El Salvador
Technical Assistance Report-The Coordinated Direct Investment Survey and the Coordinated Portfolio Investment Survey (April 21-29, 2014)

This Technical Assistance Report discusses the findings and recommendations made by the IMF mission about the compilation of Coordinated Direct Investment Survey and Coordinated Portfolio Investment Survey (CPIS) in El Salvador. The mission recommended the authorities to research the nature of the information available that may be useful for starting the CPIS. This entails discussions regarding the forms designed during the mission for requesting information from new sources. It was also recommended to continue efforts to improve the coverage of surveys applied to nonfinancial private sector enterprises, with emphasis on the largest enterprises that are still reluctant to respond to the balance-of-payments questionnaires.


This Technical Assistance Report discusses the findings and recommendations made by the IMF mission about the compilation of Coordinated Direct Investment Survey and Coordinated Portfolio Investment Survey (CPIS) in El Salvador. The mission recommended the authorities to research the nature of the information available that may be useful for starting the CPIS. This entails discussions regarding the forms designed during the mission for requesting information from new sources. It was also recommended to continue efforts to improve the coverage of surveys applied to nonfinancial private sector enterprises, with emphasis on the largest enterprises that are still reluctant to respond to the balance-of-payments questionnaires.

Executive Summary

As part of the Regional Project on Harmonizing External Sector Statistics (ESS) developed by the Technical Assistance Sector for Central America, Panama, and the Dominican Republic, a technical assistance mission on ESS visited San Salvador during the period April 21–29, 2014.

The purpose of the mission was to support the Central Reserve Bank of El Salvador (BCRES) in its efforts to make further improvements in the compilation and dissemination of balance of payments and international investment position (IIP) statistics. Particular emphasis was placed on the compilation of the Coordinated Direct Investment Survey (CDIS) and the Coordinated Portfolio Investment Survey (CPIS).

In El Salvador, ESS are processed and disseminated by the Balance of Payments Department (DBP), which reports to the Office of Economic Studies and Statistics. The BCRES does not have legal instruments empowering it to request information from the country’s various economic agents, or to apply effective sanctions for failure to comply. Currently, the legal mandate of the BCRES to request information to compile the ESS is based on Article 64 of its 2011 Organic Law, which establishes, on the one hand, the obligation of the BCRES to publish macroeconomic statistics, with specific mention of the balance of payments and, on the other hand, the obligation of public sector, financial sector, and nonfinancial private sector entities to provide the information requested of them for this purpose.

The mission gave particular attention to the coverage and compilation practices for foreign direct investment (FDI) statistics, including the CDIS. El Salvador has been participating in the CDIS since 2010, using data from late 2009. There are significant deficiencies in coverage, particularly with respect to FDI abroad, because the largest nonfinancial private sector companies are reluctant to respond to FDI surveys. Some issues were also found with the identification of fellow enterprises.

El Salvador is considering participating in the CPIS. During the mission, visits were made to the main securities custodians as well as to the securities area of the Financial System Superintendency (SSF). These meetings made it clear that detailed information exists for implementing the CPIS, particularly data kept at the SSF where all securities traded by local brokers must be properly reported. However, there are still some doubts regarding the correct recording of secondary market operations.

The mission identified three benchmark actions:

  • Research the nature of the information available that may be useful for starting the CPIS. This entails discussions regarding the forms designed during the mission for requesting information from new sources (SSF and principal custodians) (December 2014).

  • Continue efforts to improve the coverage of surveys applied to nonfinancial private sector enterprises (FDI and domestic enterprises), with emphasis on the largest enterprises that are still reluctant to respond to the balance of payments questionnaires (January 2015).

  • Make residence-based adjustments to external public debt to exclude government securities held by residents. This exercise should include treasury bills held by nonresidents under ESS (December 2014).

Action Plan

To achieve the objectives of the Regional Project on Harmonizing External Sector Statistics of the member countries of the Technical Assistance Center for Central America, Panama, and the Dominican Republic as well as to improve practices in the compilation and dissemination of balance of payments and IIP statistics, the following action plan has been developed in conjunction with the BCRES. There are three benchmark actions, which will be subject to special compliance monitoring.

article image

I. Introduction

1. As part of the Regional Project on Harmonizing External Sector Statistics, a technical assistance mission on ESS visited the Central Reserve Bank of El Salvador (BCRES) during the period April 21–29, 2014.

