## _wp07149

## Source details

**Canonical URL:** [_wp07149](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07149.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07149.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07149.pdf.json)

---

### I. Introduction — objectives, context, and headline result
- Objectives:
  - Analyze IPOs’ performance in the short- and long-run against the benchmark of the equity markets’ general price indices.
  - Identify the main determinants of IPOs’ abnormal returns.
- Context and literature:
  - Loughran, Ritter, and Rydqvist (1994, updated May 2006) find average initial returns across 39 countries ranging from 5.4 percent (Canada) to 256.9 percent (China).
  - Most empirical studies show significant initial positive returns for IPOs; evidence on long-run IPO performance is mixed.
- Sample and key overall result:
  - Sample: 47 firms that went public between 2001 and 2006 (long-run observations limited to 29 firms with at least one year of post-listing data).
  - Average initial abnormal returns: 290 percent (mean initial abnormal return for IPO sample).
  - Alternative statement: On average, IPOs yield around 315 percent on the first trading day (Table 4, Panel A).
  - One-year horizon: IPOs’ returns beat the market index benchmark when initial returns are included; excluding initial returns, IPOs yield negative abnormal returns over the one-year period.
- Interpretation:
  - Evidence consistent with investor over-optimism at IPO launch followed by increased pessimism over time.

### II. Institutional settings and market context in the GCC
- Countries in study: Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, U.A.E. (Abu Dhabi and Dubai markets specified).
- Market foundation years (Table 1 highlights):
  - Bahrain BSE — 1987; Kuwait KSE — 1977; Oman MSM — 1988; Qatar DSM — 1995 (activities started May 1997); Saudi Tadawul — 1984 (platform launched October 2001); U.A.E. ADSM — 2000; U.A.E. DFM — 2000.
- Market expansion and 2003–05 boom:
  - Excluding Oman and Bahrain, the GCC weighted price index increased by 480 percent on average in 2003–05.
  - MSCI grew by 140 percent during the same period.
  - Total GCC market capitalization rose from $117 billion (2000) to $1,135.5 billion (2005).
- Market breadth and weights (GCC weighted index, end-2003 exchange rates):
  - Saudi Arabia: 51.6 percent; Kuwait: 22.2 percent; U.A.E.: 14.4 percent; Qatar: 6.5 percent; Oman: 3.5 percent; Bahrain: 1.8 percent.
- Correction and volatility:
  - GCC weighted price index peaked in February 2006 at over 530 percent above end-2002 level, then corrected with cumulative losses of 40–60 percent from peak values through end-May 2006.
  - As of end-2006, price volatility remained high.
- IPO activity:
  - Number of IPOs: 24 in 2005 and 23 in 2006.
  - Capital subscribed increased from $5.6 billion to $7.5 billion in 2006 (increase by 1/3).
  - Total IPOs 2000–June 2006: 59 IPOs; total value: $15.5 billion.

### III. Data set, sample construction, and key descriptive statistics
- Data set and sample:
  - Dataset includes GCC firms that went public from 2001 to 2006 and were listed for trading as of June 2006.
  - There were 59 IPOs in the GCC countries from 2000 until June 2006; useful observations after exclusions: 47 IPOs; long-run sample: 29 firms with at least one year of post-listing data.
  - Missing observations aggregated: 12 observations totaling 913.9 (US$ million).
  - Available observations aggregated: 47 observations totaling 14,604 (US$ million).
  - More than 75 percent of sample firms were either state-owned or state-influenced.
  - Industry composition: majority in oil and energy, financial sector (mainly banks), and real estate and construction.
- Data sources:
  - ABQ Zawya and Gulf Base: IPO prospectuses (date, price, size, percentages of equity sold, oversubscription).
  - Shuaa Capital Limited’s database: corporate actions for effective returns.
  - Stock exchanges: daily stock prices and general indices used as benchmarks.
- Table 4 — Panel A (Initial returns, 47 firms) selected exact statistics:
  - IPO returns (ar) Mean: 3.15
  - Market returns (crp) Mean: 0.25
  - IPO returns (ar) Median: 1.52
  - Market returns (crp) Median: 0.07
  - IPO returns (ar) Maximum: 14.3
  - IPO returns (ar) Minimum: -0.3
  - IPO returns (ar) Standard deviation: 3.45
  - IPO returns (ar) Standard skewness: 4.6
  - IPO returns (ar) Standard kurtosis: 3.48
- Table 4 — Panel B (12-month CR and BHR including initial returns, 29 firms) selected exact statistics:
  - Cumulative returns (CR) — IPOs Mean: 3.9; Market Mean: 1.13; CAPM Mean: 0.80
  - Buy-and-hold returns (BHR) — IPOs Mean: 3.90; Market Mean: 1.1; CAPM Mean: 0.72
  - BHR IPOs Maximum: 14.62; Minimum: -0.01; Standard deviation: 3.70
- Table 4 — Panel C (12-month CR and BHR excluding initial returns, 29 firms) selected exact statistics:
  - Cumulative returns (CR) — IPOs Mean: 0.20; Market Mean: 0.71; CAPM Mean: 0.46
  - Buy-and-hold returns (BHR) — IPOs Mean: 0.23; Market Mean: 0.46; CAPM Mean: 0.31
  - BHR IPOs Maximum: 2.04; Minimum: -0.71; Standard deviation: 0.67

### IV. Methodology — definitions and tests (exact formulas and procedures)
- Benchmark: country general price index.
- Initial return measures:
  - Gross initial return: r_{i} = (P_{t,i} − P_{0,i}) / P_{0,i}.
  - Net raw initial return adjustment: r_{i} = [ (P_{t,i} − P_{0,i}) / P_{0,i} − (i_TC + (1−Ψ)×((ARF/365)×TDSD−)) ].
  - Abnormal initial return: ar_{i} = r_{i} − r_{crp}, with r_{crp} = (I_{t,i} − I_{0,i}) / I_{0,i}.
- Aftermarket measures:
  - AR_{i,t} = R_{i,t} − R_{crp,t}.
  - CAR_{i,s,e} = Σ_{t=s}^{e} AR_{i,t}.
  - ARAvg_{t} = (1/n_{t}) Σ_{i=1}^{n_{t}} AR_{i,t}; CAR_{s,e} = Σ_{t=s}^{e} ARAvg_{t}.
  - BHAR_{i,T} = Π_{t=1}^{T} (1 + r_{i,t}) − Π_{t=1}^{T} (1 + r_{crp,t}) with T = 252 trading days.
  - BHARAvg_{T} = (1/n) Σ_{i=1}^{n} BHAR_{i,T}.
  - WR_{T} = (1 + BHAR_{IPOs,T}^{vg}) / (1 + BHAR_{crp,T}^{vg}); WR > 1.00 implies IPOs outperform reference portfolio.
  - CAPMAR_{i,t} = R_{i,t} − R_{crp,t} − β_{i} (R_{crp,t} − R_{f,t}) with R_{f,t} proxied by one-month bank deposit rate.
- Statistical tests:
  - Parametric t-type statistics for means (equations (13)–(17)).
  - Wilcoxon signed-rank (non-parametric) for medians.
  - Mann-Whitney to compare medians across subsamples.
  - Robustness checks include chi-square and Shapiro-Walks W tests for normality.
- Notes:
  - ARF proxied by one-month bank customer deposit rates (IMF IFS line 60k for all countries except Qatar, where Qatar Central Bank website used).
  - No firm in sample was delisted during the study period.

