آیا توسعه بازار سهام همیشه تامین مالی در سطح بنگاه را بهبود می بخشد؟ مدارک و شواهد از تونس

The question of whether or not increased stock market size allows for improved financing conditions for firms in emerging markets is an important one for policy-making. This paper seeks to investigate this issue by analyzing whether increases in market-level liquidity have indeed trickled down to individual firms over the last decade of stock market development in Tunisia, a fast-growing Mediterranean emerging market. We develop time varying liquidity scores for all firms listed in the Tunisian market over the 1997–2009 period and analyze the extent to which market development, firm-level characteristics and risk exposure affect the magnitude and the distribution of liquidity using a set of fixed effect panel regressions. Our results suggest that massive increases in value traded have created market congestion, thereby increasing the costs of trading, in a context of persistently low efficiency and increased international integration. The main implications of this process are (i) market-level development and international integration are not sufficient conditions to ease access to finance for local firms, (ii) further reforms in the Tunisian market should focus on diversifying corporate ownership and improving the disclosure of information, and (iii) international investors seeking diversification in Tunisia should be aware of a significant illiquidity risk.

As part of a broader transition agenda backed by the European Union and international institutions, stock market development has for the past fifteen years been a cornerstone of Tunisian financial sector modernization reforms. As early as 1995, the legal status of the Tunis Stock Exchange shifted from a public body to a private corporation held by brokers, and it was one of the first markets in the region to adopt an electronic trading system in 1996. Commissions were reduced to less than 1% of the value of transactions, while dividend and capital gains taxes were suppressed. More importantly, market size and liquidity increased very significantly as a result of privatization programs and seasoned equity offerings. As shown in Table 1, the market capitalization of listed firms more than doubled from 2.562 to 6.381 billion dinars between 1997 and 2008. Over the same period, annual trading volume increased tenfold (from 182.5 to 1769 DM), the number of transactions increased fivefold (from 55,083 to 293,952), average transaction value was multiplied by two (from 3313.18 to 6017.99 DM), and the turnover ratio was multiplied by four (from 7.12% to 27.72%). The Tunis stock market is thus expected to play a pivotal role in the country's export-led, FDI-driven development strategy, in a context where high economic growth is needed to absorb an ongoing demographic shock.Nonetheless, Tunisian stock market development takes place in a specific institutional context. A civil law country, Tunisia adopted a state-directed financial system after independence, in a context of limited public rights (until the 2011 revolution, the Destour political movement had been in power for 50 years). Tunisian capital tended to be allocated strategically by technocrats rather than private financiers, and business lobbies relied on crony capitalist networks close to their political leaders to maintain oligopoly power ( Henry and Springborg, 2004). 1 In addition, crony banks affected resource allocation. In 1996, the banking system owned about two-thirds of stock exchange capitalization and about 90% of the securities listed on the exchange, in a context where government was an active banking shareholder. Tunisia was (along with Egypt) the only MENA country whose banking system was deemed less ‘open’ in 2000 than in 1996 by international business analysts (O’Driscoll et al., 2000). Finally, although foreign investment has been liberalized, foreigners willing to buy more than 10% of a company listed on the Tunis Stock Exchange need central bank approval, suggesting potential political interference. Market-based adjustments are hindered by bureaucratic regulations: shareholders breaching the 5, 10, 20, 33.33, 50 or 66.66% thresholds of ownership of a listed company must notify the company, the Conseil du Marché Financier, the Tunis Stock Exchange and specify their intentions, while price fluctuation cannot exceed a 6.09% daily margin. Perhaps not surprisingly, the Tunisian stock market is still considered a ‘frontier’ market by rating agency Standard and Poor's. In this context, quantitative stock market development may have had mixed effects on the actual functioning of the Tunisian financial system. Political economic theory indeed suggests that financial development erodes public and private rents. Elites with access to government power and economic rents have incentives to limit financial transactions in order to hedge against competition from new market entrants (Rajan and Zingales, 2003). More recent work suggests that financial development levels jointly determined the balance of powers between winners and losers and the underlying political context. Finally, the impact of institutions on microstructures, international integration, and the overall ability of listed firms to raise capital is well established in the emerging market finance literature (Bekaert et al., 2001). Taking this into account, this paper seeks to analyze whether increased stock market size is per se sufficient to generate firm-level financial development when the institutional context is lax. To do so, we analyze the impact of market level dynamics and firm characteristics on the magnitude and distribution of firm-level liquidity in Tunisia. Two alternative scenarios may indeed be identified, with different policy implications. According to an ‘optimistic’ scenario, market-level improvements in depth and liquidity has spilled over to firms and ease financing through a more efficient pricing, lower risk premia and greater investor participation. In this case, Tunisian policy makers should focus on pursuing the privatization program, while foreign and domestic investors might readily enter the stock market to diversify their portfolios. According to a more ‘pessimistic’ scenario, market-level development has not improved firm-level financing conditions, due to the impact of cronyism and stale pricing on asset allocation. For instance, we may argue that only a sub-segment of informed investors have benefited from stock market development, so that increased overall liquidity fostered the ‘migration’ of trading from one sub-set of firms to another (Levine and Schmukler, 2006). This would suggest that quantitative stock market development policies are be sufficient to improve the functioning of the stock market, and intertwined political, regulatory and economic reforms might also be required. Finally, it should be noted that theoretical work on liquidity suggests that the arrival of new investors has an ambiguous impact on the costs of trading by thickening the market on the one hand (thereby encouraging the participation of new investors through a size externality) and creating potential ‘market congestion’ on the other hand if the market is one sided for a given number of contracts (Afonso, 2009). To the best of our knowledge, this phenomenon has never been investigated in developing stock exchanges. From an academic point of view, this analysis will contribute to our understanding of the dynamics and determinants of emerging market microstructures. From a policy-making point of view, it should help us to assess the impact of financial sector and stock exchange development policies in Tunisia, and provide a set of recommendations. Finally, although higher liquidity has often been associated with lower absolute returns, risk-averse investors have a preference for liquid assets (Green et al., 2005). By providing contextual information on liquidity dynamics in Tunisia, this analysis may thus also be useful for investors considering diversification in this country. The remainder of the paper is structured as follows: Section 2 briefly discusses the transmission mechanisms of stock market liquidity on economic development. Section 3 reviews the literature on the Tunisian stock exchange. Section 4 presents our dataset. Section 5 investigates firm-level time-varying liquidity. Section 6 analyzes the determinants of liquidity levels and distribution. Finally, Section 7 brings together our conclusions.

