Printer Friendly

The structure of privatization plans.

In recent years, a vast transfer of state-owned assets to the private sector has occurred in many countries, irrespective of their level of development or the political affiliation of their government. Privatization is believed to improve economic incentives; attract managerial and technological resources from the private sector; broaden share ownership; and reduce public sector borrowing. In fact, privatization can be interpreted as an alternative form of public financing, a sort of "equity financing" to reduce the overhang of public debt, as the budget gains from the higher value of the firm under private ownership. The historical evidence, in fact, indicates that state-owned enterprises have systematically drained public resources rather than contributed earnings, particularly in the Third World (World Development |18~).

We document strong regularities in privatization programs across several countries. The data indicate a predominance of partial, staggered sales. Even though transfer of control typically takes place rapidly, governments tend to retain a significant stake for long intervals of time.

This paper also examines the traditional argument for gradual sales, namely limited market capacity, with a confidence-building rationale.

Even after privatization restrains government interference, a firm is still exposed to the risk of adverse policy changes, particularly when it operates in a monopolistic market. A selling government will face investors' diffidence about its policy intentions after the sale; it may therefore structure the sale as to build policy credibility and maximize proceeds. To enhance investors' confidence, a selling government may signal commitment to current policy by retaining a stake in the firm for some time (while transferring managerial control), thus showing willingness to bear some financial costs of policy changes. As time passes without a policy reversal, confidence and thus sale prices improve. In addition, early sales may be deliberately underpriced in order to convince the market to absorb larger sales, which reduce the risk borne by the government and therefore enhance policy risk.

While the two explanations have similar empirical implications, our informal evaluation of the evidence appears more favorable to the reputation-building hypothesis than to the notion of temporary market capacity constraints. Presumably, the risk and the extent of policy change is largest for monopolistic industries or firms in protected markets, while firms operating in competitive markets are less subject to the threat of quasi rent appropriation. We find evidence that firms in such policy-sensitive sectors tend to be privatized with smaller initial sales and larger underpricing, and possibly requiring a longer time horizon before the share retained by the government can be sold. Complete sales are, on average, associated with manufacturing firms in competitive markets, while sales of utilities with potential monopolistic rents are more distributed over time.

We document that retained stakes are explicitly meant to be sold gradually over a few years. Often, stakes in several firms are sold simultaneously, creating considerable government risk-sharing across industries. The profile of privatization proceeds increases over time, suggesting gradual selling calibrated to build investors' confidence. As policy credibility increases, larger initial sales become more frequent.

We also document extensive underpricing, which is, on average, greater in privatization sales than in initial public offerings (IPOs) of private firms. Underpricing appears to be largest for firms with large taxable rents, such as utilities. This is consistent with a signalling argument, since these firms are exposed to greater policy risk, and inconsistent with an asymmetric information explanation over asset values, since these firms tend to be large and well known relative to private IPOs.

Section I interprets privatization as an enhanced commitment to reduced policy interference which reinforces incentives. After a sale, a sovereign government often retains considerable ability to affect the value of a firm sold. This has implications for the structure of privatization sales: specifically, policy risk can be resolved only through a maintained policy over time; and therefore, gradual sales may maximize sale proceeds. Section II presents data from several large privatization plans, and offers an informal evaluation of the empirical support for the reputation building hypothesis.

I. A Rationale for Privatization

There is evidence that state-owned firms are less profitable than comparable private firms (e.g., Boardman and Vining |3~, Megginson, et al |13~, and Galal, et al |5~). Evidence of lower profitability does not prove by itself that public ownership is undesirable, since public firms may be pursuing worthy purposes other than profit maximization. For instance, higher profits in the private sector may derive from the exercise of market power.(1) However, it is not clear why public policy needs to be pursued through state ownership rather than by arm's-length regulation, particularly if private ownership is desirable on the ground of efficiency. Even the argument that state ownership might overcome information asymmetries relevant for regulatory purposes can at most justify partial ownership or board membership, and certainly not state control.(2) In fact, the British experience indicates that the process of public regulation actually improved because of privatization, in becoming more openly scrutinized.

Two related questions are just as important: Why should private ownership be necessary for improved efficiency? Why can't a government minimize costs while pursuing its policy?

The property rights school (Alchian |1~), interprets public ownership as nontransferable common ownership, which reduces incentives by not allowing firm stakeholders to receive the capitalization of future earnings through the sale of property rights. This view does not explain why agents could not be given transferrable income rights distinct from property rights.

