## _wp05182 - 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades

## Source details

**Canonical URL:** [_wp05182 - 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05182.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05182.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05182.pdf.json)

---

### Context and research question
- November 2001 EU enlargement announcement coincided with large stock price rises in eight Central and Eastern European accession countries: between then and July 2004, stock prices in the eight Central and Eastern European candidate countries increased in dollar terms by over 90 percent on average.
- World market index return over the same period: about 8 percent.
- Average return in local currency: 65 percent.
- Research question: Did the rise reflect repricing of systematic risk (segmentation → integration) or changes in expected future dividends?

### Theoretical measures of repricing (formulas and intuition)
- Segmentation required return: ki = rf + βiMλM, where βiM = cov(Ri, RM)/var(RM).
- Full integration required return: ki* = rf* + βiWλW, where βiW = cov(Ri, RW)/var(RW).
- DIFCOV (equation (3) specification): DIFCOV = cov(Ri, RM) − cov(Ri, RW). Firms with large local covariance relative to world covariance should see larger price increases upon integration.
- DIFBETA (equation (4) specification): DIFBETA = βiM − βiW under λW = λM = λ. Firms with larger DIFBETA should see larger price increases upon integration.
- Empirical note: average sample variance of local market returns is nearly nine times the sample variance of world returns (ratio = nine to one), motivating DIFCOV if market premium ∝ variance of market returns.

### Dating integration and empirical window
- Baseline integration date: November 2001 (motivated by November 2001 European Commission progress report and media).
- Baseline integration window length: sixteen months.
- Baseline price change calculation: November 2001 to February 2003.
- Authors acknowledge integration may be gradual; analysis jointly tests that integration occurred in months after the 2001 announcement and that markets price stocks according to systematic risk.

### Data, sample, and measurement
- Return data sources: Emerging Markets Database (EMDB) for firm-level returns (dividend-inclusive); S&P global total return indices for local market indices (log differences); MSCI world equity index for world market returns (log differences).
- Currency: current U.S. dollars using current exchange rates.
- Covariances and betas calculated using 36 months of historical data; results qualitatively similar with up to 60 months.
- Minimum data requirements: firms must have data back to at least October 1999 and remain until February 2003.
  - Original EMDB sample: 410 firms.
  - Eliminated by restrictions: 310 firms.
  - Further eliminated for data problems: 26 firms.
  - Eliminated for no earnings estimates: 23 firms.
  - Dropped for suspect price data: 2 firms.
  - Dropped due to bankruptcy: 1 firm.
  - Final sample: 74 firms.
- Countries: eight accession countries (Czech Republic, Hungary, Poland, Slovakia, Slovenia, Lithuania, Latvia, Estonia) and three control countries (Russia, Romania, Croatia).

### Empirical descriptive findings (selected statistics preserved exactly)
- Sample size after filtering: 74 firms.
- Country-level percentage stock price increases (price increase = percentage stock price increase between November 2001 and February 2003):
  - Poland: 0.16
  - Hungary: 0.11
  - Czech Republic: 0.43
  - Slovenia: 0.63
  - Slovakia: 0.61
  - Lithuania: 0.39
  - Latvia: 0.28
  - Estonia: 0.46
  - Romania: 0.54
  - Russia: 0.37
  - Croatia: 0.55
  - Total: 0.36
- Table 3 descriptive statistics (exact values):
  - Price Change: Mean 0.356; Median 0.413; Min -0.462; Max 1.085; Std. Dev. 0.366
  - DIFCOV: Mean 0.007; Median 0.006; Min -0.001; Max 0.030; Std. Dev. 0.006
  - DIFBETA: Mean 0.237; Median 0.234; Min -1.404; Max 2.091; Std. Dev. 0.761
  - Net Upgrades: Mean -0.682; Median 0.000; Min -11.000; Max 10.667; Std. Dev. 3.233
  - Size: Mean 0.111; Median 0.047; Min 0.001; Max 0.824; Std. Dev. 0.152
  - Turnover: Mean 37.414; Median 3.258; Min 0.005; Max 246.225; Std. Dev. 65.632
- DIFCOV: average positive for all 11 countries; positive for all but two firms in firm-level sample.
- DIFBETA: average positive for all but three countries (Poland, Hungary, Russia); some firms have negative DIFBETA despite positive DIFCOV due to relatively high variance of local market returns in Poland, Hungary, Russia.
- Net upgrades (analysts’ expected earnings revisions):
  - Country patterns: Poland, Hungary, Slovenia, Russia, Croatia—more analysts lowered than increased estimates; Czech Republic, Slovakia, Romania, Baltic countries—more analysts upped estimates.
  - Overall average: On average there were only 0.7 upward revisions for every downward revision (Net Upgrades Mean -0.682).
- Authors note surprise that analysts were not broadly optimistic given large stock price increases, suggesting price increases more likely from reductions in discount rates than increases in expected dividends.