### V. Empirical findings — initial abnormal returns (magnitude, distribution, and cross-sample)
- Main magnitudes:
  - Mean initial abnormal return: 290 percent (Table 5, Panel A).
  - Median initial abnormal return: around 147 percent.
  - Proportion with positive initial abnormal returns: around 98 percent (46 out of 47).
  - Mean WR ratio (first trading day): around 3.32.
  - t-statistic: mean initial abnormal return significantly different from zero at the one-percent level (t-stat: 6.14***).
  - Wilcoxon signed-rank z-statistic: 5.89***.
- Cross-sample comparisons (selected exact figures from Table 5):
  - By country (booming vs non-booming markets; Panel B):
    - Number of IPOs: 33 (booming) and 14 (non-booming).
    - Mean (median) initial abnormal return in booming stock markets: 3.80 (2.4) — corresponding to 380 (240) percent.
    - Mean (median) initial abnormal return in non-booming stock markets: 0.8 (0.6) — corresponding to 80 (60) percent.
    - Difference significant at the one-percent level (Mann-Whitney average rank 29.4-11.30***).
  - By year (booming years vs non-booming years; Panel C):
    - Number of IPOs: 34 (booming years) and 13 (non-booming years).
    - Mean (median) booming years: 3.5 (2.1); non-booming years: 1.32 (1.22).
    - Difference significant (t-stat 2.16**, Mann-Whitney average rank 26.24-18.15*).
  - By oversubscription (Panel D) and industry (Panel E): differences not significant at conventional levels in medians for oversubscription; regulated industries show higher mean/median but differences not significant in Panel E.
- Appendix I highlights (country averages of initial returns — exact figures):
  - Bahrain Average Initial Return: 44 (In percent)
  - Kuwait Average Initial Return: 330 (In percent)
  - Oman Average Initial Return: 52 (In percent)
  - Qatar Average Initial Return: 399 (In percent)
  - Saudi Arabia Average Initial Return: 349 (In percent)
  - UAE Average Initial Return: 473 (In percent)

### VI. Empirical findings — aftermarket (one-year) performance including and excluding initial returns
- Including initial returns (Table 7, Panel A; 29 firms):
  - Average aftermarket abnormal return (investor) around 280 percent using both BHAR and CAR.
  - Mean WR ratio including initial returns: 2.33; Median WR ratio: 1.98.
  - t-statistics and Wilcoxon signed-rank tests significant at the one-percent level for all models.
  - CAR mean: 2.75; BHAR mean: 2.8 (CAPM columns 3.1 and 3.2 respectively).
- Excluding initial returns (Table 7, Panel B; 29 firms):
  - Mean abnormal returns excluding initial returns: CAR -0.50; BHAR -0.23 (CAPM -0.25 and -0.07).
  - Mean WR ratio excluding initial returns: 0.84; Median WR ratio: 0.79.
  - t-statistics and z-statistics negative and significant (e.g., t-stat -3.6*** for CAR excluding initial returns).
  - Reported long-run abnormal return ranges (excluding initial returns) cited elsewhere in paper: ranging from -23 percent to -50 percent.
- Cross-group aftermarket findings (selected exact findings from Panels C–R):
  - By country: differences between booming and non-booming markets small and generally not statistically significant.
  - By oversubscription: higher oversubscription associated with higher long-run abnormal returns; significant at five-percent level only for BHAR CAPM model (Panel J Mann-Whitney significant).
  - By industry: regulated industry IPOs show worse aftermarket cumulative abnormal returns than non-regulated; significance found in risk-unadjusted models (Panel K t-stat -2.03**).
  - By initial abnormal return levels: no consistent significant difference in long-run performance between above- and below-median initial abnormal return groups.

### VII. Determinants — multivariate regression results (exact coefficients and diagnostics)
- Determinants of initial abnormal returns (Table 6; dependent variable: initial abnormal returns; 47 observations):
  - Regression (ar_i = α + β1 EXANTE_i + β2 OVRS_i + β3 PSO_i + β4 MV_i + β5 TIM_i + β6 CONT_i + β7 IND_i + ε_i).
  - Reported coefficients and t-statistics (exact):
    - Intercept: Coefficient -4.96; t-statistic (-2.17)**.
    - EXANTE: Coefficient 45.8; t-statistic (1.18).
    - PSO: Coefficient 3.1; t-statistic (1.2).
    - MV: Coefficient -2.94; t-statistic (-0.06).
    - OVRS: Coefficient 0.0004; t-statistic (0.17).
    - CONT: Coefficient 2.82; t-statistic (3.02)***.
    - TIM: Coefficient 1.96; t-statistic (2.37)**.
    - IND: Coefficient 2.32; t-statistic (2.47)**.
  - Model diagnostics:
    - R2: 43.85 (percent); Adj. R2: 32.03 (percent); F-value: 3.71***; Observations: 47.
    - Heteroskedasticity detected; White’s HCSE used.
  - Interpretation:
    - CONT (country), TIM (timing), and IND (industry) significant; traditional variables EXANTE, OVRS, PSO, MV not significant in explaining initial underpricing in GCC sample.
- Determinants of aftermarket abnormal returns (Table 8; dependent variable: aftermarket abnormal returns; sample reduced to 26 observations after excluding three influential observations):
  - Regression (AFTMARKAR_T,i = α + β1 ar_i + β2 EXANTE_i + β3 OVRS_i + β4 PSO_i + β5 MV_i + β6 CONT_i + β7 IND_i + ε_i).
  - Key significant coefficients (exact):
    - Panel A (Abnormal Results Model):
      - MV: Coefficient 65.73; t-statistic (3.6)*** (CAR); Coefficient 40.21; t-statistic (1.98)* (BHAR).
      - CONT: Coefficient -0.57; t-statistic (-2.13)** (CAR); Coefficient -0.15; t-statistic (-0.49) (BHAR).
      - IND: Coefficient -1.1; t-statistic (-4.51)*** (CAR); Coefficient -0.71; t-statistic (-2.59)** (BHAR).
      - R-squared: 43.9 percent (CAR) and 40.16 percent (BHAR). Adj. R-squared: 25.19 percent (CAR) and 20.22 percent (BHAR). F-value: 2.35* (CAR) and 2.01* (BHAR).
    - Panel B (CAPM-adjusted Model):
      - MV: Coefficient 37.78; t-statistic (2.16)** (CAR); Coefficient 21.7; t-statistic (1.15) (BHAR).
      - CONT: Coefficient -0.56; t-statistic (-2.17)** (CAR); Coefficient -0.09; t-statistic (-0.34) (BHAR).
      - IND: Coefficient -0.54; t-statistic (-2.29)** (CAR); Coefficient -0.31; t-statistic (-1.21) (BHAR).
      - R-squared: 34.72 percent (CAR) and 33.66 percent (BHAR). Adj. R-squared: 12.69 percent (CAR) and 11.55 percent (BHAR).
  - Interpretation:
    - Market volatility (MV), country sentiment (CONT), and industry classification (IND) drive one-year aftermarket performance. Negative signs on CONT and IND indicate IPOs from booming markets and regulated industries provided negative long-run abnormal returns.

### VIII. Distributional diagnostics and inferential caveats
- Many return measures exhibit standardized skewness and kurtosis outside the range -2 to +2, indicating non-normality.
- Parametric t-test results should be interpreted with caution where normality is violated; non-parametric Wilcoxon signed-rank and Mann-Whitney tests used for robustness.
- Both parametric and non-parametric tests yield similar inference in key results (e.g., significance of initial abnormal returns).

### IX. Summary magnitudes, interpretation, and implications
- Exact empirical magnitudes and summaries:
  - Initial abnormal returns averaged 290 percent (mean) and median around 147 percent.
  - On average, IPOs yield around 315 percent on the first trading day (Table 4, Panel A).
  - Aftermarket performance including initial returns: positive average abnormal returns between 275 and 280 percent.
  - Aftermarket performance excluding initial returns: negative, with reported ranges from -23 percent to -50 percent relative to the benchmark portfolio.
  - Practical implication: Investing one dollar in IPOs at the end of the first trading day and holding it for one year would have left an investor with only 77 to 50 percent of the return of each dollar invested in GCC stock exchange general indices.
- Main drivers identified:
  - Market excitement (bullish or bearish) and the regulated nature of the IPO’s economic sector are more important determinants than risk, expected demand (oversubscription), or firm signaling metrics in the GCC sample.