Stiglitz and Sappington |16~ argue that public control reduces the cost of intervention to the policymaker; however, it also limits its ability to commit. Their analysis, however, leaves unclear the nature of costs of intervention and limited commitment. In Perotti |14~, a rationale for privatization is presented, based on the residual nature of property rights. The idea is that under public ownership the government retains unconditional control over the use of firm assets. This discretionary power is very costly because it encourages rent-seeking behavior by firm insiders, who represent a more coordinated interest group than dispersed taxpayers. Even if a government may, in principle, prefer to minimize costs, it is vulnerable to political pressure to maintain established rents (such as high wages/low effort, high and secure employment, favoritism to domestic suppliers, etc.); this leads to loss of incentives. In contrast, a private owner has both an incentive and the ability to commit contractually to reward efficient behavior.

In this view, privatization establishes a firmer commitment by the state because property rights are constitutionally protected against direct state interference. However, the transfer of property rights does not eliminate arm's-length policy risk. Control over implementation of legislation still allows the government discretion to redistribute part of firm value, through new regulation, taxation, deregulation of entry, etc. A sovereign government cannot commit to maintain its current policy in the future, even when it is, in part, enshrined in detailed legislation or semi-independent legislation. Often, only a policy maintained over time will gradually eliminate the perception of political risk.(3)

This raises the issue of which instruments the government has to signal commitment. In private initial public offerings (IPOs), there is also an adverse selection problem faced by investors, since the seller presumably has better information on the value of assets. The literature on IPOs has demonstrated that partial sales may be employed as signalling devices (Leland and Pyle |10~). More recently, several authors (Allen and Faulhaber |2~, Grinblatt and Hwang |6~, and Welch |18~) have shown that underpricing may also be interpreted as a signal of high value of assets.

One difficulty with a direct application of these models to a privatization sale is the assumption that a government knows more about asset values than the private sector, which seems implausible. Vickers and Yarrow |19~ argue the opposite is likely to be true. Another hypothesis is that in the case of privatization sales, the information asymmetry over firm value concerns the preferred choice of policy of the selling government. Perotti |14~ extends the partial sale/underpricing model to the case of privatization, and shows that a committed government can improve its reputation with investors by transferring control immediately to the private sector, while initially selling only a fraction of the shares and retaining the remainder for a certain time period. As in the literature, the underpricing of the stock sale may also contribute to signal greater commitment.

The intuition gained by the model is that a government with no intention to interfere is inherently more willing to retain a (noncontrolling) stake in the firm for some time period, since it knows that it will sell it at a high price in the future once its credibility has grown. Conversely, a government which expects to change its current policy prefers a rapid sale, since it expects reduced profits from the policy change and a lower market value for the firm. Similarly, underpricing may signal commitment because an uncommitted government cannot expect higher proceeds from a subsequent sale, and is therefore not willing to underprice the initial sale.

In conclusion, investors' valuation for the firm will be lower when the sale asks them to absorb a lot of policy risk in early stages. More generally, in a large privatization plan firms will be sold gradually, according to a timetable. Over time, in the absence of changes in policy the subsequent sales will fetch higher prices.(4)

An alternative and popular rationale for gradual sales is that capital markets are capacity constrained in the short-term. This view implies a temporarily inelastic demand for stocks. A very large stock sale in a small, segmented market may require investors to invest a large fraction of their wealth in the issue, and therefore demand a large risk premium. Then, a rapid sale program would swamp the capital market, temporarily depressing share prices and reducing sale proceeds. The speed of privatization is then constrained by the gradual increase in financial wealth available for equity investment.

The two views are difficult to distinguish empirically, because the reputation-building hypothesis also suggests that the sale price is lower for larger sales, reflecting a rational inference by investors on policy intentions. We discuss the evidence in Section II.

A. The Separation of Income and Control Rights

Another rationale for underpricing arises once we realize that since stock provides voting rights, it is necessary to sell a minimum stake to offer to the private sector a reliable controlling majority. In fact, the initial sale may need to be larger than 50% if there are coordination problems among private investors. But a larger sale has the effect of shifting more risk towards the private sector, which may not be interpreted favorably. In these circumstances, underpricing of early sales allows the state to sell more of the stock while still signalling commitment. This is also important when private incentives to invest in the firm depend, in addition to credibility, also on the fraction of residual profit they receive.