### Control variables and data features (constructed following Chari and Henry)
- Size: ratio of firm’s average market capitalization to average market capitalization of entire local market during the 12 months prior to integration date. Range: 0.1 percent to 82 percent.
- Turnover: dollar value traded during 12 months prior to liberalization date divided by firm’s average market capitalization during same period. Range: 0.5 percent to 24,600 percent.
- Forecast data limitations: interim forecasts, long-term growth forecasts and secondary forecasts eliminated due to mostly being unavailable. Unavailability of long-term growth forecasts excludes part of economic impact of integration that may occur only after actual accession.

### Estimation — empirical strategy
- Objective: estimate relationship between percentage stock price increase (Nov 2001–Feb 2003) and measures of change in systematic risk: DIFCOV and DIFBETA.
- Expectation: positive relationship for accession countries (EU dummy = 1); insignificant for nonaccession countries (EU dummy = 0).
- Interaction terms: DIFCOV·EU and DIFBETA·EU used to test whether EU enlargement leads to repricing of systematic risk.
- Standard errors: heteroskedasticity robust, assuming independence across countries but not within.

### Baseline regressions — key findings (Tables 4 panels A and B)
- DIFCOV panel (panel a):
  - Intercept: positive and highly significant — firms with no change in systematic risk could expect about a 56 percent increase in stock prices following the November 2001 announcement.
  - EU dummy: insignificant when controlling for DIFCOV.
  - DIFCOV and DIFCOV·EU: statistically insignificant across specifications.
  - Net upgrades: coefficients positive and statistically significant in specifications (2a) through (4a).
  - Size and turnover: statistically insignificant in baseline specifications.
- DIFBETA panel (panel b):
  - Intercept: large and highly significant — firms with no reduction in systematic risk as measured by DIFBETA experienced large price increases.
  - DIFBETA: statistically insignificant on its own.
  - DIFBETA·EU: positive and statistically significant at 1 percent — indicates repricing of systematic risk in EU accession countries but not in nonaccession countries.
  - DIFBETA·EU significance persists after controlling for expected earnings, country effects, size and turnover.
  - Net upgrades: significant at the 1 percent level in all specifications.
  - Size and turnover: insignificant.
- Magnitude interpretation:
  - Coefficient on DIFBETA·EU ranges from about 0.3 to 0.4.
  - Average DIFBETA = 0.24.
  - Implied price increase from DIFBETA effect: 0.35 · 0.24 = 0.08 (about 8 percent).
  - Average price increase: about 36 percent.
  - Conclusion: difference in betas explains, on average, about 22 percent of the price increase during the integration window (8 percent of 36 percent).
- Selected coefficients and model summaries (exact values as reported in Table 4):
  - Intercept Panel A (1a): 0.560** (4.48)
  - Intercept Panel B (1b): 0.457** (7.42)
  - EU dummy Panel B (1b): -0.201* (-2.65)
  - DIFBETA * EU Panel B (2b): 0.349** (3.31)
  - Net Upgrades example: 0.039** (3.35) in (1a)
  - R-squared Panel A: 0.090, 0.205, 0.329, 0.345
  - R-squared Panel B: 0.296, 0.356, 0.398, 0.403
  - Number of observations: 74. T-statistics in parentheses; * and ** indicate significance at 5 and 1 percent.

