**4. The sample**

price is expected to rise later. This anomaly undermines the semi-strong form

explained sufficiently by the finance theory. The difficulty in explaining the momentum anomaly is that, as the efficient capital markets theory suggests, an increase in the price of an asset cannot be indicative of a further increase in future prices. Behavioral finance has offered some possible explanations to the existence of the momentum anomaly. In particular, investors are assumed to be irrational and, consequently, they underreact to the release of new information failing to incorpo-

ETFs must be affected by the profitability factor in a negative fashion.

define the overall buy-and-hold risk-adjusted performance of ETFs.

rate new information in the prices of their transactions.

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

The existence of momentum in asset prices is an anomaly which has not been

The Conservative Minus Aggressive and Robust Minus Weak factors correspond to the [44] investment and operating profitability factors. [44] use past investment as a proxy for the expected future investment and, based on valuation theory, they suggest that CMA implies a negative relation between the expected investment and the expected internal rate of return. Furthermore, based on the findings of [44], a negative loading is expected for the RMW factor, that is, the excess return of IPO

The usage of Eq. (7) aims at capturing the market elements that can affect the performance of IPO ETFs and considering whether these funds can produce any meaningful above market returns, which will be represented by a positive and statistically significant alpha. With respect to the latter, [44] assert that if an asset pricing model fully captures expected returns, the intercept of the model should be indistinguishable from zero in a regression of an asset's excess return on the factor

The model is successively run for six different time periods. The first period concerns the first 21 trading days of each ETF excluding the month in which the ETF began trading on the exchange. We do so as we did when we estimated the long-run performance of ETFs above to allow for the possibility of price support in the first few trading days. This month is also excluded from all the other time

The second interval assessed concerns the first 63 trading days of each ETF. The third period examined regards the first 6 months of trading, that is, the first 126 trading days of each fund. The next period taken into consideration covers the first 12 months of trading data. In our analysis, the intervals ranging up to 1 year can be considered as a short-term investment horizon. Looking for more long run, we run the model for a period covering the first 18 and the first 24 months of each ETF's trading records. Finally, we run Eq. (7) over the entire history of each ETF so as to

In the last step, we perform a "market trend" analysis of IPO ETF returns by examining how the return of ETFs responds to the decreasing or increasing swings of the overall stock market as the latter is alternatively represented by the S&P 500

In our analysis, we first sort the daily returns of benchmark and then compute the number and portion of daily returns of each ETF that are negative or equal to

descending path and during the ascending path of the stock market. If ETFs follow the market closely, they are expected to decline when the market declines and vice versa. Similarly to the short- and long-term performance analysis in the previous sections, we use three alternative types of returns, which are the absolute,

zero and the number and portion of positive ETF daily returns during the

efficiency of the stock market.

returns of the model.

intervals over which Eq. (7) is applied.

**3.4 Market trend return analysis**

Index and the S&P 600 Small Cap Index.

benchmark-adjusted and abnormal return.

**104**

The sample of the study includes the four IPO ETFs available in the US capital market. **Table 1** presents the profiles of ETFs. Presented in the table are the ticker, name of each ETF, name of tracking index, inception date, age in years, expense ratio, average daily volume in number of traded shares, and the historical tracking error of ETFs, which is the difference in returns between ETF and benchmark based on information on historical performance of ETFs before taxes and benchmark returns from each ETF's inception up to December 31, 2015. The information on ETFs' ticker, name, benchmark, inception date, and expense ratio as well as on the historical performance of ETFs and underlying indices has been found on the websites of ETFs' managing companies.

Moreover, **Table 1** reports the trading frequency of ETFs that is calculated as the fraction of trading days with nonzero volume to the total trading history (in days) for each fund, the average intraday volatility computed as the percentage fraction of the highest minus the lowest trade price of each fund on day *t* to its close trade price on the same day, and the fraction of each ETF's intraday volatility to the intraday volatility of the S&P 500 Index and the S&P 600 Small Cap Index, respectively. The last ratios help assessing whether IPO ETFs are more volatile than the market or not. The time series of daily volumes, open, high, low, and close prices of ETFs and the S&P 500 Index have been found on the website of NASDAQ. The historical data of the S&P 600 Small Cap Index have been obtained from Yahoo! Finance.