A reliable transfer of control may also be achieved simultaneously to signalling through an unbundling of the voting and residual income rights contained in corporate stock. In other words, the government could retain a significant amount of nonvoting stock while selling on the market all voting rights: this would signal willingness to bear risk while ruling out any intent to run the firm. There have been, in fact, a few examples of such arrangements, such as in the sale of Telmex (Telefonos de Mexico) and in bank privatization in Jamaica (Leeds, 1986). However, unbundling income and control rights may be impossible, perhaps because income rights alone may not be as unambiguously protected as property rights. Alternatively, the stock's voting rights may be suspended until the time of a final sale to the public.(5)

The implications of the reputation-building conjecture for the time series of sales can be summarized in a few predictions. In general, the time profile of privatization proceeds should increase over time, since demand will tend to increase with the degree of confidence in public policy. Particularly at the beginning of the program, individual firms should be sold in stages; in a privatization plan involving the sale of several firms, the government should distribute their sales over time. Government statements on the retained stakes should indicate a complete sale within a certain time interval. As reputation for commitment increases, larger initial share offerings ought to become more frequent, while discounts decrease.(6)

In general, the probability of a policy change, as well as its redistributive effect, are greater for monopolistic industries operating in a rent-earning, protected market. In contrast, firms operating in a competitive market are subject to a lower threat of appropriation. The model predicts that firms in policy-sensitive sectors (monopolies, protected or subsidized manufacturers, etc.) will tend to be privatized with smaller initial sales and larger underpricing, and possibly a longer time horizon for the share retained by the government.

The next section presents transaction data from several privatization plans, and examines whether the evidence supports the reputation-building hypothesis.

II. Empirical Evidence and Discussion

We present here suggestive evidence concerning several privatization programs in both developed and developing countries. Although the data come from only a few countries, they represent some of the most extensive and successful programs.
Exhibit 1. Privatization in the UK

Financial Net Proceeds
Year (Million |pound~s)

1979 290
1981 373
1982 611
1983 862
1984 4655
1985 1602
1986 6963
1987 3541
1988 2500
1989 5239
1990 5181
1991 5034

Source: Vickers and Yarrow |19~ and the London Stock Exchange.

The first four exhibits refer to privatization programs from economically developed countries, namely France, the United Kingdom and Spain. A common characteristic to all these cases is the presence of a well-developed domestic capital market. We can therefore evaluate the argument for gradual sales due to a limited capacity of the capital market to absorb the whole amount. The evidence suggests, however, that the structure of privatization sales does not differ across developed and underdeveloped markets. It seems also that many countries which initially had small equity markets (e.g., Mexico, Jamaica and Turkey) have succeeded in promoting their growth through progressively larger privatization sales and a stable policy.

Exhibits 1 and 2 document the history of public sales and proceeds in the UK program. There is a clear progression of sale volumes, which indicates an allocation of risk-bearing over time consistent with a model of reputation building. Exhibit 2 offers more detail on sales to the stock market. Note the tendency, particularly early in the program, to only partly privatize individual companies and retain large stakes in them for a few years. Some early partial sales are for 51% of the stock, suggesting a symbolic transfer of a majority of voting rights. Complete sales seem to become common only at the end of the privatization program, presumably when the determination of the government to allow free rein to market forces had been established.

Data on French and Spanish transactions, portrayed in Exhibits 3 and 4, also suggest that partial sales are indeed common. Unfortunately, the French sample is quite short because the program was interrupted after a change in government.



Exhibits 2, 3 and 4 offer evidence on the remarkable extent of underpricing in these privatization programs. Although there are a few instances of undersubscribed issues in the UK, the data suggest a remarkable and presumably deliberate choice of low offering prices. Note that in almost all cases where the price did not rise to a premium (excluding the sale of British Petroleum, whose price was set just prior to Black Monday for an offering on October 30, 1987), the government sold the stock through an auction (tender sale) offer. This form of sale invites bids without a fixed price, so it ensures that the final price eliminates any excess demand. Thus, the data offer evidence that when the UK Treasury chose to sell stock through a tender sale offer, it was able to capture the full market value of the company (Jenkinson and Mayer |8~). However, more often the government deliberately chose a form of sale (fixed-price offerings) and a pricing which generated enormous excess demand, a phenomenon often anticipated in the press before the sale. The data on application multiples (the ratio of demand to supply at the offer price) give an indication of the degree of excess demand at the fixed issue price. The resulting rationing was deliberately skewed in favor of smaller domestic investors, consistent with the government's policy to broaden share ownership. The goal of this policy was probably to create a class of investors which would resist policy changes adverse to the firm, thus binding future governments. This reinforces the view that the perception of policy commitment is crucial in privatization sales.

