**5.3 Risk-adjusted performance analysis**

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. 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-and-

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

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

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

<sup>5</sup> A comment that can be made with respect to the First Trust US Equity Opportunities ETF significantly

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

outperforming its peers (in the case of the S&P 500 Index, the mean benchmark-adjusted

hold absolute return, which approximates 159%.<sup>4</sup>

S&P 600 Small Cap Index, respectively.<sup>5</sup>

historical buy-and-hold absolute return of the sample be equal to 36%.

the return of aged ETFs against their young counterparts.

**112**

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 verified by the estimations of systematic risk via regression analysis.

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 reasonable finding.




estimates up to the 18-month investment window are positive and statistically significant. After this period, only the First Trust US Equity Opportunities ETF presents a stable positive relationship with the momentum factor, whereas the rest three ETFs are negatively related to this factor. This is also the case when the S&P 600 Small Cap Index is the reference market portfolio in the model. The main conclusion reached though analyzing the results about the momentum factor is that IPO ETFs follow the trends of the overall stock market in the short-run, but in the long-run, the pricing behavior of that products can deviate from the market.

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

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

When it comes to the Conservative Minus Aggressive factor, the results indicate a rather negative impact on IPO ETF performance. The majority of CMA estimates are negative when the S&P 500 Index is the market portfolio in the model. Moreover, 12 out 27 single CMA estimates are statistically significant. More or less, the same results are obtained when we use the S&P 600 Small Cap Index in regressions. The negative sign of the CMA variable is in accordance with our expectations for a negative relationship between the performance of IPO ETFs and the CMA factor based on the suggestions of Fama and French [44] about a negative relationship

Finally, as far as the impact of Robust Minus Weak factor on performance of IPO

The outcomes of the market trend return analysis are provided in **Table 5**. The results are presented for absolute, benchmark-adjusted and abnormal returns over the descending and ascending paths of the S&P 500 Index and the S&P 600 Small Cap Index, respectively. For each single ETF and over each market path, the number and percentage of days with negative (or zero) and positive returns are displayed along with the corresponding average negative and positive returns.

To begin with, when the S&P 500 Index declines, the average IPO ETFs declines too on 75.31% of the corresponding trading days. The average absolute return on these negative days amounts to �86 bps. During the negative days of the S&P 500 Index, ETFs present an average positive absolute return on 24.69% of negative trading days. When the market goes up, IPO ETFs move upward too on 62.08% of the respective positive days delivering an average return of 102 bps. Moreover, during the positive path of the S&P 500 Index, ETFs move opposite to the market on 37.92% of days. When we use the S&P 600 Small Cap Index as a proxy for the

The main conclusion that can be drawn from the discussion of absolute returns is that IPO ETFs are quite but not absolutely aligned to the overall stock market. The fact that when the market moves downward, IPO ETFs have more than 20% probability of moving against the market indicates that IPO ETFs can possibly be useful hedging tools during turbulent stock markets. However, the significant number of negative return days (i.e., 37.92%) when the stock market moves upward should be borne in mind when planning investment strategies with

ETFs is concerned, the results in **Table 4** reveal a negative such effect. In both versions of the model, the majority of the relevant RMW estimates are negative and statistically significant (15 and 14 out of 27 individual estimates in the case of the S&P 500 Index and the S&P 600 Small Cap Index, respectively). This finding is in accordance with our expectations about a negative relationship between the performance of ETFs and RMW. According to Fama and French [44], the combination of negative CMA and RMW slopes in the performance regression model (as is the case in our analysis) indicates that the returns of IPO ETFs resemble the returns of

between expected investment and expected rate of return.

those firms that invest a lot despite their low profitability.

**5.4 Market trend return analysis**

stock market return, we obtain similar results.

IPO ETFs.