### X. Policy recommendations (as presented in the source)
- Implement measures to avoid market volatility to smooth aftermarket abnormal negative results.
- Address structural shortcomings of GCC equity markets:
  - Lack or small size of traditional institutional investors.
  - Small free-float of shares.
  - Lack of widespread financial culture among retail investors.
  - Relatively new and incomplete regulatory and supervisory frameworks.
- Create incentives for development of institutional investors with a long-term investment profile (including opening markets to foreign investment where currently limited).
- Educate retail investors and create professional, accountable financial advisory firms to reduce investor over-optimism.
- Strengthen regulations and ensure compliance from investors, intermediaries, and listed companies to ensure complete and accurate information and penalize price manipulation practices.
- Reconsider policy of underpricing IPOs as a mechanism to share oil wealth with the population, because it may distort price formation; consider alternative distribution mechanisms that do not interfere with price formation.

### XI. Appendix I — selected country-level IPO performance highlights (exact examples)
- Qatar examples (selected IPOs and exact initial returns and one-year figures reported):
  - Qatar Fuel Company (Woqod) — Initial Return: 275 (In percent); Excluding 1/: 11 (In percent); Including 2/: 316 (In percent).
  - Qatar Technical Inspection — Initial Return: 572; Excluding 1/: 20; Including 2/: 706.
  - Dlala Brokerage and Investment Holding — Initial Return: 1,031; Excluding 1/: -60; Including 2/: 349.
- Saudi Arabia examples:
  - Bank Albilad — Initial Return: 1,430; Excluding 1/: 2; Including 2/: 1,462.
  - Etihad Etisalat Company — Initial Return: 500; Excluding 1/: 139; Including 2/: 1,336.
- UAE examples:
  - Finance House — Initial Return: 1,281; Excluding 1/: -31; Including 2/: 851.
  - Al Dar Properties — Initial Return: 650; Excluding 1/: 19; Including 2/: 792.
- Country average initial-return figures (exact):
  - Bahrain Average Initial Return: 44 (In percent)
  - Kuwait Average Initial Return: 330 (In percent)
  - Oman Average Initial Return: 52 (In percent)
  - Qatar Average Initial Return: 399 (In percent)
  - Saudi Arabia Average Initial Return: 349 (In percent)
  - UAE Average Initial Return: 473 (In percent)

*Italic: Source: _wp07149 - References..............................................................................................................*

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

### _wp07149 - References..............................................................................................................

### I. INTRODUCTION
- Objectives:
  - Analyze IPOs’ performance in the short- and long-run against the benchmark of the equity markets’ general price indices.
  - Identify the main determinants of IPOs’ abnormal returns.
- Context and literature:
  - Most empirical studies show significant initial positive returns for IPOs. Loughran, Ritter, and Rydqvist (1994, updated May 2006) find average initial returns across 39 countries ranging from 5.4 percent (Canada) to 256.9 percent (China).
  - Mixed evidence exists on long-run IPO performance: some studies find positive market-adjusted excess returns, but most studies report insignificant and/or negative abnormal returns.
- Sample and key overall result:
  - Sample: 47 firms that went public between 2001 and 2006 (18 IPOs lacked at least one year of post-listing data, reducing long-run observations to 29).
  - Result: IPOs yield economically and statistically significant initial abnormal returns in line with underpricing.
  - The average initial abnormal returns equal 290 percent.
  - One-year horizon: IPOs’ returns beat the market index benchmark, but yield negative abnormal returns once initial returns are excluded; on average initial abnormal returns overcompensate for the negative returns over the one-year period, yielding overall positive abnormal returns.
- Determinants:
  - Several multivariate regressions reject that common independent variables used in the literature are relevant determinants of GCC IPOs’ abnormal returns.
  - Evidence indicates country- and industry-specific characteristics, plus timing of offers, play key roles.
- Interpretation:
  - Findings support that investors are initially over-optimistic about an IPO’s performance and grow more pessimistic over time.

### II. INSTITUTIONAL SETTINGS AND LATEST DEVELOPMENTS IN THE GCC STOCK MARKETS
- GCC membership in the study: Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates (U.A.E.).
- Table 1 — Market foundations (year of establishment):
  - Bahrain: Bahrain Stock Exchange (BSE) — 1987
  - Kuwait: Kuwait Stock Exchange (KSE) — 1977
  - Oman: Muscat Securities Market (MSM) — 1988
  - Qatar: Doha Securities Market (DSM) — 1995 (activities started in May 1997)
  - Saudi Arabia: Tadawul — 1984 (organized market existed since 1984; platform launched in October 2001)
  - U.A.E. Abu Dhabi: Abu Dhabi Security Market (ADSM) — 2000
  - U.A.E. Dubai: Dubai Financial Market (DFM) — 2000
- Historical performance and episodes:
  - GCC equity markets generally underperformed industrial countries’ equity markets and were only marginally above the MSCI Emerging Markets Index (MSCI) except for short-lived speculative outbursts.
  - Notable boom-and-bust episodes: 1982 Al-Mannak bubble (Kuwait), 1991 Saudi Arabian bubble, 1997 Omani bubble.
- Market expansion and 2003–05 boom:
  - Starting early 2003, propelled by increasing oil prices, record corporate profits, and sustained structural reforms, stock prices in most GCC markets rose rapidly.
  - Excluding Oman and Bahrain, the GCC weighted price index increased by 480 percent on average in 2003–05.
  - GCC outperformed MSCI (MSCI grew by 140 percent during same period).
  - Total GCC market capitalization rose from $117 billion (2000) to $1,135.5 billion (2005). (Source table indicates GCC market cap: 117.0 and 1,135.5; MENA: 312.0 and 1,594.6.)
- Market breadth and country weights (GCC weighted index):
  - Index includes seven regional markets using market capitalization (in U.S. dollars at end-2003 exchange rates) as weight.
  - Resulting weights: 51.6 percent for Saudi Arabia; 22.2 percent for Kuwait; 14.4 percent for the U.A.E.; 6.5 percent for Qatar; 3.5 percent for Oman; 1.8 percent for Bahrain.
- Correction and volatility:
  - Market sentiment changed in late 2005; sharp correction in early 2006.
  - GCC weighted price index peaked in February 2006 at over 530 percent above end-2002 level, then corrected with cumulative losses of 40–60 percent from peak values through end-May 2006.
  - As of end-2006, price volatility remained high; upward movements short-lived despite substantial improvement in valuation ratios of many blue chip stocks.
- IPO activity:
  - Substantial increase in number of IPOs in 2005 signaled speculative bubble characteristics.
  - Number of IPOs: 24 in 2005 and 23 in 2006.
  - Capital subscribed increased by 1/3 in 2006, from $5.6 billion to $7.5 billion.
  - IPO intensity noted as a warning signal per theory that companies issue equity when equity is overpriced.

### III. DATA SELECTION AND SAMPLE CONSTRUCTION
- Sample construction notes:
  - Final sample includes 47 IPOs from 2001–2006; long-run analysis sample reduced to 29 firms due to at least one year of post-listing data requirement.
- Methodological outline (to be detailed later in paper):
  - Methodology measures initial and aftermarket abnormal returns of IPOs.
  - Statistical techniques include multivariate cross-sectional regressions to investigate determinants of initial and aftermarket abnormal returns.
  - Key descriptive statistics cover initial and aftermarket returns and abnormal returns.

### IV. KEY TABLES AND FIGURES (as referenced in source)
- Tables listed (titles only, contents in source):
  - Table 1. GCC Stock Marktes
  - Table 2. Equity Markets in Selected MENA Countries: Selected Indicators: 2000–05
  - Table 3. Classification of IPOs in the GCC Countries from 2001–06
  - Table 4. Basic Descriptive Statistics for Initial and Aftermarket Returns of IPOs
  - Table 5. Initial Abnormal Returns of IPOs in the GCC
  - Table 6. Multivariate Cross-Sectional Regression Analysis for the Determinants of Initial Abnormal Returns of IPOs
  - Table 7. Aftermarket Abnormal Returns of IPOs in the GCC
  - Table 8. Multivariate Cross-Sectional Regression Analysis for the Determinants of Aftermarket Abnormal Returns of IPOs
- Figures referenced:
  - Figure 1. Performance of GCC Stock Price Indices, 2003–05 (December 31, 2002 =100)
  - Figure 2. Performance of Stock Price Indices, January 2006–April 9, 2007 (Jan 2006=100)
  - Figure 3. Performance of GCC Stock Price Indices, 2003–February 2005 (Dec. 2002 =100)
  - Figure 4. Total Number of IPOs in GCC Equity Markets (annual counts shown for 2001–2006)

### V. EMPIRICAL FINDINGS AND INTERPRETATION (summary)
- Initial returns and underpricing:
  - IPOs in sample display pronounced underpricing, with average initial abnormal returns of 290 percent.
- One-year performance:
  - IPOs outperform market index benchmark over one-year horizon when including initial returns.
  - Excluding initial returns, IPOs display negative abnormal returns over one-year horizon.
- Determinants:
  - Common independent variables used in international IPO literature are rejected as significant determinants for GCC IPO abnormal returns in multivariate regressions.
  - Country-specific, industry-specific characteristics and timing of offers have explanatory power for IPO performance in the region.
- Behavioral interpretation:
  - Results consistent with investor over-optimism at IPO launch followed by increased pessimism over time.