This evidence on underpricing should be compared with initial returns on private public offerings, a well-documented phenomenon for private IPOs. Interestingly, there is strong evidence (Jenkinson and Mayer |8~) that underpricing in privatization sales is even greater, both in the UK and France. This is very hard to reconcile with prevailing explanations for discounting IPOs, based on the presence of better informed investors. While most private IPOs are relatively new and unknown companies, firms privatized in the UK and France were large and well known companies with a long track record. Moreover, many of those showing the largest discounts were utilities, which operate in less than competitive markets and whose revenues are fairly predictable. A gradual sale may be interpreted as suggesting that the government had superior information over the value of the assets; if the private sector were better informed, as seems natural, an auction sale would maximize proceeds. In contrast, the choice of a gradual sale and a high level of discounts suggests that the government needed to convey some strong signal. Underpricing may then be interpreted as an attempt by the government to signal political intent, and invest in credibility capital.

The signalling model suggests that as reputation for commitment increases, larger initial share offerings may be preferred (in part to improve incentives), while discounts, controlling for sale size and specific firm risk, may decrease. In the case of the UK, the country with the longest data series, sales of larger stakes seem indeed to become more common over time. More recent privatization sales, such as the sale of Rolls Royce, British Airways or BAA, were, in fact, for the entire capital stock. This could be interpreted either as an improvement in the ability TABULAR DATA OMITTED of the government to access equity markets because of improved confidence or as an exogenous increase in the share of financial wealth allocated to equity investment which relaxes the capacity constraints in the market.

The progressive decrease in discounts is quite clear in the French data, which is however a very short sample. There is less clear evidence in the UK case.(7)

Exhibit 4 contains fairly detailed data on public sales in the Spanish privatization program. This sample is only a subset of all privatization sales. Once again, the evidence indicates a preference for partial initial sales, associated with a progressive dismissal of the remaining stake held by the government. These sales were fixed-price offerings: the evidence on their pricing is consistent with deliberate underpricing, with discounts ranging as high as 100%. The data on application multiples also indicate enormous excess demand at the sale price. It is apparent that there were no binding market capacity constraints. Finally, we have no table on the time series of total privatization proceeds, since we do not have information on private sales, which were predominant in the Spanish program. However, one can infer from the column of proceeds from public sale the usual sharp progression in proceeds over time. Ultimately, the size of the stakes sold is much higher, on average, for the manufacturers such as AMPER, ACESA or ENCE than for the utilities such as GASMADRID or the oil company REPSOL, which is consistent with the reputation-building argument. Moreover, while the REPSOL was quite large, it seems hard to explain the limited size of the GASMADRID sale on the basis of its size.

Exhibit 5 presents some aggregate data on the Chilean program. Two subperiods can be identified: a first massive transfer of assets to the private sector in 1974-1981, and a second wave of sales from 1986 onwards. Also, in this case, the progression of sales appears to increase over time in both subsamples.

The particular history of the Chilean experience deserves some attention. The first wave of sales was aimed at a very rapid transfer of majority stakes in a large number of firms to the private sector, as a result, it was biased towards sales to those private firms which could raise substantial amounts. Specifically, a small number of conglomerates had access to international capital markets, and could finance large purchases through foreign borrowing. In addition, the government de facto financed a large amount of these sales by extending terms of payment over TABULAR DATA OMITTED time. A traumatic recession with very high real interest rates in 1981-1982 caused the collapse of several of these overborrowed private groups. As a result, the state found itself forced to renationalize many of the firms sold. The second wave was financed more conservatively with a broader equity base and with a greater dispersion of buyers, which included many foreign firms.

The Chilean experience offers some important insights for privatization. We have so far focused on the effect of public policy commitment on incentives. However, private commitment may be just as important. For instance, the private buyer of state assets ought to be required to contribute an adequate amount of new capital to finance its purchase, in order to avoid creating perverse incentives for risk-taking behavior which shifts the ultimate financial responsibility to the state in the form of contingent liabilities. Similarly, a commitment to new investment may be demanded from the private sector to rule out both high-risk strategies, where the private owner gambles with the acquired assets with little new capital or slows investment in capacity to maintain low output, and oligopolistic profits.

