Industry risk premia in Pakistan.

Nishat, Mohammed

1. INTRODUCTION

Industry characteristics is one of the main factors that determines a firm's business risk [Kale, Hakansson, and Platt (1991)], and a single information can affect more than one security price change, perhaps even the whole market. Lessard (1974, 1976) explains that industry plays an important role in explaining national market volatility. One of the reasons for stock index behaviour are attributed to industrial composition as some industries are internally more volatile than the other [Grinold, Rudd, and Stefek (1989)]. Moreover, some sectors show a high degree of global integration, for example, the finance sector [Roll (1992)]. Similarly, consumer goods, fuel and energy, and transportation sectors are extremely important for any country index. King (1966) suggests that if a significant difference in industry risk premia is observed, then we need to isolate the market risk premia and industry risk premia. He observed that the industry components of variance showed much less change from sub-period to sub-period. Significant differential impact of regulatory policy on cost of capital across various sectors was also observed [Isimbabi (1994); Prager (1989)].

The industry specific policies in Pakistan are observed either as a part of the reform package during 1988 and early 1990s, or as an additional policy measure to further boost the private investments in priority sectors. These policies included incentives for foreign investment through permission for repatriation of profits, the easing of investment and banking sector regulations and easy access to loans and tax exemptions on priority sectors like power, exports and agriculture based industries. In addition, the government encouraged equity participation to avoid instability through growing leverage. Some sectors like Islamic and institutional investors were regulated to make the investment more competitive during the reform period. For borrowers as well as lending units these policies are important for estimating the alternative cost of capital and comparing it with the risk premium of the firms to value their future cash flows [for details see Nishat (1999)]. Many industries enjoyed tax exemptions/holidays, additional fiscal benefits and access to concessional loans like agricultural based industries, Modarabahs and power projects. To boost foreign exchange earnings government provided concessional export funding to export dominated firms, and relief on import duties on machinery and raw materials. (1) To attract the foreign investors, a legal framework and security on capital investment was provided which included permission to remit profit and capital and relief on double taxation in the case of specific countries. In addition, foreign investors were allowed to negotiate the terms and conditions of foreign currency loans without government intervention. I expect a higher level of risk premia in industry portfolios during reform period. However, there is a possibility that industry characteristics being high or less volatile continues over time, and government policies and reforms either induce more movements in prices in all industries, or only in specific once, which are more sensitive to opening of financial market to foreign investors. On the other hand, there is evidence that in some countries the movements in stock prices are stabilised after liberalised policies [De Santis and Imrohoroglu (1997)].

The other important factor, which identifies the industry portfolio in Pakistan, is the extent of leverage observed across industries due to preferential sectoral debt policy prevailing in Pakistan. For example, the industries based on locally manufactured machinery are given loans at the interest rate almost half of the market interest rate. Moreover, during the non-reform period it was easier and cheaper to get capital from financial institutions than raising through capital market. As a result, the extent of the debt-equity ratio has been higher across industries during both the non-reform and reform period. Due to poor performance of industrial sector, in general, the loan recovery rate had been very low. This resulted in higher debt-equity ratio in the KSE, particularly during the non-reform period. The financial reforms during the 1990s attempted to reduce the extent of leverage across firms and encouraged equity participation in Pakistan. It is argued that the higher level of leverage causes higher volatility in returns of the firms [French, Schwert, and Stambaugh (1987)]. Leverage is one of the factors causing volatility in industry returns in the KSE [Nishat (2001)].

For investors the factors identified above are important to consider while calculating cost of capital and discount rates to evaluate their investments and to value the expected future cash flows. For policy-makers and lending institutions, it is vital to incorporate the risk prevailing in that sector to charge the cost of capital, which is comparable to expected risk premia or discount rate of the firms in that sector. This will help them to justify the economic costs and benefits of subsidy involved in certain sectors. Moreover, opening of the stock market also resulted in an inflow of risky foreign capital in industries like chemical, food and allied, fuel and energy, and engineering. Similar funding also came through export oriented firms, who in many cases got advance payments from their overseas clients. I expect that the financial reforms and changing industry specific policies could be another reason causing the change in individual industry risk premium share in the total market risk premia overtime. I also expect difference in risk premia pattern for the firms competing in export market, multinationals, and the industries which are domestically protected.

In this paper the following alternative hypotheses are tested:

* The industries subject to differential policies and reforms have higher risk premia.

* The share of the industry risk premium in market risk premium varied during the non-reform and reform periods.

* Export, multinational and most growth industries, had higher risk premia, and contributed more in total market risk premium than other industries.

* The relation between industry risk and return, if exists, is different during the non-reform and reform periods.

* The industry portfolio returns are more volatile and predictable during the reform period than the non-reform period.

2. ECONOMETRIC MODEL AND ESTIMATION METHOD

I estimate the industry risk factors using the standard CAPM model:

[R.sub.it]-[R.sub.ft]=[[alpha].sub.i] + [[beta].sub.i] ([R.sub.mt] = [R.sub.ft]) + [[epsilon].sub.it] ... ... ... ... (1)

where [R.sub.it] is the value-weighted return on industry portfolio i in period t, t = 1, 2, 3, .... T. [R.sub.mt] the return on market portfolio m in period t, t = 1, 2, 3 ...... T and [R.sub.ft], the risk free rate in period t, t =1, 2, 3 ..... T. [[beta].sub.i] is the risk factor of industry portfolio i and [[alpha].sub.i] is the intercept.

I used the same model to estimate the risk factor for Islamic industry. As there is no concept of risk free rate of interest in Islamic finance. [R.sub.f] is pure time value of money or compensates for time preference, and it is represented by market rate of interest. Though it is permissible in Islamic perspective to have a compensation for time value of money, it cannot be realised in the form of interest. It can only be an implicit part of the outcome of a real economic transaction [Khan (1991, 1996)]. In Islamic framework, we would have a good indicator of risk free return, as argued in literature, if we had efficiently and competitively operating Islamic bank [Khan (1991)]. Islamic banks are supposed to manage risk to the minimum possible level through diversification of their investments. The rate of return paid by them to depositors can be considered a close proxy for the pure time value and hence the risk free return. The rates of return on saving deposits of Islamic banks, however, are not readily available currently. Moreover, an Islamic portfolio is one of the alternatives for investors in Pakistan and is open for all investors. I, therefore, use the same risk free return to estimate the risk premia of Islamic stocks as I use for industries.

As argued in literature and described earlier, many of the CAPM average return anomalies are related and are captured by the three factor model of Fama and French (1996). They largely disappear except for the contribution of short run returns. I estimate the following three-factor model to estimate the industry risk premia.