*Italic: Source: _wp07149 - References..............................................................................................................*

### 10.      The data set analyzed includes GCC firms that went public from 2001 to 2006 and

### _wp07149 - 10.      The data set analyzed includes GCC firms that went public from 2001 to 2006 and

### Data set and sample
- The data set analyzed includes GCC firms that went public from 2001 to 2006 and were listed for trading in the stock exchanges as of June 2006.
- There were 59 IPOs in the GCC countries from 2000 until June 2006, with a total value of $15.5 billion (Table 3, Panel A).
- After excluding IPOs not yet tradable and those with data limitations, the number of useful observations was reduced to 47.
- Most of these 47 IPOs come from the U.A.E., Saudi Arabia, and Qatar.
- The highest number and value of IPOs took place in 2005 (peak of the bull market).
- More than 75 percent of the sample firms were either state-owned or the state intervened in the IPO process by determining the share prices because they operate in regulated industries and are subject to licensing to start operations (Table 3, Panel B).
- Industry composition (Table 3, Panel C): most IPO issuers operate in oil and energy, financial sector (mainly banks), and real estate and construction.

### Data sources
- ABQ Zawya and Gulf Base: IPO prospectuses — date, price, size of issue, percentages of equity sold, size of oversubscription.
- Shuaa Capital Limited’s database: corporate actions (stock splits, dividend payments, capital increases, stock bonuses) used to calculate effective returns.
- Stock exchanges of each GCC country: daily stock prices and general indices used as benchmarks.

### IPO classification and key statistics (as reported)
- Total IPOs (2000–June 2006): 59 IPOs; total value: $15.5 billion.
- Useful observations after exclusions: 47 IPOs.
- Missing observations aggregated: 12 observations totaling 913.9 (US$ million).
- Available observations aggregated: 47 observations totaling 14,604 (US$ million).
- Note: Table 3 panels (A–C) classify IPOs by market, ownership (state vs. private), and industry; specific country- and year-level counts and values appear in Table 3 (Panel A, B, C) as presented.

### Empirical models and methodology — overview
- Benchmarks: general market index of each corresponding country used as the reference portfolio.
- Tests:
  - Parametric test statistic and non-parametric Wilcoxon signed-rank test to examine IPO abnormal returns in the short and long run.
  - Mann-Whitney test to compare medians between subsamples split by country, year, percentage of oversubscription, industry, and initial abnormal returns.
  - Multivariate models to test predetermined drivers explaining IPO performance.
- Abnormal returns definitions: initial and aftermarket abnormal returns computed relative to country general price indices.

### Initial abnormal returns — definitions and adjustments
- Gross initial return for security i:
  - r_{i} = (P_{t,i} − P_{0,i}) / P_{0,i}  (where P_{t,i} is closing price on first trading day, P_{0,i} is offer price).
- Net raw initial returns adjustments:
  - Account for rationing/allocation percentage (Ψ), transaction costs (i_TC), time value of money using average risk-free rate (t,0 ARF), and elapsed days between last subscription day and first trading day (TDSD−).
- Net raw initial return formula used:
  - r_{i} = [ (P_{t,i} − P_{0,i}) / P_{0,i} − (i_TC + (1−Ψ)×((ARF/365)×TDSD−)) ]  (as presented in equation (2) — variables preserved as in source).
- Abnormal initial return for security i:
  - ar_{i} = r_{i} − r_{crp}  (where r_{crp} is return on country general equity market index from subscription period to closing of first trading day).
- Market index return:
  - r_{crp} = (I_{t,i} − I_{0,i}) / I_{0,i}  (I_{t,i} is price index at end of first trading day; I_{0,i} is price index at last day of subscription).

### Aftermarket abnormal returns — measurement approaches
- Two main methods used:
  - Cumulative Abnormal Returns (CAR).
  - Buy-and-Hold Abnormal Returns (BHAR).
- Additional approaches applied to both methods:
  - Wealth Relative index (WR).
  - Sharp-Lintner’s CAPM for risk adjustment.
- Aftermarket period definition:
  - One year exclusive of initial returns; initial return period is month 0, aftermarket includes 12 months defined as 21 successive trading-day periods.
  - First aftermarket month: listing days 2–22; second month: listing days 23–43; etc.
  - Each IPO truncated if delisted before its anniversary (for the entire sample, no firm was delisted over the period of the study).
- Monthly abnormal return:
  - AR_{i,t} = R_{i,t} − R_{crp,t}  (raw monthly return on security i minus raw monthly return on reference portfolio for same 21-trading-day period).
- CAR for an individual IPO:
  - CAR_{i,s,e} = Σ_{t=s}^{e} AR_{i,t}  (s = starting month after listing; e = anniversary month or delisting).
- Average abnormal return on portfolio for month t:
  - ARAvg_{t} = (1/n_{t}) Σ_{i=1}^{n_{t}} AR_{i,t}  (equally weighted arithmetic average).
- CAR on a portfolio:
  - CAR_{s,e} = Σ_{t=s}^{e} ARAvg_{t}.
- BHAR for an individual IPO over one year:
  - BHAR_{i,T} = Π_{t=1}^{T} (1 + r_{i,t}) − Π_{t=1}^{T} (1 + r_{crp,t})  (as specified in equation (9), with T = aftermarket trading day number 252 and truncation at delisting date: },min{delistingT).
- Average BHAR for sample of n IPOs:
  - BHARAvg_{T} = (1/n) Σ_{i=1}^{n} BHAR_{i,T}  (T = 252).
- Wealth Relative (WR) index:
  - WR_{T} = (1 + BHAR_{IPOs,T}^{vg}) / (1 + BHAR_{crp,T}^{vg})  (WR > 1.00 implies IPOs outperform reference portfolio).
- Risk-adjusted abnormal return (Sharp-Lintner CAPM, monthly):
  - CAPMAR_{i,t} = R_{i,t} − R_{crp,t} − β_{i} (R_{crp,t} − R_{f,t})  (expressed in source as equation (12) with variables preserved); R_{f,t} proxied by one-month bank deposit rate; β_{i} from CAPM regression.
- Risk-adjusted returns used to compute CARs and BHARs.

### Summary notes from the source
- Benchmarks chosen were the country general price indices due to limited numbers of listed firms preventing use of control-firm approaches.
- ARF (average risk-free rate) proxied by one-month bank customer deposit rates as reported in line 60k in IMF’s International Financial Statistics for all countries except Qatar (Qatar Central Bank website used for Qatar).
- No firm in the entire sample was delisted over the study period.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07149.pdf*

### 28.      On average, IPOs yield around 315 percent on the first trading day (Table 4,

### 28. On average, IPOs yield around 315 percent on the first trading day (Table 4, Panel A)

### Key empirical findings
- On average, IPOs yield around 315 percent on the first trading day (Table 4, Panel A).
- The average first-day IPO return far exceeds the average returns on the corresponding reference portfolios and exceeds what is documented for any developed or emerging economies.
- Some investors obtained initial returns as high as 1,430 percent and others achieved negative initial returns as low as 30 percent (Appendix I).
- IPOs seem to perform better than the market over the initial one-year period regardless of methods or models of calculation (Table 4, Panel B).
- Excluding initial returns, IPOs tend to perform far below their benchmarks (Table 4, Panel C).
- Conclusion: most of the total abnormal after-market returns of IPOs in the GCC were due to the underpricing of the issues; on average, investors who bought IPOs on the first trading day achieved lower returns relative to the market.