The objective of this paper was to analyze the dynamics and determinants of time-varying firm-level liquidity in Tunisia, in an effort to shed light on the microstructural implications of stock market development. Our main results can be summarized as follows. First, we observe significant improvements in market size and sectoral diversification over the last decade. In addition, firm returns display an increasing sensitivity to international industry effects, suggesting higher levels of international financial integration. However, firm-specific effects are consistently low, indicating a lack of informational efficiency, suggesting that enhanced international integration and increased market size do not necessary imply a better functioning of the stock exchange. Second, we find that market size has failed to spill over on firm-level liquidity, as both domestic and international-driven increases in market value have diminished the number of transactions and increased price pressure through a possible crowding out effect on existing stocks. This suggests that efforts should be made to enlarge the investor base and the number of stocks available for trade in an effort to smooth trading conditions. This might imply significant changes in the Tunisian corporate ownership structure. Third, trading patterns displays asymmetric response to positive and negative information, suggesting low informational efficiency and a possible collusive behaviour between government-accredited brokers and informed shareholders. Improving transparency levels should hence be a policy priority. However, we find that informationally efficient stocks are less liquid; suggesting that enforcing better information disclosure 3 might not be sufficient to improve trading conditions. Overall, these results suggest that policies seeking to improve market level liquidity are not sufficient to guarantee an improved access to finance for Tunisian firms. To maximize the impact of the stock exchange on the real economy, policy makers may need to tackle the issue of informational efficiency, while simultaneously enlarging the investor base and diversifying corporate ownership. In addition, international investors seeking diversification in Tunisia should be aware of the significant liquidity costs. Future research should seek to disentangle the impact of specific reforms on firm-level microstructures in emerging markets, in order to identify the content and sequencing of the appropriate reform agenda. We could also investigate the impact of foreign shocks on firm-level liquidity patterns in order to better understand the relationships uniting firm-level liquidity and financial integration. Finally, the relationship between different political regimes and the functioning of financial systems also needs to be investigated.