I.  Goals and Objectives of the Coding System

The principal goals and objectives of this coding system are completeness of coverage, brevity, simplicity, adaptability to automation, stability over time, and, where appropriate, extensibility. In addition, the codes must be suitable (useful) for the broadest possible range of countries, as member countries of the IMF, the OECD, and the European Union will be asked to use them to meet part of their data reporting responsibilities. Each of the above objectives had some impact on the structure of the codes.

The scope of the codes is quite narrow. They include the standard components for balance of payments and international investment position data as defined in the 5th edition of the Balance of Payments Manual; the Extended Balance of Payments Services (EBOPS) classification, as defined in the Manual on Statistics of International Trade in Services; and the classification of foreign currency liquidity, as shown in the Data Template on International Reserves and Foreign Currency Liquidity: Operational Guidelines.

The coding scheme makes no attempt to address dates or periodicity, currency, country or partner country, economic activity, or a number of other related topics, as these are the concern of a much broader audience. By avoiding any overlap with these topics, the balance of payments coding system may be seen to be compatible with any international standards developed for these topics. In addition, the use of the coding system is not limited to any specific environment; rather, it is designed to be sufficiently neutral to allow its use with any number of manual and electronic environments.

The coding system consists of three parts: a single digit position code, a three digit topic code, and a tag code that may be any number of digits in length (i.e., <position><topic><tag>). The position and topic codes are always required, while the tag code is optional. The position and topic codes are completely defined in this document; that is, agreement has been reached on the set of positions and topic codes shown in this document. Proposed additions or changes to the position or topic codes may be made by applying to the IMF and will take effect when agreement is obtained from the parties involved in the original agreement.

In contrast to the position and topic codes, the tag codes are not subject to any international agreements and, thus, have no restrictions on their use. Tag codes are expected to be defined by countries, groups of countries, or international organizations to suit their own needs and purposes. As such, tag codes do not follow any worldwide standards. Their purpose is to provide extensibility to the coding as needed by various user communities.

The topic codes were chosen for their brevity; also, the limitation of the codes to numerals is designed to contribute greatly to the ease of use. Short numbers are easy to read and recall. Their brevity and memorability will help minimize the burden placed upon respondents who need to fill out data collection forms, improve the efficiency of those involved in entry of the data into automated systems, and improve the transparency of the data when it is exchanged. Leading zeros, blank (null), and punctuation characters have been avoided to ensure the adaptability of the codes to any computing software environments that accept simple integers (e.g., the topic codes are drawn from the set of integers from 1 to 998).

The codes were also chosen to provide some guidance as to the subject covered. Thus, the single-digit position code defines the type of item being described in the balance of payments or the international investment position (e.g., flows, stocks, or adjustments). The first digit of the topic code determines the broad section of the balance of payments accounts being addressed. While asset and liability classifications and institutional classifications are treated consistently throughout the codes, no attempt was made to incorporate additional intelligence in the codes to support sorting. This was thought to be unnecessary because balance of payments data are being entered into computing environments in virtually every country in the world, and the sorting needs of balance of payments compilers are easily defined. Once in a computing environment, the data may be sorted by provision of programming instructions, which, once defined, may be continuously reused. The small increase in the potential burden on computers by this coding structure will be trivial in terms of operational costs and, presumably, will be overshadowed by the savings in human costs owing to its simplicity.

The codes resulting from the above-described criteria are compact. In the services and other investment sections, there is almost no room for additional codes. However, these sections are also the most thoroughly covered, and, therefore, the least in need of additional space. All other sections of the topic codes have ample room for expansion. Nevertheless, it is not expected that this space will be used casually. The stability afforded by this very process is what will make the topical classification and the associated codes useful.

II.  Definition of the Coding System

The coding system incorporates all the standard components and supplementary information lines of the 5th edition of the Balance of Payments Manual, all the components and supplementary items contained in the Extended Balance of Payments Services classifications in the Manual on Statistics of International Trade in Services, and the data items contained in the Sample Form for Presenting Data in the Template on International Reserves/Foreign Currency Liquidity in the Data Template on International Reserves and Foreign Currency Liquidity: Operational Guidelines.

The code consists of three components or sections as follows:

<position>	one alphanumeric character describing the position of the subject in the IIP/BOP accounts or the use of the topic in foreign affiliate statistics;
<BOP topic>	three decimal digits (selected from the range of integers from 1 to 998) identifying all the BOP, trade-in-services, reserves template, and selected supplementary information components;
<tag>	an alphanumeric user-defined component that may be of any length.

The complete code would take the form <position><topic><tag>. However, the tag component is entirely optional. In contrast, the position and the topic components of the code are always required. Thus, a common implementation of the code will take the form <position><topic>.

The first section of the code describes the position in the international investment position and balance of payments accounts and is defined as follows:

The first digit of the topic component identifies the section of the balance of payments as follows:

The second and third digits of the component are generally sequential counts of the components with some gaps to allow for additional codes to be included at a later time. In addition, with the exception of the direct investment accounts, the second digit of the topic component takes the values 0, 1, 2, 3, or 4 for assets and 5, 6, 7, 8, or 9 for liabilities in the financial accounts.

III.  Examples of Usage of the Coding System

As described in the earlier sections of this document, the coding system consists of three components: the position, topic, and tag. The codes for the position and topic components have been defined in this document. However, the tag component is to be user-defined. For purposes of providing an example of the usage of tag codes, we assume the following definition of tag codes for exports. These codes are based upon the components published in the quarterly publication Balance of Payments Australia and thus may be presumed to be user defined:

A number of options may be chosen by the user for the presentation of the position, topic, and tag codes. For example, the exports of wool and sheepskins may be coded as 2 110 ad (i.e., 2 110 for general merchandise credits, and ad for wool and sheepskins). However, this may also be presented as 2110ad, 2 110ad, or 2110 ad in various documents to suit the convenience of the user. In electronic environments, it is generally recommended that the position, topic, and tag components be stored in separate fields within the data record. A complete coding of the above table is given below.

For a second example, let us consider the coding of a small portion of the other investment section of the financial accounts. This example will not include a tag component:

When considering credits and debits for loans, the language that is used commonly is drawings and repayments. The BPM recommends reporting of all drawings and repayments for long-term loans as a supplementary classification. Most other financial account items are currently collected by the IMF on a net basis. Nevertheless, the coding system provides for the identification of all flows on a credit, debit, and net basis.

The example below shows the coding for the international investment position for the same set of topics used in the previous example. The blank column labeled BOP Flows would normally contain the three columns for the balance of payments flows shown in the previous example.

For more examples of coding, one may look at the three sets of codes that the IMF uses for the collection of balance of payments, international investment position, and international reserves/foreign currency liquidity data.