Short to mid-term projections for Credit Unions in the United States

Abstract

Credit Unions are financial co-operatives owned and controlled by their members. Credit unions are very popular in the United States (US) where they operate both on state as well as on a national level. Since these organisations are, in the US, in direct competition with retail banks, short to mid-term extrapolations are essential in order to visualize the trends for access to credit in the future. In the present study we use published data for six key financial figures from ten states in the US. An Expert Forecasting Support System, selecting via a competition among classic extrapolative techniques, has been employed in order to prepare one-year as well as five-years ahead forecasts and confidence intervals for the times series under consideration.

Full text

SHORT TO MID-TERM PROJECTIONS FOR CREDIT UNIONS
IN THE UNITED STATES
Dr Konstantinos Nikolopoulos, Decision Sciences and Operations Management Group,
Manchester Business School, Booth St. East, Manchester M15 6PB, United Kingdom
kostas.nikolopoulos@mbs.ac.uk
Mr Michael C. Handrinos, Department of Applied Social Science, Faculty of
Arts and Social Sciences, Lancaster University, Lancaster LA1 4YD, United Kingdom
m.handrinos@lancaster.ac.uk
ABSTRACT
Credit Unions are financial co-operatives owned and controlled by their members. Credit
unions are very popular in the United States (US) where they operate both on state as well as
on a national level. Since these organisations are, in the US, in direct competition with retail
banks, short to mid-term extrapolations are essential in order to visualize the trends for
access to credit in the future. In the present study we use published data for six key financial
figures from ten states in the US. An Expert Forecasting Support System, selecting via a
competition among classic extrapolative techniques, has been employed in order to prepare
one-year as well as five-years ahead forecasts and confidence intervals for the times series
under consideration.
Keywords: Credit Unions, Forecasting, Time series, Method Selection
INTRODUCTION
Credit Unions are financial co-operatives owned and controlled by their members. In most
countries in the world they serve the ones who can not get access to credit from retail banks
due to bad credit history or low credit scoring. Credit unions are very popular in many
countries as for example in Ireland and in the United Kingdom. In the United States (US)
Credit Unions do play a significant role in the economic environment. In the state of
California, for example, one in five residents trust his savings and counts for access to credit
to such institutions. Since those organisations are, in the US, in direct competition with retail
banks, short-to mid-term extrapolations are essential in order to visualize the trends for access
to credit in the future.
Time series methods have been proven very successful for extrapolating basic trends into the
future (11). Among these methods, Exponential Smoothing approaches have been
consistently performed significantly well in terms of out-of-sample forecasting accuracy
(10,12,13). In the present study we use published data acquired by permission from Callahan
and Associates Inc. for the following ten states in the US: Alabama, Alaska, Arizona,
Arkansas, California, Colorado, Connecticut, Delaware, District of Columbia (DC) and
Florida. Six key figures for each state are examined: number of Credit Unions, number of
Members, Total Assets, Total Shares, Total Loans and Total Capital.
An expert software, selecting via a competition among classic extrapolative techniques as
exponential smoothing, moving averages, etc, is employed in order to provide one-year and
five-year ahead forecasts as well as 95% confidence intervals for the selected times series.
1303