Exhibit 6 offers an interesting perspective on gradual sales. It describes in detail the plan for sequential sales of stakes in several firms to the private sector. It also contains information which allows |us~ to verify that the original plan was, in fact, executed rather precisely. Although there are some deviations from sale targets, perhaps reflecting some attention to market conditions, it is remarkable to see how the sale plan indicated in advance a gradual dismissal policy, to which it adhered quite closely.

Exhibits 7 through 11 present extensive data on the Nigerian, Turkish and Malaysian programs. Although our Nigerian data lack detailed information on the timing of sales, the extent of partial sales and their progression over time document a significant but temporary risk-bearing role of the government. This is particularly evident in Exhibits 7 and 9. However, this privatization program has certain specific features. In the first place, Exhibits 7 and 9 indicate that the government intends to sell its entire shareholdings in many firms (although we do not know whether this represents a final sale, implemented in stages). It can be argued that a sequence of complete sales of stock in individual firms, when distributed over time, is equivalent to a sequence of partial sales in a larger number of firms from the point of view of aggregate retained shareholdings. This may not establish policy credibility on a sector-by-sector basis, but a comparison of firms which are the object of complete and partial sales reveals that there is significant overlap of sectors. In addition, partial shareholding by the state appears to have been already important in the past; thus these complete divestitures may simply represent a final sale of firms in sectors over which the government has already established a certain policy credibility over time.

The information available on pricing of these sales is not very satisfactory for our purposes. The data presented on post-sale market prices come from the same point in time for all firms, so we cannot measure underpricing, since post-sale changes in market valuation presumably have been different for different firms. It is significant, TABULAR DATA OMITTED TABULAR DATA OMITTED however, that all these later prices indicate a very strong value appreciation, which is suggestive of consistent and significant underpricing.
Exhibit 8. Privatization in Nigeria -- Enterprises in Which
State Holdings are To Be Partially Privatized

 Present Intended State
 State Participation as %
Enterprise Holding of Equity After Sale

Federal Bank of Nigeria 100 70
Niger. Ind. Dev. Bank 100 70
N. Bank for Com. & Industry 100 70
Federal Savings Bank 100 70
Unipetrol 100 40
Nat'l Oil & Chem. Markt. Co. 60 40
African Petroleum Ltd. 60 40
Jos Steel Roll. Mill 100 40
Katsina Steel Roll. Mill 100 40
Oshogbo Steel Roll. Mill 100 40
Nigeria Airways Ltd. 100 40
N. Nat'l Shipping Line 100 40
N. Superphosphate Fertilizer 100 40
Nat'l Fertilizer Co. 70 40
N. Nat'l Paper Manuf. Co. 86.5 40
N. Newsprint Manuf. Co. 90 40
N. Paper Mills Ltd. 90 40
Savannah Sugar Co. Ltd. 75.4 40
Sunti Sugar Co. Ltd. 90 40
Lafiagi Sugar Co. Ltd. 70 40
Ashaka Cement Co. Ltd. 72 30
Benue Cement Co. Ltd. 39 30
Calabar Cement Co. Ltd. 68 30
Cement Co. of Northern N. 31.53 30
N. Cement Co. Ltd., Nkalagu 10.72 10

Source: Privatization Committee, Government of Nigeria.

Exhibit 10 presents data from the Turkish privatization program. We are able to determine the timing of sales, which indicate the usual progression in the amount sold.

Here, as before, partial sales appear to be the norm. Some earlier partial sales are of firms with considerable exposure to public policy changes, such as the telecommunications monopoly. Such a firm has very large capital investment and supplies the entire population; the political risk of redistribution of their quasi rents is evident. Therefore, an early sale of such firms may establish credibility faster; as a populist government will be eager to intervene in these firms, the development over time of a reputation for commitment to current policy would be faster. The fact that no further sale of stock in these firms was made in the following two years seems consistent with a desire to establish a track record. Moreover, its initial sale exhibits the strongest underpricing. On the other hand, this is also a much larger sale than average, and it may be made gradually to avoid swamping the market. Furthermore, it is not clear that the sale of monopolies should come earlier than the sale of enterprises in more competitive industries.
Exhibit 9. Privatization in Nigeria -- Nigerian Enterprises in
Which the Whole Stake Currently Held By the Government Will Be