### Robustness checks and diagnostics
- Scatter plots (Figure 2 summaries):
  - DIFCOV vs. stock price change: weak relationship for accession and nonaccession countries.
  - DIFBETA vs. stock price change: relatively strong positive relationship for accession countries; much weaker for nonaccession countries.
  - Net upgrades vs. stock price change: positive for both accession and nonaccession countries.
  - Outliers: relationships are not driven by outliers.
- Varying integration date and window length (Table 5 and specification (4b)):
  - Shifting window one month forward/back: DIFBETA remains statistically significant for accession countries.
  - Using November 2000 as start (one year before announcement): DIFBETA insignificant for both accession and nonaccession countries; net upgrades remain significant → DIFBETA significance linked to announcement period.
  - Shortening window:
    - 16 months → 10 months: DIFBETA significant at 5 percent; net upgrades positive but marginally insignificant.
    - Reducing to 4 months: DIFBETA no longer significant → integration effects evolve over longer horizons, consistent with gradual integration.
  - Comparison: Chari and Henry (2004) find significance with 1–2 month windows in other regions; difference attributed to gradual nature of EU accession integration.
- High-beta upmarket check (Table 6):
  - Regressions of price changes on local market beta and local beta·EU: coefficients insignificant.
  - Net upgrades: significant.
  - Interpretation: results driven by changes in systematic risk (DIFBETA), not by high-beta stocks simply performing differently in an upmarket.

### Limitations, interpretation, and policy implication
- Limitations and cautions:
  - Small sample: 74 observations and 12 explanatory variables (including country effects) limit degrees of freedom and coefficient precision.
  - Reliance on historical betas: assumes investors treat historical betas as accurate guides to future betas despite structural changes.
  - Timing of integration: uncertainty complicates estimation of repricing effects.
- Interpretation of investor behavior:
  - Evidence consistent with investors applying CAPM mechanically: discounting future cash flows with local betas prior to the announcement and world betas after the announcement.
  - Limited historical data in these markets may lead investors to avoid estimating market premia via historical variances, explaining why DIFCOV is less informative.
- Policy implication:
  - Findings support further capital market integration: after the enlargement announcement, investors re-valued firms according to their systematic risk, and firms benefited from integration according to their capacity to diversify risk for global investors.

### Tables and figures (reported exactly as in source)
- Table 1: Opening Dates and Legal Restrictions (selected country entries preserved in source).
- Table 2: Means of Key Variables by Country (selected exact values reproduced in empirical descriptive findings).
- Table 3: Descriptive Statistics (exact summary statistics reproduced above).
- Table 4: Baseline Regression Results (selected exact coefficients and R-squared values reproduced above).
- Table 5: Varying Integration Date and Window (selected exact coefficient examples and R-squared reproduced above).
- Table 6: Price Changes and Beta (selected intercepts, net upgrades and BETAM * EU coefficient examples and R-squared reproduced above).
- Figure 1: Stock Market Indices in Accession Countries (“Total U.S. dollar return indices from the Emerging Markets Database for each country are scaled to equal 100 in January 1999.”)
- Figure 2: Scatter Plots of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades (“Source: Authors’ calculations.”)

*Source: _wp05182 - 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades (IMF working paper PDF).*

### References..............................................................................................................

### _wp05182 - References................................................................................................................................18

### Tables
- 1. Opening Dates and Legal Restrictions.................................................................................22
- 2. Means of Key Variables by Country ...................................................................................23
- 3. Descriptive Statistics............................................................................................................24
- 4. Baseline Regression Results ................................................................................................25
- 5. Varying Integration Date and Window ...............................................................................26
- 6. Price Changes and Beta .......................................................................................................27

### Figures
- 1. Stock Market Indices in Accession Countries .....................................................................20

*Source: _wp05182 - References..............................................................................................................*

### 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV,

### 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades

### Context and research question
- November 2001 announcement of the European Union (EU) enlargement coincided with a large rise in stock prices in the eight Central and Eastern European accession countries: between then and July 2004, stock prices in the eight Central and Eastern European candidate countries increased in dollar terms by over 90 percent on average.  
- World market index return over the same period: about 8 percent.  
- Average return in terms of local currency: 65 percent.  
- Research question: Did the rise reflect repricing of systematic risk because of increased integration of local stock markets with the world market (a move from segmentation to integration), or did it reflect changes in expected future dividends?

### Theoretical measures of repricing (key formulas and intuition)
- Under segmentation, required return: ki = rf + βiMλM, where βiM = cov(Ri, RM)/var(RM).  
- Under full integration, required return: ki* = rf* + βiWλW, where βiW = cov(Ri, RW)/var(RW).  
- DIFCOV (equation (3) specification): change in required return depends on the change in the risk-free rate and DIFCOV = cov(Ri, RM) − cov(Ri, RW). Empirical interpretation: firms with large local covariance relative to world covariance should see larger price increases upon integration.  
- DIFBETA (equation (4) specification): alternative measure assuming a constant market premium (λW = λM = λ). Change in required return depends on DIFBETA = βiM − βiW. Empirical interpretation: firms with larger DIFBETA should see larger price increases upon integration.  
- The authors note a theoretical argument for DIFCOV based on market premium proportional to the variance of market returns and report that the average sample variance of local market returns is nearly nine times the sample variance of world returns (ratio = nine to one), implying a potentially large drop in market premium upon integration if proportionality holds. They also note practitioners commonly use rule-of-thumb market premium estimates (e.g., consensus U.S. market premium ≈ 7 percent from Welch (2000)) and therefore treat DIFBETA as a relevant alternative.