In terms of structure the present paper includes the following: section 2 consists of a short
introduction to Credit Unions, while section 3 describes the current state of Credit Unions in
the US. Section 4 presents the data analysis as well as the projections for 2006 to 2010. The
paper concludes with the discussion of the results and with some suggestions for future
research. Finally, the appendix includes all the tables with the projections for each state
separately.
A SHORT INTRODUCTION TO CREDIT UNIONS
A credit union is a mutual financial co-operative which provides convenient and accessible
savings and loans to its members; it is a democratically owned co-operative society which is
built on membership and on the principles of equity (1). It is owned, operated and controlled
by the members, a group of people who share a ‘common bond’ of, for instance,
employment, profession, locality, religion, or ethnic group. It uses members’ savings to
provide loans at reasonable rates of interest. Credit unions rely heavily on volunteer
commitment; they can help foster community spirit and play an educational role in money
management (14). Hermann Schulze-Delitzsch and Friedrich Wilhelm Raiffeisen created the
first true credit unions in Germany in 1852 and 1864. In 1849, Raiffeisen founded a credit
society in Flammersfeld, but it depended on the charity of wealthy men for its support. In
1864, Raiffeisen organized a new credit union along principles still fundamental today (5).
Credit Unions’ foundation are small pre-cooperative groups who federate together to form the
‘union’. This union is not controlled or owned by non-borrowers. It challenges conventional
‘banking for the poor’ in which institutions are built by professionals who by and large lead a
life very different from the poor borrowers who they seek to serve. Credit unions are an
important mechanism through which people themselves can own and control resources
(Albee and Gamage, 1996).
Credit unions are financial co-operatives which are open to people wishing to save and lend
on the basis of a common bond. There are threads which unite credit unions world-wide, but
the details of how any given union functions is determined locally. Credit unions most often
operate without the bureaucracy and institutional apparatus of centralised and hierarchical
organisations such as banks. Worldwide there are an estimated 136 million individual
members of Credit Unions in 90 countries, where it is estimated that their impact reaches an
average of four additional household members for each Credit Union member. Thus, they
have an effective outreach of 680 million persons (16).
Key elements of the Credit union movement, and their difference from retail banks, include:
 Decentralised development in which the base of the organisation is the heart of
activities. This contrasts with a banking approach in which a hierarchy of decision-
making and management exists, and where resources are delivered downwards to
beneficiaries rather than managed by them;
 Saving is promoted before loans are issued to members thus building the self-reliant
base of lending capital from the members themselves. Again, this is contrary to
banks’ way of doing business as saving is promoted as unrelated to ones ability to
borrow;
 A common bond between members is emphasized. This bond can form on the basis of
employment, profession, locality, religion, or ethnic group. Retail banks have no such
pre-requisites; as long as a person wishes to become a customer they would
investigate upon a person’s financial profile rather than any of the above-mentioned;
1304

 Common ownership by the members is promoted. This is done through the issuing of
shares which are sold to members. This also contrasts with the banking approach as
retail banks are Public Listed Companies where the shareholders are anonymous and
can hold uneven numbers of shares;
 Equity is another important element. That is voting takes place on the basis of
membership not on unequal ownership of shares as banks do.
Various studies have indicated that the vast majority of people joined their credit union
specifically to gain access to the services it offered, particularly as they had found it difficult
to gain access to them through retail providers. In particular, people liked the combination of
saving and credit that both encouraged them to put money aside and also amplified their
borrowing power. Around half, however, had drawn a distinction between saving and
borrowing at the time they joined their credit union, and had been attracted primarily by one
or the other (15). It is important to note, however, that whatever people’s initial motivation
for joining the credit union, the majority made full use of the services on offer once they had
become members.
However, despite the fact that their philosophy and characteristics are very simple, Credit
Unions sometimes confuse people who misunderstand their purpose and methods and who
want to borrow a large amount on the day they join. In order to qualify for a loan, the
borrower must be a member of the credit union, and therefore within the common bond, must
have saved for a certain period of time (although this requirement can be waived in some
cases) and must retain savings in the particular ratio to the amount borrowed.
CREDIT UNIONS IN THE US
The credit union industry in the US, since the 1977 change of legislation of the financial
services industry, has been affected by a number of forces such as innovation, technology,
deregulation and competition. Due to the complex interplay of these forces various types of
financial intermediaries with similar functions have been created but, at the same time, a
separate and distinct legal and regulatory structure remains (9). The credit union industry has
been affected by all these changes as this deregulation was accompanied by the introduction
of a less restrictive interpretation of the common bond requirement for membership, which
created new opportunities for growth and merger (6).
This relaxation of the previously restrictive common bond requirements has inevitably
brought Credit Unions into closer competition with other financial institutions, especially
retail banks. Thus, one could well argue that they act as mini-banks as they operate in a more
market-oriented approach than they do in some other parts of the world by offering a range of
products and services that are similar to that of retail banks. They are a success story in the
movement’s history as they have grown significantly in recent years. With more than 8,880
credit unions now serving 86.5 million members the Credit union industry continues to attract
shareholders/depositors because of its generally lower cost services and higher returns on
savings (2). This represents a penetration of the economically active population of over 50
per cent making Credit unions a strong player in the US financial services’ market. Average
membership per union stands at 9,741 almost nine and a half times that of Britain and double
that of Ireland (3,4). In February 2006, the average membership number in British Credit
Unions was 814,538 in a total number of 779 Credit Unions and £900 million in assets (7).
Nevertheless, despite their small number, the credit union movement is now the fastest-
growing, fully mutual co-operative sector in Britain (8).
1305

