تنظیم حاشیه سود و کیفیت بازار : تجزیه و تحلیل ریزساختار

We find that trading volume increases and market liquidity remains unchanged, while the adverse selection and order-processing cost components of the spread increase and decrease, respectively, after margin levels decline when stocks become margin-eligible. This evidence indicates that the information content of trades has increased, thereby improving market quality. However, no changes were detected after the 1997 regulatory reforms. These results have implications across a broad swath of corporate finance dimensions, including the (1) cost of capital, (2) public vs. private financing decision, (3) form of managerial compensation, (4) type of ownership structure, and (5) degree of shareholder monitoring.

The Board of Governors of the Federal Reserve System (hereafter, the “Fed”) is empowered to set the percent of a security transaction's value that can be financed by a lender such as the initiating broker.1 For example, if the margin level is set at m%, an investor can borrow up to (100−m)% from the broker when purchasing stock, with the stock serving as collateral for the loan. The effectiveness of margin regulation, in general, and its alleged influence on a stock's volatility, in particular, have been extensively examined with conflicting results. 2 In this study, we analyze the relation between margin levels and market quality, which we define as the liquidity and the informativeness of stock prices. We investigate links between margin levels and liquidity—namely, spreads, depths, and number of market-makers—and relate these links to a number of corporate finance issues.3Seguin (1990) shows that lower margin levels are associated with lower return volatility and larger trading volume. His findings suggest that lower levels of m should lead to some combination of lower spreads, greater depths, and more market-makers, thereby improving market liquidity. According to, for example, Amihud and Mendelson, 1986 and Amihud and Mendelson, 2001, greater liquidity should lower the cost of capital for a firm. 4 In contrast, Hardouvelis and Theodossiou (2002) find that lower margin levels are associated with greater return volatility during bull and normal markets. Under their paradigm, lowering the level of m can lead to a deterioration in market liquidity, thereby raising the firm's cost of capital. We document that lower margin levels lead to a significant increase in daily trading volume, but detect no significant changes in liquidity.5 Specifically, after considering the impact of certain control variables, we find no significant changes in quoted and effective spreads or in the number of market-makers, but we do find a significant decrease in depths. However, this decrease is economically immaterial. These results indicate that lower margin levels do not change a stock's liquidity. Consequently, the first corporate finance implication of our results is that a lower margin level does not materially affect the firm's cost of capital. We next investigate possible links between margin levels and the informativeness of stock prices by estimating the components of the spread using the decomposition model of Madhavan et al. (1997). We find that lower margin levels are associated with significant increases in the proportion of the spread attributable to adverse selection. Combining this result with our findings that spreads and depths do not change while trading volume increases, we conclude that the information content of trading and, therefore, the informational environment of the firm and the efficiency of the firm's share price improves when the margin level is lowered. Because our period of analysis runs from 1993 through 1998 and involves a sample of Nasdaq Small Cap stocks, we investigate the impact of two major market regulatory reforms that were instituted in 1997 on our results. The first reform, the implementation of the Order Handling Rules, mandated the public display of limit orders on Nasdaq. The second regulatory reform is the introduction of “teenies” that lowered the minimum tick size from eighths to sixteenths.6 After examining the entire sample period, we calculate and compare the results from two subsamples—one involving the portion of the sample period before the first of these reforms and the second involving the portion after the second reform was initiated. We find that our aggregate results are driven by the first subsample, because there were no significant changes in liquidity or spread components in the second subsample. Thus, the regulatory reforms appear to have removed any impact of changes in margin levels on market quality. Finally, while examining the impact of margin regulation on volatility is not our primary objective, we do provide evidence in support of the view that lower margin levels do not significantly increase volatility. Coupled with the liquidity and spread decomposition results, we conclude that lower margin requirements led to improvements in the quality of the market for the affected shares before the 1997 regulatory reforms, but to no change afterwards. Our conclusion that lower margin levels enhance the informational environment surrounding the firm has numerous and widespread implications across a number of corporate finance paradigms. For example, Subrahmanyam and Titman (1999) argue