[R.sub.it] - [R.sub.ft] = [a.sub.i] + [b.sub.i] ([R.sub.mt] - [R.sub.ft]) + [s.sub.i](SMB) + [h.sub.i](HML) + [[mu].sub.it] ... ... (2)

where [R.sub.it] - [R.sub.ft] is the return on portfolio i in excess of the risk free rate in period t, t = 1, 2, 3,..... T. [R.sub.mt] - [R.sub.ft] is the excess return on market portfolio in period t, t = 1, 2, 3, .... T. SMB is the difference between the returns on a portfolio of small stocks and returns on a portfolio of high stocks in period t, t = 1, 2, 3, ....... T and HML is the difference between returns on portfolios of high and low book-to-market stocks in period t, t = 1, 2, 3, ..... T. [b.sub.i], [s.sub.i] and [h.sub.i] are the slopes in the above time series regression, [a.sub.i] is the intercept.

Time-varying Risk Premia

In the CAPM estimation described earlier I assumed that the industry risk premia are stationary, normally distributed and serially uncorrelated, in which case the error process will be ND(O, [[sigma].sup.2]). I analyse the empirical performance of the CAPM and test for the following implications:

* the disturbances, [[epsilon.sub.it], in regression (1) should be serially uncorrelated, homoskedastic and normal,

* the systematic relationship between portfolio return and market returns should be linear, and

* the [[beta].sub.i]'s in regression (1) should be time invariant.

For examining the industry risk premia during the non-reform and reform periods, the following GARCH-M model, is estimated:

[y.sub.t] = [y.sub.0] + [y.sub.1][x.sub.t] + [theta][h.sup.1/2.sub.t] + [[mu].sub.t] ... ... ... ... ... (3)

[[mu].sub.t] = [[epsilon].sub.t] = [phi][[epsilon.sub.t-1], ... ... ... ... (4)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)

Note that [y.sub.t] is the excess return on industry portfolio on week t, t = 1, 2, 3 ..... T and the single explanatory variable [x.sub.t] is the excess return on market portfolio on week t, t = 1, 2, 3 ..... T. The error term ulis assumed to be MA(1). A dummy variable (Dr =1 for reform period, and 0 otherwise) is included in Equation (5) to capture the impact of institutional developments and reforms on risk return relation in GARCH-M framework. A significant coefficient for the dummy variable will identify a shift in reward for risk across non-reform and reform periods. I also test for the difference between second sub-period of reform and non-reform period to distinguish the impact of aggressive and frequent policy measures observed during the later period of reforms.

Within the framework of the basic GARCH-M model any institutional or reform news may affect directly the level of share prices/industry returns through an independent news effect. Or it may affect the variance of the industry return through a GARCH process and then only affect the level through the effect of the variance on the mean via the notion of a risk premia effect. Conventional likelihood ratio or Wald tests may be constructed to test for the significance of these effects. Under the mean-variance hypothesis, [theta] > 0, so that large values for the conditional variance are expected to be associated with large returns. The coefficient [alpha] indicates the ARCH effect and [beta] explains the non-synchronous trading effect in the model. An estimate of [alpha] + [beta] close to 1 indicates a high degree of persistence in volatility movements, that is the long run effect of unit innovation shock, in [h.sub.t]. This shows that today's volatility in industry returns affects the forecasts of volatility in industry returns into the indefinite future. The persistence phenomenon is important in pricing options and futures as well as consumption/savings and portfolio decisions. The GARCH-M model is used to estimate time-varying conditional second moments and a mean/variance ratio. This ratio is a proxy for the risk-return trade-off or the market price of volatility. Since over time the incentives for investment opportunities and industry level policy have changed, I expect that the risk-return trade-off will also change, as will the investors' preference towards risk.

3. DATA

The firm level weekly share prices, dividend, capital issues, and paid-up capital data on KSE is collected and computerised by the author using the original "Daily List" and "List of Daily Trading Documents" published by the KSE during January 1980 to December 1994. The data consists of weekly share prices adjusted for dividend and capital issues. The value-weighted industry returns are calculated for non-reform (January 1980 to June 1988) and reform period (July 1988 to December 1994) to test the hypothesis of differential industry risk premia in Pakistan. For comparison industry returns are also calculated for two non-reform sub-period (January 1980 to June 1985 and July 1985 to June 1988) and reform sub periods (July 1988 to June 1991 and July 1991 to December 1994. For further details [see Nishat (1999)].

4. EMPIRICAL RESULTS AND DISCUSSION

In this section I present the estimated results to highlight the changing behaviour of industry risk return relationship which could either be due to industry characteristics or induced by reforms, or due to both factors. The risk returns relationship are compared during non-reform and reform periods. The industry risk premia is also compared during above periods. The time-varying risk premia and return relationship estimated through GARCH-M process is presented in this section.

Industry Risk and Returns

In order to test that the industry returns are significantly different during non-reform and reform periods, we consider the null hypothesis of no difference in industry returns before and after the institutional development and reforms. The t-tests conducted to see if returns on industry portfolios during the non-reform and reform periods are significantly different. On the basis of t-tests at 0.05 significance level, we can not reject the null hypothesis and therefore conclude that the mean returns of most industry portfolios are statistically the same during the non-reform and reform period. The t-tests are also conducted to compare if the returns on industry portfolios during non-reform period and the second sub-period of reforms are different. On the basis of t-tests at 0.05 level, we reject the null hypothesis for several industry portfolios and therefore accept the alternative hypothesis that for these industry portfolios the average return are different during the second subperiod of reforms and the non-reform period. In most cases the industries with growth firms, foreign equity component and export-based firms have higher average returns than the industries with domestic firms.

Table 1 compares the industry returns and risk factors for the overall study period, January 1980 to December 1994. The expected average returns indicate a lot of variation across industries. The average weekly return varies from 0.157 to 0.836 percent. The maximum expected return is observed in the food and allied industries, and the minimum average return is observed in the jute industry. The industries dominated with multinational and export-based firms have higher average returns than the industries with domestic base. The correlation between industry size and industry return is positive (0.380). The correlation between industry risk beta and industry return is positive (0.588). The results support the theoretical relationship that high returns are associated with risky industries during the overall study period. The risk return relationship is more significant (indicated by higher t-statistics values for risk parameter beta) for industries with growth firms, foreign equity component and export-based firms than the industries with domestic firms. The intercepts for industry portfolios are not different than zero as the [[alpha].sub.i]'s are statistically insignificant at the 5 percent significance level, except for textiles for which the intercept is positive and statistically significant. This indicates that for most of the industries portfolio pricing is in equilibrium. For the textile case the zero-beta portfolio return is higher than the risk free return which supports the zero-beta version as suggested in other markets [Jensen (1968)].