DATA AND FORECASTS
Annual data for five consecutive years from 2001 to 2005 has been collected from the
published database of Callahan and Associates Inc. for the following ten states in the US:
• Alabama , Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware,
District of Columbia, Florida
These ten datasets are the first part of a larger dataset with data for all 50 states in the US.
The analysis of the whole dataset will be the subject of a future study where regional
differences will also be investigated. For each of those states six time series were constructed
analysed and extrapolated for the following six key figures for each state respectively:
• Number of Credit Unions, Number of Members, Total Assets, Total Shares, Total
Loans, Total Capital
In total sixty time series were analysed and a projection for up to 5 years ahead has been
prepared with 95% confidence intervals. For each individual series the best method was
selected via a forecasting competition over available data (2001-2005) among the following
exponential smoothing methods (11):
• Simple Exponential Smoothing
tmt
ttt
ttt
LF
eLL
FYe
=
+=
−=
+
−
α
1
where et: the error of the model, Yt the data, Ft the forecasts, Lt the level of the series as
estimated by the model and a the level smoothing factor for each period t=1..5.
• Holt – Linear Trend Exponential Smoothing
ttmt
ttt
ttt
tttt
ttt
mTLF
TLF
eTT
eTLL
FYe
+=
+=
+=
++=
−=
+
+
−
−−
1
1
11
β
α
where et: the error of the model, Yt the data, Ft the forecasts, Lt and Tt the level and trend
of the series as estimated by the model and a and β the level and trend smoothing factors
for each period t=1..5.
• Damped Trend Exponential Smoothing
∑
=
+
+
−
−−
+=
+=
+=
++=
−=
m
i
t
i
tmt
ttt
ttt
tttt
ttt
TLF
TLF
eTT
eTLL
FYe
1
1
1
11
φ
φ
βφ
αφ
where et: the error of the model, Yt the data, Ft the forecasts, Lt and Tt the level and trend
of the series as estimated by the model and a, β, φ the level, trend and damping smoothing
factors for each period t.
1306

Box Jenkins or decomposition methods (11) have not been tested due to the lack of sufficient
number of observation for each individual series (five data points in total). In the majority of
the cases Holt – Linear Trend Exponential Smoothing method is selected by the expert
method.
The forecasts for all ten states are presented in the appendix (tables 1-10). For all states the
same pattern from 2006 to 2010 can be seen:
• The number of Credit Unions will be decreasing annually 3.6 % on average for the
next five years;
• The number of Members will be increasing annually 1.3 % on average for the next
five years;
• Total Assets will be increasing annually 5 % on average for the next five years;
• Total Shares will be increasing annually 4.8 % on average for the next five years;
• Total Loans will be increasing annually 8.4 % on average for the next five years;
• Total Capital will be increasing annually 5.6 % on average for the next five years.
The case of California is of great interest as 20% of the population trust Credit Unions for
savings and access to credit. Figure 1 shows the decrease of Credit Unions in the state of
California while Figure 2 shows the increase of total capital.
Figure 1. Number of Credit Unions for the state of California
Confidence
Intervals*
1307

