1.

Introduction
Working capital analysis and management is a standard subject in managerial accounting and
financial management. The matter is treated virtually in all textbooks on management accounting,
financial management, corporate finance, and other similar areas (Ross et al., 1993; Van Horne,
1998; Brigham and Houston 1996; Gitman, 1994). Although conventional working capital analysis is
still in use nowadays, a new proposal for dealing with this topic arose in Brazil in the 80’s. In 1980,
French professor and researcher Michel Fleuriet (Fleuriet et al. 1980, 2003) introduced in the country
a new model for analyzing and managing working capital. In Brazil, the method became attached to
his name, but it is also known as Advanced or Dynamic Analysis of Working Capital. It is a method
proposing advances with respect to the extant conventional analysis. In this model, subdivisions of
current assets and liabilities are created for the sake of analysis, according to the nature of their
components: financial (erratic) current accounts must be separated from current operating (or
cyclical) accounts. The model establishes, through certain relationships between those classes of
current assets and liabilities, whether a firm is in financial equilibrium or not. According to Assaf
Neto and Silva (2002), a similar method was developed and published in English by Cox and
Shulman (1985). According to Braga (1991), “this methodology permits to rapidly assess
firms’financial situation by classifying balance sheets in one of the six possible types of
configurations based on certain accounting elements. It is a step forward with respect to the
traditional scheme based on financial ratios. Although some Brazilian scholars have reproduced the
methodology in textbooks and academic work, we think that it has not been sufficiently divulged in
our country”.
It should be mentioned that no reference is made either in Fleuriet (1980, 2003) or in later textbooks
and articles to any kind of empirical verification of the model’s assumptions, which leads to the
impression that the model’s assumptions were based on intuition or other untested ideas lacking
theoretical and empirical backing.
The paper’s purpose is to test empirically a fundamental assumption of Fleuriet’s model. This
assumption states that there is no relationship between current financial assets and liabilities and the
firm’s operating activities. Since that model rests on this assumption, its rejection leads to rejecting
the model itself.
The paper’s remainder is divided into Sections from 2 to 6. Section 2 summarizes the conventional
method of working capital analysis; Section 3 presents the bases of Fleuriet’s model; Section 4
explains the methodology used in this paper and the empirical results obtained; Sections 5 and 6
show the conclusions and references, respectively.
2. The Conventional Method
By and large, conventional working capital management is based on the observation of the net
working capital behavior, and of its components inside the current assets and liabilities, in liquidity
ratios or metrics (working capital, current ratio, cash ratio, and acid test or quick ratio), and on firm’s
operating and financial cycles. In general, one expects that a firm with a sound liquidity situation has
a positive net working capital CCL and liquidity ratios equal or superior to certain benchmarks,
which may be established on an industry basis or on firms of similar type or size. It is also
recommended to observe the evolution of those ratios trough time, so as to determine trends.
More recently, other metrics have been used, such as duration, which an average of maturities of
current-asset and liability accounts weighted by their respective present values (Assaf Neto and Silva
2002). The determination of each of the working capital components is done by disaggregating these
components and treating them separately. For cash management,for example, there are Baumol’s
(1952), Miller and Orr’s (1966), and Gallinger and Healey’s (1991) models. For inventories’
management, there are the Economic Order Quantity (EOQ) model (Wilson, 1934), the ABC curve,
just-in-time, etc.
3. Fleuriet’s model
Fleuriet’s model (Fleuriet et al., 1980, 2003) was intended to bring a new method for working capital
management. Initially, the model proposes a new managerial classification for current asset and
liability accounts, according to their financial or operating nature, with this segregation being
essential to the process of assessing firms’ working capital requirements. Current assets (CA) are
thus divided into current financial assets (CFA) and current operating assets (COA). The former
includes essentially financial elements such as cash, banks, short-term financial investments and
bonds. This group “does not show, as a consequence, any pre-established behavior, varying more
strictly as a function of the economic situation and of a higher or lower risk the firm whishes to take”
(Assaf Neto and Silva, 2002). Current Operating (or Cyclical) Assets (COA) is composed of
accounts related to the firms’ operating activities, such as inventories, accounts recei vable, and bad
debt provision, being influenced by the business volume or by characteristics of the operating cycle
phases, such as inventory management decisions or sales policy.