The magnitude of industry returns during the first sub-period of non-reform is lower than the overall period (see Table 2). The highest average return during this period is for engineering (a growth industry which also has higher weight in KSE) that is 0.568 percent, and the lowest is in the leather/tanneries industry that is -0.653 percent. The correlation between industry return and industry risk factor is positive (0.447). In many cases industry risk factor betas are higher for industries which yield higher average returns. However, some industries have the higher value for risk factor beta but are not compensated with higher average returns. In this sub-period we do not find much difference in domestic and multinational firms' risk factors. The risk returns relationship during this sub-period is less strong than the overall study period (indicated by t-statistics of beta). No correlation is observed between industry size and industry return as the coefficient of correlation is negligible. All intercept terms are not different from zero, which indicates that industry portfolio stock pricing is in equilibrium during this period.

The correlation between industry return and beta (risk factor) is very low (0.047) during the second sub-period of non-reform. The higher returns are not attributable to higher risk factors (see Table 3), which is not consistent with theory. The reason could be that there was less opportunity to diversify the risk during this sub-period than the reform period. The risk return relationship is weaker during this sub-period than the overall period as the t-statistics for risk beta (risk factor) are lower than the overall study period. The correlation between industry size and respective average returns is positive but very low. During this period the [[alpha].sub.i]'s are also not different from zero except textiles, which supports the hypothesis that the pricing of the industry portfolios is in equilibrium. (2)

During the overall reform period the average industry returns are higher than during both the sub-periods of non-reform (see Table 4). The risk return relationship is stronger and more consistent with theory during reform period than the two sub-periods of non-reform. The industries with high (low) risk factors have high (low) average returns. The correlation between industry returns and risk factors is positive. The coefficient of correlation between risk and returns is higher (0.511) during the reform period than the non-reform period. However, some industries are exceptions during this period like transport and communications for which the average return is low but the risk factor is higher. Similarly, for leather and tanneries the average return is higher but the attributed risk factor is lower than expected. Industry size and industry returns have a positive relationship but the coefficient of correlation is only 0.369. All intercept terms are not different from zero (except for miscellaneous firms portfolio) which indicates equilibrium pricing for industry portfolios during the reform period. The coefficient for miscellaneous industries is positive and almost two times the borrowing rate, which supports the zero-beta portfolio during the reform period.

As presented in Table 5, in most cases the average returns on industry portfolios during the first sub-period of reforms are lower than the overall reform period, except for food and allied industries. However, the correlation between industry risk and returns are higher (0.765) than the non-reform period. This indicates that the empirical relationship between risk and return is stronger during this sub-period than the non-reform period. There are some cases where the theoretical relationship is not observed. For example, the transport and communication portfolio has a high beta but a negative return. Similarly, construction and vanaspati and allied have very low beta factors but the average returns are comparatively higher. The correlation between industry size and portfolio returns is low (0.382) during this sub-period but higher than in the non-reform period. All intercept terms are statistically not different from zero except for Islamic portfolios. The value of the intercept for Islamic portfolios is higher than the prevailing markup rate of 15 to 17.5 percent per annum. The other explanation could be that Islamic stocks had some degree of non-equilibrium in their pricing during this period, possibly due to many new flotations of Islamic firms during this sub-period.

Both the average returns and the respective risk factors are higher for most of the industry portfolios during the second sub-period of reform than the non-reform period (Table 6). However, there are cases of low returns attributed to high risk and vice versa. The correlation between industry risk factor beta and average return is low (0.275), which indicates a weak theoretical relationship between industry risk and return during this period. The correlation between industry size and industry return is positive, but low. The intercept terms are statistically not different from zero in all cases except cement and miscellaneous industries. For the cement and miscellaneous industries the intercepts are much higher than the risk free return and support the zero-beta version portfolio for these industries. This is probably due to non-equilibrium of cement and miscellaneous stocks pricing during the second sub-period of reforms. The above analysis indicates that the average returns for most industries are higher during the reform period than the non-reform period, particularly during the second sub-period of reform. However, statistically the average return is higher only for few industries during the overall and the second sub-period of reforms. The theoretical relationship between risk and return, that the higher returns are attributed to higher risk, is stronger during the reform period than the non-reform period, particularly during the first sub-period of reform. Most industries indicated their pricing in equilibrium during both the non-reform and reform periods.

Industry Risk Premia

The market risk premia in the KSE increased significantly after liberalised policies in Pakistan. The volatility is returns is more evident, persistence and more predictable during the reform period than non-reform period [Nishat (2000)]. In this section I compare the industry portfolios if they are equally sensitive to government policies or their industry characteristics prevail during non-reform and reform periods. I test the alternative hypotheses that:

* The industry risk premia are higher during the reform period than the non-reform period.

* The risk premia are higher for industries with foreign capital component than for the domestic industries.

In order to test the above hypotheses I estimate the risk premia on industry portfolios using cross-sectional regression procedure described in Section 2. The explanatory variables for cross-sectional regression are obtained for each week t, t = 1, 2, 3,.... T through CAPM and three factor model, given earlier in Equations 1 and 2. I conducted Chow tests of whether the industry risk premia estimated during the non-reform and reform periods are governed by the same relationship. The null hypothesis is that there is no difference in the coefficients of regressions in the two periods. The acceptance of the alternative hypothesis will establish that there is a significant difference in risk premia during non-reform and reform periods. I also distinguish the second sub-period of reform (July 1991 to December 1994) from the non-reform period to see if there is any significant difference in risk premia due to frequent policy measures observed during later period of reform.

Industry risk premia estimates of the CAPM and the three-factor model (3) are presented in Tables 7 to 12.* The Chow test statistics (reported for CAPM case only) shown in Table 7 indicate that out of 22, 12 industries indicated the risk premia being estimated through different relationship during non-reform and reform periods. Similarly, 8 out of 22 industries indicated that the risk premia estimated during the second sub-period of reforms and non-reform are governed by different relationships.

In most cases the risk premia estimated through both CAPM and three-factor model have almost the same pattern across industries. As presented in Table 7, the export-based industries such as textile and synthetic rayon and growth industries (which are subject to international factors as well as domestic policy changes) have higher risk premia. For example, the food and allied industries indicated a higher risk premium mainly due to couple of large multinational firms in this sector. This supports the hypothesis that the industries with foreign capital component in their operations are more risky and require higher risk premia than the domestic industries. During this period the risk premia estimated through CAPM varies between 0.036 and 0.535 percent per week (three factor model estimates the risk premia between 0.046 to 0.492 percent per week).

As evident in Table 8, during the first sub-period of non-reform the risk premia varies between 0.038 to 0.324 percent per week (0.040 to 0.314 per cent for the three factor model). The food and allied industry contributed the highest risk premia in total market risk premium. A similar pattern is observed during the second sub-period of non-reform (see Table 9). Again the industries which have multinational component in their equity share have higher risk premia and greater contribution to market risk premium during the second sub-period of non-reform. During this period the risk premia varied between 0.006 and 0.328 percent per week.

