Journal of Money Investment and Banking
Journal of Money Investment and Banking
Journal of Money Investment and Banking
76
Keywords: Corporate finance, firm size, bank relationship, cluster analysis, discriminant
analysis.
1. Introduction
Treasury management is one of the tasks involved in the current culture of cash management. As such
it means assuming a number of responsibilities concerned with the control of monetary flows of
organisations and liquidity positions that will lead to improvements in the results of the treasury
department and in those of the remaining departments. The tasks entrusted to this department must
centre on obtaining profit by maximising short-term profitability obtained through surpluses in
liquidity and through cost cuts in the management of treasury deficits, all of which will help to achieve
the general objective of maximising the value of the organisation. But effective cash management
involves the use of tools and techniques, and also covers developing and maintaining successful bank
relationships. The role of the bank is to provide services that include payment lines, overdraft and
credit facilities, information flows and investments, etc. From the point of view of the corporation, the
banks pricing structure is very important, but bank customers point to other important factors besides
price, i.e. qualitative variables (Tse et al., 1998b, p.284).
Treasury management thus becomes a highly important part of corporate strategy, as it means
implementing the philosophy of cash management at the treasury department. A direct link is
established between treasury management and the concepts of liquidity and profitability. The treasury
department ceases to be considered merely as a cost centre and becomes a profit centre, as for other
departments, which implies an active, autonomous, independent concept of corporate liquidity
management.
A treasurer has various functions, depending largely on the size of the corporation and his
position within its structure, so there should always be a clear definition and distribution of tasks. In his
empirical study on Spanish corporations (2002, p. 23), Fernndez Pascual finds that it is generally
large, internationally oriented corporations that have treasury departments, while only 24% of small
and medium-sized enterprises which account for most of the enterprise sector have such a
department. The remaining 76% allocate treasury department functions to other departments, such as
accounting (36%), administration (34%) and finance (25%). However, various studies within the
specialist literature on the topic, e.g. Emmanuel et al (1990) and Dent (1996), evidence the need to
foster treasury departments within corporate organisation systems and to centralise the tasks involved
in treasury management, regardless of company size.
However, there are some responsibilities that are allocated to this department in practically all
cases. The research studies conducted under the auspices of the Treasury Management Association
(Phillips, 1997, p. 69-70) stress liquidity management in particular among the functions of treasury
management. This is identified with actual tasks such as determining treasury positions and the
management of net cash requirements and surpluses.
Treasurers are generally specialists with expertise in the monetary and credit market who are
skilled in managing relations with financial institutions. They are in charge of administering the
corporate treasury dynamically, effectively, consistently and prudently. From our point of view, taking
the conclusions of the study by Iturralde, Maseda & San Jos (2004) as our basis, treasury management
involves the following tasks: Liquidity monitoring and cash management, Management of short-term
needs and surpluses, Financial risk management and Management of relations with financial
institutions.
Corporations have to negotiate a very large number of variables with financial institutions, so it
is advisable to draw up an overall package of services and features that reflects the percentage of the
corporations activities that is to be managed by the bank and converted to net assets, and to establish
terms of agreement (Marco, 2001). The maturity dates of sources of financing, the value dates of the
various financial instruments, interest rates, exchange rates, charges, etc. are just some of the
quantitative variables that must be set in bargaining agreements so that they can be monitored via
77
regular analysis of the agreements entered into. This enables corporations to considerably improve the
management of their disposable assets.
But these are not the only variables that must be considered: qualitative variables are also
highly significant in corporate banking policies, and therefore in choosing the right bank. Standards, of
service, the speed of response of the bank, personal service, the extent of its branch network, the
frequency of clerical errors, the standard of its advice and its capability for problem solving are some
of the main variables of this type.
In this regard the study by Soenen (1989) was one of the first to make clear the importance of
non monetary variables in choosing banks. This study groups factors for assessing bank selection
criteria according to a quantitative variable (price), behind which variables such as charges, interest
rates, etc. are concentrated, and according to variables that are significant to bank users but cannot be
measured in price terms, such as standard of service (speed, efficiency, clerical errors, number of
branches, opening hours, offices abroad, etc.), personal relations and special services, among others.