### Dating integration and empirical window
- Baseline dating: November 2001 is treated as the beginning of the window in which integration occurred (motivated by the November 2001 European Commission progress report and associated media coverage).  
- Baseline integration window length: sixteen months.  
- Change in stock prices is calculated over the integration window from November 2001 to February 2003.  
- Authors acknowledge dating integration is difficult and integration may be gradual; analysis can be viewed as a joint test that integration occurred in the months after the 2001 announcement and that markets price stocks according to their systematic risk.

### Data, sample, and measurement
- Return data sources: Emerging Markets Database (EMDB) for firm-level returns (dividend-inclusive), S&P global total return indices for aggregate local market indices (log differences), and MSCI world equity index for world market returns (log differences).  
- Currency: All returns and price changes calculated in current U.S. dollars using current exchange rates (assumes global investor approximated by dollar-based investor).  
- Covariances and betas calculated using three years of historical data; authors note results are qualitatively similar when allowing up to five years.  
- Minimum data requirements: firms must have data back to at least October 1999 and remain until February 2003. Original EMDB sample: 410 firms; 310 firms eliminated by these restrictions; further 26 eliminated for data problems; additional 23 eliminated for no earnings estimates; 2 dropped for suspect price data; 1 dropped due to bankruptcy. Final sample: 74 firms.  
- Countries in study: eight Central and Eastern European accession countries (Czech Republic, Hungary, Poland, Slovakia, Slovenia, Lithuania, Latvia, Estonia) plus three control Eastern European countries (Russia, Romania, Croatia).

### Empirical descriptive findings (selected statistics preserved exactly)
- Sample size after filtering: 74 firms.  
- Average percentage stock price increase during the integration window (by country): shown in Table 2 (text summary: stock prices went up substantially in the eight countries as well as Romania, Russia, and Croatia; highest increase in Slovenia, with stock prices rising 62 percent; lowest increases occurred in Poland and Hungary).  
- For the full sample of firms (Table 3 summary): average share price went up 36 percent in the 15 months following the integration, ranging from a 46 percent decrease to a 108 percent increase.  
- DIFCOV: average DIFCOV is positive for all 11 countries (empirical counterpart to cov(Ri, RM) − cov(Ri, RW)); DIFCOV is positive for all but two firms in the firm-level sample.  
- DIFBETA: average DIFBETA is positive for all but three countries (exceptions: Poland, Hungary, and Russia). For some firms DIFBETA is negative even though DIFCOV is typically positive. Explanation offered: relatively high variance of local market returns in Poland, Hungary, and Russia makes local betas small relative to world betas despite local covariance exceeding world covariance.  
- Net upgrades (analysts’ expected earnings revisions): measured as number of analysts who increased estimates minus number who lowered estimates, averaged across forecast periods. Country-level patterns: in Poland, Hungary, Slovenia, Russia, and Croatia more analysts lowered rather than increased earnings estimates; in the Czech Republic, Slovakia, Romania, and the Baltic countries more analysts upped their estimates. Overall average: On average there were only 0.7 upward revisions for every downward revision. Authors note surprise that analysts were not more optimistic during the period of dramatic stock price increases, suggesting price increases may be more likely due to reductions in the discount rate than to increases in expected dividends.

### Methodological choices emphasized
- Use of DIFCOV as theoretically appealing measure (accounts for possible change in market premium upon integration), but recognition that DIFBETA is useful because practitioners price by betas and because estimating market premium via sample variances (local vs world) may be unreliable given short historical series and the observed nine-to-one variance ratio.  
- Control for changes in expected dividend growth via analysts’ earnings estimates (IBES) rather than actual dividend/earnings changes, arguing this directly captures investors’ expectations at announcement time and differs from prior firm-level studies that used realized earnings or dividends.

