**5. Empirical results**

#### **5.1 Short-term return analysis**

**Table 2** presents the estimations of IPO ETFs'short-term performance. Specifically, the table reports the three types of initial returns, that is, the first-day absolute, benchmark-adjusted, and abnormal return along with the corresponding average returns over the first 2, 3, 4, 5, 21, and 63 trading days.

As far as absolute returns are concerned, **Table 2** shows that the average initial return of the sample is positive amounting to 29 bps. This positive mean term indicates a favorable response to the launch of these alternative investing tools on behalf of investors. However, it should be noted that, when focusing on the performance of individual funds, we can see that the absolute initial return can be either negative or positive and ranges from �0.35% for the Renaissance International IPO ETF to 1.45% for the Renaissance IPO ETF.

After the first trading day, the average absolute return of the sample remains positive up to the first 5 days and becomes negative when the 1- and 3-month periods are assessed. In addition, after the third day, average absolute return starts deteriorating. At the fund level, most of the longer-term average returns are positive for the Renaissance International IPO ETF and the First Trust US Equity Opportunities ETF, they are steadily negative for the Renaissance International IPO ETF, while returns are mixed in the case of the First Trust International IPOX ETF.

When it comes to benchmark-adjusted returns, **Table 2** reports a positive average initial return for the sample, either when the S&P 500 Index or the S&P 600 Small Cap Index is the benchmark into consideration. In addition, when the latter index is used, the average initial return of the sample is about double the respective return when the S&P 500 Index is used to compute the benchmark-adjusted returns of IPO ETFs. At the fund level, only the First Trust International IPOX ETF produces negative benchmark-adjusted return, whereas the rest of ETFs can beat the market on their first trading day.

benchmark by 12 bps. As a comment on tracking error, we should point out that the literature has accentuated that tracking error is an unavoidable event for ETFs given that their returns are usually calculated free of expenses and transaction costs while

**Ticker Name Benchmark Inception**

*Linear and Non-Linear Financial Econometrics - Theory and Practice*

IPO Index

Renaissance International IPO Index

IPOX®-100 US Index

IPOX International Index

IPO Index

Renaissance International IPO Index

IPOX®-100 US Index

IPOX International Index

IPO Index

Renaissance International IPO Index

IPOX®-100 US Index

IPOX International Index

**Mean 0.815 0.844 0.689** *This table presents the profiles of IPO ETFs, which include their ticker, name, benchmark, inception date, age as at October 31, 2016, expense ratio, average daily volume, historical tracking error, i.e., difference in performance between ETF and its benchmark, since each ETF's inception up to December 31, 2015, trading frequency calculated as the fraction of trading days with nonzero volume to the total trading history (in days) for each fund, average intraday volatility calculated as the percentage fraction of the highest minus the lowest trading price of each fund on day t to its close price on the same day, and the fraction of each ETF's intraday volatility to the intraday volatility of*