Subsequent studies have confirmed this tendency in identifying the criteria on which decisions
on bank/corporation relations are based. Although price is the main determinant variable in the
contracting of financial services, other aspects such as speed of response and standard of service are
now becoming increasingly important.
In the study by Kennington, Hill & Rabowska (1996), price, the reputation of a bank and its
standard of service are considered to be the three leading factors that influence the choice of banks. On
the basis of data from the biannual Global Cash report for 1998. Santom (1999, p. 31) concludes that
price and standard of service are by far the most important factors in distributing business among the
various banks. The most ambitious study on this topic is that of Mols, Bukh & Blenker (1997), whose
sample covers 1129 firms in twenty European cities. Their results indicate that standard of service is
the main variable considered in selecting a bank (31.7%), followed closely by price (29.8%) and
personal treatment (23.2%). The remaining variables (commitment to the corporation (11.3%),
classification of the bank (7.9%), number of branches (6.7%), offsetting of other services (5.5%),
reputation in treasury management (3.8%) and technology (3.7%) are far less important to firms in
their decisions on which banks to use.
In that context, this article seeks analyse the factors that guide the actions of the treasurer in
selecting sources of financing, in handling situations that result in treasury deficits and in selecting
investment instruments if surplus liquidity is generated. On the other hand, the factors that influence
the choice of the financial institution with which the treasury department works preferentially are also
analysed.
Our results show that in selecting these sources of financing, corporations basically take two
elements into account: the cost of financing and the advice provided by their bank, and among the
innate qualities of financial assets, security and liquidity are prioritised over profitability in the choice
of investment instruments for treasury surpluses, with the most widely used instruments being bank
deposits and government debt. Finally, our results confirms the trend for qualitative variables to
become determinant in the choice of banks. For corporations, standard of service and speed are by far
the most highly valued factors in their relations with banks.
The rest of the paper is organised as follows. First of all, objectives of the study are described.
In Section 3 the sample and the questionnaire used are presented and the analysis procedure followed
in carrying out the empirical study is explained. The main results of the investigation are explained in
Section 4. Finally, Section 5 present the main conclusions.
78
To look at this area in more depth, we have performed an empirical analysis based on
corporations in the Basque Country (Spain). Our specific objectives were:
1. to identify the financing sources used by corporations to cover short-term treasury
deficits;
2. to identify the investment instruments used to manage short-term treasury surpluses;
3. to identify the factors that determine the choice of financing method for treasury deficits;
4. to identify the factors that determine the choice of investment for treasury surpluses;
5. to analyse the relationship between the size of corporations and the way in which they
manage their treasury surpluses, and the factors that influence their choice of banks;
6. to analyse the relationship between the training of corporate treasurers or decision-makers
on treasury matters and the management of treasury needs and surpluses and the factors
that influence their choice of banks;
7. to establish sufficiently homogenous groups of companies from the point of view of bank
negotiations;
8. to detect the variables that influence the choice of the financial institution with which the
treasury department works preferentially.
3. Research Method
3.1. The Sample
The database used comprises 4699 Spanish corporations with more than 15 employees. Our selection
ignores the so-called "micro-corporations", i.e. those with less than 15 employees, because one of their
identifying characteristics is precisely their lack of a major organisational structure, especially in
financial matters.
1,500 of these 4,699 corporations were sent questionnaires distributed proportionally
according to the populations of the various provinces. A response rate of 14.5% was obtained, i.e. 217
questionnaires were returned. 14% of the corporations that returned the questionnaire are classed as
small, 62% as medium and 24% as large.
We divide these firms into two groups on the basis of whether their treasury decision-maker has
a university degree or not: the split was 69% with a degree and 31% without.
3.2. The Questionnaire
The core objective for which the questionnaire was drawn up was to analyse management of treasury
needs and surpluses in Basque Country corporations with more than 15 employees. To that end, the
questionnaire was divided into four blocks of questions: Financing of treasury deficits, Investment
of treasury surpluses, Negotiations with banks and General information.
The first block -Financing of treasury deficits- seeks to identify the sources of financing used
and also to analyse what factors lead them to choose those particular sources.