### Contribution and interpretation
- The paper follows firm-level strategies of prior work (Errunza and Miller (2000), Chari and Henry (2004)) but in a different setting: an increase in actual integration linked to EU enlargement rather than removal of legal restrictions.  
- Novelty points emphasized by authors: (i) explicit control for changes in expected earnings using IBES data; (ii) testing repricing effects in accession countries where legal openness predated the enlargement announcement but investor behavior may have changed with clearer EU prospects; (iii) using both DIFCOV and DIFBETA as complementary measures of changes in systematic risk.  
- Authors frame the analysis as informative for assessing benefits of EU integration (lower cost of capital from reduced risk-free rates and systematic risk), for testing whether markets price changes in systematic risk, and for using EU enlargement as a natural experiment on asset pricing.

*Source: _wp05182 - 2. Scatter Plot of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades, IMF working paper PDF.*

### 11. Following Chari and Henry, we construct a number of additional control variables.

### _wp05182 - 11. Following Chari and Henry, we construct a number of additional control variables.

### Control variables and data features
- Size: ratio of a firm’s average market capitalization to average market capitalization of the entire local market during the 12 months prior to the integration date.
  - Range: 0.1 percent to 82 percent.
- Turnover: dollar value traded during the 12 months prior to the liberalization date divided by the firm’s average market capitalization during the same time period.
  - Range: 0.5 percent to 24,600 percent.
- Forecast data: interim forecasts, long-term growth forecasts and secondary forecasts are eliminated due to mostly being unavailable.
  - Consequence: unavailability of long-term growth forecasts means part of the economic impact of integration—likely to come only after actual accession—is excluded from the analysis.

### Estimation — empirical strategy
- Objective: estimate relationship between the increase in stock prices following the enlargement announcement and two measures of changes in systematic risk: DIFCOV and DIFBETA.
- Expectation:
  - Positive relationship for the EU accession countries (EU dummy = 1 for eight accession countries).
  - Insignificant relationship for three nonaccession countries (EU dummy = 0).
- Interaction terms: DIFCOV·EU and DIFBETA·EU used to test whether EU enlargement (integration) leads to repricing of systematic risk.
- Dependent variable in baseline regressions: percentage stock price increase from November 2001 until February of 2003.
- Standard errors: heteroskedasticity robust, assuming independence across countries but not within.

### Baseline regressions — key findings
- DIFCOV panel (panel a):
  - Intercept: positive and highly significant — firms with no change in systematic risk could expect about a 56 percent increase in stock prices following the November 2001 announcement.
  - EU dummy: insignificant — controlling for DIFCOV, accession countries did not experience greater price increases than nonaccession countries.
  - DIFCOV and DIFCOV·EU: statistically insignificant throughout (including when controlling for expected earnings, country fixed effects, size and turnover).
  - Net upgrades: coefficients positive and statistically significant in specifications (2a) through (4a).
  - Size and turnover (when included): statistically insignificant in baseline specifications.
- DIFBETA panel (panel b):
  - Intercept: large and highly significant — firms with no reduction in systematic risk as measured by DIFBETA experienced large price increases.
  - DIFBETA: statistically insignificant on its own.
  - DIFBETA·EU: positive and statistically significant at 1 percent — indicates repricing of systematic risk in EU accession countries but not in nonaccession countries.
  - Significance of DIFBETA·EU persists after controlling for changes in expected earnings, country effects, size and turnover.
  - Net upgrades: significant at the 1 percent level in all specifications.
  - Size and turnover: insignificant.
- Magnitude interpretation:
  - Coefficient on DIFBETA·EU ranges from about 0.3 to 0.4.
  - Average DIFBETA = 0.24.
  - Implied price increase from DIFBETA effect: 0.35 · 0.24 = 0.08 (about 8 percent).
  - Average price increase: about 36 percent.
  - Conclusion: difference in betas explains, on average, about 22 percent of the price increase during the integration window (8 percent of 36 percent).