### Table 4 — Basic descriptive statistics (as presented)
- Panel A: Initial returns (47 firms)
  - IPO returns (ar) Mean: 3.15
  - Market returns (crp) Mean: 0.25
  - IPO returns (ar) Median: 1.52
  - Market returns (crp) Median: 0.07
  - IPO returns (ar) Maximum: 14.3
  - Market returns (crp) Maximum: 2.09
  - IPO returns (ar) Minimum: -0.3
  - Market returns (crp) Minimum: -0.3
  - IPO returns (ar) Standard deviation: 3.45
  - Market returns (crp) Standard deviation: 0.53
  - IPO returns (ar) Standard skewness: 4.6
  - Market returns (crp) Standard skewness: 6.47
  - IPO returns (ar) Standard kurtosis: 3.48
  - Market returns (crp) Standard kurtosis: 7.43

- Panel B: 12-month cumulative returns and buy-and-hold returns including initial returns (29 firms)
  - Cumulative returns (CR) — IPOs Mean: 3.9; Market Mean: 1.13; CAPM Mean: 0.80
  - Cumulative returns (CR) — IPOs Median: 3.0; Market Median: 1.12; CAPM Median: 0.72
  - Cumulative returns (CR) — IPOs Maximum: 14.40; Market Maximum: 4.65; CAPM Maximum: 2.70
  - Cumulative returns (CR) — IPOs Minimum: 0.25; Market Minimum: -0.59; CAPM Minimum: -0.65
  - Cumulative returns (CR) — IPOs Standard deviation: 3.53; Market Standard deviation: 1.23; CAPM Standard deviation: 0.80
  - Cumulative returns (CR) — IPOs Standard skewness: 3.40; Market Standard skewness: 2.52; CAPM Standard skewness: 0.60
  - Cumulative returns (CR) — IPOs Standard kurtosis: 2.60; Market Standard kurtosis: 2.85; CAPM Standard kurtosis: -0.16
  - Buy-and-hold returns (BHR) — IPOs Mean: 3.90; Market Mean: 1.1; CAPM Mean: 0.72
  - Buy-and-hold returns (BHR) — IPOs Median: 2.50; Market Median: 0.77; CAPM Median: 0.53
  - Buy-and-hold returns (BHR) — IPOs Maximum: 14.62; Market Maximum: 7.45; CAPM Maximum: 4.30
  - Buy-and-hold returns (BHR) — IPOs Minimum: -0.01; Market Minimum: -0.20; CAPM Minimum: -0.22
  - Buy-and-hold returns (BHR) — IPOs Standard deviation: 3.70; Market Standard deviation: 1.50; CAPM Standard deviation: 0.90
  - Buy-and-hold returns (BHR) — IPOs Standard skewness: 3.34; Market Standard skewness: 7.04; CAPM Standard skewness: 5.33
  - Buy-and-hold returns (BHR) — IPOs Standard kurtosis: 2.24; Market Standard kurtosis: 13.90; CAPM Standard kurtosis: 9.4

- Panel C: 12-month cumulative returns and buy-and-hold returns excluding initial returns (29 firms)
  - Cumulative returns (CR) — IPOs Mean: 0.20; Market Mean: 0.71; CAPM Mean: 0.46
  - Cumulative returns (CR) — IPOs Median: 0.27; Market Median: 0.94; CAPM Median: 0.40
  - Cumulative returns (CR) — IPOs Maximum: 1.32; Market Maximum: 3.09; CAPM Maximum: 1.80
  - Cumulative returns (CR) — IPOs Minimum: -1.0; Market Minimum: -0.96; CAPM Minimum: -0.98
  - Cumulative returns (CR) — IPOs Standard deviation: 0.58; Market Standard deviation: 0.98; CAPM Standard deviation: 0.79
  - Cumulative returns (CR) — IPOs Standard skewness: -0.41; Market Standard skewness: 0.17; CAPM Standard skewness: -0.26
  - Cumulative returns (CR) — IPOs Standard kurtosis: -0.22; Market Standard kurtosis: -0.13; CAPM Standard kurtosis: -0.86
  - Buy-and-hold returns (BHR) — IPOs Mean: 0.23; Market Mean: 0.46; CAPM Mean: 0.31
  - Buy-and-hold returns (BHR) — IPOs Median: 0.19; Market Median: 0.51; CAPM Median: 0.24
  - Buy-and-hold returns (BHR) — IPOs Maximum: 2.04; Market Maximum: 2.30; CAPM Maximum: 1.33
  - Buy-and-hold returns (BHR) — IPOs Minimum: -0.71; Market Minimum: -0.46; CAPM Minimum: 0.45
  - Buy-and-hold returns (BHR) — IPOs Standard deviation: 0.67; Market Standard deviation: 0.64; CAPM Standard deviation: 0.50
  - Buy-and-hold returns (BHR) — IPOs Standard skewness: 2.20; Market Standard skewness: 1.63; CAPM Standard skewness: 0.85
  - Buy-and-hold returns (BHR) — IPOs Standard kurtosis: 1.25; Market Standard kurtosis: .98; CAPM Standard kurtosis: -0.82

### Distributional diagnostics and test implications
- Standardized skewness and kurtosis tests are used to assess normality; values outside the range of -2 to +2 indicate significant departures from normality.
- The results show that standardized skewness and kurtosis for some return measures are outside the range of -2 to +2; these variables are not normally distributed.
- Resulting implication: the parametric test for significant difference in means should be interpreted with caution.
- Robustness: a chi-square test and a Shapiro-Walks W test for normality were run; both yielded similar results.

### Test statistics and inferential approach
- The null hypothesis that cross-sectional average initial and aftermarket abnormal returns are equal to zero is tested using parametric t-type statistics defined in equations (13)–(17), where sample averages and cross-sectional sample standard deviations of initial abnormal returns, aftermarket CAR and BHAR (risk unadjusted and CAPM-adjusted) are used.
- Under normality, these statistics follow a Student’s t-distribution.
- Given positive skewness in some measures, a non-parametric Wilcoxon signed-rank test is also used to test the null that the median of abnormal returns is zero.
- The Mann-Whitney test is used to compare medians across IPO sub-samples.
- Both parametric and non-parametric tests are used for robustness; however, findings from the parametric t-test should be treated with caution when corresponding return measures are not normally distributed.

### Determinants of initial abnormal returns (conceptual and model)
- Ex-ante uncertainty (expected sign: positive)
  - Measured by the standard deviation of daily returns of each individual IPO over twenty-five days following listing.
- Oversubscription (expected sign: positive)
  - Measured by the multiple of IPO shares’ demand on supply.
- Proportion of shares offered (expected sign: negative)
  - Measured as the percentage of shares offered through the IPO over the total shares of the company.
- Market volatility (expected sign: positive)
  - Measured by the standard deviation of daily market returns over the two-month period prior to the closing of the IPO subscription period.
- Market sentiment (expected sign: positive)
  - Controlled by Timing (TIM) dummy: equals one if IPO in a booming year (2003–05 bullish period) and zero otherwise.
  - Controlled by Country (CONT) dummy: equals one for firms affiliated to booming markets (Qatar, Saudi Arabia, U.A.E.) and zero otherwise (Bahrain, Kuwait, Oman).
- Type of industry (expected sign: positive for regulated industries)
  - Controlled by IND dummy: equals one if firm is classified as regulated industry and zero otherwise.
- Initial abnormal returns regression (equation 18):
  - ar_i = α + β1 EXANTE_i + β2 OVRS_i + β3 PSO_i + β4 MV_i + β5 TIM_i + β6 CONT_i + β7 IND_i + ε_i

### Determinants of aftermarket (long-run) abnormal returns (conceptual and model)
- Similar independent variables as for initial returns, with three notable differences:
  - Oversubscription: expected sign negative for long-run abnormal returns.
  - Country and industry dummies: expected sign to reverse as sentiment cools and actual profitability is revealed.
  - Initial abnormal return (ar_i): expected inverse relationship with long-run performance.
- Aftermarket regression (equation 19):
  - AFTMARKAR_T,i = α + β1 ar_i + β2 EXANTE_i + β3 OVRS_i + β4 PSO_i + β5 MV_i + β6 CONT_i + β7 IND_i + ε_i
  - Ti_AFTMARKAR,i denotes aftermarket abnormal return of firm i over T period (one-year).