Similarly, Current Liabilities divides into Current Financial Liabilities (CFL) and Current Operating
Liabilities (COL). CFL includes short-term debt, bank loans, discounted duplicates, short-term
portions of long-term debts, and dividends. OCL is formed by the firm’s short -term obligations
directly identifiable with its operating cycle (suppliers’ credits,salaries and benefits, and taxes).
Accounts inside Long-Term Assets and Fixed Assets compose a block called Permanent or Non-
Cyclical Assets. On the other side of the balance sheet, there is a corresponding block: Permanent (or
Non-Cyclical) Liabilities, which includes equity and debt.
This reformulation is based on the following Fleuriet’s assumptions: operating or cyclical accounts
are those related to the firm’s operating activities, whereas financial accounts are not connected to its
operating activities. It is worthwhile to stress the definition of erratic, in Fleuriet’s words: “erratic,
from the Latin erratic - errant, vagrant, random, wayward. In other words, this implies the non-
linking of those accounts to the firm’s operating cycle” (Fleuriet 2003, p.7).
From the balance-sheet segmentation, the concepts of Working Capital Requirements (WCR),
Working Capital, and Treasury Balance arise, which according to the model, are economic concepts,
in opposition to the CCL concept, which is a legal definition. “When the firm’s operating activity
generates a cash outflow that is faster than the cash inflows (payments for production inputs occur
prior to receiving the proceeds from sales, for example), a permanent working capital requirement is
identified, which is computed through the difference between cyclical assets and liabilities” (Assaf
Neto and Silva, 2002). It should be stressed that WCR is different from Net Working Capital (NWC),
since WCR is partly composed of current assets and liabilities (operations related accounts):
WCR COA- COL where COA and COL stand for Current Operating Assets and Liabilities,
respectively. It can be seen that WCR can be negative, which characterizes an excess of operating
activities, i.e.cash inflows are on average generated later than cash outflows. In this case, it is implied
that operating liabilities are financing more than current operating liabilities, meaning that operating
liabilities become larger than operating assets, which represents a source of funds for the firm.
Another concept used by Fleuriet is Working Capital (WC), which represents a permanent source of
funds for the firm, with the purpose of financing its WCR. The working capital has the same value as
CCL, but it is calculated differently:
WC Permanent Liabilities - Permanent Assets
One of the WC’s characteristics is it is relatively stable through time, changing when the firm makes
new investment, which may be implemented either through self financing (cash generated by the
firm’s operations), long -term loans, or equity issues. Finally, Treasury Balance (T) is obtained
through the difference between current financial assets and current financial liabilities, i.e.
T CFA- CFL (3)
Treasury Balance can also be represented by the residual value resulting from the difference between
WC and WCR:
T WC - WCR (4)
This variable represents “a firm’s financial reserve capable of cushioning eventual expansions in
working capital investment requirements, especially those with a seasonal nature” (Assaf Neto and
Silva, 2002). The fundamental condition required for the firm to be in financial equilibrium is that its
treasury balance is positive.
4. Methods
The paper’s aim is to test the basic assumptions of Fleuriet’s model. The first assumption states that
the Current Financial Assets (CFA) and Current Financial Liabilities (CFL) are erratic, thus bearing
no direct relation to the firms’ operations. The second assumption states that Current Operating
Assets (COA) and Current Operating Liabilities (COL) keep close relationships with the firms’
operating activities, being directly influenced by the business activity. Having in mind that if CFA
and CFL, as well as COA and COL are not actually erratic, then Current Assets and Current
Liabilities, i.e. CA = CFA + COA and CL = CFL + COL, will not be erratic either.
Aiming at testing those assumptions, three groups of hypotheses were formulated:
H0A: CFA and CFL are unrelated to the firms’ operations.
H1A: CFA and CFL are related to the firms’ operations.
H0B: COA and COL are unrelated to the firms’ operations.
H1B: COA and COL are related to the firms’ operations.
H0C: CA and CL are unrelated to the firms’ operations.