2. The objectives of the harmonization project are to compile exhaustive and consistent external sector statistics for the entire region, in line with the basic principles, standards, and internationally accepted compilation and dissemination practices contained in the sixth edition of the Balance of Payments and International Investment Position Manual (BPM6).

3. The main tasks of the mission were to:

  • Review methodological aspects of the Coordinated Direct Investment Survey (CDIS).

  • Evaluate the regulatory framework and the availability of information for launching the Coordinated Portfolio Investment Survey (CPIS).

  • Review the information available on external debt incurred through portfolio investment instruments, in order to adapt that information to the residence criteria of the BPM6.

  • Review the information available on insurance, in order to suggest requirements for achieving improvement in the collection and recording of this item.

  • Evaluate the Balance of International Transactions information to determine whether some deficiencies in the delivery of information can be remedied in this way, particularly with respect to the national airline.

  • Advise the Balance of Payments Department (DBP) on other topics that may arise during the mission.

II. Background

4. The DBP reports to the BCRES Office of Economic Studies and Statistics and is the area responsible for the compilation, generation, and quarterly dissemination of balance of payments and international investment position (IIP) statistics. In that capacity, the DBP also concerns itself with submitting information to the IMF with respect to other international statistical undertakings, such as the CDIS, the International Reserves and Foreign Currency Liquidity Template and, potentially, if defined accordingly, the CPIS. El Salvador is a subscriber to the Special Data Dissemination Standard.

5. Since April of this year, with dissemination of the figures corresponding to the fourth quarter of 2013, El Salvador’s balance of payments and IIP statistics are prepared on the basis of the principles and methodological guidelines prescribed in the BPM6. Thus, in the balance of payments, a distinction is made between goods and services for goods entering the country for processing and subsequent export as part of another good, based on whether or not a change in ownership occurs. Similarly, changes were made in the balance of services to include, as part of financial services, financial intermediation services indirectly measured, which were previously implicit in the interest figures. FDI is shown according to the criterion of classification of assets and liabilities and includes related operations with fellow enterprises abroad.

6. Obtaining the information that is used as the basis for constructing the balance of payments and IIP statistics starts with the quarterly application of a variety of questionnaires directed primarily to nonfinancial private sector economic units, as well as access to administrative and accounting information from public and financial sector entities.

III. Coordinated Direct Investment Survey

7. El Salvador has been participating in the CDIS since this exercise was promoted by the IMF in 2010 for reporting year-end 2009 figures. The information reported to the IMF includes data on positions in direct investment received from abroad (inward) as well as investments made by residents abroad (outward). The positions are in turn classified as direct capital contributions (equity) and those related to debt instruments.

8. For purposes of complying with the requirements of the CDIS and the BPM6, the DBP has expanded the questionnaire it has traditionally used to capture DFI information (Survey of Foreign Investment Companies and/or Participation Abroad). Thus, the new form now includes a specific section intended to obtain data on the positions, flows, and revaluations associated with resident enterprises’ operations with nonresident fellow enterprises.

9. FDI statistics are compiled and produced by the BCRES because none of the official agencies related to the phenomenon are effectively able to do so. The Ministry of Economy lacks operational capacity to force companies to comply with the obligation to record foreign capital contributions, while El Salvador’s Exports and Investment Promotion Agency only has information regarding intended investment commitments but not the amounts actually invested nor the timing thereof.

10. In principle, the above-mentioned survey seeks to obtain information on the universe of enterprises with FDI that are operating in the country. However, given the lack of legal provisions requiring companies to respond and the already noted lack of effective support from other public entities, the identification and subsequent collection of data on such enterprises becomes difficult. To maintain good coverage, the DBP carries out frequent exercises to compare its own directory of enterprises with the directory used by the General Directorate of Statistics and Census in its economic surveys. In this way, DBP staff ensures that at least the largest companies are being contacted.

11. Based on the exercise cited in the previous paragraph, it is estimated that in the case of the survey taken in 2012, about 80 percent of enterprises were surveyed, including most of the largest enterprises, so that it is reasonable to think that in value terms the coverage should exceed that percentage. For that same year, the survey response rate was approximately 70 percent, which means that the FDI figure yielded by the survey should have covered between 55 percent and 60 percent of the total. An inference is made in order to compensate for non-response, based on repeating the final position figure reported by each enterprise. It should be noted that reported flows are not affected (i.e., the inference only affects the positions data).