Fin 1
Fin 2
Fin 3
Fin 4
Fin 5
Fin 6
Fin 7
Confin 1
Confin 2
Confin 3
Confin 4
Confin 5
Confin 6
The second block covers the investment instruments used by corporations to manage treasury
surpluses and the factors that lead them to choose those instruments.
79
Inv 1
Inv 2
Inv 3
Inv 4
Inv 5
Inv 6
Inv 7
Inv 8
Inv 9
To analyse how heads of treasury departments see the services provided by banks in this area
and determine their main quality criteria in choosing banks, we use a number of quality factors selected
on the basis of the proposals in the Global Cash report drawn up every two years by The Bank
Relationship Consultancy. These factors are covered in the third block of questions Negotiations with
banks.
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Factor 6
Price
Service & speed
Personal treatment
Number of branches
Electronic banking
Loyalty
The fourth block General information covers the data required to identify corporations:
annual turnover, total assets on the balance sheet, number of employees and level of training of the
decision-maker on treasury matters, among others.
The questionnaire contains a total of 34 specific questions on the areas related to the study,
which require equally specific answers. Most of the questions are presented with a Likert type scale
that enables a quantitative score to be allocated to the qualitative opinion of the corporation, so that the
required empirical comparisons can be made. A number of additional questions are also asked to
determine the level of training of the treasurer and the size of the corporation.
The levels of agreement/ disagreement used in the questions are as follows.
3.3. Statistical Tests
The effect of corporation size and treasurer training on the management of treasury needs and surpluses
and on negotiation between banks and firms was studied via a Kruskal-Wallis variation analysis. This
analysis enables us to check whether corporations behave differently depending on their size and the
level of qualifications of their treasurer.
The major variables in relations between banks and firms were sought using the following
process:
80
Mean
Not important
Of little importance
Of some importance
Quite important
Extremely important
Fin1
3.45
23.81
8.10
7.62
20.00
40.48
Fin2
1.47
79.02
8.29
3.90
4.39
4.39
Fin3
3.29
25.00
6.25
15.38
21.15
32.21
Fin4
1.94
53.92
19.12
12.75
735
6.86
Fin5
1.29
88.61
2.48
347
248
297
Fin6
246
4203
1014
1643
2222
918
Fin7
171
6031
1753
1598
309
309
Analysing the scores obtained by each answer, we can confirm the differences found in the use
of these sources of financing, especially if we consider answers with scores of 4 and 5 (Often and of
considerable significance and Regularly and of great importance) together. These results show a
clear preference for discounting commercial paper and for credit lines, which can be considered major
sources of financing since they both score over 50%. On the other hand the issuing of promissory notes
and factoring score less than 9% in all cases.
Having determined the relative significance of each of the short-term financing sources
habitually used by corporations, let us now look at the factors that lead them to select those sources.
In this regard, the data in Table II show that the foremost consideration is cost, with an average
score of 2.89 out of 5, followed by advice from banks with a score of 2.43. In regard to estimated
variations in interest rate, the corporations surveyed value the opinion of financial analysts more highly
than that of their own management, though the difference is minimal. Finally, we note the low score
obtained by fixed interest rate loans to avoid exposing firms to additional interest rate variations:
71.12% of those questioned did not consider this point as significant, even though some risk of interest
rate increases exists.
Table II:
Mean scores & frequencies of factors leading to the choice of financing sources
Mean
Not important
Of little importance
Of some importance
Quite important
Extremely important
confin1
2.43
33.33
19.49
23.08
19.49
4.62
confin2
2.07
45.55
19.37
19.90
12.57
2.62
confin3
1.97
52.11
15.79
17.37
12.11
2.63
confin4
1.99
48.95
18.95
17.89
12.11
2.11
confin5
2.89
27.75
8.90
24.61
24.08
14.66
confin6
1.52
71.12
12.30
12.30
2.14
2.14
81
Mean
Not important
Of little importance
Of some importance
Quite important
Extremely important
Inv1
2.1
59.26
5.29
12.17
12.70
10.58
Inv2
2.41
47.64
9.42
13.09
13.61
16.23
Inv3
1.74
75.81
2.69
4.84
5.37
11.30
Inv4
1.55
80.65
3.23
3.23
5.91
6.99
Inv5
1.84
67.03
6.49
8.65
11.35
6.49
Inv6
1.44
81.72
6.45
5.30
4.38
2.15
Inv7
2.18
51.60
11.70
15.43
10.11
11.17
Inv8
1.13
94.05
2.70
1.08
0.54
1.62
Inv9
1.25
89.19
3.24
3.24
1.62
2.70
Investment decisions are directly and decisively influenced not only by the characteristics of
financial assets themselves (security, liquidity, profitability) but also and to a very great extent by
advice from banks (see Table IV). The significance of this variable in the management of surplus assets
supports the idea of the second part of our study: to show the importance of relations between
corporations and banks in treasury management.