### Robustness checks and diagnostics
- Scatter plots (Figure 1):
  - DIFCOV vs. stock price change: weak relationship for both accession and nonaccession countries.
  - DIFBETA vs. stock price change: relatively strong positive relationship for accession countries; much weaker for nonaccession countries.
  - Net upgrades vs. stock price change: positive for both accession and nonaccession countries.
  - Outliers: relationships are not driven by outliers.
- Varying integration date and window length (specification (4b) from Table 4):
  - Shifting window one month forward (October 2001–January 2003) or one month back (December 2001–March 2003): DIFBETA remains statistically significant for accession countries.
  - Using November 2000 as start (one year before announcement): DIFBETA is insignificant for both accession and nonaccession countries; net upgrades remain significant. Interpretation: DIFBETA significance is linked to the integration announcement period.
  - Shortening integration window:
    - From 16 months (November 2001–February 2003) to 10 months (November 2001–August 2002): DIFBETA statistically significant at the 5 percent level; net upgrades positive but marginally insignificant.
    - Reducing to 4 months: DIFBETA no longer significant. Interpretation: integration effects evolve over longer horizons, consistent with gradual market integration rather than instantaneous repricing.
  - Comparison with Chari and Henry (2004): they find significance with 1–2 month windows in other regions; differences attributed to gradual nature of EU accession integration versus legal barrier removals in Latin America and Asia.
- High-beta upmarket check (Table 6):
  - Regressions of price changes on local market beta and local beta·EU: coefficients insignificant.
  - Net upgrades: significant.
  - Interpretation: results are driven by changes in systematic risk (DIFBETA), not by high-beta stocks simply performing differently in an upmarket.

### Conclusion — synthesis and implications
- Two tested measures of change in systematic risk:
  - DIFCOV (difference in covariances): does not appear related to stock price changes.
  - DIFBETA (difference in betas): explains part of the stock price increase; stocks with high local beta but low world beta experienced higher price increases.
- Expected earnings:
  - Changes in expected earnings are consistently and positively related to stock price changes; upward revisions in expected earnings increase stock prices.
- Interpretation of investor behavior:
  - Evidence consistent with investors applying CAPM mechanically: discounting future cash flows with local betas prior to the announcement and world betas after the announcement.
  - Limited historical data in Central and Eastern European markets may lead investors to avoid estimating market premia via historical variances, explaining why covariances (DIFCOV) are less informative.
- Limitations and cautions:
  - Small sample: 74 observations and 12 explanatory variables (including country effects) limit degrees of freedom and coefficient precision.
  - Reliance on historical betas assumes investors treat historical betas as accurate guides to future betas, which may be strong given ongoing structural changes in the economies.
  - Timing of integration: considerable uncertainty about timing complicates estimating repricing effects.
- Policy implication:
  - Findings support further capital market integration: following the enlargement announcement, investors re-valued firms according to their systematic risk, and firms benefited from integration according to their capacity to diversify risk for the global investor.

*Source: _wp05182 (excerpt).*

### References

### _wp05182 - References

### Major cited works
- Adjounte, Kpate, and Jean-Pierre Danthine, 2003, “European Financial Integration and Equity Returns,” The Transformation of the European Financial System, ed. by V. Gaspar and others (Frankfurt: European Central Bank).
- Baele, Lieven, Annalisa Ferrando, Peter Hordahl, Elizaveta Krylova, and Cyril Monnet, 2004, “Measuring Financial Integration in the Euro Area,” Occasional Paper Series No. 14 (Frankfurt: European Central Bank).
- Bekaert, Geert, and Campbell Harvey, 1995, “Time-varying World Market Integration,” Journal of Finance, Vol. 50, No. 2, pp. 403–44.
- Bekaert, Geert, and Campbell Harvey, 2000, “Foreign Speculators and Emerging Equity Markets,” Journal of Finance, Vol. 55, No. 2, pp. 565–613.
- Bekaert, Geert, Campbell R. Harvey, and Christian T. Lundblad, 2003, “Equity Market Liberalization in Emerging Markets,” The Journal of Financial Research, Vol. 26, No. 3, pp. 275–299.
- Chari, Anusha, and Peter Blair Henry, 2004, “Risk Sharing and Asset Prices: Evidence From a Natural Experiment,” Journal of Finance, Vol. 59, No. 3, pp. 1295–1324.
- Clauss, James, and Jacob Thomas, 2001, “Equity Premia as Low as Three Percent? Evidence from Analysts’ Earnings Forecasts for Domestic and International Stocks,” Journal of Finance, Vol. 56, No. 5, pp. 1629–1666.
- Errunza, Vihang R., and Darius P. Miller, 2000, “Market Segmentation and the Cost of Capital in International Equity Markets,” Journal of Financial and Quantitative Analysis, Vol. 35, No. 4, pp. 577–600.
- Fama, Eugene F., and Kenneth R. French, 2002, “The Equity Premium,” Journal of Finance, Vol. 57, No. 2, pp. 637–59.
- Henry, Peter Blair, 2000, “Stock Market Liberalization, Economic Reform, and Emerging Market Equity Prices,” Journal of Finance, Vol. 55, No. 2, pp. 529–64.
- Rouwenhorst, K. Geert, 1999, “European Equity Markets and the EMU,” Financial Analysts Journal, Vol 55, pp. 57–64.
- Sentana, Enrique, 2002, “Did EMS reduce the cost of capital?,” The Economic Journal, Vol. 112, pp. 786–809.
- Siegel, Jeremy J., and Richarch H. Thaler, 1997, “The equity risk premium puzzle,” Journal of Economic Perspectives, Vol. 11, pp. 191–200.
- Stulz, Rene, 1999, “Globalization of Equity Markets and the Cost of Capital,” NBER Working Paper No. 7021 (Cambridge, MA: National Bureau of Economic Research).
- Welch, Ivo, 2000, “Views of Financial Economists on the Equity Premium and on Professional Controversies,” Journal of Business, Vol. 73, No. 4, pp. 501–530.
- Additional working papers and unpublished manuscripts cited include Hardouvelis, Malliaropulos, and Priestley, 2004; Christoffersen, Chung, and Errunza, 2002; and others.