12. The questionnaire used to capture information on (inward and outward) FDI consists of a broad form that additionally considers other information needs, such as data on enterprises with FDI corresponding to the balance of services, as well as other financial items that do not represent FDI. The questionnaire requires enterprises to report data related to initial and final asset and liabilities positions, related transactions, as well as variations due to revaluation of the various FDI concepts: equity, debt, and reinvested earnings. In addition, to comply with the requirements of the BPM6, a section was added for recording FDI concepts related to fellow enterprises.

13. Although everything indicates that the staff of the surveyed companies responding to the questionnaire seem to understand the concept of fellow enterprises well, the response to the related section has presented problems primarily due to difficulties that these respondents face when identifying all fellow companies in the corporate group to which they belong, and being able to separate out operations with each of them. This means that the limitations on having this information have been primarily practical in nature. An example is the case of a large transnational enterprise that in its list of fellow enterprises identifies some 300 companies all over the world, which it would have to track individually in order to comply with the questionnaire requirements. Specifically, this company offers to identify only the principal movements in each reporting period. Thus, besides the fact that this type of case would involve a constant under-estimation of the phenomenon; for the CDIS, this would lead to the assumption that positions with fellow enterprises would be reported only for the principal countries in each period, which could change frequently over time.

14. Independently of the CDIS, DBP staff has identified a problem with the existence of holding companies. It has not yet decided how to resolve the problem and plans to study it further. The study should focus on determining what information is available for putting together the FDI statistics, or whether the information is on the holding company, which is often established in the country to channel investment flows from that company to other resident enterprises, or the information is on these latter companies. In principle, it was suggested during the discussion that since the holding company receives the FDI, this is the only information that need be captured. However, for purposes of the sectorization of FDI, this would mean not being able to identify where the flows entering the country will go and everything would be classified as investment in the preponderant activity of the enterprise group. Consequently, and provided it is possible to obtain information from all the enterprises in the group, the conclusion was that the ideal would be to obtain data from both sources, the holding company and the other resident enterprises in the same group.

15. Ultimately, if the above is too ambitious a plan, an effort should be made to at least identify whether the contacted respondent is reporting data on the holding company or data on some of the enterprises in the group. Thus, in each case, the procedure should consider:

  • Holding company report: The response should specify which enterprises are covered by the data, so that the appropriate purges can be done to avoid duplications when information is received from some other enterprise in the group.

  • Report on enterprises other than the holding company: Identify whether there is a holding company in the country so as to attend to duplications and also to cover for omissions by other enterprises in the group that do not report.

16. As already mentioned, it should be noted that the collection instrument used to construct the FDI statistic includes asking for information on outward FDI. The response to the respective questionnaire by resident enterprises is very low and thus the data represent a very small part of the phenomenon. In particular, the figure reported to the IMF in the context of the CDIS for this type of investment amounts in each of the years reported (2009–2012) to six million dollars, five million of which correspond to a single enterprise.

17. There was discussion during the mission on the possibility of using mirror statistics [“partner economy statistics”] (i.e., information that other countries report to the CDIS as FDI received from El Salvador). Although this information does not show a slot by sectors, like what the BCRES shows on its website for FDI received, it would be useful to include it in the IIP and in the balance of payments, recording flows as the difference in balances. However, this information could not be used as input for the CDIS on outward FDI since El Salvador would be reporting in the survey the figure that the other country reported in its inward survey (i.e., it would be reproducing a figure already included in the CDIS system without providing any additional information).

18. The priority for the immediate future in terms of outward FDI should be to identify enterprises and, above all, to establish contacts to persuade them to respond to the survey. It should be noted that in many cases the largest national enterprises and thus those that would potentially make the largest investments abroad are routinely contacted and respond to some questionnaires related primarily to trade in services. However, they are still reluctant to provide information on asset-related subjects, which are those that preponderantly constitute relevant information for constructing FDI statistics.

