testable functional form:

Cost function linked to adjustment:

where, c is a parameter representing the significance of adjustment costs, and.

Function linked to agency costs :

where r is the average rate of investment, and r ≥ 0.

Generally speaking, the firm does issue new stocks and does not pay dividends simultaneously. It does not finance its debt and does not issue stocks, and the dividend is strictly positive.

With these assumptions, the maximization of Equation (1) subject to constraints of Equations (2) and (5) permits to specify Euler’s equation which is given in the following form:

where, the estimated values of the coefficients and which respectively capture the presence (or not) of a premium of access to external financing and a level of the debt ceiling.

If and, then the debt ceiling constraint is absent. The observed ratio of the debt to the capital stock representing the debt ceiling (see constraint (5) of the maximization of the firm’s value) does not exist. For this reason, imperfections on the credit market are linked to agency problems. This situation constrains the firm in its access to external indebtedness, which is explained by the premium to be paid in addition to the interest rate set by the financial institution.

On the other hand, if and, then the firm is subject to an indebtedness ceiling constraint, and an increase in the ratio of the debt to the capital stock reduces the advantage of financing through the debt, owing to the premium to be paid in addition to the interest rate without risk ( see constraint (2)).

The disturbance term of the investment equation, which is considered as being random, contains a variety of specification errors:

The differences between firms (particularly, in the areas of technology, the rate of depreciation, measures used etc.) affects the specification of the model to be estimated. These effects are captured by the presence of individual and temporal, as well as by the term of idiosyncratic errors (the common),.

The model to be estimated here is characterized by the presence of lagged endogenous variables among the explanatory variables of the model. This is an autoregressive model with a causality of the bias that must be considered: the correlation between explanatory variables, and individual specific effects. The origin of the bias is due to the fact that the rate of investment, is a function of, and hence is correlated with individual specific coefficients.

Moreover, the effects may be correlated with other explanatory variables, in particular, the profitability term.

This technique boils down to highlighting, according Arellano and Bond [

where, is the first difference operator such that

, , X= matrix of explanatory variables, except for the lagged endogenous variable, and b = the vector of corresponding parameters.

The data we use in this study show a great variability in individual decisions. They help to capture better the behaviour of investment in its detailed and structural aspects. These data take into consideration the accounting and financial data listed, and extracted from the stock markets of Nigeria, Ghana, and Libreville Gabon).

Note that our sample is composed of a panel of 73 manufacturing firms of different countries (more particularly the synthesis of the firms of Nigeria, Côte d’Ivoire, Ghana, Gabon, etc.), and different activity sectors such as (metallurgy, food processing, chemical, electricity, etc.) over the period 1999-2009.

Our approach carries out an estimation of the Euler investment equation over a ten-year period (1999-2009). In the model, we take account of the existence of variables in first differences, as well as exogenous variables lagged one period. Thus, we are forced to eliminate the first two years (1999 and 2000) of the study period used. In order to decide in favour of the use of the model with a fixed individual effect or with a random effect, we carried out the Hausman [

The Hausman test is specified in the following manner:

The null hypothesis of the absence of correlation between the characteristics of manufacturing firms and explanatory variables is rejected, if the theoretical χ^{2} with k (number of exogenous variables) degrees of freedom is lower than the determined χ^{2}.

If the theoretical χ^{2} with k degrees of freedom is higher than the calculated χ^{2}, then the null hypothesis Ho is accepted and the specific effects are considered as being random; and furthermore, there are no specification errors.

Generally speaking, our model tests two constraints which may affect the investment decisions of a firm, namely: the setting of a ceiling and the increase in the cost of the debt as a function of the ratio of the debt to the capital stock.

This model may then be written as follows:

whereI = investment in equipment or (or the investment process of the firm = investment expenditure);

Stock K or; a measure of the stock K; = gross operating result of the firm at time t; = medium and long-term debt;

= the turnover of the firm; = the error term;

Ratio of investment to the capital stock K = rate of accumulation of K; Ratio of output to the capital stock K; Ratio of profit to the capital stock; change in the ratio of investment to the stock of capital lagged 2 periods; change in the ratio of the debt to the capital stock lagged 2 periods.

By generating the model with the synthesis of the available data of different Sub-Saharan countries, the interpretation of the estimated results takes account of two approaches: before the decomposition of the samples, and with the taking account of the dimension (size), and the participative relations of financial institutions.

They concern two aspects:

^{2 is} equal to 11.9287. This value is lower than the theoretical χ^{2} (read) at the 5% significance level and 7 degrees of freedom (that is to say, 16.088). The Hausman test does not show the existence of a correlation between the individual effect and the explanatory variables. For this reason, this may be a model with purely random effects.