H1C: AC and PC are related to the firms’ operations.
H0A, H0B, and H0C are the null hypotheses and H1A, H1B and H1C are the alternative hypotheses. It was
assumed that Net Operating Revenue (NOR) is an appropriate proxy to measure the firms’ operating
activity.
It must be emphasized that only hypothesis H 0A is actually being questioned in the study, since the
other relations are not subject to doubt. Actually, it is expected that all null hypotheses above are to
be rejected. However, the rejection of H0A means that Fleuriet’s basic assumption concerning CFA
and CFL is not valid, which invalidates, as a consequence, the model itself.
In order to test empirically Fleuriet’s model assumptions, several procedures were carried out.
Initially, we sought to verify the model’s plausibility by verifying which percentage of the sample’s
firms would be in financial equilibrium, according to what is established by that model. Since
Fleuriet’s model proposes that firms with negative treasury balances are in financial disequilibrium,
one can verify the plausibility of this condition. With this purpose, the treasury balance was
calculated for a random sample containing 80 firms between 1995 and 2002. This result was
confronted with the calculation of NCC for the same firms, in order to verify what percentage of
these firms has positive NCC.
In second, a correlation analysis was performed involving variables belonging to current assets and
current liabilities with net operating revenue. If Fleuriet’s model is correct, operating assets and
liabilities should be significantly correlated to net operating revenue, whereas financial assets and
liabilities should not.
Thirdly, cross-section linear regressions were performed for each year from 1996 to 2002. Again, if
Fleuriet’s model is correct, we can expect that cyclical assets and liabilities present significant
correlation with net operating revenue, whereas financial assets and liabilities do not.
Finally, with the purpose of confirming the previous results, panel-data regressions were carried out.
The variables analyzed in all tests, with their corresponding abbreviations, are:
CA Current Assets
CFA Current Financial Assets
COA Current Operating Assets
CL Current Liabilities
CFL Current Financial Liabilities
COL Current Operating Liabilities
NOR Net Operating Revenue.
5. Conclusion
As a conclusion of the analyses performed in this study, we affirm Fleuriet’s model of working
capital management cannot be empirically validated, as far as Brazilian firms are concerned, and
must be consequently rejected. Based on financial statements pertaining to a sample of 80 Brazilian
non-financial firms listed in Bovespa, referring to the 1995-2002 period, we verified that applying
Fluorite’s model to these firms results that 3/4 of them were in financial disequilibrium, which is an
implausible result. Contrary to this, when we used the concept of positive NCC utilized in the
conventional liquidity analysis, about 2/3 of the sampled firms are in financial equilibrium in that
period, which is much more plausible. The null hypotheses established in the study were based on the
untested Fluorite’s model’s assumptions that financial assets and liabilities are erratic, whereas
operating assets and liabilities are not. Current assets and liabilities were also tested for their
relationships with the firms’ operations. The null hypotheses of non -existence of relationship with
firms ‘operations were empirically tested against the alternative that these relations exist and are
significant. The correlation, cross-section and panel-data methods led to the rejection of the null
hypotheses. In particular, the rejection of H 0A indicates empirical inconsistency of a fundamental
assumption and, therefore, the model’s inconsistency. The null H 0A that current financial assets and
liabilities are erratic and unrelated to the firms’ operations is rejected in all instances.
In Fleuriet’s model, this assumption is essential to justify the segregation of financial assets and
liabilities from operating assets and liabilities, respectively, permitting the formulation of the
concepts of treasury balance and working capital requirements. We found in this study that the basis
for such segregation is not valid, since all variables considered are related to the firms’ operations. If
it is not feasible to make the segregation proposed by Fleuriet, then current assets and liabilities must
be analyzed as entire blocks. However, this is exactly what is done in the conventional analysis of
working capital. In other words, the rejection of Fleuriet’s model naturally leads to the conventional
model. The paper demonstrates the dangers of model building based on intuitive ideas that were not
proven by empirical scrutiny. Fleuriet’s mo del has being taught in Brazilian undergraduate and
raduate university courses since the 80’s as a modern and sophisticated tool for working capital
analysis and management, without concern that it could eventually be wrong.