 Current State
Enterprise Ownership Stake

Nigeria Hotels Ltd. 51
Durbar Hotels Ltd. 100
Aba Textile Mills 70
Central Water Trans. Co. 100
Nat'l Cargo Handling Ltd. 100
N. Nat'l Fish Co. Ltd. 55
N. Food Co. Ltd. 56
Nat'l Grains Prod. Co. Ltd. 100
Nat'l Root Crops Prod. Co. 100
N. Nat'l Shrimps Co. Ltd. 86
New Nigerian Salt Co. Ltd. 100
Nat'l Salt Co. Ltd., Ijoko 100
Specomill Nigeria Ltd. 60
South East R.Wood Ind. 16.27
Niger-Rumanian Wood 25
Nigerian Film Co. 100
Opobo Boat Yard 35
Ore/Irele Oil Palm Co. 60
Road Constr. Co. of Nigeria 60
Impresit Bakolori Nigeria Ltd. 60
North Breweries Ltd. 50
West African Distilleries Ltd. 100
N. Engineering Constr. Co. 60
Tourist Company of N. Ltd. 100
Elect. Meters Co. Ltd. 60
United N. Insurance Co. 42
United N. Life Insurance 33
Mercury Assurance Co. 40
Ayip-Eku Oil Palm Co. 60
Ihechiowa Oil Palm Co. 60
Sokoto I. Livestock Co. 80
Motor Engineering Serv. Co. 100
Nichemtex Industries Ltd. 10

Source: Privatization Committee, Government of Nigeria.

The data on pricing indicate that many sales traded at small or even negative premia, although the situation appears to have been reversed in more recent sales. Since we do not know whether these sales were tender-offer rather than fixed-price sales, it is impossible to tell whether the full pricing was deliberate.

Exhibit 11 collects the information available on the Malaysian privatization program, one of the oldest in the developing world. The data is here quite complete, and therefore offers some scope in interpreting its evolution. TABULAR DATA OMITTED The extent of underpricing is truly remarkable: the market price immediately after the sale jumps to an average premium of roughly 80% from the offer price, while the application multiples suggest a demand on average ten times larger than supply. We find that the smaller stakes sold refer to sales of firms with a certain degree of domestic monopoly power, such as the national telecommunication company and the airline affiliate; presumably, these companies are particularly vulnerable to the risk of changes in domestic regulatory policy, unlike the cement manufacturers or the hotel company. The case of the sale of the highly subsidized automobile manufacturer is probably similar, since the company relies heavily on a protective attitude by the government. Annual proceeds follow a rising progression, culminating in recent years in the sale of the Malaysian Telecom Syarikat. The timing of this delicate sale, six years after the onset of the program, is probably not accidental.

A. The Early Forms of Privatization in Eastern Europe

A major test for the reputation-building approach presented here will be its ability to predict the form that the privatization plans in Eastern Europe will take in the next years. However, privatization in Eastern Europe is certainly a more complex process for two reasons: the fact that most state-owned firms are unprofitable, and the sheer TABULAR DATA OMITTED enormity of the task. These elements suggest that a gradual approach may not be desirable. First, failure to attain rapidly a critical mass of privatized firms could lead to a much too slow improvement in productivity and to a collapse of the reform program (Roland and Verdier |15~). Second, for countries in difficult financial conditions, only a clear separation of firms' liabilities and the budget can avoid fiscal collapse. Finally, speed may be required to overcome the effect of a control vacuum over the corporate sector. As plan discipline has disappeared, labor and management are de facto in charge of most decisions; in the end-game period after privatization is announced but prior to firm private control, they are led to decapitalize the firms by appropriating or liquidating assets. A rapid transition to private owners and a clear separation from the government budget seem necessary to introduce some discipline and avoid further subsidization of inefficient production.

There is, however, a set of firms in these countries which are profitable, often because they command a strong market position and have attracted foreign investors, for whom the risk of policy changes is certainly a major consideration. The following exhibits describe the structure of sales to foreigners of these potentially profitable firms. Since these are private sales, data on pricing are unavailable. However, the evidence on the gradual transfer of shares and residual income rights is evident, even though it is well known that managerial control transfer is quite complete from the outset.

In addition, the mass privatization programs in these countries have elements consistent with both capital scarcity (most certainly a major consideration) and confidence building. In Poland, for instance, the mass privatization scheme involves a combination of partial sales with underpricing. The plan calls for distributing 30% of stock free to the general population through a voucher program, with an additional deeply discounted sale of 20% of the stock to the firms' employees, while the state sector would retain the remaining 50% of stock for some time (Dhanij |4~). The structure of this privatization program is similar to programs announced later in other Eastern European countries, such as Romania and Bulgaria. It is noteworthy that while these governments are retaining significant stakes, they have already indicated their intention to sell them in just a few years' time.