Those questioned awarded the highest score for innate features of financial assets to security,
which was considered significant by 66.66% (Often and of considerable significance and Regularly
and of great importance). In second place we find liquidity with 57.89% and third comes profitability
with 53.65%. So security and liquidity are valued more highly than profitability.
Table IV: Mean scores & frequencies of factors leading to the choice of investment
Mean
Not important
Of little importance
Of some importance
Quite important
Extremely important
coninv1
2.75
25.65
18.32
23.04
20.94
12.04
coninv2
2.22
42.55
17.02
21.28
14.36
4.79
coninv3
2.04
49.74
15.34
19.58
12.17
3.17
coninv4
3.32
21.58
4.74
15.79
35.79
22.11
coninv5
3.24
20.31
4.17
21.88
38.02
15.63
coninv6
3.64
16.67
3.13
13.54
33.33
33.33
82
Price
3.69
4.22
Personal
treatment
3.66
Number of
branches
2.65
Electronic
banking
3.42
Loyalty
3.41
By far the most highly valued factor for corporations in their relations with banks is the
standard of service provided. This confirms the trend marked by previous studies that have shown a
rise of qualitative factors such as speed and standard of service as determinants in the contracting of
banking services. Price obtains practically the same score as personal treatment. These percentages are
very similar to those reported in Mols, Bukh & Blenker (1997), which enables us to assert that
corporations in the Basque Autonomous Community behave similarly to those in the European cities
where the said authors conducted their study.
When level four (Often and of considerable significance) and level five (Regularly and of
great importance) scores are added together for each of the six factors, the differences can be clearly
seen (Graph I). The service offered by banks, the price of financial products and personal treatment can
be considered as significant, as all three were considered by over 63% of those questioned to be
fundamental to their choice.
The score obtained by electronic banking services merits further comment: advances in
technology in recent years have enabled corporations to maintain much more fluid relations with banks
by providing fast, up-to-the-minute information on their banking situation. This has favoured
considerable progress in treasury management.
In spite of the progress made in electronic banking, it is not seen as a major criterion in
choosing banks and there is little evidence of consumer demand for internet banking services
(Davidson, 1998). The critical question is whether the spread of electronic banking is determined more
by customer acceptance than by seller offerings (Mols et al., 1999). However this may become a factor
for differentiation in the future. It seems logical that this factor should carry ever greater weight in
choosing banks, as truly substantial progress in internet banking has been made recently, and more will
be made in the future, making this a highly competitive area for differentiating between banks.
Graph I: Significance of factors used in negotiations with banks
85,43
65,68
63,78
55,83
56,93
27,60
Price
Personal
treatment
Number of
branches
Electronic
banking
Loyalty
83
Inv8
1.105
.676
Inv9
8.613
.013*
Loyalty
3.671
.160
The results of this variance analysis enable us to reject the null hypothesis of equal means for
most cases (those indicated in Table VI). In other words, the use of credit lines as a source of
financing, investment in Repurchase agreement operations, euro deposits and corporate promissory
notes, consideration of advice from banks, security of investments and number of branches are the
elements for which there is a statistically significant link between corporate size and the management
of treasury needs and surpluses and the banking negotiation.
Treasurer training levels only show up as influential in some cases, as can be seen in Table VII.
The use of credit lines, corporate financial management expertise, the cost of financing and personal
treatment of firms in their dealings with banks are all elements in which there is a significant difference
in analysed factors depending on the training of the decision-maker.
