De-Internationalisation Patterns in Hungary

The debate on firm internationalisation has predominantly focused on different aspects connected to growth. However, the notion of de-internationalisation is not as popular, although it could also contribute significantly to our understanding of internationalisation. This paper focuses on de-internationalisation, its different modes and patterns followed by companies in the Hungarian context. Three hypotheses are tested: that de-internationalisation is a mass phenomenon, after de-internationalisation most companies are terminated, and de-internationalisation does not mean the end of international exposure. To test these hypotheses the Hungarian Corporate Tax Database was used with which the whole population of Hungarian companies in the years from 2009 to 2014 was analysed. The database consists 385,723 companies in 2009 and 422,500 companies in 2014, which is the whole Hungarian private sector. Among these companies 73,442 companies were registering export revenues, but this seems to be stable only for a smaller amount of companies. De-internationalisation is uncovered in this paper with different patterns followed by companies in the Hungarian context.


INTRODUCTION
There is a lively debate in international business literature about firm internationalisation which focuses on several aspects connected to growth.A distinct aspect of de-internationalisation, or the reverse of internationalisation, is not as popular, however, although it could also contribute significantly to our understanding about internationalisation.This paper aims to focus on deinternationalisations, its different modes and patterns followed by companies in the Hungarian context.The Hungarian Corporate Tax Database is used in the present study, whereby the whole population of Hungarian companies in the years between 2009 and 2014 was analysed.The database consists of 385,723 companies in 2009 and 422,500 companies in 2014, which is the whole Hungarian private sector.Among these companies significant amount is conducting international trade, but this seems to be stable only for a smaller amount of companies.De-internationalisation however seems to be a -so-far hiding -natural process.The aim of this paper is to uncover de-internationalisation and show its patterns followed by companies in the Hungarian context.

LITERATURE REVIEW
Although foreign market entry seems to be much more fancy (and therefore more researched) in the literature, but it has to be reconsidered in the light of foreign market exit knowledge as exporting longevity is far from ideal.Bonaccorsi (1992) examined Italian exporting companies in the 70's and the 80's and found that only a small portion of companies were stable exporters.In the seven-year period of 1978-1984 45.2% of exporting companies, namely 104,910 companies exported for only one year, and only 28.5% exported for at least four years in the given seven year period.
"One third of firms exporting in one year did not export the previous year.There is therefore a very high turnover of exporting companies" (Bonaccorsi, 1992: 617) As entering foreign market is frequent, exiting foreign market is also a mass phenomenon, which opens the floor for de-internationalisation research.Benito and Welch (1997) defined de-internationalisation in their seminal paper as: "any voluntary or forced action that reduce a company's engagement in or exposure to current crossborder activities.In extreme case, of course, a company may withdraw completely from international operationswhat may be termed full or complete deinternationalisation". (Benito and Welch, 1997: 9) Turcan ( 2011) focused on the different modes of deinternationalisation and created a typology as it can be seen in Figure 1.According to modes of de-internationalisation Turcan (2011) distinct total and partial withdrawal from export market.Total withdrawal can be because of ceasing trading (i.e.dissolving the company) or focusing on home market, whereas partial withdrawal can be because of optimising number of markets, reducing operations or changing entry mode.Benito and Welch (2007) are not focusing as much on termination, but they also list different forms of de-internationalisation, such as  "reduction of operations, in whatever form, in a given market or withdrawal from that market;  switching to operation modes that entail a lower level of commitments;  sell-off or closure of foreign sales, service or manufacturing subsidiaries;  reduction of ownership stake in a foreign

DATA AND METHODS
In this paper the de-internationalisation patterns of Hungarian companies are analysed, according to their business performance.The whole Hungarian private sector was analysed with the full Hungarian Tax Database from the Hungarian Tax Authority for six years (2009)(2010)(2011)(2012)(2013)(2014).This database include data from basic financial statements and some additional tax data for all companies following the double-entry bookkeeping system.The period taken for analysis is 6 years (from 2009 to 2014).
All companies which recorded exporting revenues in the period of analysis were included in the research.The population of the Hungarian private sector and the number of exporters can be seen in table 1.It has to be noted however that internationalisation is not only about exporting relationships, different methods (licensing, franchising, joint-venturing and strategic alliances) should have been taken into account as well as importing relationships.Financial data are, however, only available from export revenues, therefore in this level of explorative research only these data could have been considered.In a latter phase with more in-depth analysis other methods can be analysed as well.
According to Bernard et al. (2007), while 4% of the companies are exporters in the US, this figure is doubled in the Hungarian context, which can be explained by the openness and the integration of the Hungarian economy into the European Union.Companies are registered with an ID number in the database which can be tracked in each and every year of the database, therefore longitudinal analysis is conducted according to several aspects.

DISCUSSION OF KEY FINDINGS
The number of exporters in Hungary is steadily growing with around 4.37% CAGR in the period of analysis which strongly outperforms the growth of the number of companies (1.84% CAGR) which can be seen in Table 2.As in 2013 the government has made stricter regulations for founding new firms the negative growth in 2014 is understandable.Even more important that new registry and authorised share capital regulations hit small and/or non-functioning SMEs the hardest, therefore more exit is expected from 2014-2017.Numbers of exporting companies are rising, however, and the mentioned effects are not expected to take their toll on exporters.Source: Author's calculation Benito and Welch (1997) stated that in extreme cases companies can even withdraw from international operations.Based on the evidence from Hungarian exporters, after each and every year in the period of analysis around 30-33% of the exporting companies are ceasing their exporting operation, which is not including the partial withdrawal scenarios.
Only 33.6% (9,533) of the 28,336 exporting companies from 2009 were exporting throughout the six year period of analysis.From the exporters of 2010, 38.4% (11,571) were exporting in every year of the five year period.44.4% of the exporters from 2011 were able to export throughout their four year period, 54.8% of 2012's exporters were exporting in the three year period and only 69.4% of 2013's exporters were exporting in 2014 as well.
The fluctuation in Table 3 can be viewed from a different perspective as well.In 2009 28,336 companies were exporting, out of which only 19,163 companies were exporting in 2010, but an additional 10,954 companies were exporting as well in 2010.Most of these companies were newcomers to export, but some of them were already de-internationalised exporters coming back to export markets.
In 2014 there were 13,429 companies exporting from the 28,336 exporters of 2009, but only 9,533 were exporting in each and every year of the given period, which means at least 3,896 companies (13,75% of 2009's exporters) had a pause in export revenues in the six year period of analysis.In addition to that it can be seen in Table 4. that from 5 to 14 thousand companies had different pauses in export sales.
Bonaccorsi (1992) coined those companies stable exporters who were exporting for at least 4 years in the analysis period (which was 7 years by this argument).
It can be seen in table 4 that in the period of this analysis 73,442 companies were conducting export in Hungary.40% of them 29,453 companies were adventurer exporters only exporting for one year (although because of the period of analysis, it cannot be seen from the database if a company ceased export in 2010, but was an exporter in 2008 or earlier neither the number of second time exporters in 2015 could be forecasted).19.15% of companies were exporting for two years in the period of analysis, 11.95% for three years, 8.5% for four years, 7.32% for five years and 12.98% for six years.With the most amicable definition (Bonaccorsi's four year) stable exporters are only 28.8% of exporters, however I would rather stick to stricter definition of stable exporters and only use it for the 9,533 (12,98%) companies exporting in each and every year of the analysis.
Several authors like Bernard and Jensen (1999), Greenaway and Kneller (2007), Grazzi (2012) and Stocker (2014) made comparisons between exporting and nonexporting companies across different performance metrics.From Table 3 and 4 however it can be claimed that exporting companies has to be segmented, as there are huge, different segments of the exporting companies.Stable exporters could be much more different from adventurer exporters than the latter group from most of the non-exporters (not to mention that surely non-exporter groups can be also very different).
De-internationalisation is the reason behind the shrinking number of companies exporting in the long term.Table 5 shows the number of companies deinternationalised in the given years of the period of analysis.It is very interesting that the number of deinternationalisation is increasing year by year, although the number of foreign entry overcompensate the increasing number of de-internationalisation annually.According to these data the first hypothesis is supported as de-internationalisation -even in its strictest definition -is a mass phenomenon concerning more than 52% of the exporting companies in the period of analysis.
Hypothesis 2: After de-internationalisation most companies are terminated As foreign market entry is usually connected with the positive notion of growth, foreign market exit is usually connected with the negative notion of decline or defensive strategy.
According to Turcan's typology of deinternationalisation (which was shown in table 1) there are three valid segments of de-internationalisation, total withdrawal from international activities, and yet, in business, partial withdrawal from international activities and total withdrawal from international activities, and ceased trading at or right after (Turcan, 2011).
In table 6 the number of previously exporting companies terminated can be seen in the period of analysis.16.8% of the exporting companies (namely 12,337 company) were terminated in the period of analysis (however it cannot be decided from the database which companies were terminated in 2014, therefore this number is supposed to be even higher).Number of termination is a significantly high number, however it can be seen that deinternationalisation is not strictly connected with the terminus of the company.8 the number (and percentage) of companies can be seen who were terminated in the period of analysis from the de-internationalised companies with focus on the year of de-internationalisation and terminus.
In 2009 5,911 companies de-internationalised, which is 15.4% of the total number of de-internationalised companies.1,582 companies out of the 5,911 deinternationalised companies were terminated in 2009 whilst 585 companies terminated in 2010, 452 companies in 2011, 321 companies in 2012 and 250 companies in 2013.Altogether 3,190 companies, 54% of those deinternationalised in 2009 were terminated in the period of analysis and 2,721, 46% were still existing in 2014.
In 2010 6,329 companies de-internationalised, which is 16.5% of the total number of de-internationalised companies.1,411 companies out of 6,329 went out of business in the year of their de-internationalisation, 687 companies in 2011, 419 in 2012 and 330 in 2013.Altogether 2,847 companies, 45% of those deinternationalised in 2010 went out of business in the period of analysis, whilst 3,482 (55%) were still in operation in 2014.

Table 7 De-internationalisation in 2009 and 2010 and terminus of de-internationalised companies
Source: Author's calculation In 2011 7,660 companies de-internationalised, which is 20% of the total number of de-internationalised companies.1,943 companies out of 7,660 went out of business in the year of their de-internationalisation, 606 in 2012 and 412 in 2013.Altogether 2,961 companies, 38.7% of those de-internationalised in 2011 went out of business in the period of analysis, whilst 4,699 (61.3%) were still in operation in 2014.
In 2012 7,949 companies de-internationalised their activities, which is 20.7% of the total number of deinternationalised companies.1,300 companies out of 7,949 went out of business in the year of their deinternationalisation and 627 in 2013.Altogether 1,927 companies, 24.2% of those de-internationalised in 2012 went out of business in the period of analysis, whilst 6,022 (75.8%) were still in operation in 2014.

Table 8 De-internationalisation in 2011 to 2013 and terminus of de-internationalised companies
Source: Author's calculation Interestingly in 2013 were the largest number of deinternationalised companies in the period of analysis with 10,498 companies, registering 27,4% of all deinternationalisation.1,412 companies, 13.5% of them went out of business in the year of their de-internationalisation and the remaining 86.5%, 9,086 were operating in 2014.
From these data the total de-internationalisation part of Turcan's de-internationalisation typology can be calculated, as it can be seen in table 9 Although with total de-internationalisation tens of thousands of companies were losing their export markets, the vast majority of them survived total deinternationalisation, therefore hypothesis 2 is rejected.
Hypothesis 3: De-internationalisation does not mean the end of overall international exposure.
As most of the companies are surviving the exit from their export market it is interesting to examine with which pattern do they export and de-internationalize.Exporting and de-internationalisation patterns of stable exporters can be seen in table 10 and table 11.The most stable exporters were exporting in each and every year, therefore there is no de-internationalisation pattern for them.
In table 10 the different patterns (and number of companies following these patterns) can be seen of companies who are exporting for 5 years out of the six years of analysis period.Pattern one is obvious as these companies were entering foreign market in 2010 and are exporting from that date.This pattern is followed by 2,038 companies which is 38% of the whole group of companies exporting for 5 years in the period of analysis.Pattern six is also obvious, which shows that after five years of exporting these companies are de-internationalising with ceasing export sales, this pattern is followed by 1,235 companies (23% of the group).Pattern two, three, four and five are much more interesting as those companies following these patterns had one year pause in their export sales.These companies were ceasing their export sales only for one year but after that they were re-entering foreign market accordingly.Altogether 2,100 companies are following these patterns (39% of the group) which means these companies could become rich soil for deinternationalisation research, according to their situation, capabilities, business performance and re-entry to foreign market.

Source: Author's calculation
In table 11 those patterns can be seen which are followed by the companies exporting in 4 years in the period of analysis.Pattern one is the most obvious, those companies are following this pattern who entered foreign market in 2011 and are stable exporters from that point.The 2,071 companies are 33.2% of the whole group.
Pattern fifteen is the most obvious de-internationalisation pattern as these companies were ceasing their export operation in 2012 and were not re-entering foreign market the year after.1,145 companies were following this pattern which is 18.3% of the whole group.

Table 11 Exporting and De-internationalising patterns of stable exporters (exporting in 4 years out of 6)
Source: Author's calculation Pattern five is a sad pattern as these companies entered foreign market in 2010 but were ceasing exporting in 2013, hopefully most of them will be registering export sales later.This pattern is followed by 540 companies, which is 8.6% of the whole group.
The remaining 39.8% (2,488 companies) are much more interesting from de-internationalisation point of view, as one or two years after their de-internationalisation they re-entered the foreign market.Together with the reinternationalising companies of the 5 year exporter group, there were 4,588 companies out of the "stable" exporters who were re-entering foreign market after deinternationalisation.Exporting in year …

Patterns
In table 12 the exporting and de-internationalisation patterns of those companies can be seen which are exporting in 3 years in the period of analysis.Pattern one, four, seven and twenty are obvious patterns again, pattern one shows new exporters which is followed by 2,638 companies (30% of the group).Pattern twenty shows those companies exporting in the early period but completely deinternationalising in 2011, this pattern is followed by 1,768 companies (20% of the group).Pattern four (744 companies, 8,5%) and seven (627 companies, 7,1%) is about companies entering foreign market in the period of analysis but de-internationalising in 2013 and 2012 respectively.All other patterns which is followed by a combined of 3,001 companies (34,2% of the group) are very interesting as they shows patterns of companies entering end exiting foreign markets in any conceivable way.Pattern eight and fourteen seem to be the most interesting patterns as companies following these patterns were opportunistic exporters, exporting and deinternationalising in even and odd years (vice versa for pattern fourteen).
In table 13 the patterns of those companies can be seen which were exporting only for two years in the period of analysis.
Table 13 Exporting and de-internationalising patterns of companies exporting in 2 years out of 6 Source: Author's calculation Pattern one and fifteen are the most obvious patterns in this group.Pattern one is followed by 2,448 (17,4%) companies showing those which were completely deinternationalising in 2010.Pattern fifteen is followed by 3,944 (28,1%) companies which were entering foreign market in 2013.All other patterns are showing different types of foreign market entry and exit and are followed by a combined of 7,669 companies which is 54,5% of this group.Pattern five can be the most interesting here, the 221 company in this group were exporting in 2009 then had four year brake in exporting and re-entered export market in 2014.

Table 14 Exporting and De-internationalising patterns of companies exporting in a single year in the period of analysis
Source: Author's calculation In table 14 the exporting and de-internationalisation patterns of those companies can be seen which were only exporting in one year in the period of analysis.Altogether 29,453 companies were exporting for only one year which is 40% of the 73,442 companies conducting export in the period of analysis.It can be supposed however that some Exporting in year … Nr. of Companies of the 5,911 companies who were last registering export sales in 2009 were exporting in 2008 or before and it can be hoped as well that the majority of the 7,516 companies who were entering foreign market in 2014 will export in 2015 and further, as well.
According to the Hungarian exporting and deinternationalising companies hypothesis 3 is supported, as de-internationalisation not only does not mean the end of international exposure, but by thousands of companies it is followed by re-entry to foreign market.

CONCLUSION
Based on the 73,442 Hungarian companies which were registering export revenues in at least one year from 2009 to 2014, it can be observed that de-internationalisationeven in its strictest definition -is a mass phenomenon concerning more than 52% of the exporting companies in the period of analysis.
Although de-internationalisation is often connected with the negative notion of decline or even termination of business, the vast majority of the tens of thousands of companies getting through total de-internationalisation with the loss of their export markets survived total deinternationalisation.
Thousands of companies not only survived deinternationalisation, but were re-entering foreign markets with more or less success.
Based on the several different exporting and deinternationalisation patterns and the significant number of companies following these patterns, it is important to use a better segmentation in the literature than the popular exporter vs. non-exporter distinction, since stable exporters, opportunistic exporters and adventurer exporters are very different groups of the exporting companies.
From the de-internationalisation patterns, policy makers should derive the conclusion that it is not enough to support entry to foreign markets, but those capability building processes has to be supported, which will enable the companies to compete in the international market for the long run, and if for some external reasons they have to abandon some of their markets they can utilize these capabilities and even the earned experience to enter different markets or even re-enter the given market with stronger proposals.

Table 1
Population of Hungarian firms in the private sector and number of Exporters

Table 4
Number of companies exporting by years of conducting export .