### Figures reported in source
- Figure 1. Stock Market Indices in Accession Countries
  - “Total U.S. dollar return indices from the Emerging Markets Database for each country are scaled to equal 100 in January 1999.”
- Figure 2. Scatter Plots of the Relationship Between Stock Price Changes and DIFCOV, DIFBETA, and Net Upgrades
  - “Source: Authors’ calculations.”

### Table summaries and key statistics

- Table 1. Opening Dates and Legal Restrictions
  - Presents Stock Market Established, Restrictions Lifted, Start of EMDB Coverage, First ADR for listed countries (examples preserved exactly as shown):
    - Czech Republic: June 1992; September 1994 1/; January 1994; June 1995
    - Hungary: July 1990; 1996 2/; December 1992; July 1992
    - Poland: January 1991; February 1997; December 1992; February 1997
    - Slovenia: December 1988; 1999 3/; January 1996; June 1997
    - Notes: 1/ More restrictions lifted in 1999. 2/ More restrictions lifted in 1998. 3/ Until 1999 foreign sales within 7 years taxed 12 percent. 25 percent foreign ownership limit. (Further country footnotes preserved in table.)

- Table 2. Means of Key Variables by Country
  - Variables: # of Firms; % Price Increase; DIFCOV; DIFBETA; Net Upgrades
  - Selected country means (exact values):
    - Poland: 15; 0.16; 0.0070; -0.19; -1.91
    - Hungary: 10; 0.11; 0.0056; -0.24; -1.70
    - Czech Republic: 8; 0.43; 0.0071; 0.41; 0.08
    - Slovenia: 5; 0.63; 0.0023; 0.59; -0.20
    - Slovakia: 4; 0.61; 0.0035; 0.59; 0.00
    - Lithuania: 4; 0.39; 0.0010; 0.46; 0.25
    - Latvia: 4; 0.28; 0.0028; 0.06; 0.00
    - Estonia: 5; 0.46; 0.0041; 0.52; 0.70
    - Romania: 10; 0.54; 0.0074; 1.43; 0.07
    - Russia: 7; 0.37; 0.0231; -0.86; -1.14
    - Croatia: 2; 0.55; 0.0054; 0.66; -0.83
    - Total: 74; 0.36; 0.0071; 0.24; -0.68
  - Definitions preserved: Price increase is the percentage stock price increase between November 2001 and February 2003. DIFCOV and DIFBETA calculated using 36 months of historical returns from November 2001 and prior. Net upgrades defined as average number of upward revisions minus downward revisions between November 2001 and December 2002.

- Table 3. Descriptive Statistics
  - Exact summary statistics:
    - Price Change: Mean 0.356; Median 0.413; Min -0.462; Max 1.085; Std. Dev. 0.366
    - DIFCOV: Mean 0.007; Median 0.006; Min -0.001; Max 0.030; Std. Dev. 0.006
    - DIFBETA: Mean 0.237; Median 0.234; Min -1.404; Max 2.091; Std. Dev. 0.761
    - Net Upgrades: Mean -0.682; Median 0.000; Min -11.000; Max 10.667; Std. Dev. 3.233
    - Size: Mean 0.111; Median 0.047; Min 0.001; Max 0.824; Std. Dev. 0.152
    - Turnover: Mean 37.414; Median 3.258; Min 0.005; Max 246.225; Std. Dev. 65.632
  - Definitions preserved: Size and Turnover definitions and calculation windows as stated in source.