**Ticker Name Benchmark Intraday**

IPO Renaissance IPO ETF Renaissance US

**Mean 16,339** �**0.56% 83.12%**

**volatility**

**Mean 4.418 0.68% Ticker Name Benchmark Volume Tracking error Trading freq.**

IPO Renaissance IPO ETF Renaissance US

International IPO ETF

International IPOX ETF

International IPO ETF

International IPOX ETF

International IPO ETF

International IPOX ETF

*the S&P 500 Index and the S&P 600 Small Cap Index, respectively.*

IPO Renaissance IPO ETF Renaissance US

IPOS Renaissance

FPXI First Trust

IPOS Renaissance

FPXI First Trust

IPOS Renaissance

FPXI First Trust

**Table 1.** *Profiles of ETFs.*

**106**

FPX First Trust US Equity Opportunities ETF

FPX First Trust US Equity Opportunities ETF

FPX First Trust US Equity Opportunities ETF

**date**

**Age Expense ratio**

10/14/13 3.049 0.60%

10/6/2014 2.071 0.80%

4/12/2006 10.562 0.60%

11/4/2014 1.992 0.70%

13,273 �1.12% 99.61%

461 0.12% 55.36%

37,016 �0.68% 98.87%

14,605 �0.56% 78.64%

**ETF Int. Vol/ S&P 500 Int. Vol**

1.101 1.232 0.976

0.486 0.568 0.515

1.098 0.934 0.740

0.574 0.644 0.527

**ETF Int. Vol/ S&P 600 Int. Vol**


The abnormal returns of ETFs behave similarly to benchmark-adjusted returns. The average first-day abnormal return of the sample is positive irrespective of the index incorporated in the market model to estimate abnormal returns. In addition, the average abnormal return of the sample based on the S&P 600 Small Cap Index is more than three times the respective average return when the S&P 500 Index is taken into consideration. Scanning through the single ETF returns, we observe that three out of four funds deliver significant first-day profits, which range from 17 bps for the Renaissance International IPO ETF to 20 bps for the Renaissance IPO ETF. At the longer-run level, abnormal returns mimic the benchmark-adjusted returns quite closely. For instance, similarly to benchmark-adjusted returns, the average 2-day abnormal returns based on the S&P 500 Index is negative for three out of four ETFs (the same ETFs as in the case of benchmark-adjusted performance). Returns over other frequencies up to 5 days are either negative or positive without allowing us to detect any specific pricing trend. However, when it comes to

*IPO ETFs: An Alternative Way to Enter the Initial Public Offering Business*

*DOI: http://dx.doi.org/10.5772/intechopen.90269*

longer periods reaching one or 3 months, average returns become negative.

investors with a short-term horizon, which does not exceed 3 months.

cannot compute returns over a 24-month period for this fund.

absolute return only over the first 6 months of trading.

underperform the S&P 500 Index (the S&P 600 Small Cap Index).

**5.2 Long-term return analysis**

form.

**109**

The main conclusion that can be reached by the analysis of the various types of short-term returns is that, on average, significant gains can be realized on the first trading day of IPO ETFs. However, gains diminish or even vanish when longer periods, such as 1 month or 3 months of trading, are assessed. Based on these findings, we could claim that IPO ETFs may be suitable for day traders but not for

The long-run performance of ETFs is assessed in this section. The relevant estimations of cumulative and buy-and-hold returns over a 6-month, 12-month, 18 month, and 24-month period are presented in **Table 3** along with the corresponding returns over the entire trading history of ETFs. Cumulative returns are presented from an absolute, benchmark-adjusted, and abnormal perspective. On the other hand, buy-and-hold returns are reported in their absolute and benchmark-adjusted

We note that, as shown in **Table 1**, the Renaissance International IPO ETF is about 2 years old. This means that the total trading history of this fund coincides with the 24-month subperiod considered in our analysis. Moreover, the trading history of the First Trust International IPOX ETF is less than 2 years and, thus, we

When it comes to cumulative absolute returns, the results in **Table 3** indicate that IPO ETFs can produce positive such returns. In particular, the average return over the first 6 months of trading approximates 30 bps. Average returns are also positive for the rest intervals considered as well as over the entire trading history of ETFs. In addition, except for the First Trust International IPOX ETF, the funds of the sample provide positive cumulative returns over the several time periods assessed. The First Trust International IPOX ETF presents a positive cumulative

When the cumulative benchmark-adjusted returns are considered, **Table 3** reports mixed results. For example, the average historical return (i.e., over the total trading history up to October 31, 2016) of the sample is positive when the S&P 500 Index is taken into consideration, but it becomes negative when the S&P 600 Small Cap Index is the reference portfolio. Moreover, three (two) out of four funds

In the case of cumulative abnormal returns, the results in **Table 3** indicate that the average IPO ETF underperforms the S&P 500 Index, but it performs better than the S&P 600 Small Cap Index up to 24 trading months. However, the average IPO

*This table presents the return of IPO ETFs on their first trading day, i.e., the first day with positive trading volume, calculated in percentage terms as the fraction of the close trade price minus the opening trade price to the opening trade price, as well as the average absolute (daily) return over the first 2, 3, 4, 5, 21 (i.e., 1 month), and 63 (i.e., 3 months) days of each ETF's trading history. Then, the table presents the benchmark-adjusted returns of ETFs, i.e., ETF return minus benchmark return, over the same intervals using as alternative benchmarks the S&P 500 Index and the S&P 600 Small Cap Index. Finally, the table presents the abnormal returns of IPO ETFs estimated with the market model.*