that the choice between public and private financing depends on the efficiency of the firm's share price. Holmstrom and Tirole (1993, p. 678) argue that the structure of executive compensation depends on “the amount of information contained in the stock price”. Finally, a branch of research (e.g., Bolton and von Thadden, 1998, Kahn and Winton, 1998, Maug, 1998 and Yu, 2002) argues that there are links between the information content of stock prices, institutional ownership (or ownership concentration), and the degree of monitoring. Although this research suggests that the signs of these relations change depending on various factors, the pertinent studies nonetheless agree that the information content of the publicly traded equity is an important determinant of these links. Rather than investigating differences in market quality surrounding changes in market-wide alterations in margin levels, our methodological design follows Seguin (1990). We use firm-by-firm changes in the level of m associated with firms becoming margin-eligible. 7 We believe that this design provides a statistically more powerful experiment to conduct our analysis for at least four reasons. First, margin-eligibility essentially lowers the level of m for these stocks from 100% to 50%, whereas historically, the largest change in the overall margin level was 25%. 8 Thus, our method provides inherently greater statistical power, because the changes in margin levels associated with margin-eligibility are of much larger magnitude than the changes in the overall margin level. Second, the Fed has changed the margin level for stock purchases only 23 times since it was empowered to do so in 1934, and it has remained at 50% since January 3, 1974. In contrast, during our sample period of 1993–1998, the Fed published a list of newly margin-eligible stocks every quarter, yielding a sample of changes that is relatively current and chronologically diverse. Both the number and magnitude of the changes, yielded by our experimental design, econometrically dominate a design based on market-wide changes by offering inherently greater statistical power. Thus, the likelihood of discerning any impact of changes in margin levels on market quality should be greater and more timely in our margin-eligible sample than in a sample of changes in the overall margin requirement. Third, because changes in margin-eligibility affect only a small subset of equities at any one time, we can control for market-wide factors that impact market quality, including, but not limited to, secular shifts in aggregate volatility, a control conspicuously absent in the work of Hardouvelis and Theodossiou (2002). As a result, we are confident that our approach provides a greater degree of ceteris paribus than studies that examine Fed-mandated changes for all traded stocks. Finally, there are no publicly traded derivative contracts on Nasdaq Small Cap stocks, the set of stocks from which our sample is drawn. In addition, these stocks are, relative to Nasdaq NMS or exchange-listed stocks, thinly traded. Hence, it is reasonable to expect that any impact on market quality associated with changes in margin levels is more likely to be observed for the stocks included in our sample. Our study is organized as follows. Section 2 contains a review of the literature and identifies the hypotheses, while Section 3 discusses the data. Section 4 presents our aggregate results. In this section, we also demonstrate that any latent selection bias due to the fact that the Fed does not randomly change margin-eligibility status does not affect our results. Section 5 presents the results for the two subsamples straddling the 1997 regulatory reforms, while Section 6 concludes.

In this study, we investigate the impact of margin-eligibility on the market quality of the affected equities. First, we find that the market liquidity of stocks is unchanged when they become margin-eligible, as there are insignificant changes in spreads, depths, and the number of market-makers. Second, we find that the portion of the spread that is attributable to adverse selection increases after margin-eligibility, suggesting there are more informed investors, while the portion attributable to order-processing costs decreases. While this increase should typically lead to a widening of spreads, the observed increase in trading volume appears sufficient to offset the increase in adverse selection so that, on balance, spreads are unaffected. In sum, once firms become margin-eligible, the quality of the market for their shares improves. These findings suggest that margin-eligibility has no impact on a firm’s cost of capital, as liquidity is unaffected. However, the observed increase in the information content of market prices has potential repercussions across a broad swath of other corporate finance dimensions. Examples include the public vs. private financing decision, the form of managerial compensation, the type of ownership structure, and the degree of shareholder monitoring. However, the recent introduction of two regulatory reforms—the introduction of teenies and the implementation of Order Handling Rules—appear to have eliminated the impact of margin-eligibility on all variables under consideration. Hence, on balance, a reduction in margin is a non-event for corporations in the current regulatory environment