19. At a later stage, once the FDI coverage has been effectively expanded, it will be advisable to add to Section A of the survey a request that relevant enterprises report the names of the enterprises in which they have equity ownership both in the country and abroad. In particular, the list of subsidiaries, affiliates, branches, and partners in the country would be useful for identifying resident enterprises with some relationship to enterprises abroad based on a single Salvadoran matrix. In this way, in the case of those enterprises that already respond to the survey covering local enterprises without foreign equity (Survey of Companies with External Debt), it would be possible to identify those that are tied to some enterprise abroad, so they would report in a separate section their positions and transactions with such enterprises and the corresponding entry would be made under the respective FDI items, and not as been happening so far, under the functional portfolio investment and/or other investment categories.

20. Thus, following conversations with technical DBP staff, it is clear that, although the questionnaire being used satisfies the requirements of the BPM6 and the CDIS and there are no apparent deficiencies in the respondents’ understanding of FDI concepts, it is the low response rate that affects the quality of the FDI information, a weakness that is particularly serious in the case of direct investment abroad.

21. To a large extent, the information availability problems noted are due to the fact that current legal provisions (Article 64 of the Organic Law of the Central Reserve Bank), on which the DBP bases its requests to the private sector for information, are insufficient because there are still no regulations governing the terms for the reporting of information and, in particular, establishing effective penalties for non-compliance. However, it should be noted that the mission was able to conclude that much progress has already been made on the draft regulations in question.

22. As already noted, although the response level is the largest problem encountered when preparing the FDI statistics, it is also important to pay attention to the identification of enterprises with FDI. On this subject, there was discussion of the possible usefulness of information available in the National Records Center regarding details on the ownership of resident enterprises in the country, particularly with respect to the partners’ nationality (as a proxy for residence). However, in response to previous inquiries the DBP has made in this regard, the National Records Center has indicated it does not have this type of information.

23. Consideration was also given to the possibility that, once the balance of international transactions information is systematically processed, it could provide data for expanding the directory of resident enterprises that receive FDI, but not so much for domestic enterprises that invest abroad since there are few such enterprises and they are well known. As already noted, in this latter case, the challenge is to overcome their reluctance to respond.

24. As examples of the response problems encountered when compiling FDI statistics, reference can be made to two of the country’s largest enterprise groups, the national airline TACA (now AVIANCA) and the Poma Group:

  • AVIANCA-TACA: The DBP has surveyed this company for years using a questionnaire that, like all the other surveys, includes sections for obtaining information on transactions abroad under the services items, as well as transactions and positions under the financial items. However, so far, the response has been limited to reporting on trade in services, despite the DBP’s insistence through visits to the company in recent years seeking to expand the response to cover financial and equity items as well. This means that a transaction as important as the acquisition of a significant portion of the company’s capital by the Colombian airline AVIANCA is not being included in the FDI statistics.

  • Poma Group: Some of the enterprises that make up the group respond, but a review of the sizeable amounts involved makes it clear that the information received does not cover all of the group’s operations abroad.

25. There was discussion during the mission regarding the Central American Monetary Council’s proposal that the countries of the region design a single questionnaire format that meets the requirements of the CDIS. In this regard, the view of the DBP’s specialists at the BCRES is that it is more practical to make some additions to the FDI questionnaire they have been working with for some time than to develop a new questionnaire. This is because the respondents are already accustomed to the current questionnaire but also because a completely new questionnaire would entail significant changes to existing IT systems already in operation. Although not in the case of El Salvador, this proposal may represent additional complications for countries with data collection instruments designed to cover a wide range of external accounts items and not just the specific FDI statistics needs. In any case, with the current questionnaire or a new one, the results must be adapted to the CDIS forms submitted to the IMF (i.e., the presentation of the data has a standard format for all countries), which guarantees comparability of figures, which is one of the aspects of greatest concern in the survey.

IV. Coordinated Portfolio Investment Survey

26. El Salvador does not currently participate in the CPIS and implementation of this survey was discussed during the mission. For this purpose, the information that the DBP currently has for putting together the asset side of the portfolio investment category in the external accounts was evaluated. Working meetings were also held during the mission with staff of the Central Securities Depository (CEDEVAL) and the Financial System Superintendency (SSF).

27. The information currently available comes from three different sources:

  • CEDEVAL: Foreign debt securities holdings of various sectors of the economy such as government, nonfinancial private enterprises, households, nonbank financial intermediaries, pension funds, and nonprofit institutions serving households. The data include the name of the issuer, the total amount of the issue, the annotated amount (value of securities held in custody by CEDEVAL), and issue and maturity dates. This information contains a code that does not correspond to the ISIN code, but rather to the CEDEVAL’s own classification.