Table VII: Kruskal-Wallis variance analysis. Training levels of decision-makers
Factor: Training levels of decision-makers (university degree or not)
Dependent variable
Fin1
Fin2
Fin3
Fin4
Fin5
Fin6
Fin7
2e
1.883
.092
5.210
.017
.958
1.104
1.246
stadstic
p-value
.170
.762
.022*
.896
.328
.293
.264
Dependent variable
confin1 confin2 confin3 confin4 confin5 confin6
.420
2.626 7.091 1.818
9.555
.443
2 estadstic
p-value
.517
.105 .008**
.178
.002**
.506
Dependent variable
Inv1
Inv2
Inv3
Inv4
Inv5
Inv6
Inv7
Inv8
2.020
.508
2.489
.792
1.994
.003
.011
1.901
2 estadstic
p-value
.155
.476
.115
.373
.158
.957
.916
.168
Dependent variable
coninv1 coninv2 coninv3 coninv4 coninv5 coninv6
.652
3.161 4.834 2.400
1.026
1.711
2 estadstic
p-value
.419
.075
.028*
.121
.311
.191
Factor: Training levels of decision-makers (university degree or not)
Service &
Personal
Number of
Electronic
Loyalty
Dependent variable
Price
speed
treatment
branches
banking
1.707
.098
9.945
.370
.129
.953
2 estadstico
p-value
.191
.754
.002**
.543
.719
.329
**, * denotes a significant difference at the 1%, 5% level, respectively
Inv9
.074
.785
84
This results reinforce the idea that the tasks of treasury departments are integrated into the
culture of cash management, and that the use of new net asset management techniques is not directly
linked to the size of a corporation or to the standard of training of its treasurer, but rather depends on
the initiative of the financial managers themselves.
4.3. Significant Variables in Negotiations between Banks and Corporations
4.3.1. Classification of Corporations
This section sorts corporations into homogenous groups according to the standard of their treasury
management. Using statistical methods, this grouping can be carried out via cluster analysis. We apply
such analysis to individual corporations in our sample using the agglomerative hierarchy method and
the square of the Euclidean distance.
The results of cluster analysis and the relevant descriptive statistical analyses indicate that there
are two main groups of corporations: Group I, containing 108 corporations, and Group II, containing
84.
The table below shows the average score for each function in each of the two groups resulting
from cluster analysis.
Table VIII:
Characteristics of groups
Price
Service & speed
Personal treatment
Number of branches
Electronic banking
Loyalty
Group I
3.93
4.52
4.06
3.18
4.06
3.94
Group II
3.37
3.86
3.10
1.98
2.54
2.73
Differentiation
0.56
0.66
0.96
1.20
1.52
1.21
The mean scores for Group I are higher, which means that these corporations grant greater
significance to negotiations with banks.
An analysis of the groups in absolute terms shows that standard of service and speed of banks,
personal treatment and electronic banking are the factors that obtain the highest scores among Group I
corporations.
Price as a quantitative variable, bringing together factors such as interest rates, charges and
exchange rates among others, which overall was the second biggest factor in choosing banks, maintains
a high position in the ranking resulting from analysis by groups, but one lower than the qualitative
variables. This confirms the tendencies pointed out in the previous section. The same goes for the score
obtained by all the services that come together in the Electronic Banking variable.
In Group II all aspects linked to treasury management are less significant, indicating a lower
level of inVolvement of treasurers in the new trends in cash management. In this group the standard of
service and speed of banks remains the most highly valued factor, with price taking second place. The
biggest difference between the groups is in electronic banking, which obtains a very low score among
Group II corporations.
4.3.2. Discriminant Analysis: Main Functions
Having grouped our corporations, we now apply the second stage of this two-stage process: linking the
groups with the factors that influence relations between banks and corporations. This enables us to
check whether the groups obtained are consistent, and establish what variables contribute most to
sorting corporations into groups. It is therefore advisable to validate the classification obtained via
cluster analysis via explanatory techniques such as discriminant analysis, using validation samples.
The following table shows the discriminant loads or structural correlations for the variables
included in the discriminant function. As can be seen, the Electronic Banking variable has the greatest
85
discriminant load, i.e. it is this variable that most characterises firms in Group I as distinct from those
in Group II. The Number of Branches, Loyalty to Bank and Standard and Speed of Bank Services
variables have discriminant loads but they are lower than that of Electronic Banking.