#### **Table 2.**

*Short-term return analysis.*

Furthermore, **Table 2** reports mixed results about longer-run benchmark-adjusted returns. For instance, when the 2-day period is considered and the S&P 500 Index is used as the benchmark, the returns of three ETFs are negative (two return estimates in the case of the S&P 600 Small Cap Index). In the case of the 3-day investment window, returns are mixed. In the case of the 4-day period and the S&P 500 Index, all the average returns are positive while over longer periods, whatever the benchmark may be, the majority of benchmark-adjusted returns are negative.

#### *IPO ETFs: An Alternative Way to Enter the Initial Public Offering Business DOI: http://dx.doi.org/10.5772/intechopen.90269*

The abnormal returns of ETFs behave similarly to benchmark-adjusted returns. The average first-day abnormal return of the sample is positive irrespective of the index incorporated in the market model to estimate abnormal returns. In addition, the average abnormal return of the sample based on the S&P 600 Small Cap Index is more than three times the respective average return when the S&P 500 Index is taken into consideration. Scanning through the single ETF returns, we observe that three out of four funds deliver significant first-day profits, which range from 17 bps for the Renaissance International IPO ETF to 20 bps for the Renaissance IPO ETF.

At the longer-run level, abnormal returns mimic the benchmark-adjusted returns quite closely. For instance, similarly to benchmark-adjusted returns, the average 2-day abnormal returns based on the S&P 500 Index is negative for three out of four ETFs (the same ETFs as in the case of benchmark-adjusted performance). Returns over other frequencies up to 5 days are either negative or positive without allowing us to detect any specific pricing trend. However, when it comes to longer periods reaching one or 3 months, average returns become negative.

The main conclusion that can be reached by the analysis of the various types of short-term returns is that, on average, significant gains can be realized on the first trading day of IPO ETFs. However, gains diminish or even vanish when longer periods, such as 1 month or 3 months of trading, are assessed. Based on these findings, we could claim that IPO ETFs may be suitable for day traders but not for investors with a short-term horizon, which does not exceed 3 months.

#### **5.2 Long-term return analysis**

The long-run performance of ETFs is assessed in this section. The relevant estimations of cumulative and buy-and-hold returns over a 6-month, 12-month, 18 month, and 24-month period are presented in **Table 3** along with the corresponding returns over the entire trading history of ETFs. Cumulative returns are presented from an absolute, benchmark-adjusted, and abnormal perspective. On the other hand, buy-and-hold returns are reported in their absolute and benchmark-adjusted form.

We note that, as shown in **Table 1**, the Renaissance International IPO ETF is about 2 years old. This means that the total trading history of this fund coincides with the 24-month subperiod considered in our analysis. Moreover, the trading history of the First Trust International IPOX ETF is less than 2 years and, thus, we cannot compute returns over a 24-month period for this fund.

When it comes to cumulative absolute returns, the results in **Table 3** indicate that IPO ETFs can produce positive such returns. In particular, the average return over the first 6 months of trading approximates 30 bps. Average returns are also positive for the rest intervals considered as well as over the entire trading history of ETFs. In addition, except for the First Trust International IPOX ETF, the funds of the sample provide positive cumulative returns over the several time periods assessed. The First Trust International IPOX ETF presents a positive cumulative absolute return only over the first 6 months of trading.

When the cumulative benchmark-adjusted returns are considered, **Table 3** reports mixed results. For example, the average historical return (i.e., over the total trading history up to October 31, 2016) of the sample is positive when the S&P 500 Index is taken into consideration, but it becomes negative when the S&P 600 Small Cap Index is the reference portfolio. Moreover, three (two) out of four funds underperform the S&P 500 Index (the S&P 600 Small Cap Index).