**117**

*This table presents the results of a six-factor performance regression model. The daily excess return of IPO ETFs is successively regressed on the excess return of the S&P 500 Index or the S&P 600 Small Cap Index, and the Fama&French SMB (small minus big) factor, the Fama&French HML (high minus low book-to-price ratio) factor, the Carhart UMD (momentum) factor, the Fama&French CMA (conservative minus aggressive) factor, and the Fama&French RMW (robust minus weak) factor. The model is run over the first month, 3 months, 6 months, 12 months, 18 months, and 24 months of each ETF's trading history excluding the month of each ETF's launch on the stock exchange. The model is also run over the entire trading history of each ETF up to October 31, 2016 also excluding the month of each ETF's launch on the stock exchange. <sup>a</sup>*

*indicates statistical significance at 1% level.*

*b indicates statistical significance at 5% level.*

*c indicates statistical significance at 10% level.*

#### **Table 4.**

*Risk-adjusted performance analysis.*

portfolio, the results about the SMB factor are opposite to those just discussed. All the individual estimates are negative with the majority of them being statistically significant. This pattern was not expected but it could possibly be considered by the correlation of the SMB factor with the S&P 600 Small Cap Index.<sup>7</sup> This means that the effect of the size factor may be expressed by the positive slope of the market index.<sup>8</sup>

On the impact of the value factor on performance of IPO ETFs, the relevant estimates of the HML variable are all negative with most of them being statistically significant at the 10% or better, especially when the 12-month or longer estimation windows are considered. This finding applies to both versions of the model, namely either when the S&P 500 Index or the S&P 600 Small Cap Index is used. Based on Fama and French [44], the strongly negative slope of the HML factor indicates that IPO ETFs may be deemed as to resemble low B/M (i.e., book-to-market) growth stock portfolios. This is true given that the stocks that comprise the underlying indices of IPO ETFs are usually small cap companies that go public with strong perceived potential for significant growth in the future.

The next variable considered is the momentum factor of Carhart. The majority of the relevant sample average estimates are positive, especially in the short-run, namely over periods up to 12 months. In the long run, the average momentum coefficients are negative. A negative sample average is obtained when the entire trading history of each ETFs is taken into consideration when running the performance regression model. At the fund level, when the first version of the model is assessed (i.e., the one with the S&P 500 Index), about half of the momentum

<sup>7</sup> We have computed an average correlation between the SMB and the S&P 600 Small Cap Index of 0.55 over the various time intervals that correspond to the trading history of each ETFs under examination. <sup>8</sup> To verify that the results reported with the usage of the S&P 600 Small Cap Index are not spurious, we run performance regressions after detracting the SMB variable from the model. The results obtained do not differ significantly from those reported in **Table 4**.

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

estimates up to the 18-month investment window are positive and statistically significant. After this period, only the First Trust US Equity Opportunities ETF presents a stable positive relationship with the momentum factor, whereas the rest three ETFs are negatively related to this factor. This is also the case when the S&P 600 Small Cap Index is the reference market portfolio in the model. The main conclusion reached though analyzing the results about the momentum factor is that IPO ETFs follow the trends of the overall stock market in the short-run, but in the long-run, the pricing behavior of that products can deviate from the market.

When it comes to the Conservative Minus Aggressive factor, the results indicate a rather negative impact on IPO ETF performance. The majority of CMA estimates are negative when the S&P 500 Index is the market portfolio in the model. Moreover, 12 out 27 single CMA estimates are statistically significant. More or less, the same results are obtained when we use the S&P 600 Small Cap Index in regressions. The negative sign of the CMA variable is in accordance with our expectations for a negative relationship between the performance of IPO ETFs and the CMA factor based on the suggestions of Fama and French [44] about a negative relationship between expected investment and expected rate of return.

Finally, as far as the impact of Robust Minus Weak factor on performance of IPO ETFs is concerned, the results in **Table 4** reveal a negative such effect. In both versions of the model, the majority of the relevant RMW estimates are negative and statistically significant (15 and 14 out of 27 individual estimates in the case of the S&P 500 Index and the S&P 600 Small Cap Index, respectively). This finding is in accordance with our expectations about a negative relationship between the performance of ETFs and RMW. According to Fama and French [44], the combination of negative CMA and RMW slopes in the performance regression model (as is the case in our analysis) indicates that the returns of IPO ETFs resemble the returns of those firms that invest a lot despite their low profitability.