During the overall and the reform period the risk premia varied between 0.044 to 0.834 percent per week (Table 10). The higher industry risk premia during the reform period supports the hypothesis that due to regulatory policies and institutional development the industry risk premia are higher during the reform period than the non-reform period. Again the industries dominated by exports, growth firms and those with multinational connections have higher risk premia. However, industry risk premia during the first sub-period of reforms are lower than the overall reform period and varied between 0.000 to 0.819 percent per week (see Table 11). The industries with multinational capital, export orientation, and growth have comparatively higher risk premia during the first sub-period of reforms. The risk premia are significantly higher for most of the industries during the second sub-period of reforms as compared to both sub-periods of non-reform (see Table 12). One of the reasons is that the variability in portfolio returns increased significantly after liberalisation. These findings are consistent with the results observed in other emerging markets [Harvey (1995)]. The industry risk premia varied between 0.668 to 1.150 percent per week, and almost all industries indicated higher risk premia during the later period of reforms. Again during this period the industries with export potential, growth and with multinational covering have higher risk premia.

In conclusion, the industry portfolio analysis indicates that industry characteristics prevailed across both the non-reform and reform periods. The results support the hypothesis that liberalised policies have induced a higher risk premia during the reform period than the non-reform period, particularly during the second sub-period of reforms. The results also support the hypothesis that industries with foreign capital component have higher risk premia than the domestic industries during both the non-reform and reform periods.

Time-varying Risk Premia

In the last section while comparing industry risk premia it was established that industry risk premia are higher during the reform period than the non-reform period, particularly during the second sub-period of reform. However, in estimating the risk return relationship I assumed that the risk factor is invariant of time. Now by using the GARCH-M model, I allow the conditional expected industry return to vary over time (and hence market risk premia and market betas also to vary over time). In this case the conditional volatility depends on lagged residuals. I test the alternative hypotheses that:

* The relation between market risk and industry expected returns is different during the non-reform and reform periods.

* The industry portfolio returns are more volatile during the reform period than the non-reform period.

As expected, in most cases the higher average returns appear to be associated with a higher level of volatility It is also evident that for the market as well as industry portfolios, the average return and volatility is higher during the reform period. Another interesting pattern is identified through kurtosis. The index of kurtosis is considerably higher during the reform period. Moreover, the higher values of kurtosis also suggest big surprises of either sign in industry portfolio returns, at least unconditionally, particularly during the reform period. In the CAPM estimation described earlier, I assumed that the industry risk premia are stationary, normally distributed and serially uncorrelated, in which case the error process will be NID(0,[[sigma].sup.2]). I analyse the empirical performance of the CAPM and test for the following implications: the residuals of the regression (1) should be serially uncorrelated, homoskedastic and normal, the systematic relationship between portfolio return and market returns should be linear, and the estimate of beta should be time invariant. The results presented in Table 13* indicate a greater evidence of non-linearity, non-normality and parameter non-constancy. This is probably a reflection of the view that betas are time-varying and are better modeled within the autoregressive conditional heteroskedasticity (ARCH) model framework introduced by Engle (1982). The ARCH framework explicitly models the time varying conditional variances by relating them to variables known from the previous period. (4)

In order to test the above hypotheses, the GARCH(1,1)-M model, as described earlier in Equations 3 to 5, is estimated during non-reform and reform period. In this case dependent variable [y.sub.t] is the excess return on industry portfolio on week t, and the explanatory variable [x.sub.t] is the excess return on market portfolio on week t. The results of the GARCH(1,1)-M model are presented in Tables 14 to 19. The Box Pierce portmanteau test statistics 0(12) and [Q.sup.2](12) are given for an autoregressive or moving average process of order 12 in residuals, and for an ARCH(12) process of order 12 in squared residuals respectively. Both test statistics are asymptotically equivalent to Lagrange multiplier test statistics and have asymptotic chi-squared distribution with 12 degrees of freedom under the null hypothesis of residuals being uncorrected. The procedure followed is described in Baillie and DeGennaro (1990).

I have included a dummy variable ([D.sub.t] =1 for reform period, and 0 otherwise) in Equation 5 to capture the effect of liberalisation on industry risk premia through GARCH(1,1)-M process. I also distinguish the effect of frequent policy measures during the second sub-period of reforms separately. As presented in Table 14 the coefficient of the dummy variable for the reform period indicated a significant shift in risk premia in 5 industries (3 upward and 2 downward). Similarly, four industries indicated a significant increase in risk premia and a decline in one industry during the second sub-period of reform compared to the non-reform period.

In the overall study period, out of 22 industries five indicate a significant relation between risk and return at the 0.05 significance level and one other industry has significant relationship between risk and return

at the 0.10 significance level (see table 14). The reward to risk (indicated by parameter 8) varied between 0.000 and 0.173 percent per week during the long run study period. Most of the export oriented and growth industries with multinational equity capital indicated significant and higher coefficient for risk aversion during this period. The industries which do have a significant relation between risk and returns indicated a significant coefficient of volatility, oi|, or ARCH effect which causes an increase in future volatility. The estimated coefficient of the ARCH effect, a,, is less than one for all industries. Unconditional variance of excess holding yield does not indicate any fat tailed return distribution during this period for any of the industries. None of the industries indicated any significant persistence in volatility movements indicated by [[alpha].sub.1], + [beta] value. Only one industry indicated a significant impact of non-synchronous trading during this period as the estimated coefficient of moving average, (3, is significant for this industry.

As presented in Table 15, only five out of 22 industries have a significant relationship between risk and return during the first non-reform period. The coefficient of risk aversion varied between 0.000 to 0.232 percent per week across industries. Most of the industries with export and with foreign capital component indicated no significant relationship between risk and return. Locally owned industries with a domestic market indicated a significant relation between risk and return and a higher coefficient of risk aversion during this period. The spread of risk premia across industries is higher than the entire study period. Ten industries displayed a significant ARCH effect, which indicates that these industries have surprises and increased future volatility in returns. The ARCH effect coefficient values for all industries are less than one, which indicates no fat tailed return distribution or stationary process for unconditional variance of excess holding yields. Only textiles, fuel and energy, glass and cement industries have persistence in volatility movements during this period. These industries indicated a significant impact of non-synchronous trading in this period as the moving average component is significant.

In the second period of non-reform only four out of 22 industries indicated a significant relationship between risk and return, as risk aversion coefficients are significant for these industries (see Table 16). These industries are either export based or consist of infrastructural related firms. The magnitude of coefficients of risk aversion is higher than in the overall and the first sub-period of non-reform. Most locally owned industries indicated significant ARCH effects and had surprises and increased future volatility in returns. None of the industries indicated any fat tailed return distribution in unconditional variance of excess return holding yield. Only one industry indicated persistence in volatility movements in returns as the estimated coefficient of persistence, [[alpha].sub.1] + [beta], is large 0.936. This industry also indicated a significant impact of non-synchronous trading as the moving average component is significant with a estimated coefficient value of 0.826 during this period.