The idea of free distribution of vouchers clearly represents an extreme form of underpricing. On the other hand, a voucher program is probably the only option available, given the extent of the transfer.

III. Future Research and Discussion

We have provided a set of data on privatization transactions which document striking similarities across countries. Sales tend to be gradual; the government provides a temporary risk-bearing role even well after it transfers control; and they are often underpriced. One explanation is the existence of temporary market capacity constraints. Another is based on a confidence-building strategy by the selling government, in which willingness to retain a minority stake (as well as underpricing) signals a more reliable future policy and reassures investors, a crucial element for the success of a privatization program. However, the data available does not easily distinguish between the two hypotheses. In part, it is because they have similar empirical implications, that gradual sales are a means of avoiding depressing prices by large sales, although for different reasons: in one case, because of exogenous capacity constraints; in the other, because investors make some inference about stock value from the amount sold.

A formal test would only become possible once more datapoints are collected. However, the limited evidence presented here seems more supportive of the reputation-building hypothesis. The predominance of partial sales is true across all countries, whatever the degree of development of the domestic capital market. Sale proceeds tend to TABULAR DATA OMITTED increase sharply over time rather than match the growth of private savings. Sales are deliberately underpriced, and to a greater degree than private IPOs. A casual evaluation of the evidence suggests that smaller initial sales, a long time interval between stock issues, and large underpricing of sales seem to be features more common for firms which are more policy-sensitive, such as natural or legal monopolies (utilities, concessions) or heavily subsidized/protected manufacturers.

Moreover, experience has shown that a stock market is able to absorb very large stock sales. The first British Telecom sale was three times larger than the previous largest share issue, but it was still oversubscribed several times. The Telmex sale was an even larger sale relative to the sale of the domestic market, but was a huge success.

It is puzzling that public sales are made through fixed-price offerings even though tender (auction) sales are, in principle, better at maximizing proceeds. These prices are set well below market values, as our data from application multiples and one-day returns indicate. The recent literature on underpricing has suggested that it fulfills a signalling role; and while a government is not likely to have better information over asset values than the market, it certainly does over its policy intentions. To the extent that early credibility of policy also has an important effect on private investment, underpricing may be justified. Thus, the pricing and sequencing of sales are substitutes: slower sales at full price would achieve the same result over a longer period than discounted sales. In general, when policy credibility has an important effect on investment, a more rapid privatization may be justified even if associated with greater discounts.

Although the confidence-building view may explain the extent of underpricing, the issue is probably complex. The choice of underpricing offers the government the opportunity to determine the allocation of shares, which may be employed to induce wider share ownership. This policy goal is also consistent with a commitment motivation. A committed government may ration the allocation of shares with a bias towards the smaller investors, with the intention to create a large constituency vested in the success of privatization in order to reduce future political risk (and even bind a future government's actions). However, this also requires the introduction of incentives to maintain these shareholdings for the long-term in order to avoid immediate resale. Consistent with this point, the 1984 offering of British Telecom included a bonus to be paid in 1987 to buyers who still held the shares; similar vouchers were distributed at other UK sales.

1 It is, at best, unclear whether state firms are more restrained in exploiting lack of competition, although monopoly power may manifest itself in lower quality and efficiency rather than high prices. Many state firms rely on exclusive licenses: public monopolies are common in telecommunications, oil, electricity, railways, and utilities. Competition is often explicitly prohibited.

2 An opposite view holds that public ownership actually aims at suppressing information about the extent of benefits to favored constituencies. The literature on rent-seeking behavior maintains that economic rents associated with control over public policy are easily captured by interest groups. Privatization can then ensure that regulation is more exposed to public scrutiny.

3 Confidence may take time to develop, in part because a government may dissimulate their policy intent for some time, in order to obtain higher sale prices or more private investment in the firms sold. Thus, there is a minimum holding period necessary to achieve higher confidence.

4 The possibility of political changes implies that political uncertainty is not entirely resolved by confidence in the current government. However, knowledge of preferences may establish the policy regime likely to prevail in the future.

5 There are other contractual arrangements that allow such a separation. A prime example is the award to the private sector of a management contract, while the state retains ownership of the assets. However, this is, at best, a temporary solution, since it does not provide long-term incentives.