Figure 2. Total Capital for the state of California
* In these projections, there is a 95% probability that the actual value will be between those
intervals.
Table 1 shows the detailed 5-years ahead forecasts for the state of California while Table 2
shows the forecasts for the same period from aggregate data from all ten states.
Table 1. State of California
DATE 2006 2007 2008 2009 2010
Number of
Credit Unions (L) 524 493 467 442 419
Number of
Credit Unions 551 535 520 504 489
Number of
Credit Unions (U) 577 577 573 566 559
Number of Members (L) 9.650.142 9.643.802 9.680.293 9.736.290 9.804.130
Number of Members 9.917.451 10.065.147 10.212.843 10.360.539 10.508.236
Number of Members (U) 10.184.760 10.486.492 10.745.393 10.984.788 11.212.342
Total Assets (L) 102.665.136 102.855.216 104.882.544 107.743.624 111.110.256
Total Assets 113.979.768 120.747.944 127.516.120 134.284.304 141.052.480
Total Assets (U) 125.294.400 138.640.672 150.149.696 160.824.992 170.994.704
Total Shares (L) 87.346.216 86.886.680 88.051.280 89.955.832 92.309.712
Total Shares 97.498.280 102.882.760 108.267.232 113.651.712 119.036.192
Total Shares (U) 107.650.344 118.878.840 128.483.184 137.347.584 145.762.672
Total Loans (L) 71.182.912 73.549.472 78.065.072 83.403.056 89.213.232
Total Loans 78.342.384 87.064.552 95.786.720 104.508.888 113.231.048
Total Loans (U) 85.501.856 100.579.632 113.508.368 125.614.720 137.248.864
Total Capital (L) 11.391.244 11.481.502 11.770.172 12.147.917 12.579.465
Total Capital 12.564.713 13.355.891 14.147.069 14.938.247 15.729.425
Total Capital (U) 13.738.182 15.230.280 16.523.966 17.728.578 18.879.384
Confidence
Intervals*
1308

Table 2. Aggregate data from ten states
DATE 2006 2007 2008 2009 2010
Number of Credit Unions
(L) 1088 1008 939 873 814
Number of Credit
Unions 1163 1123 1085 1043 1004
Number of Credit Unions
(U) 1235 1239 1229 1214 1198
Number of Members (L) 6.384.565 6.377.983 6.398.744 6.431.963 6.472.773
Number of Members 6.597.377 6.690.281 6.783.184 6.876.087 6.968.991
Number of Members (U) 6.810.190 7.002.579 7.167.623 7.320.214 7.465.212
Total Assets (L) 47.802.786 47.326.073 47.703.524 48.472.368 49.479.392
Total Assets 53.263.860 55.883.507 58.503.153 61.122.802 63.742.446
Total Assets (U) 58.724.933 64.440.941 69.302.780 73.773.235 78.005.500
Total Shares (L) 41.390.885 40.836.837 41.030.978 41.568.799 42.315.962
Total Shares 46.219.316 48.386.023 50.552.729 52.719.435 54.886.142
Total Shares (U) 51.047.748 55.935.210 60.074.481 63.870.071 67.456.322
Total Loans (L) 31.192.781 31.457.293 32.548.623 33.962.450 35.563.682
Total Loans 34.398.735 37.184.133 39.969.529 42.754.927 45.540.324
Total Loans (U) 37.604.686 42.910.973 47.390.435 51.547.403 55.516.966
Total Capital (L) 5.600.973 5.612.117 5.722.519 5.875.208 6.053.213
Total Capital 6.140.120 6.486.214 6.832.310 7.178.406 7.524.501
Total Capital (U) 6.679.266 7.360.311 7.942.100 8.481.603 8.995.789
CONCLUSIONS AND FUTURE RESEARCH
A number of conclusions from the presented projections are the following:
• Merging
As the total number of Credit Unions will be decreasing annually with an average of 3.6
% while all the other key figures are steadily increasing, we speculate that major merges
are about to happen in the following years. Therefore, industry consolidation is a trend
that will continue to prevail.
• Blooming of all key financial figures
All the key financial figures increase more than 5% on average per year. Thus there is a
strong potential for Credit Unions in the US in terms of market penetration and growth.
As these projections have been made using time series models (11), the assumption that the
environmental variables will remain constant has to be made. Thus it is a necessity that the
financial environment, competition, legislation etc have to remain constant for the next five
years. If these variables do not remain constant then econometric models should be employed
in order to take account for these major environmental changes that could potentially happen
in the future. Such major changes could be embedded as dummy variables in an econometric
model.
Future research should focus on:
• The inclusion in the study of all 50 US states;
• The geographical analysis via a Geographic Information System in order to identify
any regional differences either intra-state or across states;
1309