Again during the overall reform period, the pattern of the relationship between industry risk and return is the same as observed during the second sub-period of non-reform. Most locally owned industries indicated a significant relationship between risk and return in this period. Only four industries indicated a significant ARCH effect during the overall reform period (see Table 17). The domestically owned industries displayed an evidence of volatility clustering and indicated big surprises of either sign, which causes an increase in future volatility. The coefficient of the ARCH effect is less than one for all industries, which indicates no fat tailed distribution of unconditional variance of excess holding yield. Only one industry indicated persistence in volatility movements and also impact of non-synchronous trading during this period.

During the first sub-period of reform five industries indicated a significant risk and return relationship (see Table 18). These industries included growth and Islamic sector firms. As expected, 11 out of 22 industries indicated an ARCH effect or significant impact of conditional variance on excess return. Locally owned industries with a domestic market were more volatile than export or multinational dominated industries during the first period of reform. These industries displayed evidence of more surprises and increased future volatility in returns. The coefficient of ARCH effect is less than one in all cases means no evidence of unconditional variance of excess holding yield during this period. Only two industries indicated a high degree of persistence in volatility ([alpha].sub.1] + [beta] of 0.924 and 0.783). These industries also indicated a significant impact of non-synchronous trading in this period.

Only one industry had a significant relationship between risk and return during the second sub-period of reform. Five industries, mainly domestically owned, indicated a significant ARCH effect as the coefficients of volatility are significant for these industries and displayed evidence of surprises in their returns which cause an increase in future volatility in returns (see Table 19). The coefficient of ARCH is less than one in all cases and indicates no evidence of fat tailed unconditional variance of excess holding yield during this period. Only one industry indicated a high degree of persistence in volatility with estimated coefficient of persistence, [[alpha].sub.1], + [beta], as 0.869 in this period. Only two industries indicated significant impact of non-synchronous trading.

As summarised in Table 20, * the industry risk premia estimated through GARCH(1,1)-M process also indicated an upward shift after the liberal policies and financial reforms, particularly during the second sub-period of reforms. The industry portfolio analysis indicates that more industries have significant relationships between risk and return during the non-reform period, but the volatility in industry returns was lesser during this period. The volatility in industry returns was more evident during the reform period after the financial market was opened to foreign investors, particularly during the first period of reform. The impact of non-synchronous trading and the degree of persistence are only significant in a few industries during both the non-reform and reform period. Leverage at the industry level has been historically high in Pakistan, hence the consistent negative and significant relationships between return and volatility change are observed. In most cases, highly levered industries had a stronger negative relationship between return and volatility change than the less levered industries [Nishat (2001)].

5. SUMMARY AND CONCLUDING REMARKS

The above findings based on industry portfolios analysis also support the hypothesis that he opening financial markets resulted in an increase in price movements and higher risk premia rather than stabilising the stock prices during the reform period, particularly during the second sub-period of reforms. On average the industry returns are higher during the reform period than the non-reform period. However, only a few industries have statistically different average returns during the non-reform and reform periods. The higher return associated with higher risk phenomenon was stronger during the reform period than the non-reform period, particularly during the first sub-period of reforms. The results indicate that the industry characteristics prevailed during both the non-reform and reform periods. However, liberalised policies induced a higher risk premia in most industries.

The results suggest that when risk factor beta is allowed to be varying over time, more industries showed evidence of the theoretical relationship between risk and return prior to the reform period. The volatility in industry return was more pronounced during the reform period, particularly during the first period of reform. The impact of non-synchronous trading and the degree of persistence in volatility movements were significant only in few industries during both the non-reform and reform periods.

REFERENCES

De Gennaro. R. P. (1990) The Effect of Payment Delays on Stock Prices. Journal of Financial Research 13:2, 133-146.

De Santis, G., and S. Imrohoroglu (1997) Stock Returns and Volatility in Emerging Financial Markets. Journal of International Money and Finance 16:4, 561-79.

Engle, R. F. (1982) Autoregressive Conditional Heteroskedasticity with Estimates of the Variance of United Kingdom Inflations. Econometrica 50:4, 987-1008.

Engle, R. F., D. M Lilien, and R. P. Robins (1987) Estimating Time Varying Risk Premia in the Term Structure: The ARCH-M Model. Econometrica 50, 987-1007.

Fama, E. F., and K. R. French (1996) Multifactor Explanations of Asset Pricing Anomalies. Journal of Finance 51, 55-84.

French, K. R., G. W. Schwert, and R. F. Stambaugh (1987) Expected Stock Returns and Volatility. Journal of Financial Economics 19, 3-29.

Grinold, R., A. Rudd. and D. Slefek (1989) Global Factors: Fact or Friction. Journal of Portfolio Management 16:1, 79-89.

Harvey, C. R. (1995) Predictable Risk and Returns in Emerging Markets. Review of Financial Studies 8, 773-816.

Isimbabi, M. J. (1994) The Stock Market Perception of Industry Risk and the Separation of Banking and Commerce. Journal of Banking and Finance 18, 325-349.

Jensen, M. (1968) The Performance of Mutual Funds in the Period 1945-1964. Journal of Finance 23, 389-416.

Kale, J. K., N. H. Hakansson, and G. W. Piatt (1991) Industry vs. Other Factors in Risk Prediction. Walter A Haas School of Business Research Programme in Finance Working Paper Series, Institute of Business and Economic Research, University of California at Berkley. (Finance Working Paper No. 201.)

Khan, M. F. (1996) Cost of Capital for an Islamic Firm: The Case of Islamic Development Bank (IDB). Islamic Studies 4:1, 67-72.

Khan, M. S. (1993) The Securities Market in Pakistan. Karachi: Royal Book Company.

King, B. F. (1966) Market and Industry Factors in Stock Prices Behaviour. Journal of Business 39, 139-190.

Lessard, D. R. (1974) World, National, and Industry Factors in Equity Returns. Journal of Finance 29:2, 379-391.

Lessard, D. R. (1976) International Diversification. Financial Analyst Journal 32, 32-38.

Mandlebrot, B. (1963) The Variation of Certain Speculative Prices. Journal of Business 36:4, 394-419.

Nishat, M. (1999) The Impact of Institutional Development on Stock Prices in Pakistan. Unpublished Doctoral Dissertation, Auckland Business School, University of Auckland.

Nishat, M. (2000) Institutional Development and Risk Premia in Pakistan. Paper presented at Asia-Pacific Finance Association Conference 2000, held in Shanghai, China, July 2000.

Nishat, M. (2000) The Systematic Risk and Leverage Effect in the Corporate Sector of Pakistan. The Pakistan Development Review 39:4, 951-962.

Prager, R. A. (1989) Using Stock Price Data to Measure the Effects of Regulation: The Interstate Commerce Act and the Railroad Industry. Rand Journal of Economics 20:2, 280-293.

Roll, R. (1992) Volatility in US and Japanese Stock Market: A Symposium. Journal of Applied Corporate Finance 5:1, 25-35.