- Table 4. Baseline Regression Results (Panels A and B)
  - Dependent variable: percentage stock price increase between November 2001 and February 2003.
  - Key coefficient and model summary points (values preserved exactly as presented in table):
    - Intercept Panel A (1a): 0.560** (4.48)
    - Intercept Panel A (2a): 0.556** (4.59)
    - Intercept Panel A (3a): 0.508** (4.47)
    - Intercept Panel A (4a): 0.512** (4.41)
    - Intercept Panel B (1b): 0.457** (7.42)
    - Intercept Panel B (2b): 0.477** (8.54)
    - Intercept Panel B (3b): 0.563** (4.92)
    - Intercept Panel B (4b): 0.532** (3.70)
    - EU dummy coefficients include -0.201* (-2.65) in (1b); -0.19** (-2.77) in (2b).
    - DIFBETA * EU significant in Panel B: 0.349** (3.31) in (2b); 0.314** (3.09) in (3b); 0.414* (2.12) in (4b); 0.402* (2.21) in (4b).
    - Net Upgrades positive and significant across specifications: e.g., 0.039** (3.35) in (1a); 0.028** (3.32) in (1b).
    - R-squared values Panel A: 0.090, 0.205, 0.329, 0.345; Panel B: 0.296, 0.356, 0.398, 0.403.
    - “The number of observations is 74.” T-statistics in parentheses; * and ** indicate significance at 5 and 1 percent.

- Table 5. Varying Integration Date and Window
  - Integration windows and reported model outcomes:
    - Example Intercept values: 0.716** (6.58); 0.639** (6.28); 0.128 (0.83); 0.481** (5.58); 0.230** (3.37)
    - DIFBETA * EU: 0.418* (2.35); 0.468** (2.81); 0.066 (0.38); 0.317* (2.15); 0.018 (0.17)
    - Net Upgrades: 0.017 (1.48); 0.034** (3.51); 0.022** (3.95); 0.015 (1.51); 0.031* (2.47)
    - Turnover includes significance in some specifications: -0.000* (-2.27); -0.000* (-2.34); -0.002 (-2.16).
    - R-squared reported: 0.287; 0.453; 0.391; 0.291; 0.207.
    - Number of Obs.: 69; 72; 65; 84; 84.

- Table 6. Price Changes and Beta
  - Four specification columns with country effects/no country effects:
    - Intercept values: 0.444 (1.65); 0.503 (1.98); 0.591** (3.01); 0.504** (3.19)
    - Net Upgrades remain positive and significant in several columns: 0.040** (3.53); 0.031* (2.98); 0.031* (2.79)
    - Turnover significant in one specification: 0.000** (5.18)
    - BETAM * EU coefficients negative but not uniformly significant: examples -0.269 (-1.09); -0.395 (-1.81)
    - R-squared values: 0.075; 0.198; 0.351; 0.364.
    - Number of Obs. 74 in all columns.

### Variable and methodology notes preserved from source
- Price increase / Price change: percentage stock price increase between November 2001 and February 2003 (or during the specified integration window where stated).
- DIFCOV: covariance of firm returns with local market returns minus the covariance of firm returns with world returns; covariances calculated using 36 months of historical returns from November 2001 and prior (or from the beginning of the integration window and prior when specified).
- DIFBETA / BETAM: difference between local market beta and world market beta (or local market beta calculated using 36 months of historical returns), with same historical-return windows as DIFCOV.
- Net Upgrades: average number of upward revisions minus average number of downward revisions over the specified period (definitions vary by table: November 2001–February 2003; November 2001–December 2002; or during integration window).
- Size: firm’s average market capitalization as a percentage of average total domestic market capitalization during the 12 months prior to November 2001 (or prior to the beginning of the integration window as specified).
- Turnover: dollar value traded during the 12 months prior to November 2001 (or prior to the beginning of the integration window) as percentage of a firm’s average market capitalization.
- Country effects constrained to sum to zero separately for accession and nonaccession countries where noted.
- T-statistics calculated using robust and country “clustered” standard errors. A * and ** indicate significance at 5 and 1 percent.

*Source: _wp05182 - References (PDF content provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05182.pdf_