Regarding the underlying assets of ETFs, we note that the first fund tracks the Renaissance US IPO Index, which reflects approximately the top 80% of newly public firms based on full market capitalization. The second ETF follows the Renaissance International IPO Index, which is a portfolio of the top 80% non-USlisted newly public companies, prior to their inclusion in global core equity portfolios. The third ETF seeks to replicate the return of the IPOX®-100 US Index. This index measures the performance of the top 100 largest, typically best performing and most liquid US IPOs during their first 1000 trading days. The last IPO ETF examined tracks the IPOX International Index, which measures the performance of the 50 largest and typically most liquid companies domiciled outside the US within the IPOX Global Composite Index during their first 1000 trading days. All the indices above are reconstituted and adjusted quarterly and companies that have been public for 2 years (in the case of the Renaissance indices) or 1000 days (in the case of the IPOX®-100 US Index and IPOX International Index) are removed.

The average age of ETFs is equal to 4.42 years with the oldest one being the First Trust US Equity Opportunities ETF, which was launched in April 2006. The rest funds are 3 years old at a maximum indicating that this niche of the ETF market is very young but possibly very prosperous. The average expense ratio is modest being equal to 0.68%. In addition, the ETFs tracking non-US-listed IPOs are more expensive than their domestically allocated peers. This cost superiority of domestic ETFs is not surprising as it has been observed in the case of the "traditional" ETFs both in and outside the US.

Moreover, **Table 1** shows that an average number of about 16,000 ETF shares are traded every day with the First Trust US Equity Opportunities ETF being the more tradable fund in the sample. The concentration of trading to the most aged fund may be the result of the advantage of this ETF in terms of information availability relative to the younger funds and may indicate that investors deem this ETF as more prosperous based on its amassed trading experience.

Going further, the average raw tracking error of the sample is equal to �0.56%. The negative sign means that the average ETF underperforms its benchmark by 56 basis points (bps). Among the four ETFs in the sample, only one outperforms its



index returns reflect no costs at all. In addition, other factors, such as the different time schedules between the stock exchanges on which the shares of ETFs and the

nondomestic market indices, can hamper the effort of ETFs to efficiently replicate

When it comes to trading frequency, **Table 1** reports an average term of 83%. This percentage indicates that there is a considerable amount of days on which ETFs

**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

As far as absolute returns are concerned, **Table 2** shows that the average initial

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

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

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

average returns over the first 2, 3, 4, 5, 21, and 63 trading days.

underlying securities are traded, especially in the case of ETFs tracking

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

present nil trading activity. These findings are in line with the low volumes discussed above. The next trading feature concerns the intraday volatility of ETFs, which is equal to 0.815 on average terms. Surprisingly enough, the international IPO ETFs are less volatile than their domestically allocated peers. This is a new finding as the relevant ETF literature has provided strong evidence that ETFs tracking international indices are riskier than those that track indices from the local stock market. In any case, however, both the average and the individual intraday volatility calculations are quite low showing that IPO ETFs can be a relatively safe haven for equity investors when the overall capital market is in turbulence. This claim can be verified by the ratios of ETFs' intraday volatility to those of the two Standard and Poor's indices taken into consideration. In both cases, the average ratios are below unity, whereas only one out of eight single ratios is greater than unity indicating that IPO

the performance of their benchmarks.

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

ETFs are less volatile than the market.

ETF to 1.45% for the Renaissance IPO ETF.

market on their first trading day.

**5. Empirical results**

IPOX ETF.

**107**

**5.1 Short-term return analysis**

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

#### **Table 1.**

*Profiles of ETFs.*

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

index returns reflect no costs at all. In addition, other factors, such as the different time schedules between the stock exchanges on which the shares of ETFs and the underlying securities are traded, especially in the case of ETFs tracking nondomestic market indices, can hamper the effort of ETFs to efficiently replicate the performance of their benchmarks.

When it comes to trading frequency, **Table 1** reports an average term of 83%. This percentage indicates that there is a considerable amount of days on which ETFs present nil trading activity. These findings are in line with the low volumes discussed above. The next trading feature concerns the intraday volatility of ETFs, which is equal to 0.815 on average terms. Surprisingly enough, the international IPO ETFs are less volatile than their domestically allocated peers. This is a new finding as the relevant ETF literature has provided strong evidence that ETFs tracking international indices are riskier than those that track indices from the local stock market. In any case, however, both the average and the individual intraday volatility calculations are quite low showing that IPO ETFs can be a relatively safe haven for equity investors when the overall capital market is in turbulence. This claim can be verified by the ratios of ETFs' intraday volatility to those of the two Standard and Poor's indices taken into consideration. In both cases, the average ratios are below unity, whereas only one out of eight single ratios is greater than unity indicating that IPO ETFs are less volatile than the market.