(1) These concessional loans consist of export financing of both pre-shipment and post shipment finances for locally manufactured machinery. The incentives and additional tax exemptions are part of the industrial package announced in most national budgets to boost the sectoral priorities to uplift the industries, which have both forward and backward linkages in the economy. During the reform period particularly, industrial packages were announced to boost the industries in rural areas. Special emphasis has been given on export oriented industries. Recently the power and energy sectors are on the priority list and have an additional fiscal incentive compared to other industries.

(2) For textiles the intercept is positive and statistically significant. The size of the coefficient of textile industry is much larger than the risk free return and could be interpreted as the support for zero-beta version only if the borrowing rate in the market is substantially higher by a very large margin. The intercept value suggests a borrowing rate of five times the risk free rate of return, which is not plausible at least during this period. Therefore, there may have been some degree of non-equilibrium pricing of textile stocks during this period [Jensen (1968)].

(3) The estimated results for CAPM and three factor model for 22 value weighted industy portfolios during overall and different non-reform and reform periods (the results are not provided but available on request) indicate that the three factor model does not provide any significant improvement over the CAPM model to explain the industry excess returns either during the non-reform or reform periods as the adj-[R.sup.2] are, in most cases, not different for both CAPM and three factor models.

* These tables are available with the author.

* The table is available with the author.

(4) In its standard form the ARCH model expresses the conditional variance as a linear function of past squared innovations; in markets where the price changes are innovations, large changes tend to be followed by large changes and small changes are followed by small changes of either sign [Mandlebrot (1963); Engle, Lilien, and Robins (1987)].

(5) These tables are available with the author.

* The table is available with the author.

Mohammad Nishat is Professor and Chairman, Department of Finance and Economics, Institute of Business Administration, University of Karachi, Karachi.

Table 1

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during overall period, January 1980-December 1994. The following
model gives the estimates for intercept and risk factor beta
[R.sub.it] - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.m] - [R.sub.ft]) + [[epsilon].sub.it] where [R.sub.it]
is the industry portfolio return, [R.sub.f] is the risk free
return on 6-month bond and [R.sub.m] is the return on
value-weighted market portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic 0.581 0.215
Inv Co. and Banks 0.487 0.097
Insurance 0.476 0.148
Textile 0.533 0.178 *
Woolen 0.273 0.086
Syn. and Rayon 0.685 0.242
Jute 0.157 -0.107
Sugar 0.374 0.087
Cement 0.637 0.214
Tobacco 0.373 0.089
Fuel and Energy 0.530 0.050
Engineering 0.587 0.173
Cab. and Electric. 0.521 0.173
Tran. and Comm. 0.525 -0.051
Chem and Pharm 0.491 0.057
Paper and Board 0.455 0.133
Vana. and Allied 0.447 0.198
Constructions 0.402 0.103
Leather and Tan. 0.726 0.513
Food and Allied 0.836 0.152
Glass and Cera. 0.552 0.210
Miscellaneous 0.493 0.168
Correlation (r, beta) (a) = 0.588

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic 0.545 6.694
Inv Co. and Banks 0.774 9.210
Insurance 0.575 7.294
Textile 0.661 15.082
Woolen 0.117 2.305
Syn. and Rayon 0.947 10.838
Jute 0.365 6.288
Sugar 0.441 9.466
Cement 0.883 12.716
Tobacco 0.433 5.421
Fuel and Energy 1.068 22.778
Engineering 0.855 11.258
Cab. and Electric. 0.638 7.315
Tran. and Comm. 1.378 13.035
Chem and Pharm 0.916 21.623
Paper and Board 0.554 7.876
Vana. and Allied 0.320 5.187
Constructions 0.481 4.321
Leather and Tan. 0.202 1.281
Food and Allied 1.731 13.197
Glass and Cera. 0.621 8.152
Miscellaneous 0.564 10.217
Correlation correlation (r, size) (b) =

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic 0.081 178091
Inv Co. and Banks 0.097 122203
Insurance 0.062 222649
Textile 0.227 486205
Woolen 0.004 14061
Syn. and Rayon 0.131 145647
Jute 0.046 35943
Sugar 0.102 212436
Cement 0.172 190665
Tobacco 0.034 65046
Fuel and Energy 0.403 728564
Engineering 0.140 125540
Cab. and Electric. 0.063 188156
Tran. and Comm. 0.179 233315
Chem and Pharm 0.326 626427
Paper and Board 0.072 85208
Vana. and Allied 0.031 26522
Constructions 0.021 8389
Leather and Tan. 0.001 92580
Food and Allied 0.183 577142
Glass and Cera. 0.077 56805
Miscellaneous 0.117 57115
Correlation 0.380

* Significant at 0.05 level.

** Significant at 0.10 level.

(a) Correlation between average industry return and the
respective risk factor beta.

(b) Correlation between average industry return and the
size (ME) of the industry.

Table 2

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during non-reform sub-period I, January 1980-June 1985. The
following model gives the estimates for intercept and risk
factor beta [R.sub.it] - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.mt] - [R.sub.ft]) + [[epsilon].sub.it] where [R.sub.it] is the
industry portfolio return, [R.sub.f] is the risk free return on
6-month bond and [R.sub.m] is the return on value-weighted market
portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic n.a n.a
Inv Co. and Banks 0.178 n.a
Insurance 0.556 0.289
Textile 0.240 0.026
Woollen 0.142 -0.072
Syn and Rayon 0.514 0.191
Jute 0.304 0.003
Sugar 0.412 0.144
Cement 0.422 -0.011
Tobacco 0.228 -0.074
Fuel and Energy 0.290 -0.035
Engineering 0.568 0.231
Cab and Electric. 0.201 -0.126
Tran and Comm 0.391 0.040
Chem and Pharm 0.345 0.004
Paper and Board 0.493 0.177
Vana and Allied 0.533 0.223
Constructions 0.245 -0.079
Leather and Tann -0.653 0.369
Food and Allied 0.521 0.022
Glass and Cera 0.266 -0.084
Miscellaneous 0.275 -0.026
Correlation (r, beta) (a) = 0.470

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic n.a n.a
Inv Co. and Banks n.a n.a
Insurance 0.547 3.319
Textile 0.534 8.027
Woollen 0.224 1.947
Syn and Rayon 0.883 6.226
Jute 0.746 8.355
Sugar 0.554 5.726
Cement 1.538 9.775
Tobacco 0.748 6.048
Fuel and Energy 0.888 12.611
Engineering 0.971 5.540
Cab and Electric. 0.895 9.455
Tran and Comm 1.040 7.394
Chem and Pharm 0.982 13.384
Paper and Board 0.837 5.303
Vana and Allied 0.804 5.919
Constructions 0.882 3.881
Leather and Tann 0.492 2.549
Food and Allied 1.932 9.163
Glass and Cera 1.037 8.656
Miscellaneous 0.746 5.963
Correlation correlation (r, size) (b) =

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic n.a n.a
Inv Co. and Banks n.a 9610
Insurance 0.034 69959
Textile 0.185 135421
Woollen 0.009 12173
Syn and Rayon 0.118 15923
Jute 0.198 33338
Sugar 0.102 112356
Cement 0.253 71930
Tobacco 0.112 38138
Fuel and Energy 0.361 324331
Engineering 0.095 64304
Cab and Electric. 0.240 166387
Tran and Comm 0.161 129773
Chem and Pharm 0.389 247712
Paper and Board 0.082 53455
Vana and Allied 0.108 19849
Constructions 0.041 15751
Leather and Tann 0.019 29546
Food and Allied 0.229 170084
Glass and Cera 0.209 26556
Miscellaneous 0.109 35201
Correlation 0.005

* Significant at 0.05 level.