In the case of cumulative abnormal returns, the results in **Table 3** indicate that the average IPO ETF underperforms the S&P 500 Index, but it performs better than the S&P 600 Small Cap Index up to 24 trading months. However, the average IPO

Furthermore, **Table 2** reports mixed results about longer-run benchmark-adjusted returns. For instance, when the 2-day period is considered and the S&P 500 Index is used as the benchmark, the returns of three ETFs are negative (two return estimates in the case of the S&P 600 Small Cap Index). In the case of the 3-day investment window, returns are mixed. In the case of the 4-day period and the S&P 500 Index, all the average returns are positive while over longer periods, whatever the benchmark

**Absolute returns**

**Benchmark-adjusted returns**

**Period IPO IPOS FPX FPXI Mean IPO IPOS FPX FPXI Mean** t1 0.209 1.041 0.222 �0.285 **0.297** 1.500 0.925 0.056 �0.274 **0.552** t2 �0.012 �0.362 0.283 �0.299 �**0.098** 0.512 �0.492 0.027 �0.308 �**0.065** t3 0.247 �0.162 �0.099 0.146 **0.033** 0.409 �0.155 �0.528 0.154 �**0.030** t4 0.070 0.008 0.048 0.127 **0.063** 0.225 �0.006 �0.483 �0.050 �**0.078** t5 �0.026 0.286 0.113 0.091 **0.116** 0.151 �0.005 �0.236 0.133 **0.011** t21 �0.133 �0.171 �0.085 �0.097 �**0.122** �0.054 �0.398 �0.077 0.001 �**0.132** t63 0.029 �0.058 �0.065 �0.127 �**0.055** 0.051 �0.124 �0.009 �0.144 �**0.056 Abnormal returns**

**Benchmark: S&P 500 Index Benchmark: S&P 600 Small Cap Index**

**Benchmark: S&P 500 Index Benchmark: S&P 600 Small Cap Index**

**Period IPO IPOS FPX FPXI Mean IPO IPOS FPX FPXI Mean** t1 0.199 0.167 0.194 �0.236 **0.081** 1.495 0.043 0.083 �0.236 **0.346** t2 �0.017 �0.252 0.252 �0.183 �**0.050** 0.580 �0.275 0.028 �0.180 **0.038** t3 0.244 �0.521 �0.120 0.244 �**0.038** 0.516 �0.553 �0.321 0.253 �**0.026** t4 0.070 �0.442 0.025 0.200 �**0.037** 0.302 �0.490 �0.269 0.125 �**0.083** t5 �0.025 �0.281 0.089 0.161 �**0.014** 0.223 �0.425 �0.086 0.179 �**0.028** t21 �0.126 �0.095 �0.114 �0.032 �**0.092** �0.029 �0.150 �0.114 0.008 �**0.071** t63 0.038 �0.037 �0.096 �0.087 �**0.046** 0.071 �0.050 �0.075 �0.101 �**0.039** *This table presents the return of IPO ETFs on their first trading day, i.e., the first day with positive trading volume, calculated in percentage terms as the fraction of the close trade price minus the opening trade price to the opening trade price, as well as the average absolute (daily) return over the first 2, 3, 4, 5, 21 (i.e., 1 month), and 63 (i.e., 3 months) days of each ETF's trading history. Then, the table presents the benchmark-adjusted returns of ETFs, i.e., ETF return minus benchmark return, over the same intervals using as alternative benchmarks the S&P 500 Index and the S&P 600 Small Cap Index. Finally, the table presents the abnormal returns of IPO ETFs estimated with the market model.*

**Period IPO IPOS FPX FPXI Mean** t1 1.446 �0.348 0.299 �0.250 **0.287** t2 0.944 �0.184 0.175 �0.125 **0.202** t3 1.103 �0.731 0.398 0.285 **0.264** t4 0.714 �0.706 0.471 0.213 **0.173** t5 0.604 �0.614 0.475 0.171 **0.159** t21 0.092 �0.046 �0.072 �0.029 �**0.014** t63 0.163 �0.021 �0.119 �0.111 �**0.022**

*Linear and Non-Linear Financial Econometrics - Theory and Practice*

may be, the majority of benchmark-adjusted returns are negative.

**Table 2.**

**108**

*Short-term return analysis.*


**Buy-and-hold**

**Period**

**111**

6 months

12 months

18 months 24 months Total period

**Period** 6 months 12 months 18 months

24 months Total period *This table presents the cumulative absolute returns of IPO ETFs over the first 6, 12, 18, and 24 months of their trading history as well as over their entire trading history up to October 31, 2016. Cumulative*

*benchmark-adjusted*

*and*  **Table 3.** *Long-term*

 *return analysis.*

*benchmark-adjusted*

 *returns of ETFs over the several periods considered.*

 *and cumulative abnormal returns relative to the S&P 500 Index and the S&P 600 Small Cap Index, respectively,*

**IPO** 7.294 14.888

18.953

19.784

21.042

 5.357

 62.226

 2.834

 5.357

 1.639

 17.577

5.721

N/A

**10.287** **22.865**

14.830

 6.341

 *are presented too. Finally, the table presents the* 

 83.129

 6.763

**27.766**

*buy-and-hold*

 *absolute*

11.559

 6.341

 4.262

 13.405

 0.621 1.537

**9.926**

11.450

 1.004

 8.434 �3.974

 N/A

**4.642**

 4.919

**6.452**

3.539

5.987

2.033

**4.713** **8.294**

7.982

 3.313

 7.147

 4.422

**5.716**

3.428

 3.211

 0.213

 5.966

**IPOS**

**FPX**

**FPXI**

 **Mean**

 **IPO**

 **IPOS**

 **FPX**

 **FPXI**

 **Mean**

*IPO ETFs: An Alternative Way to Enter the Initial Public Offering Business*

*DOI: http://dx.doi.org/10.5772/intechopen.90269*

**3.204**

**IPO** 5.468

13.251

17.409

7.241

�0.246

�4.078

 158.585

�4.078

 19.417

�4.058

 25.445

 0.040

 14.773

 12.174

 3.261

**IPOS**

**FPX**

**FPXI**

2.549 �11.786 �15.965

N/A �11.786 **Buy-and-hold**

**Benchmark:**

 **S&P 500 Index**

**benchmark-adjusted**

 **returns** **Benchmark:**

 **S&P 600 Small Cap Index**

**35.619**

**7.527**

**5.708**

**4.070**

**5.863**

 **Mean**

 **absolute returns**


*IPO ETFs: An Alternative Way to Enter the Initial Public Offering Business DOI: http://dx.doi.org/10.5772/intechopen.90269*

> **Table 3.**

 *Long-term return analysis.*

**Cumulative**

**Period**

**110**

6 months 12 months 18 months 24 months Total period

**Period** 6 months

12 months

18 months 24 months Total period

**Period** 6 months 12 months 18 months

24 months Total period

**IPO** 0.004

0.107 0.180

�0.229

�0.434

�0.170

�1.301

 0.141

�0.170

�0.074

 N/A

�0.117

�0.178

�0.244

�**0.090** �**0.158** �**0.441**

�0.388

�0.048

�2.297

�0.119

�**0.713**

0.014

�0.048

 0.106

0.430

�0.013

 0.128

0.029

�0.262

�0.199

�**0.081**

0.369

0.117

�0.034

�0.368 �0.452

N/A

**0.024**

**0.024**

**0.021**

 0.483

**IPOS**

**FPX** �0.278

 0.276

**FPXI**

**Mean** **0.121**

0.151

0.528

�0.078

 0.157

**IPO**

**IPOS**

**FPX**

**FPXI**

**Mean**

**0.189**

�0.824

�0.337

 2.350

**Benchmark:**

 **S&P 500 Index**

**IPO** �0.058

�0.018

�0.008

�0.486

�0.337

 0.625

�0.182

 0.338

�0.105

 0.081

 0.380

**IPOS**

**FPX** �0.106

 0.027 �0.626

�0.899

N/A

�0.712 **Cumulative**

 **abnormal returns**

**0.119**

�0.623

�0.406

**Benchmark:**

 **S&P 600 Small Cap Index**

 1.287

�**0.066**

�0.141

�0.406

 1.012

�**0.188**

0.278

�0.170

 0.678

�**0.167**

0.281

�0.067

 0.315

**FPXI**

**Mean** **0.061**

0.105

0.387

0.134

�0.159 �0.808

�1.073

N/A �0.925

�**0.167**

**0.155**

�**0.072**

�**0.070**

**0.117**

**IPO**

**IPOS**

**FPX**

**FPXI**

**Mean**

*Linear and Non-Linear Financial Econometrics - Theory and Practice*

 0.219

**IPO** 0.292 0.621 0.927

0.404

 0.007 0.007

5.718

�0.478 **Cumulative**

**Benchmark:**

 **S&P 500 Index**

**benchmark-adjusted**

 **returns** **Benchmark:**

 **S&P 600 Small Cap Index**

**1.366**

 1.009

�0.021

 1.168

0.147

0.707

�0.541 �0.741

N/A

**0.474**

**0.333**

**0.233**

0.564

0.186

0.143

**0.296**

**IPOS**

**FPX**

**FPXI**

**Mean**

 **absolute returns** ETF as well as the individual funds fails to derive positive abnormal returns over their entire trading history irrespective of the market index used to estimate abnormal returns. At the fund level, three and two ETFs provide investors with a positive cumulative abnormal return against the S&P 500 Index over the 6- and 12-month period, respectively, but returns are basically negative over the rest time intervals examined. In the case, of the S&P 600 Small Cap Index, only one ETF can produce consistent cumulative abnormal returns over a 24-month investment horizon.

**5.3 Risk-adjusted performance analysis**

*DOI: http://dx.doi.org/10.5772/intechopen.90269*

The results of the six-factor regression model are provided in **Table 4**. The table includes the alpha coefficient along with the estimates of the explanatory variables of the model. Probabilities on the statistical significance of estimates are provided too along with R-squared on the sufficiency of the model to explain the performance of IPO ETFs. Finally, the results are presented for each ETF over the several estimation windows considered and against the two different market indices employed in our analysis. When it comes to excess performance, **Table 4** shows that most of the average alpha estimates over the several subperiods examined are positive either when the S&P 500 Index is the benchmark in the model or the S&P 600 Small Cap Index is used as a proxy for the market return. However, at the fund level, most of the individual alphas are insignificant both in statistical and economic terms. Based on this element, we conclude that IPO ETFs fail to deliver any above market return. However, this conclusion does not apply to the First Trust US Equity Opportunities ETF. Over the 6-month estimation window or longer, the alpha estimates of this fund are positive and statistically significant at the 10% or better indicating that this ETF can beat the market. We remind that this ETF is the oldest among the funds in the sample. Thus, the assertion about the positive relationship between the age of IPO ETFs and their long-run performance is verified by the results of regression analysis. With respect to systematic risk, the beta estimates presented in **Table 4** are all positive with the majority of them being significant at the 5% or better. Moreover, a wide fluctuation in betas is observed among the various subperiods examined. However, in each single period as well as over the entire trading history of each ETF, there is a convergence in betas obtained from using the two alternative market benchmarks. At the sample level, the average beta coefficients are below unity indicating that IPO ETFs are more conservative than the market. Conservativeness implies that investors choosing IPO ETFs are relatively protected during declining paths of the overall stock market. This finding is in line with the ratios of ETFs' intraday volatility to that of benchmarks in **Table 1**, where we saw that ETFs are less volatile than the market indices. Therefore, our assertion about IPO ETFs standing as a relative safe haven for equity investors during turbulent markets is

*IPO ETFs: An Alternative Way to Enter the Initial Public Offering Business*

verified by the estimations of systematic risk via regression analysis.

reasonable finding.