  • SSF: Balance of assets at the end of each quarter based on foreign financial securities (debt securities and equity securities) recorded on the balance sheets that domestic banks submit to the SSF.

  • Surveys: Responses to the concepts “investments in debt securities abroad” and “investments in shares of foreign enterprises” contained in the various surveys that the DBP applies to non-financial private enterprises.

28. Review of the above information suggests that entries for portfolio investment assets are under-estimated in ESE. This is because the CEDEVAL data only include securities held in custody by that institution, which does not hold in custody all foreign securities belonging to residents as there is no obligation to do so, except in the case of pension fund administrators and insurance companies. CEDEVAL staff members themselves indicated that some resident investors may choose to have their securities held in custody outside the country because this is generally less expensive, although greater risks are often incurred. The presumed under-recording of portfolio investment assets is also the result of the general response problems already mentioned in previous paragraphs.

29. Further evidence that the available data have coverage constraints may be found in a comparison made between the value of the position under this item and the values of some countries that report data under the CPIS framework. Specifically, information from Costa Rica, one of the countries in the region that participates in the CPIS, was reviewed and indicated that Costa Rica’s figure is about six or seven times higher than El Salvador’s. The other country participating in the CPIS is Panama, where the financial sector has particular characteristics that make comparison difficult.

30. The visit to the CEDEVAL allowed the mission to establish that, as already noted, CEDEVAL’s information is partial because nationals are not required to place their foreign securities in its custody. According to CEDEVAL staff, there is no way to determine what percentage of foreign securities it holds in custody. Although the percentage of foreign securities held in CEDEVAL’s custody may be known at some particular point in time, this percentage may vary significantly later. However, to evaluate this information and determine whether it should be used in the absence of any other information, it was requested that the database the DBP receives quarterly be expanded to include:

  • Equity securities

  • Banking system information

  • ISIN code (for country assignment and reference for consulting international databases)

  • Original currency of the issue

31. The visit to the SSF indicated that current regulations for resident financial institutions cause this superintendency to have information with greater coverage than CEDEVAL’s information. The superintendency requires local brokerage firms to record all issues of foreign financial securities they seek to offer to resident investors. Thus, the SSF website provides public information regarding a wide range of foreign securities, including the issue’s original prospectus, as well as a summary with basic information in Spanish.

32. Brokerage houses must also provide data on all securities of this kind that they have purchased on behalf of their clients in the country. This means that, besides the accounting information that various intermediaries (banks, insurers, pension funds, investment funds, and others), as institutions regulated by the SSF, must submit regarding foreign securities transactions on their own behalf, the information available at the SSF covers a significant portion of the transactions carried out by the nonfinancial sector. However, it should be specified that there is no impediment that would prevent a resident investor from acquiring foreign securities through financial agents abroad. As a result, such operations, when carried out by a nonfinancial institution, are not included in the information available at the SSF.

33. In order to formalize the request for information from the SSF, the following form was designed in conjunction with DBP staff:

Table 1.

List of Financial Securities Issued by Nonresidents and Acquired by Residents

article image

The sectors to be classified are: central bank, banks (deposit corporations in general), other financial intermediaries (break down insurers, pension funds, investment funds, and other intermediaries), general government, nonfinancial sector (break down public enterprises, private enterprises, households, and nonprofit institutions serving households.

34. A point that could not be clarified during the mission with SSF staff was the issue of entries in the case of operations on the secondary market. Everything seems to indicate that if a local investor acquires a security registered by the SSF, whether in a primary issue or on the secondary market, the corresponding information would be known by the SSF and would thus be included in the form that it would eventually submit to the DBP. However, doubt remains as to whether the SSF is informed when a resident investor gets rid of an international security and that security passes to a nonresident (i.e., whether the SSF is advised of this operation so that the security in question is eliminated from the record).

V. Financial Derivatives

35. The assets recorded as part of the country’s international reserves include some derivative instruments consisting of futures and/or forwards to cover various risks. The information available shows data on realized gains and losses, changes in the margin deposit account, and commissions paid. Although the positions generated by these instruments are generally included in the international reserves balances that are reported in the balance of payments and the IIP, the specific entries indicating that these instruments are derivatives have not been made to date.