** Significant at 0.10 level.

(a) Correlation between average industry return and the respective
risk factor beta.

(b) Correlation between average industry return and the size (ME)
of the industry.

Table 3

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during non-reform sub-period 11, July 1985-June 1988. The following
model gives the estimates for intercept and risk factor beta
[R.sub.it] - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.mt] - [R.sub.ft]) + [[epsilon].sub.it] where [R.sub.it]
is the industry portfolio return, [R.sub.f] is the risk free return
on 6-month bond and [R.sub.m] is the return on value-weighted market
portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic 0.630 0.469
Inv Co. and Banks 0.785 0.536
Insurance 0.229 -0.241
Textile 1.051 0.727 *
Woollen 0.179 0.028
Syn and Rayon 0.640 0.422
Jute 0.196 -0.040
Sugar 0.564 0.395
Cement 0.629 0.416
Tobacco 0.226 0.034
Fuel and Energy 0.404 0.132
Engineering 0.424 0.119
Cab and Electric. 0.587 0.354
Tran and Comm 0.614 0.187
Chem and Pharm 0.460 0.174
Paper and Board 0.220 -0.082
Vana and Allied 0.330 0.096
Constructions 0.029 -0.247
Leather and Tann 0.268 0.108
Food and Allied 0.305 -0.083
Glass and Cera 0.223 0.020
Miscellaneous 0.409 0.094
Correlation (r, beta) (a) = 0.047

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic 0.033 0.113
Inv Co. and Banks 0.541 1.475
Insurance 1.647 6.853
Textile 0.925 5.339
Woollen 0.019 0.204
Syn and Rayon 0.375 0.917
Jute 0.450 2.747
Sugar 0.118 0.588
Cement 0.347 1.924
Tobacco 0.227 1.058
Fuel and Energy 0.639 5.684
Engineering 0.815 3.639
Cab and Electric. 0.445 1.015
Tran and Comm 1.444 3.405
Chem and Pharm 0.712 5.727
Paper and Board 0.791 4.608
Vana and Allied 0.445 2.301
Constructions 0.651 2.523
Leather and Tann 0.065 0.304
Food and Allied 1.230 3.781
Glass and Cera 0.284 1.655
Miscellaneous 0.863 4.698
Correlation correlation (r, size) (b) =

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic 0.000 31197
Inv Co. and Banks 0.007 23382
Insurance 0.230 342106
Textile 0.151 254612
Woollen 0.000 17656
Syn and Rayon 0.001 23786
Jute 0.040 40775
Sugar 0.004 223400
Cement 0.017 171230
Tobacco 0.005 66104
Fuel and Energy 0.168 686148
Engineering 0.074 129548
Cab and Electric. 0.005 197243
Tran and Comm 0.064 279845
Chem and Pharm 0.171 687795
Paper and Board 0.116 85636
Vana and Allied 0.026 37438
Constructions 0.033 7454
Leather and Tann 0.000 57403
Food and Allied 0.079 244670
Glass and Cera 0.011 86544
Miscellaneous 0.120 54981
Correlation 0.006

* Significant at 0.05 level.

** Significant at 0.10 level.

(a) Correlation between average industry return and the respective
risk factor beta.

(b) Correlation between average industry return and the size (ME)
of the industry.

Table 4

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during overall reform period, July 1988-December 1994. The following
model gives the estimates for intercept and risk factor beta
[R.sub.it] - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.mt] - [R.sub.jt]) + [[epsilon].sub.it] where [R.sub.it] is
the industry portfolio return, [R.sub.f] is the risk free return on
6-month bond and [R.sub.m] is the return on value-weighted market
portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic 0.577 0.159
Inv Co. and Banks 0.605 0.009
Insurance 0.521 0.184
Textile 0.534 0.084
Woollen 0.427 0.254
Syn and Rayon 0.847 0.215
Jute 0.012 -0.237
Sugar 0.250 -0.102
Cement 0.819 0.327
Tobacco 0.564 0.262
Fuel and Energy 0.791 0.095
Engineering 0.676 0.149
Cab and Electric. 0.758 0.351
Tran and Comm 0.594 -0.247
Chem and Pharm 0.626 0.052
Paper and Board 0.531 0.193
Vana and Allied 0.428 0.223
Constructions 0.709 0.425
Leather and Tann 0.999 0.814
Food and Allied 1.351 0.391
Glass and Cera 0.949 0.568
Miscellaneous 0.715 0.365 *
Correlation (r, beta) (a) = 0.511

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic 0.607 8.431
Inv Co. and Banks 0.972 11.522
Insurance 0.433 4.420
Textile 0.665 11.735
Woollen 0.090 1.324
Syn and Rayon 1.047 10.832
Jute 0.245 2.859
Sugar 0.461 9.910
Cement 0.755 8.931
Tobacco 0.359 2.985
Fuel and Energy 1.180 16.400
Engineering 0.827 9.127
Cab and Electric. 0.579 5.692
Tran and Comm 1.478 10.295
Chem and Pharm 0.925 15.335
Paper and Board 0.433 4.792
Vana and Allied 0.157 2.150
Constructions 0.323 2.086
Leather and Tann 0.120 0.451
Food and Allied 1.729 8.693
Glass and Cera 0.526 4.501
Miscellaneous 0.459 7.343
Correlation correlation (r, size) (b)=

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic 0.178 251539
Inv Co. and Banks 0.282 242372
Insurance 0.053 271642
Textile 0.294 824163
Woollen 0.002 13898
Syn and Rayon 0.261 288819
Jute 0.021 35791
Sugar 0.228 275445
Cement 0.193 280989
Tobacco 0.024 83290
Fuel and Energy 0.449 1015099
Engineering 0.200 170081
Cab and Electric. 0.087 20438
Tran and Comm 0.242 293811
Chem and Pharm 0.416 853824
Paper and Board 0.063 107775
Vana and Allied 0.011 27655
Constructions 0.010 3500
Leather and Tann 0.000 152482
Food and Allied 0.179 1004020
Glass and Cera 0.055 65155
Miscellaneous 0.139 72856
Correlation 0.369

* Significant at 0.05 level.