### Transition to empirical results
- The paper proceeds to present empirical results, beginning with section V.A Initial Abnormal Returns.

*Source: IMF working paper content unit (excerpt provided).*

### 42.      The mean initial abnormal return for the IPO sample is 290 percent (Table 5,

### 42. The mean initial abnormal return for the IPO sample is 290 percent (Table 5, Panel A)

### Initial abnormal returns: magnitude and distribution
- Mean initial abnormal return for the IPO sample is 290 percent.
- Median initial abnormal return is around 147 percent.
- t-statistic: mean initial abnormal return is significantly different from zero at the one-percent level (t-stat shown in Table 5, Panel A: 6.14***).
- Wilcoxon signed-rank (non-parametric) test obtains similar results; median test statistic rejects the null of no significant median initial abnormal returns at the one-percent level (z-stat shown in Table 5, Panel A: 5.89***).
- Proportion of IPOs with positive initial abnormal returns: around 98 percent (46 out of 47).
- Mean WR ratio: around 3.32 (implying an investor would have had to invest 70 percent less in each IPO than in each corresponding reference portfolio to achieve the same wealth after the close of the first trading day).
- Evidence indicates GCC IPOs are underpriced and the level of underpricing is much higher than in other countries.

### Cross-sample comparisons of initial abnormal returns (Table 5 summaries)
- By country (booming vs non-booming markets; Panel B):
  - Number of IPOs: 33 (booming) and 14 (non-booming) as reported in Table 5 Panel B header.
  - Mean (median) initial abnormal return in booming stock markets: 3.80 (2.4) — corresponding description in text: 380 (240) percent.
  - Mean (median) initial abnormal return in non-booming stock markets: 0.8 (0.6) — corresponding description in text: 80 (60) percent.
  - Tests: both t-statistic and Mann-Whitney confirm significant difference at the one-percent level (Table 5, Panel B; Mann-Whitney average rank reported as 29.4-11.30***).
- By year (booming years vs non-booming years; Panel C):
  - Number of IPOs: 34 (booming years) and 13 (non-booming years) per Panel C header.
  - Mean (median) initial abnormal return in booming years: 3.5 (2.1) — described as higher in text.
  - Mean (median) initial abnormal return in non-booming years: 1.32 (1.22).
  - Difference significant (t-stat and Mann-Whitney), with Mann-Whitney average rank 26.24-18.15* and t-stat marked 2.16**.
- By oversubscription (Panel D):
  - Number of IPOs above (below) median oversubscription: 25 (22).
  - Mean initial abnormal returns above: 3.04 (median 1.93); below: 2.73 (median 0.97).
  - Difference in medians: 0.31 (0.96); test statistics do not find significant differences (Panel D).
- By industry regulation (Panel E):
  - Number of IPOs in regulated (non-regulated) industries: 30 (17).
  - Mean (median) initial abnormal return regulated: 3.39 (1.73); non-regulated: 2.03 (1.23).
  - Difference in medians: 1.40 (0.5); difference not significant at conventional levels (Panel E).

### Determinants of initial abnormal returns — multivariate results (Table 6)
- Dependent variable: Initial abnormal returns.
- Regression results (coefficients and t-statistics reported exactly as in Table 6):
  - Intercept: Coefficient -4.96; t-statistic (-2.17)**.
  - EXANTE: Coefficient 45.8; t-statistic (1.18).
  - PSO: Coefficient 3.1; t-statistic (1.2).
  - MV: Coefficient -2.94; t-statistic (-0.06).
  - OVRS: Coefficient 0.0004; t-statistic (0.17).
  - CONT: Coefficient 2.82; t-statistic (3.02)***.
  - TIM: Coefficient 1.96; t-statistic (2.37)**.
  - IND: Coefficient 2.32; t-statistic (2.47)**.
- Model fit and diagnostics:
  - R2 (percent): 43.85.
  - Adj. R2 (percent): 32.03.
  - F-value: 3.71***.
  - Number of observations: 47.
  - Heteroskedasticity detected by Breusch-Pagan LM and White’s test; White’s heteroskedasticity consistent standard errors (HCSE) used for robust variance estimates.
- Interpretation (textual findings preserved exactly):
  - Traditional explanatory variables identified in the literature do not have a measurable impact on GCC IPO underpricing.
  - Country-specific characteristics (CONT), timing of offers (TIM), and industry nature (IND) have a significant impact on underpricing at the one- and five-percent levels.
  - Firms that went public in bull markets, during booming years, and from regulated industries experience higher underpricing.
  - The level of oversubscription (OVRS) and proportion of shares offered (PSO) do not seem to affect excess initial returns.
  - Market volatility/ex-ante uncertainty (EXANTE, MV) does not seem to affect the level of excess initial returns.
  - Conclusion: GCC investors appear mainly driven by market trends (booming or not) and the regulated nature of the firm’s sector rather than by risk, expected demand, or signaling.

### Aftermarket abnormal performance — objectives and key one-year results (Table 7)
- Research questions addressed:
  - Do IPOs sustain initial abnormal returns and provide positive abnormal returns over a long period?
  - What exogenous variables explain long-run IPO performance?
- Panel A (12 months cumulative and buy-and-hold abnormal returns including initial returns; 29 firms):
  - Firms with > 0 abnormal return: CAR model counts 23; BHAR counts 27 (Table 7 Panel A headers show variations across models).
  - Firms with ≤ 0 abnormal return: CAR 6; BHAR 2.
  - Mean abnormal returns: CAR 2.75; BHAR 2.8 (CAPM column shows 3.1 and 3.2 respectively).
  - t-statistics: 4.03***, 4.2*** (CAPM columns 4.7***, 4.9***).
  - Median abnormal returns: 1.1 and 1.8 (CAPM medians 1.45 and 2.0).
  - z-statistics: 3.90***, 4.2*** (CAPM 4.7***, 4.6***).
  - Mean WR ratio (including initial returns): 2.33; Median WR ratio: 1.98.
- Panel B (12 months cumulative and buy-and-hold abnormal returns excluding initial returns; 29 firms):
  - Firms with > 0 abnormal return: CAR 7; BHAR 9 (CAPM 10 and 14).
  - Firms with ≤ 0 abnormal return: CAR 22; BHAR 20 (CAPM 19 and 15).
  - Mean abnormal returns excluding initial returns: CAR -0.50; BHAR -0.23 (CAPM -0.25 and -0.07).
  - t-statistics: -3.6***, -2.1** (CAPM -2.4**, -0.82).
  - Median abnormal returns: -0.47, -0.25 (CAPM -0.21, -0.08).
  - z-statistics: -3.1***, -2.00** (CAPM -2.20**, -0.73).
  - Mean WR ratio excluding initial returns: 0.84; Median WR ratio: 0.79.
- Cross-group aftermarket comparisons (Panels C–R highlights):
  - By country (Panels C–F): differences between booming and non-booming markets in cumulative and buy-and-hold abnormal returns are small and generally not statistically significant (example values: Panel C means -0.51 and -0.48; difference -0.03).
  - By oversubscription (Panels G–J): differences between below- and above-median oversubscription groups show mean/median differences often negative (example Panel G means -0.64 vs -0.38; difference -0.26); Panel J shows a Mann-Whitney result significant at 5 percent (average rank 11.6-18.1**).
  - By industry (Panels K–N): regulated industry IPOs show worse aftermarket cumulative abnormal returns than non-regulated in several panels (Panel K mean regulated -0.68 vs non-regulated -0.1; t-stat for difference -2.03**; Panel L buy-and-hold difference t-stat -1.84*).
  - By initial abnormal returns median (Panels O–R): differences in aftermarket performance between below- and above-median initial abnormal return groups are generally small and not significant (example Panel O means -0.53 vs -0.48; difference -0.06).
- Overall aftermarket conclusion (preserved from text and tables):
  - Including initial returns, 12-month cumulative and buy-and-hold abnormal returns are positive and significant.
  - Excluding initial returns, 12-month cumulative and buy-and-hold abnormal returns are negative and, for several models/statistics, significantly below zero.
  - Aftermarket underperformance (excluding initial returns) is evident in the sample over the one-year horizon.