35. In recent months, the BCRES purchased a put option abroad to cover the value of its reserve gold holdings, anticipating a possible decline in the international price. Since this is the first time the BCRES has used this type of transaction and since balance of payments and IIP statistics have not recorded transactions and positions with financial derivatives to date, a hypothetical exercise with DBP staff was suggested to illustrate how the corresponding entry is made in both statistics.

37. From a broader perspective, it is recommended that the DBP conduct an investigation of the country’s largest enterprises, whether public or private, to determine which of them use financial derivatives to mitigate the risks inherent in their activities. In particular, it is likely that commodities export companies (coffee, sugar, etc.) cover their risks with instruments that are similar in nature to those the BCRES uses to protect the value of its gold holdings. The various questionnaires the DBP currently applies to (public and private) nonfinancial enterprises include a request for information on the use of financial derivatives. However, it does not receive any response to that request.

38. It is also conceivable that domestic banks and the government have coverage through swap, future, or forward contracts with nonresidents to deal with the risks associated with the variability of international interest rates or movements in the exchange rate for the dollar against other currencies in which some of their assets and liabilities may be denominated.

39. In the case of the banks, to date they have been known to engage in derivatives operations for insignificant amounts and, according to staff of the Financial System Office, they have done so with domestic counterparts. In any case, it will be important to bear in mind that a new chart of accounts has been defined that the banks will have to report to the SSF; this chart already includes positions for financial derivatives abroad and will be subject to mandatory reporting as of 2015. On this point and in anticipation of the likely increase in the use of financial derivatives in the near future, some cases illustrating the use of forward or futures type contracts were presented during the mission to show how these operations are recorded and the type of information that will be needed.

VI. External Debt

40. The data on the public sector’s external debt include some adjustments to take into account the residence criteria used for ESS.

41. General government debt securities issued on foreign markets must be adjusted to purge the portion thereof that is acquired by investors residing in El Salvador. For this purpose, there are two sources of information that may be useful. However, they intersect and thus, depending on whether both or just one of them is used, there is a risk of purging twice or omitting some purges. These sources are:

  • The BCRES Department of Financial and Fiscal Statistics: This information reports the amounts of debt instruments placed on international financial markets by the central government, with a breakdown of the amounts acquired by each investor. This can be used to determine which value should be excluded from the primary placement because it corresponds to resident investors.

  • CEDEVAL: This entity receives and holds in custody a portion of the securities in question, regardless of whether they were purchased during the primary placement or later on the secondary market. Thus, at least a portion of residents’ holdings can be identified.

42. Possible scenarios for excluding holdings of securities issued abroad and held by residents would be:

Table 2.

Purging Government Debt Securities Placed on International Markets

article image

43. Consequently, there is no way to do a complete purge or a purge without risk of duplication. To determine which of the different possible combinations of sources would lead to the most effective purge, it would be desirable to meet with the areas of the BCRES that are associated with the placement of these securities and that are familiar with the structure of the holdings and the investors’ most common purchase and custody methods.

VII. Insurance

44. The information on insurance that is recorded in the balance of payments is based on resident insurance companies’ responses to a specific questionnaire they are asked to complete by the DBP. The response to this questionnaire satisfactorily addresses statistical requirements in the case of flows associated with premiums and compensation paid, as well as information on premiums attributable to policyholders needed to make the counterpart entries in the primary income account. However, the data on technical reserves balances show extreme variability, as well as large-scale revisions of past data and some negative data, aspects which in combination raise doubts among compilers regarding the accuracy of the data and complicate the interpretation thereof. As a result, it is recommended that DBP staff approach the insurance companies to discuss the doubts created by the information received.

45. It is also important for the BCRES to continue insisting to the SSF that future expansions of the chart of accounts that the various financial intermediaries are required to use by the SSF must include a slot for information on insurance companies based on residence. Given that the required expansion is already planned for 2015 and that it does not include the referenced request, it is advisable to ensure that the expansions planned for 2016 take this point into account.

El Salvador: Technical Assistance Report-The Coordinated Direct Investment Survey and the Coordinated Portfolio Investment Survey (April 21-29, 2014)
Author: International Monetary Fund. Statistics Dept.