(a) Correlation between average industry return and the respective
risk factor beta.

(b) Correlation between average industry return and the size (ME)
of the industry.

Table 5

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during reform sub-period 1, July 1988-June 1988. The following
model gives the estimates for intercept and risk factor beta
[R.sub.it] - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.mt] - [R.sub.ft]) + [[epsilon.].sub.it], where [R.sub.it],
is the industry portfolio return, [R.sub.f] is the risk free
return on 6-month bond and R", is the return on value-weighted
market portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic 0.855 0.580 *
Inv Co. and Banks 0.341 0.025
Insurance 0.141 0.135
Textile 0.441 0.184
Woollen 0.175 0.011
Syn and Rayon 0.705 0.356
Jute 0.098 -0.162
Sugar 0.273 0.006
Cement 0.157 -0.154
Tobacco 0.382 0.175
Fuel and Energy 0.521 0.194
Engineering 0.428 0.100
Cab and Electric. 0.171 -0.063
Tran and Comm 0.021 -0.362
Chem and Pharm 0.512 0.119
Paper and Board 0.253 -0.038
Vana and Allied 0.412 0.251
Constructions 0.838 0.650
Leather and Tann 0.288 0.128
Food and Allied 1.932 0.965
Glass and Cera 0.676 0.394
Miscellaneous 0.410 0.198
Correlation (r, beta) (a) = 0.765

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic 0.504 3.526
Inv Co. and Banks 0.699 3.618
Insurance 0.515 3.442
Textile 0.599 6.130
Woollen 0.019 0.258
Syn and Rayon 0.864 4.819
Jute 0.447 2.796
Sugar 0.481 4.716
Cement 0.676 4.965
Tobacco 0.215 1.014
Fuel and Energy 0.754 7.144
Engineering 0.758 5.790
Cab and Electric. 0.331 1.758
Tran and Comm 0.803 4.300
Chem and Pharm 1.053 10.856
Paper and Board 0.590 2.844
Vana and Allied 0.011 0.091
Constructions 0.099 0.248
Leather and Tann 0.002 0.014
Food and Allied 3.669 6.837
Glass and Cera 0.559 2.244
Miscellaneous 0.238 2.189
Correlation correlation (r, size) (b) =

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic 0.076 193294
Inv Co. and Banks 0.074 142553
Insurance 0.066 357634
Textile 0.000 857947
Woollen 0.000 18792
Syn and Rayon 0.128 172191
Jute 0.043 47004
Sugar 0.123 367002
Cement 0.135 288770
Tobacco 0.004 110584
Fuel and Energy 0.249 1125247
Engineering 0.177 184998
Cab and Electric. 0.019 232540
Tran and Comm 0.104 347863
Chem and Pharm 0.436 876945
Paper and Board 0.044 118144
Vana and Allied 0.000 38608
Constructions 0.000 4821
Leather and Tann 0.000 264708
Food and Allied 0.232 969735
Glass and Cera 0.025 87472
Miscellaneous 0.024 84076
Correlation 0.382

* Significant at 0.05 level.

** Significant at 0.10 level.

(a) Correlation between average industry return and the
respective risk factor beta.

(b) Correlation between average industry return and the size
(ME) of the industry.

Table 6

Industry Portfolio Average Weekly Returns and Other Characteristics

This table presents the average weekly returns and other
characteristics for value-weighted industry portfolios of the KSE
during reform sub-period 11, July 1991-December 1994. The following
model gives the estimates for intercept and risk factor beta
[R.sub.it], - [R.sub.ft] = [[alpha].sub.i] + [[beta].sub.i]
([R.sub.mt] - [R.sub.ft]) + [[epsilon].sub.it], where [R.sub.it]
is the industry portfolio return, [R.sub.f] is the risk free return
on 6-month bond and R", is the return on value-weighted market
portfolio.

 Mean Weekly
Industry Portfolio Return (%) [[alpha].sub.i]

Islamic 0.334 0.000
Inv Co. and Banks 0.830 0.001
Insurance 0.846 0.465
Textile 0.610 0.028
Woollen 0.641 0.473
Syn and Rayon 0.964 0.094
Jute -0.061 -0.308
Sugar 0.228 -0.199
Cement 1.383 0.758 *
Tobacco 0.718 0.344
Fuel and Energy 1.020 0.018
Engineering 0.885 0.194
Cab and Electric. 1.258 0.726
Tran and Comm 1.121 0.000
Chem and Pharm 0.720 -0.012
Paper and Board 0.766 0.395
Vana and Allied 0.438 0.202
Constructions 0.603 0.232
Leather and Tann 1.605 1.427
Food and Allied 0.844 -0.170
Glass and Cera 1.178 0.722
Miscellaneous 0.972 0.520 *
Correlation (r, beta) (a) = 0.275

Industry Portfolio [[beta].sub.i] t([[beta].sub.i])

Islamic 0.649 7.507
Inv Co. and Banks 1.037 11.087
Insurance 0.400 3.003
Textile 0.684 9.194
Woollen 0.096 0.979
Syn and Rayon 1.096 8.914
Jute 0.201 1.850
Sugar 0.462 8.573
Cement 0.752 6.693
Tobacco 0.389 2.485
Fuel and Energy 1.284 13.177
Engineering 0.841 6.720
Cab and Electric. 0.619 4.813
Tran and Comm 1.632 8.134
Chem and Pharm 0.898 11.111
Paper and Board 0.386 3.847
Vana and Allied 0.192 1.995
Constructions 0.385 2.558
Leather and Tann 0.118 0.292
Food and Allied 1.299 8.177
Glass and Cera 0.511 3.701
Miscellaneous 0.503 6.215
Correlation Correlation (r, size) (b) =

 Mkt. Cap.
Industry Portfolio [[bar.R].sup.2] (Rs Mill)

Islamic 0.244 309784
Inv Co. and Banks 0.413 342193
Insurance 0.049 185651
Textile 0.326 790379
Woollen 0.005 9004
Syn and Rayon 0.312 405448
Jute 0.019 24578
Sugar 0.296 183888
Cement 0.204 273207
Tobacco 0.034 55995
Fuel and Energy 0.498 904951
Engineering 0.205 155165
Cab and Electric. 0.117 174335
Tran and Comm 0.274 239759
Chem and Pharm 0.414 830704
Paper and Board 0.078 97405
Vana and Allied 0.022 16703
Constructions 0.036 2180
Leather and Tann 0.000 40257
Food and Allied 0.276 1038305
Glass and Cera 0.073 42837
Miscellaneous 0.181 61635
Correlation 0.051

* Significant at 0.05 level.

** Significant at 0.10 level.

(a) Correlation between average industry return and the respective
risk factor beta.

(b) Correlation between average industry return and the size (ME)
of the industry.