• The analysis of data from other countries in order to see if these US trends can be
generalised on a global scale. An analysis of the UK Credit Union movement would
be of great value as it appears that it is on a very high growth rate thus making it a
good area for research;
• An analysis between the different kinds of Credit Unions (employee-based,
community-based etc) and the different patterns of their growth and expansion.
ACKNOWLEDGEMENTS
The authors would like to thank Chip Filson and Scott Patterson from Callahan and
Associates, Inc. (http://www.creditunions.com) for allowing us to use the organisation’s data
in our analysis.
REFERENCES
1. Albee, A. and Gamage, N. (1996) Our money, our movement: building a poor people’s
credit union. London: Intermediate Technology Publications.
2. Boldin, R. J., Leggett, K. and Strand, R. (1998) Credit union industry structure: an
examination of potential risks. Financial Services Review 7 (1998) 207-215.
3. Callahan and Associates (2006) Directory Online.
http://www.creditunions.com/directoryonline. Accessed 18/02/2006
4. Callahan and Associates (2005) Callahan’s 2006 Credit Union directory.
5. CUNA (2006) Early cooperative activites. Credit Union National Association.
http://www.creditunion.coop/history/1stcoops.html Accessed 23.03.2006, 3.15pm.
6. Goddard, J. A., McKillop, D. G., Wilson, J. O. S. (2002) The growth of US credit unions.
Journal of banking and finance, Vol. 26, pp. 2327-2356.
7. Goth, P., McKillop, D. and Ferguson, C. (2006) Building better Credit Unions. Bristol:
The policy press.
8. Jones, P. A. (1999) Towards sustainable credit union development. ABCUL, Manchester.
9. Kaushik, S. K. and Lopez, R. H. (1994) The structure and growth of the Credit Union
Industry in the United States: meeting challenges of the market. American Journal of
Economics and Sociology, Vol. 53, No.2, April.
10. Makridakis S. and Hibon M. (2000). The M3-Competition: Results conclusions and
implications, International Journal of Forecasting 16 451-476.
11. Makridakis S., Wheelwright S. and Hyndman R. (1998). Forecasting methods and
applications 3rd Edition Wiley.
12. Makridakis S., Chatfield C., Hibon M., Lawrence M., Mills T., Ord K. and
Simmons L. (1993) The M2-Competition - A real-time judgmentally based forecasting
study International Journal of Forecasting 9 5-22.
13. Makridakis S., Andersen A., Carbone R., Fildes R., Hibon M., Lewandowski R.,
Newton J., Parzen E. and Winkler R. (1982). The accuracy of extrapolation (time-
series) methods - results of a forecasting competition, Journal Of Forecasting 1 111-153
14. National Consumer Council (1994) Saving for credit: the future for credit unions in
Britain. York: Joseph Rowntree Foundation.
15. Whyley, C. Collard, S. and Kempson, E. (2000) Saving and borrowing: use of the Social
Fund budgeting loan scheme and Community Credit Unions. London: Department of
Social Security.
16. WOCCU (World Council of Credit Unions) (2004) Annual Report
(http://www.woccu.org/pdf/annrpt_04.pdf)
1310