*Source: _wp07149 — content as supplied from the IMF PDF excerpt (Tables 5–7 and accompanying text).*

### 48.      Including initial returns, GCC IPOs yielded positive returns over a one-year period

### 48.      Including initial returns, GCC IPOs yielded positive returns over a one-year period

### Aftermarket performance including initial returns
- Including initial returns, GCC IPOs yielded positive returns over a one-year period (Table 7, Panel A).
- Both the parametric test statistics and the non-parametric Wilcoxon signed-rank test are significant at the one-percent level for all models.
- The average aftermarket abnormal return for the investor is around 280 percent using both BHAR and CAR.
- The average WR ratio is around 2.33, implying an investor would have had to invest 57 percent less in each IPO than in each corresponding reference portfolio to achieve the same wealth after one year of official listing.

### Aftermarket performance excluding initial returns
- Excluding initial returns, aftermarket abnormal returns of GCC IPOs over a one-year period have been significant and negative (Table 7, Panel B).
- The null hypothesis that the mean (median) abnormal returns are not different from zero is rejected at the one and five percent levels for most models.
- Both the t-statistic and z-statistic are negative, implying IPOs underperform the market in the long run.
- The IPOs average WR ratio is 0.84, implying an investor would have to invest 19 percent more to get the same performance as the market.
- Reported long-run abnormal return ranges (excluding initial returns) cited elsewhere in the paper: ranging from -23 percent to -50 percent (see VI. Summary and concluding remarks).

### Subsample splits and robustness checks
- Market climate (country): Firms that went public in booming markets (Qatar, Saudi Arabia, and the U.A.E.) underperformed those that went public in less “hot markets,” although the parameters are not significant at any level and under no model (Table 7, Panels C–F).
- Oversubscription levels: Firms that experience higher oversubscription levels provide investors with higher long-run abnormal returns; this relationship is significant at the five-percent level only for the BHAR method using the CAPM model (Table 7, Panels G–J).
  - Authors note this contradicts theoretical expectations and some literature (Agarwal, Liu, and Rhee (2006)).
  - Possible explanation: large oversubscription multiples may reflect relatively small firm or offering size vis-à-vis excess liquidity, not necessarily large underpricing.
- Industry classification: Firms classified as regulated industries perform worse than those related to unregulated industries. The relationship is significant only using the risk-unadjusted models at the one- and five- percent levels (Table 7, Panels K–N). This may result from higher levels of underpricing of IPOs in regulated industries (Table 5, Panel E).
- Initial abnormal returns level: No significant difference between firms that experienced higher initial abnormal returns and those with lower initial abnormal returns (Table 7, Panels O–R).

### Determinants of aftermarket abnormal returns (multivariate regression results)
- Regression specification: Equation (19) estimated by OLS after excluding three influential observations, reducing sample from 29 to 26 observations.
- Exogenous variables: same as used for initial abnormal returns, adding initial abnormal returns and dropping the timing variable.
- Key significant coefficients from Table 8 (Panel A: Abnormal Results Model; dependent variable: Aftermarket abnormal returns):
  - MV (market volatility): Coefficient 65.73 with t-statistic (3.6)*** for CAR; Coefficient 40.21 with t-statistic (1.98)* for BHAR.
  - CONT (market sentiment / country): Coefficient -0.57 with t-statistic (-2.13)** for CAR; Coefficient -0.15 with t-statistic (-0.49) for BHAR.
  - IND (industry classification): Coefficient -1.1 with t-statistic (-4.51)*** for CAR; Coefficient -0.71 with t-statistic (-2.59)** for BHAR.
  - R-squared: 43.9 percent (CAR) and 40.16 percent (BHAR). Adjusted R-squared: 25.19 percent (CAR) and 20.22 percent (BHAR). F-value: 2.35* (CAR) and 2.01* (BHAR).
  - Number of observations: 26.
- Key significant coefficients from Table 8 (Panel B: CAPM; dependent variable: Aftermarket abnormal returns):
  - MV: Coefficient 37.78 with t-statistic (2.16)** for CAR; Coefficient 21.7 with t-statistic (1.15) for BHAR.
  - CONT: Coefficient -0.56 with t-statistic (-2.17)** for CAR; Coefficient -0.09 with t-statistic (-0.34) for BHAR.
  - IND: Coefficient -0.54 with t-statistic (-2.29)** for CAR; Coefficient -0.31 with t-statistic (-1.21) for BHAR.
  - R-squared: 34.72 percent (CAR) and 33.66 percent (BHAR). Adjusted R-squared: 12.69 percent (CAR) and 11.55 percent (BHAR). F-value: 1.6 (CAR) and 1.52 (BHAR).
  - Number of observations: 26.
- Interpretation: Aftermarket performance over one year appears driven by market volatility (MV), market sentiment (CONT), and industry classification (IND). Negative signs on CONT and IND indicate firms that went public in booming markets and from regulated industries provided investors with negative long-run abnormal returns.
- Statistical significance notation: ***, **, and * Significant at the 1, 5 and 10 percent levels, respectively.

### Summary of empirical magnitudes and overall conclusions
- Initial abnormal returns averaged 290 percent.
- Aftermarket performance including initial returns: positive average abnormal returns between 275 and 280 percent.
- Aftermarket performance excluding initial returns: negative, ranging from -23 percent to -50 percent relative to the benchmark portfolio.
- Practical implication: Investing one dollar in IPOs at the end of the first trading day and holding it for one year would have left an investor with only 77 to 50 percent of the return of each dollar invested in GCC stock exchange general indices.
- Main drivers identified: market excitement (bullish or bearish) and the regulated nature of the IPO’s economic sector, rather than risk, expected demand, or firm signaling.
- Authors note potential sample-period influence: majority of IPOs in multivariate regressions took place during booming years (particularly 2005), while their one-year anniversaries coincided with a severe correction in most GCC equity markets. Further research extending the sample period beyond one year and widening the study to other equity markets is recommended.

### Policy recommendations
- Implement measures to avoid market volatility to smooth aftermarket abnormal negative results.
- Address structural shortcomings of GCC equity markets:
  - Lack or small size of traditional institutional investors.
  - Small free-float of shares.
  - Lack of widespread financial culture among retail investors.
  - Relatively new and incomplete regulatory and supervisory frameworks.
- Emphasize creating incentives for development of institutional investors with a long-term investment profile to counterbalance short-term euphoria.
  - Note: this includes opening investment markets to foreign investment, currently limited to a few countries and economic sectors.
- Educate retail investors and create professional, accountable financial advisory firms to reduce investor over-optimism.
- Strengthen regulations and ensure compliance from investors, intermediaries, and listed companies to ensure complete and accurate information and penalize price manipulation practices.
- Reconsider policy of underpricing IPOs as a mechanism to share oil wealth with the population, since it may have negative side effects on market stability and price formation; consider alternative distribution mechanisms that do not interfere with price formation.

*Source: _wp07149 - 48.      Including initial returns, GCC IPOs yielded positive returns over a one-year period*

### APPENDIX I: DETAILED INFORMATION ON THE PERFORMANCE OF IPOS BY COUNTRY

### APPENDIX I: DETAILED INFORMATION ON THE PERFORMANCE OF IPOS BY COUNTRY

### Overview
- This appendix table provides information on initial returns and buy-and-hold aftermarket performance of the IPOs over a one-year period excluding and including initial returns.
- Information is provided on each country and each IPO, and a simple average return of each country.
- Notes:
  - 1/ BHR Excluding refers to one-year buy-and-hold return excluding initial returns.
  - 2/ BHR Including refers to one-year buy-and-hold return including initial returns.