6 In an econometric investigation of discounts, it is necessary to adjust for the risk profile of each firm (the sensitivity of its value to policy changes) as well as control for the amount of stock sold, since the formal analysis suggests that underpricing should increase with sale size.

7 However, the need for retaining a large stake will appear every time a firm for sale is in a sector which presents large quasi rents (as is the case for firms in oligopolistic markets requiring large fixed investment, such as utilities). A track record of noninterference in the manufacturing industry (the object of many early sales) may not be sufficient evidence of a reluctance to tax away rents in firms such as British Telecom or the water utilities.


1. A. Alchian, "Some Economics of Property Rights," Il Politico 30, No. 816, 1965.

2. F. Allen and G.R. Faulhaber, "Signalling by Underpricing in the IPO Market," Journal of Financial Economics (August 1989), pp. 303-323.

3. A.E. Boardman and A.R. Vining, "Ownership and Performance in Competitive Environments: A Comparison of the Performance of Private, Mixed and State-Owned Enterprises," Journal of Law and Economics (April 1989), pp. 1-33.

4. F. Dhanij, "Transformation Programs: Content and Sequencing," American Economic Review Papers and Proceedings (May 1991), pp. 323-328.

5. A. Galal, L. Jones, T. Pankaj, and I. Vogelsang, "Welfare Consequences of Selling State Enterprises: Case Studies from Chile, Malaysia, Mexico and the UK," Book Manuscript, World Bank, 1992.

6. M. Grinblatt and C.Y. Hwang, "Signalling and the Pricing of New Issues," Journal of Finance (June 1989), pp. 393-420.

7. S. Grossman and O. Hart, "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration," Journal of Political Economy (August 1986), pp. 691-719.

8. T. Jenkinson and C. Mayer, "The Privatization Process in France and the UK," European Economic Review (March 1988), pp. 482-490.

9. S. Kikeri, J. Nellis, and M. Shirley, "Privatization: The Lessons of Experience," Country Economics Department Report, World Bank, 1992.

10. H. Leland and D. Pyle, "Information Asymmetries, Financial Structure and Financial Intermediation," Journal of Finance (May 1977), pp. 371-387.

11. R.J. Luders, "Privatization in Chile: Lessons from a Massive Divestiture Program in a Developing Country," Mimeo, 1990.

12. M. Marcel, "La Privatizacion De Empresas Publicas En Chile, 1985-88," Notas Tecnicas No. 125 (January 1989), CEPLAN.

13. W. Megginson, R. Nash, and M. Van Randerborgh, "Efficiency Gains from Privatization: An International Empirical Analysis," presented at the Financial Management Meetings, 1992.

14. E. Perotti, "Credible Privatization," Working Paper No. 91-12, Boston University, 1991.

15. G. Roland and T. Verdier, "Privatization in Eastern Europe: Irreversibility and Critical Effects," paper presented at a CEPR Conference on "Privatization of Public Enterprises," Milan, May 1991.

16. D. Sappington and J. Stiglitz, "Privatization, Information and Incentives," Journal of Policy Analysis and Management (Vol. 6, 1987), pp. 567-582.

17. The Financial Times, Various Issues, London, 1979-1991.

18. I. Welch, "Seasoned Offerings, Imitation Costs, and the Underpricing of Initial Public Offerings," Journal of Finance (June 1989), pp. 421-449.

19. J.S. Vickers and G. Yarrow, Privatization: An Economic Analysis, Cambridge, MIT Press, 1988.

20. G. Yarrow, "Privatization in Theory and Practice," Economic Policy (April 1986), pp. 323-364.

21. World Development, Washington, D.C., May 1989.

Enrico C. Perotti is a Professor of Finance at the Graduate School of Management, and Serhat E. Guney is a Doctoral Student in the Department of Economics, Boston University, Boston, Massachusetts.
COPYRIGHT 1993 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Perotti, Enrico C.; Guney, Serhat E.
Publication:Financial Management
Date:Mar 22, 1993
Previous Article:IPO underpricing, insider selling and subsequent equity offerings: is underpricing a signal of quality?
Next Article:The effect of the trading system on the underpricing of initial public offerings.

Related Articles
Globalizing through privatization.
Overcoming public employee opposition to privatization.
Privatization: the long road to success in Argentina.
The rush to privatize in the Asia-Pacific region.
Twenty myths about privatization.
The economic effects of privatization: evidence from a Russian panel.
Privatization, finance key to water sector.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters