Abstract
The mass privatization (MP) program in Bulgaria was implemented in 1996–1997. Following programs in countries like the Czech Republic, more sophisticated regulatory bodies were put into place to prevent the kind of abuses observed elsewhere. This study finds that Bulgaria avoided some of the extreme problems that manifested themselves in these other countries, but there were still serious problems of dilution of shareholder value. We find that dilution in Bulgaria is similar in both MP firms and non-MP firms. After a number of years have passed, MP firms have performed better than firms privatized by other means or firms that were never privatized.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See Miller and Petranov (2000) for an analysis of the early outcomes of the MP program.
Denisova et al. (2010) evaluate attitudes towards privatization and the market system in transition economies. In their Table 1 they record responses across countries to the question: ‘In your opinion, what should be done with most privatized companies? They should be …’. In Bulgaria 7.2% of respondents checked ‘Left in the hands of current owners with no change’ (across the entire sample the response was 19.4%). In Bulgaria 48.3% checked ‘Left in the hands of current owners provided that they pay privatized assets’ worth’ (across the entire sample the response was 34.8%). This suggests that there was support for the market system in Bulgaria but not the privatization process. The MP process was only part of the privatization process, but it was an important part.
In fact the program was underway when some of the worst problems of the Czech program were becoming clearer.
See Petranov and Miller (1999) for an analysis of the problems in the development of the stock exchange. This monograph also contains an analysis of the early securities’ laws and how they were implemented in the early years of the stock market. Atannassov et al. (2006) provides an updated review and shows how later changes in the law were designed to prevent further dilution of the interests of the original voucher shareholders. While there are other means of diluting shareholder wealth, it is clear from these legal changes that dilution was a major problem.
The 18 largest privatization funds controlled 60% of the vouchers in the auctions.
Table 1 illustrates the difficulty of doing these cross country studies. The European Bank for Reconstruction and Development data used in these studies list Bulgaria as a country where the principle privatization method was sales of enterprises starting in 1993. As can be seen in the table, significant privatization did not really take place in Bulgaria until the MP occurred in 1997. Following the MP program there was significant privatization through sales to outside owners and management-employment buyouts. This record was not that unusual. Transition countries often used various methods of privatization.
The number of MP firms with negative net worth from 1996–2001 was: 29, 17, 32, 49, 69 and 94, respectively.
These changes included the issuing of warrants to existing shareholders, which had to be purchased when new shares are issued. This meant that shareholders could no longer be passive observers during new share issues. For a more complete description of these changes see Atannassov et al. (2006).
An example of a compound dilution is provided in Appendix (A).
We assume that whenever total equity per share falls, we assume this is 100% dilution, that is the largest shareholder buys it all. Up to 2002, it was common practice, because of the flaw in the Bulgarian corporate governance law, to hold stockholder meetings in some remote place that was not easily accessible to the interested shareholders so that the biggest shareholder could initiate and buy the entire new issue. There were also no requirements for warrant issues, or any minimum pricing rules or minimum minority shareholder approval rules to seriously prohibit such schemes. In a few cases some interested current shareholders would show up, but they were quickly dissuaded from complaining by thick-neck security guards. Atannassov et al. (2006) using the actual number of shares for many firms of our sample confirms this finding that current shareholders were very far from acquiring even their break-even (initial) stake in the firm and so it is a reasonable assumption that dilution was close to 100% whenever total equity per share fell.
According to the Bulgarian National Accounting Standards, the liabilities side of the balance sheet has Total Equity and Total Debt. Total Equity consists of Shareholder Equity, Profit/Loss for the current year, Profit/Loss for the previous year (which we accounted for our dilution calculations), and Reserves. ‘Shareholder Equity’ is the item on the balance sheet that we use as a proxy for the number of shares for the vast majority of firms (855) in our sample have a share par value of one Bulgarian lev (BGN). The other cases of non-unitary par value are adjusted by the par value of shares.
Atannassov et al. (2006) argues that the legal changes in 2002 reduced the amount of dilution. (They calculate dilution differently by evaluating the share held by the majority shareholder before and after the issuance of new shares.) We can neither confirm nor deny this since our sample does not extend far enough beyond the implementation of these changes.
Share buybacks and splits were not a common practice during that period. Only a few firms in the entire period bought back any shares and only a few did share splits. We adjusted our equity measures to account for these events before turning to the calculation of dilution.
There are too few listed non-MP firms to make statistical comparisons between listed MP and non-listed MP firms.
Because the number of listed non-MP firms and the number of unlisted holding companies is so small, reasonable statistical comparisons between listed and unlisted cannot be made for these categories.
At the time of the auctions some people were said to have purchased Multigroup shares because they thought it would be good to have the mafia on their side. They were clearly mistaken in this view!
For all firms and owners, we have snapshots at the end of 1997, 1999, 2001 and 2003. For 1996 we assume the ownership data are all zeros. For years 1998, 2000 and 2002 we assume that ownership has not changed from the previous year and have used the previous year. Although variations in ownership from one year to the next are small, they are sufficient to identify the coefficient in ownership. While it might be acceptable from an econometric perspective, using means for the missing years of the two adjacent years this would violate the discreteness of the occurrence of the process of purchasing shares by which the ownership changes. Changing the analysis to use every 2 years (eg 1997–1998, 1999–2000, etc) could lead to a loss of much of the intrapanel variation and significantly detract from the reliability of the analysis.
In this regression and in the one below on the impact of MP we do not interact P and D as this question of the dilution level of privatized versus non-privatized firms is already investigated at length, including a breakdown according to listing status by Miller (2006).
See Jakubson (1991).
See, for example, Salinger's (1984) widely cited article where he finds that the coefficient on the minimum efficient scale variable is not statistically different from zero.
Results show we do not reject the null of validity at the 5% level for all three equations.
Even if the exogeneity of the suspect Size instrument failed, which it did not, a Hansen (Sargan) C-statistic, with p-values of 0.10, 0.94 and 0.78 for the three equations, is still safely within the non-rejection range where the comparative p-values are 0.22, 0.19 and 0.22, respectively.
While it might be argued that investment funds would obtain large holdings in only the best firms, the economic crisis situation that existed at the time of the MP was so extreme that evaluating the situation of all but a few firms was extremely difficult.
See Jakubson (1991).
Full specification of the econometric model is summarized in Appendix (B).
References
Atanassov, B . 2003: Privatization and corporate governance in a transition economy: The case of Bulgaria. PhD dissertation, University of Delaware, December 2003.
Atannassov, V, Black, B, Ciccotello, C and Gyoshev, S . 2006: How does law affect finance? An examination of financial tunneling in an emerging market. September, 2006 (College of William and Mary, unpublished).
Barnow, BS, Glen, GC and Arthur, SG . 1981: Issues in the analysis of selectivity bias. In: Stromsdorfer, E and Farkas, G (eds). Evaluation Studies Review Annual, Vol. 5. Sage Publications: Beverly Hills, CA, pp. 43–59.
Bennett, J, Estrin, S and Urga, G . 2007: Methods of privatization and economic growth in transition economicies. Economics of Transition 15 (4): 661–683.
Cameron, AC and Trivedi, PK . 2005: Microeconometrics. Cambridge University Press: New York, NY, pp. 879–881, 883–884.
Denisova, I, Markus, E, Timothy, F and Ekaterina, Z . 2010: Everyone hates privatization, but why? Survey evidence from 28 post-communist countries. Centre for Economic and Financial Research at New Economic School, Working Paper No. 143, May 2010.
Djankov, S and Murrell, P . 2002: Enterprise restructuring in transition: A quantitative survey. Journal of Economic Literature XL (September): 739–792.
Estrin, S, Jan, H, Evzen, K and Jan, S . 2009: The effects of privatization and ownership in transition economies. Journal of Economic Literature 47 (3): 699–728.
Frydman, R, Gray, C, Hessel, M and Rapaczynski, A . 1999: When does privatization work? The impact of private ownership on corporate performance in the transition economies. The Quarterly Journal of Economics 114 (4): 1162–1163, 1182–1184.
Gouret, F . 2007: Privatization and output behavior during the transition: Methods matter. Journal of Comparative Economics 35: 3–34.
Hansen, L . 1982: Large sample properties of generalized method of moments estimators. Econometrica 50 (3): 1029–1054.
Heckman, J . 1979: Sample selection bias as a specification error. Econometrica 47: 153–161.
Hsiao, C . 2003: Analysis of panel data. In: Paris, S (ed). Econometric Society Monographs. Cambridge University Press: Cambridge, UK, p. 85.
Jakubson, G . 1991: Estimation and testing of the union wage effect using panel data. The Review of Economic Studies 58 (5): 976–976.
Megginson, W . 2005: The financial economics of privatization. Oxford University Press: New York.
Miller, J . 2006: Evolution of mass privatization in Bulgaria. Working Paper 2006–2010, University of Delaware.
Miller, J and Petranov, S . 2000: The first wave of mass privatization in Bulgaria and its immediate aftermath. Economics of Transition 8 (1): 225–250.
Petranov, S and Miller, J . 1999: Bulgaria's capital markets in the context of EU accession: A status report. Center for the Study of Democracy: Sofia, Bulgaria.
Salinger, MA . 1984: Tobin's q, unionization, and the concentration-profits relationship. The Rand Journal of Economics 15 (2): 159–170.
Sargan, J . 1958: The estimation of economic relationships using instrumental variables. Econometrica 26 (3): 393–415.
Stock, J and Yogo, M . 2005: Testing for weak instruments in linear IV regression. Ch. 5, In: Stock, JH and Andrews, DWK (eds). Identification and Inference for Econometric Models: Essays in Honor of Thomas J. Rothenberg. Table 2. Originally published as NBER Technical Working Paper No. 284. Cambridge University Press: Cambridge, UK.
Van der Klaauw, W . 2003: Estimating the effect of financial aid offers on college enrollment: A regression-discontinuity approach. International Economic Review 43: 1249–1287.
Wooldridge, JM . 2002: Econometric analysis of cross section and panel data, Ch. 10. MIT Press: Cambridge, MA.
Acknowledgements
This project was funded in part through a grant provided by the United States Department of State's Program for Study of Eastern Europe and the Independent States of the Former Soviet Union (Title VIII) and administered by the William Davidson Institute. The opinions, findings and conclusions or recommendations expressed herein are those of the authors and do not necessarily reflect those of the Department of State or the William Davidson Institute. We thank two anonymous referees, Evangelos Falaris for his many hours of technical support, and Saul Hoffman. Stefan Petranov, Anastasia Miteva and Lubomir Evstatiev were helpful in framing the project and collecting data.
Author information
Authors and Affiliations
Appendices
APPENDIX A
An example of how compound dilution is determined for the entire period
Firm 15 in our sample had its ‘shareholder equity’ item on the balance sheet (which we use as a proxy for the number of shares) changed five times in the entire 1996–2003 period. The first time was at the end of 1998 but then the total equity per share increased rather than decreased, so no dilution was recorded.
The second time that the firm increased the number of shares was in 1999. At this point, total equity per share dropped from 42.6 to 11.5, that is dilution occurred, which reduced the value of shares to only 27% of their former worth. New shares were issued a third time in 2001. Total equity per share was 18.33 before the new issue and 8.87 after, at which time the shares were worth less than 50% of their former value. Then next year further dilution caused the value to 7.2, that is another drop of 19%. The fifth issuance of new shares did not cause further dilution. In total, the investors who did not manage to buy new shares had the value of their original fall, and suffered a cumulative loss of 89.35%. Thus, investors in Firm 15 were holding at the end of year 2003 only 10% of their original investment.
All these calculations rest on the assumption that practically all shareholders (except the few who initiated the issuance of new shares) were prevented from buying the new shares. This was a widespread practice in Bulgaria. General Assemblies of shareholders were held in remote places where not many current shareholders could exercise their rights, or when they showed up they were prevented from entering by the ‘security’ guards).
APPENDIX B
The main econometric model estimated is:
Step I
-
1
I it =Y it−2′φ 1+Y it−3′φ 2+(d t P i )′ϕ 1+D it ′ψ 1+ξ it , Fixed Effects to obtain fitted Y it−1;
-
2
D it =S it ′θ +Y it−1′φ 3+(d t P i )′ϕ 2+ϑ it , Fixed Effects to obtain fitted D it ;
-
3
P itk =μ i1+B itk ′κ 1k +Y it−1′φ 4+D it ′ψ 2+ζ it , (7 Probits), k=1997–2003 to obtain fitted P it ;
-
4
P it =μ i2+B itk ′κ 2k +Y it−1′φ 5+D it ′ψ 3+υ it , (Probit), k=1997–2003 to obtain λ it =φ (ω′η̂)/Φ (ω′η̂).
Step II
Y it =α i +Y it−1′β+P itk ′γ k +D it ′δ+λ it ′π+ɛ it , Fixed Effects
where υ
it
, ɛ
it
∼(0, 0, σ
υ
=1, σ
ɛ
, ρ); π=ρσ
ɛ
, I
it
=Y
it−1−Y
it−2; S
it
–Shareholder equity; B
itk
–Very big and very unprofitable firms in 1996 only; d
t
–dummies equal to 1 only if t is the respective year; ω=(1, B
itk
, Y
it
−1, D
it
); η=(μ
i2, κ
2k
, φ
5, ψ
3); 
; Φ (ω′η)=∫−∞
ω′η
φ(t)dt.
Rights and permissions
About this article
Cite this article
Miller, J., Lazarov, V. Evaluation of Mass Privatization in Bulgaria. Comp Econ Stud 53, 621–646 (2011). https://doi.org/10.1057/ces.2011.23
Published:
Issue Date:
DOI: https://doi.org/10.1057/ces.2011.23