Moreover, we note that the first two exogenous variables (and) have significant coefficients, but their signs are opposite. The sign of the coefficient of the first variable is negative, while that of the second variable is positive. We note that the investment rate of the previous period negatively affects the present rate at an increasing rate.

In effect, the insignificance of the coefficients of variable clearly illustrates the fact that the profitability variable is not important enough. Similarly, the estimation result indicates that the capital accumula-

^{***}Significant at 5%, ^{**}Significant at 1%.

tion behaviour of the activity of the previous year, which is taken into account by the ratio of the turnover to the capital stock, has an impact on the rate of investment. This result helps us to distinguish as Daoud and Kammoun [

If we get interested in variables which measure financial constraints, the variable has a significant coefficient, but it has a negative sign. This demonstrates the effectiveness of setting a ceiling on indebtedness by financial institutions. The increase in the ratio of the debt to the capital stock reduces the advantage of financing through the debt because of the premium to be paid; in addition to the interest rate without risk (an effect which is due to the agency cost which is captured by the ratio

and which therefore re duces the firms’ investment expenditures.

The coefficient estimate of is positively significant. It also shows the effectiveness of the existence of costs linked to information asymmetries between the firm and the financial institution of credit. On that account, the higher (B) is, the more incentive problems are severe. These results corroborate with the works of Jaramillo et al. [

Therefore, it appears that Sub-Saharan firms are experiencing an environment with quite a strong information asymmetry between the borrowers and lenders. In addition, these firms are constrained in their access to external indebtedness owing to the levelling-off of indebtedness.

In order to integrate the heterogeneity of the firms in the analysis of the impact of financial constraints on the investment decisions of manufacturing firms, we have carried out a categorisation of the firms according to two criteria: the dimension and the participative relationship of financial institutions.

The financial literature states the fact that small firms are put at a disadvantage by the banks, either by higher guarantees, or by high borrowing costs.

Ydriss Ziane [

The dimension of the firm is considered as a factor of differentiation of the impact of financial constraints on the rate of investment. Several criteria are used to appreciate the firm’s dimension, namely the number of employees in the firm, the turnover achieved, and total assets.

In the present study, we consider a large firm as any structure making a turnover of higher than 2 billions CFA Francs, and a small firm as a firm whose turnover is lower that this value.

On the basis of this criterion (the participative relations of financial institutions), we have distinguished the firms where financial institutions appear in their property structure, and those where financial institutions do not have participative relations. In order for financial variables to be different according to their dimension (a participative relation of financial institutions), we have defined a dimension control variable (DIME) which is equal to 1 if the firm is large (there exists a relation between firm and financial institution) and zero if not.

For that reason, to estimate the coefficient of

for large firms we estimate the coefficient of the variable (× DIME). This has helped us estimate the following equations respectively:

According to the Hausman test results, the χ^{2} determined by the first and the second equation is equal to 11.989 and 11.402 respectively. These values are lower than that of the theoretical χ^{2} which is equal to 17.889 at the 5% significance level.

From the preceding presentation, we can maintain the lack of correlation between the individual effect and explanatory variables. Both models have purely random effects.

The examination of

As concerns the variable, it has a positive and significant effect on the investment expenditures of small firms. On the other hand, for large firms the coefficient of this variable is negative and insignificant. Similarly for the second variable which represents the indebtedness ceiling constraint, its coefficient is negative and significant for small firms. Both of these results imply that information problems and the indebtedness ceiling influence the rate of investment.In addition, these results agree with those obtained by Jaramillo

Sources: The figures between parentheses represent the values of t-student statistics.

et al. [

The purpose of this paper was to explain the influence of financial constraints on the investments of manufacturing firms in the context of Sub-Saharan countries. In this framework, we empirically tested two constraints which may affect the investment decisions of a firm. These constraints correspond to a levelling-off of the indebtedness level and the increase in the cost of the debt according to the amount borrowed reflecting the asymmetry of information which exists between the lender and the borrower. The estimation results of Euler’s capital accumulation equation derived by Jaramillo et al. [

In addition, the effect of information problems and of the levelling-off of indebtedness on the rate of investment according to the dimension criterion, show that small firms are put to a disadvantage by the financial institutions either by high borrowing costs, or by high guarantees. Similarly, the distinction of the effect of two financial constraints on investment according to the criterion of a participative relation of financial institutions indicates that the firms where the financial institutions appear in their property structure are more capable of finding external financial resources there, thanks to the relations of trust between financial institutions and the firms. Generally speaking, the firms are less constrained in obtaining financing on the credit market.