**113**

Going further, when the S&P 500 Index represents the stock market in the model, the effect of the size factor on the performance of ETFs seems to be significant only in the case of the Renaissance IPO ETF and the First Trust US Equity Opportunities ETF. For these funds, the coefficients of the SMB factor are constantly positive, while most of them are significant at the 10% or better. The positive and significant effect of the size factor on the performance of at least two ETFs in the sample is in line with our expectations given that, according to Fama and French (2015), the SMB slopes are strongly positive for small stocks (and slightly negative for big stocks), and that the ETFs examined are indeed small cap or even very small cap.<sup>6</sup> When the S&P 600 Small Cap Index is used as the market

<sup>6</sup> Based on the definition of "small capitalization" offered by Investopedia, a small cap firm is a company with a capitalization of between \$300 million and \$2 billion (http://www.investopedia.com/terms/s/sma ll-cap.asp). As of 5 January, 2017, the ETFs in our sample have a minimum of Net Assets of \$1.9 million in the case of the Renaissance International IPO ETF and a maximum of assets of \$633.6 million in the case of the First Trust US Equity Opportunities ETF (according to information found on the website of ETFs' managing companies). Based on these figures, it is obvious that the ETFs in the sample stand as small cap portfolios and, consequently, the positive sign of the SMB estimates for at least two funds is a

After discussing cumulative returns, we focus on the buy-and-hold returns of IPO ETFs. One element revealed in **Table 3** is that the average ETF derives significant buy-and-hold absolute returns, which range from 4.07% over the 12-month period to 35.62% over the entire trading history examined. Moreover, all the individual ETFs present positive buy-and-hold absolute returns over the first 6-month trading period, three funds offer positive returns during the first year of their trading, and two ETFs achieve positive returns during the 18- and 24-month periods. However, when the entire trading history of ETFs is considered, just the First Trust US Equity Opportunities ETF provides a significant positive buy-andhold absolute return, which approximates 159%.<sup>4</sup>

On the question of how the buy-and-hold benchmark-adjusted returns of ETFs behave, **Table 3** reports significant such returns over the several intervals investigated. With only one exception, all the return estimates of each single ETF are positive. Moreover, the average IPO ETF produces a mean buy-and-hold benchmark-adjusted of 23 and 28% against the S&P 500 Index and S&P 600 Small Cap Index, respectively, over the whole trading history of ETFs up to October 31, 2016. At the fund level, the First Trust US Equity Opportunities ETF is the most profitable ETF in the sample. The historical buy-and-hold benchmark-adjusted return of this fund amounts to 62 and 83% in the case of the S&P 500 Index and the S&P 600 Small Cap Index, respectively.<sup>5</sup>

Overall, the analysis of long-run performance reveals that IPO ETFs can be suitable investment choices for investors looking for substantial long-term profits from entering the IPO business. More importantly, the findings on buy-and-hold benchmark-adjusted returns indicate that IPO ETFs can beat the overall stock market over shorter or longer periods. This pattern should be highly welcome by investors who always seek for alternative investment tools to perform above the market.

<sup>4</sup> The rest ETFs present significantly negative buy-and-hold absolute returns. However, the magnitude of the positive return of the First Trust US Equity Opportunities ETF is that big so that the average historical buy-and-hold absolute return of the sample be equal to 36%.

<sup>5</sup> A comment that can be made with respect to the First Trust US Equity Opportunities ETF significantly outperforming its peers (in the case of the S&P 500 Index, the mean benchmark-adjusted outperformance of this fund over other ETFs in the sample is equal to 45.81% whereas, in the case of the S&P 600 Small Cap Index, average outperformance approximates 67%), concerns the age of this ETF. In particular, as we have seen in the previous sections, this ETF was the pioneer in the IPO ETF business and has more than 10 years of trading records. The performance superiority of the oldest fund in the sample over its younger peers from a buy-and-hold benchmark-adjusted perspective resembles the longrun performance advantage of the companies going public after several years of operation as private non-listed firms. The findings of several studies such as those of [23, 45–47] provide strong evidence of a positive relationship between a firm's age and its long-run performance. Obviously, an ETF has no operating history before its inception on a stock exchange. That said, after inception, the trading experience accumulated to an ETF seems to be a decisive factor that can affect its performance. In our study, the very small size of the sample (just four funds) does not allow running a cross-sectional regression of ETFs' long-run performance on their age to obtain statistical support of our assertion about the return of aged ETFs against their young counterparts.
