STD/NA(2002)21
1
STD/NA(2002)21
The measurement of unquoted shares[1]
I.Introduction
1.When one deals with unquoted shares one has to do with a difficult problem. This has to be clear from the beginning. Although ESA95 contains some paragraphs on the issue, centring around the idea of approximating market value, one is left with many practical problems.[2] The issue is even more complex is one wants to eschew national solutions and wants to achieve a high degree of international comparability.
2.When one adds the international dimension, one moves rapidly from a purely theoretical approach to an organisational approach. The role of Eurostat is to provide the facilities to make a common organisational approach workable. A first round of discussions by a group of experts was done when preparing the Manual on Sources and Methods.[3] However it appeared that a more systematic approach would have been necessary.
3.In March 2002 the Financial Accounts Working Party (FAWP) approved the creation of the Working Group on Unquoted Shares (WGUS), with the mandate of exploring the methodology of unquoted shares and of issuing a set of recommendations. Attention will be paid to the complete methodology of unquoted shares, in other terms: all sectors; transactions as well as stocks; assets as well as liabilities.
4.The participation to the WGUS is on a voluntary basis. Still, the participation to its first meeting, in May 2002, has been very large.[4] The main objectives of the WGUS are the following:
a)Assess the current methodologies.
b)Test in detail the most important elements of the methodology.
c)Highlight best practices in areas in which, for various reasons, a common test exercise is not feasible.
d)Create the premises to produce the split of shares into quoted and unquoted for all institutional sectors.
II.The questionnaire
5.During the first meeting of the WGUS a paper was presented to describe the general framework of unquoted shares. It was discussed in detail and a certain number of ideas and clarifications were added to it. This led to another version of the paper and to the writing of the questionnaire.
6.The questionnaire has four purposes:
a)Describe current methodologies.
b)Let emerge important differences in methodologies.
c)Bring out general problems that need to be addressed.
d)Make a complete list of problems, including those of lesser importance
7.The questionnaire is made of 26 boxes, each of them with a standard structure. There is at first a set of 13 boxes on stocks, then 13 boxes on transactions. In order to facilitate the description of the methodology, it was established a typical order of compilation of unquoted shares, by sector and side of the account. It is the following:
a)Non-financial corporationsliabilities
b)Monetary financial institutionsliabilities
c)Other financial intermediariesliabilities
d)Insurance corporations and pension fundsliabilities
e)Rest of the Worldliabilities
f)Rest of the Worldassets
g)Central and State Governmentassets
h)Social security fundsassets
i)Monetary financial institutionsassets
j)Other financial intermediariesassets
k)Insurance corporations and pension fundsassets
l)Non-financial corporationsassets
m)Households and NPISHassets
8.Each box of the questionnaire is made of eight items, which are the following:
a)Reference population (Total of the population; Evolution of the population over time; Stratification by branch; Stratification by size)
b)Sample (Coverage; Evolution of the sample over time; Stratification by branch; Stratification by size)
c)Source of information on the units in the sample (Balance sheet; Supervisory report; Administrative data, including Fiscal Data; Statistics on the issuance of securities; Foreign Direct Investment Statistics; Survey)
d)Adequacy of the content of the basic information (Mixture of more types of equity; Mixture of various sectors)
e)Valuation(s) in the basic information
f)Large unquoted corporation (definition, if any)
g)Relation stock/flow (Stocks as starting point for transactions; Transactions as starting point for stocks)
h)Other remarks
9.The first two items (Reference population; Sample) attempt to understand which are the possibilities to expand the results of a sample to the whole population, and whether work can be done for sub-populations with different characteristics.
10.The third item (Source of information on the units in the sample) is devoted to understand if the information on individual units in the sample comes from reports that are similar or not.
11.The fourth item (Adequacy of the content of the basic information) deals explicitly with the ability to split the basic information by type of equity (quoted shares/unquoted shares/ other equity) and/or by branch of economic activity.
12.The fifth item (Valuation(s) in the basic information) tries to understand if the basic information can be used straightforwardly or requires some type of transformation.
13.The sixth item (Large unquoted corporation) deals with the possibility of limiting the transformation, that might be necessary, from book values to market values to the largest corporations. The idea is that of defining large unquoted corporations as those similar in size to the smaller quoted corporations.
14.The seventh item (Relation stock/flow) tries to establish which is the practical computational link between stocks and flows, if any.
15.The last item (Other remarks) is for comments and information that do not fit well in the above scheme, but are considered relevant.
III.The common test exercise
16.In parallel with devising the questionnaire, it has been decided to explore a totally different approach. Instead of letting the current national approaches play a role, it has been decided to test some methodologies where everything is as standardised as possible across countries. This approach is very costly. It was decided to apply it at first to what was considered the most important problem: the valuation of stocks of unquoted shares of non financial corporations.