Table IX: Discriminant function. Structural matrix
Electronic banking
Number of branches
Loyalty
Service & speed
Discriminant charge
.675
.497
.410
.329
The last indicator of the effectiveness of the discriminate function is the proportion of
individual corporations correctly sorted. The figures obtained in this regard are 86.7% for the analysis
sample and 78.4% for the validation sample. But to check the goodness of these results, they should be
compared with the proportion of individual corporations that might be sorted correctly on a random
basis, without the aid of the discriminant function. A priori, the random sorting figure for the analysis
sample would be 50.88% of cases ((98/173)2 + (75/173)2), and for the validation sample it would be
50.1% ((10/19)2 + (9/19)2). Given the number of cases actually sorted correctly a posteriori as
indicated above, the improvement obtained via this analysis is significant.
5. Conclusions
This study has focused on the analysis of cash management, providing evidence using Spanish
companies with more than 15 employees. There are several important features of our analysis which
extend the literature on cash management routines.
The short-term sources most widely used to fund shortfalls in liquidity are the discounting of
commercial paper and credit lines. The high flexibility of management of these methods and the
relative ease with which they can be obtained influence this choice. However we consider that there are
also other factors that lead corporations to prefer these methods, such as maintaining corporate
financing separate from the image of a corporation in the eyes of its suppliers and customers.
In selecting these sources of financing, corporations basically take two elements into account:
the cost of financing and the advice provided by their bank. In large corporations, however, advice
from banks may be a reference point, but the choice of financing sources is backed up by in-house
analyses. This reinforces the importance of having a consolidated financial department within the
structure of a corporation.
Another of the functions entrusted to treasurers is to make the most of temporary treasury
surpluses generated by corporations. However the firms surveyed do not value this function in the
same way: the replies obtained indicate that they allocate little importance to maximising the
profitability of surpluses. The idea of treasury departments as profit centres has as yet hardly begun to
develop among these corporations.
Among the innate qualities of financial assets, security and liquidity are prioritised over
profitability in the choice of investment instruments for treasury surpluses, with the most widely used
instruments being bank deposits and government debt. Repos, euro deposits and corporate promissory
notes are used very little, even though they offer high levels of liquidity and security. This may be due
to unawareness of their existence and to a lack of information on them from banks.
If treasurers are to seek financing to carry on corporate business at the lowest cost and obtain
the maximum possible profit from liquidity surpluses then negotiations with banks must be a key
aspect of treasury management, due to the importance of maintaining stable relations with banks.
86
For corporations, standard of service and speed are by far the most highly valued factors in
their relations with banks. This confirms the trend for qualitative variables to become determinant in
the choice of banks.
The relationship between the determinant factors in the choice of investment/ financing
instruments and banks on the one hand and size of corporations and the training levels of treasurers on
the other proves to be significant only in some cases. This reinforces the idea that the tasks of treasury
departments are integrated into the culture of cash management, and that the use of new net asset
management techniques is not directly linked to the size of a corporation or to the standard of training
of its treasurer, but rather depends on the initiative of the financial management themselves.
To form corporations into homogenous groups according to their use of similar treasury
management routines, we performed a cluster analysis that placed them in two blocks.
Electronic banking is the variable that best distinguishes corporations that allocate more
importance to treasury management. These corporations seek to incorporate new integrated
management models into their systems. The emerging trend throughout corporate cash management is
towards more use of technology, making cash management more effective and efficient. It is logical
that this factor should carry increasing weight as a basis for choosing banks, because Internet banking
has recently achieved major advances, and seems set to continue doing so in the future. This will
certainly be a highly competitive point for distinguishing between banks.
On the other hand it is surprising how little significance is attributed to new technologies by the
firms in the second group, i.e. those that do not believe in the philosophy of cash management. We can
only assume that their attitude will change in the not too distant future.
For the purpose of estimating the research models for hypotheses testing first, a sample of
companies listed in Tehran Stock Exchange for the time period of 2001-2003 is used. Second, a sample
of state companies is used. We estimate the research models with pooled data for three years, and
overall 647 years-firm. Then, similarly the models are estimated for sample companies in different
industrial groups. Finally, we estimate the research models using cross-sectional data for each year
(2001 to 2003). We estimate the research models for the sample of state companies in the same way.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
87
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]