### Bahrain
- Firms and performance:
  - Al Khaleej Development Company — Date of subscription: Feb-05 — Initial Return (In percent): 100 — Excluding 1/ (In percent): 53 — Including 2/ (In percent): 205
  - Nass Corporation — Date of subscription: Oct-05 — Initial Return (In percent): 14 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Banader Hotels Company — Date of subscription: Oct-05 — Initial Return (In percent): -30 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Ithmaar Bank — Date of subscription: Feb-06 — Initial Return (In percent): 0 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Al Salam Bank — Date of subscription: Feb-06 — Initial Return (In percent): 138 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 44
  - Average Excluding 1/ (In percent): 53
  - Average Including 2/ (In percent): 205

### Kuwait
- Firms and performance:
  - First Takaful Insurance Company — Date of subscription: Jun-03 — Initial Return (In percent): 180 — Excluding 1/ (In percent): 27 — Including 2/ (In percent): 255
  - Boubyan Bank — Date of subscription: Apr-04 — Initial Return (In percent): 480 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 330
  - Average Excluding 1/ (In percent): 27
  - Average Including 2/ (In percent): 255

### Oman
- Firms and performance:
  - Al Maha Petroleum Products Marketing — Date of subscription: Apr-04 — Initial Return (In percent): 58 — Excluding 1/ (In percent): 38 — Including 2/ (In percent): 118
  - Al Kamil Power Company — Date of subscription: Jul-04 — Initial Return (In percent): 71 — Excluding 1/ (In percent): -5 — Including 2/ (In percent): 63
  - Dhofar Power PRE — Date of subscription: Apr-05 — Initial Return (In percent): 32 — Excluding 1/ (In percent): -9 — Including 2/ (In percent): 20
  - AES Barka — Date of subscription: Jan-05 — Initial Return (In percent): 88 — Excluding 1/ (In percent): 34 — Including 2/ (In percent): 153
  - Oman Telecommunications Company (Omantel) — Date of subscription: Jun-05 — Initial Return (In percent): 91 — Excluding 1/ (In percent): -48 — Including 2/ (In percent): -1
  - Tageer Finance — Date of subscription: Sep-05 — Initial Return (In percent): 9 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Sahara Hospitality — Date of subscription: Aug-01 — Initial Return (In percent): 18 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 52
  - Average Excluding 1/ (In percent): 2
  - Average Including 2/ (In percent): 71

### Qatar
- Firms and performance:
  - Qatar Fuel Company (Woqod) — Date of subscription: Mar-02 — Initial Return (In percent): 275 — Excluding 1/ (In percent): 11 — Including 2/ (In percent): 316
  - Qatar Technical Inspection — Date of subscription: May-02 — Initial Return (In percent): 572 — Excluding 1/ (In percent): 20 — Including 2/ (In percent): 706
  - Industries Qatar — Date of subscription: Jun-03 — Initial Return (In percent): 511 — Excluding 1/ (In percent): -12 — Including 2/ (In percent): 440
  - Qatar Company for Meat and Livestock Trading — Date of subscription: Sep-03 — Initial Return (In percent): 109 — Excluding 1/ (In percent): 34 — Including 2/ (In percent): 180
  - Gulf Warehousing Company — Date of subscription: Dec-03 — Initial Return (In percent): 131 — Excluding 1/ (In percent): 174 — Including 2/ (In percent): 534
  - Qatar Gas Transport (Nakilat) — Date of subscription: Jan-05 — Initial Return (In percent): 609 — Excluding 1/ (In percent): -55 — Including 2/ (In percent): 222
  - Dlala Brokerage and Investment Holding — Date of subscription: May-05 — Initial Return (In percent): 1,031 — Excluding 1/ (In percent): -60 — Including 2/ (In percent): 349
  - Barwa Real Estate Company — Date of subscription: Dec-05 — Initial Return (In percent): 286 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Rayan Bank — Date of subscription: Jan-06 — Initial Return (In percent): 70 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 399
  - Average Excluding 1/ (In percent): 16
  - Average Including 2/ (In percent): 392

### Saudi Arabia
- Firms and performance:
  - Saudi Telecom Company — Date of subscription: Jan-03 — Initial Return (In percent): 39 — Excluding 1/ (In percent): 77 — Including 2/ (In percent): 147
  - Sahara Petrochemical Company — Date of subscription: Jun-04 — Initial Return (In percent): 200 — Excluding 1/ (In percent): 204 — Including 2/ (In percent): 812
  - Etihad Etisalat Company — Date of subscription: Dec-04 — Initial Return (In percent): 500 — Excluding 1/ (In percent): 139 — Including 2/ (In percent): 1,336
  - National Company for Cooperative Insurance — Date of subscription: Jan-05 — Initial Return (In percent): 81 — Excluding 1/ (In percent): 87 — Including 2/ (In percent): 239
  - Bank Albilad — Date of subscription: Mar-05 — Initial Return (In percent): 1,430 — Excluding 1/ (In percent): 2 — Including 2/ (In percent): 1,462
  - Saudia Dairy and Foodstuff Company — Date of subscription: May-05 — Initial Return (In percent): 94 — Excluding 1/ (In percent): -39 — Including 2/ (In percent): 18
  - Almarai Company — Date of subscription: Jul-05 — Initial Return (In percent): 53 — Excluding 1/ (In percent): 39 — Including 2/ (In percent): 112
  - National Petrochemicals Company (Yansab) — Date of subscription: Dec-05 — Initial Return (In percent): 1,120 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Aldrees Petroleum and Transport Services Company — Date of subscription: Jan-06 — Initial Return (In percent): 87 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Saudi Research and Marketing Group — Date of subscription: Apr-06 — Initial Return (In percent): 106 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Saudi Paper Manufacturing — Date of subscription: May-06 — Initial Return (In percent): 130 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 349
  - Average Excluding 1/ (In percent): 73
  - Average Including 2/ (In percent): 590

### UAE
- Firms and performance:
  - Dubai Islamic Insurance and Reinsurance Company — Date of subscription: Oct-02 — Initial Return (In percent): 387 — Excluding 1/ (In percent): 17 — Including 2/ (In percent): 473
  - Amlak Finance — Date of subscription: Jan-04 — Initial Return (In percent): 152 — Excluding 1/ (In percent): 67 — Including 2/ (In percent): 320
  - Arab Technical Construction — Date of subscription: Aug-04 — Initial Return (In percent): 392 — Excluding 1/ (In percent): 49 — Including 2/ (In percent): 633
  - Arab International Logistics Company — Date of subscription: Mar-05 — Initial Return (In percent): 294 — Excluding 1/ (In percent): -12 — Including 2/ (In percent): 250
  - Al Dar Properties — Date of subscription: Oct-04 — Initial Return (In percent): 650 — Excluding 1/ (In percent): 19 — Including 2/ (In percent): 792
  - Emirates Foodstuff — Date of subscription: Jan-05 — Initial Return (In percent): 641 — Excluding 1/ (In percent): -71 — Including 2/ (In percent): 117
  - Finance House — Date of subscription: Apr-04 — Initial Return (In percent): 1,281 — Excluding 1/ (In percent): -31 — Including 2/ (In percent): 851
  - Abu Dhabi National Energy — Date of subscription: Jul-05 — Initial Return (In percent): 611 — Excluding 1/ (In percent): -60 — Including 2/ (In percent): 185
  - Sorouh Real Estate Company — Date of subscription: May-05 — Initial Return (In percent): 564 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - RAK Properties — Date of subscription: Apr-05 — Initial Return (In percent): 320 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Dana Gas — Date of subscription: Sep-05 — Initial Return (In percent): 307 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Aabar Petroleum Investments Company — Date of subscription: Apr-05 — Initial Return (In percent): 421 — Excluding 1/ (In percent): — Including 2/ (In percent):
  - Emirates Integrated Telecommunications Co. — Date of subscription: Mar-06 — Initial Return (In percent): 125 — Excluding 1/ (In percent): — Including 2/ (In percent):
- Country average:
  - Average Initial Return (In percent): 473
  - Average Excluding 1/ (In percent): -3
  - Average Including 2/ (In percent): 453

*APPENDIX I: DETAILED INFORMATION ON THE PERFORMANCE OF IPOS BY COUNTRY*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp07149.pdf_