17.In this field a preliminary test exercise was made in 1998, with the participation of France, Spain and Italy.[5] A method to which much space was devoted was that of capitalisation of own funds at book value (COFBV). The basic idea of COFBV is that of computing at first the ratio of market value to own funds at book value for quoted corporations, and then applying it to unquoted corporations. The COFBV method was applied in different versions: alternative definitions of the average value; different populations to which to apply it (all unquoted or large unquoted). However it was tested only for two branches.
18.The method was however not very plausible in branches with very few quoted corporations. This of course depends both on the number of branches that one decides to use and on the characteristics of the national economy, like its degree of specialisation.[6]
19.In the discussion in the first meeting of the WGUS the COFBV method seemed the most appealing one. So it was decided to apply it with a number of new features:
a)The treatment of all branches of non-financial corporations at the same time.
b)The use of a limited number of branches (between 6 and 10).
c)The use of two alternative sets of branches: 'traditional NACE' and 'ICT activities on their own'.
d)The creation of a European data base of quoted corporations by branch.
e)The participation of a large group of countries to the test exercise.
f)The participation of large and small countries.
20.Seven countries have accepted to participate to the common test exercise.[7] During the summer all the preliminary steps have been taken to define the branches and a create a first European data base of quoted corporations by branch.[8] Using the quotation indices by branch provided by the European data base on quoted shares, countries are now proceeding in the computations on their national data bases on corporations.
IV.First outcomes from the questionnaire
21.The questionnaire was circulated in late June. By mid August 12 countries have provided replies to it. These replies have been summarised in two steps:
- A stocktaking document. It has the same structure of the original questionnaire, and in 26 boxes it describes, for each item, the prevalent pattern and the exceptions. This makes up a large document of about 60 pages.
- Six short summary documents, that contain the main finding by institutional sector (about 2 pages each).
22.All this material needs to be checked with the participants of the WGUS. In some cases there are requests for further clarification, in other cases there might have been some misunderstanding in interpreting the information. It is expected that making available to the participants of the WGUS the streamlined methodology of all the countries will stimulate the discussion, and help in selecting the methods that seem more valid.
23.The last outcome from the Questionnaire is an open issues document. It deals with large areas that seem worth exploring. We will present here some of them, dealing mostly with households.
Ratios on stocks and transactions
24.Sometimes the methodology to obtain splits in the information is to use ratios derived from phenomena similar to the one for which one has no direct observations. (This can apply to different phenomena, like the split between quoted and unquoted shares of a sector, the split of unquoted shares between different sectors etc.).
25.Are there any practical cases for which the same ratio is applied to both stocks and transactions? Do we think that this is a good methodology? Could we improve it?
26.For example we could examine the pros and cons of substituting the first chain below with the second chain:
Given ratio on stocks stocks
Given ratio on stocks transactions
Given ratio on stocks stocks difference of stocks new ratio on difference of stocks transactions
27.The main issue here seems to be the appropriateness of deriving 'approximate' transactions from stocks. This depends largely on the methods for valuing stocks.
Securities deposited with banks
28.In some cases securities deposited with banks are used to estimate the assets of households (and of other sectors).
29.How many banks provide this information? How frequently do customers deposit unquoted shares with banks? Do corporate clients deposit unquoted shares in the same way as non-business customers? What is the valuation of securities deposited with banks? Do we only have data on outstanding amounts?
Securities deposited with banks and residual calculation.
30.If we can compute households' holdings unquoted shares both from securities deposited and as a residual, how do we reconcile the two estimates?
Residual calculation of stocks and transactions
31.Often households' holdings of and transactions in unquoted shares are derived from two independent residuals, one for stocks and the other for transactions. How do we analyse the coherency of the two residuals? Are the two calculations consistent? How do we treat possible inconsistencies?
1
[1]Note by Riccardo Massaro, Eurostat, Unit B4. Prepared for the OECD meeting on National Accounts of 10 October 2002.
[2]The main references in ESA95 are §§ 7.54-7.55.
[3]See §§ 2.1-2.2 of Issue N°2, in the 'Manual on Sources and Methods for the Compilation of ESA95 Financial Accounts', Eurostat 2002.
[4]Thirteen EU Member States, three candidate countries, Switzerland, OECD and ECB.
[5]The work was carried out by a sub-group of the WG MUFA (EMI, and then ECB).
[6]As a matter of fact, Spain is currently using a method that uses the net operating profits of past years to derive the current valuation of unquoted shares.
[7]The countries are: Belgium, Denmark, Finland, France, Italy, Portugal and Sweden.
[8]Banque de France has kindly accepted the role of collecting the information to create the European data base on quoted shares.