Initial Public Offerings: Status, Flaws and Dysfunctions

3. Medium-term Performance

The performance and possibly even the survival of IPO securities is a major topic of study for two sets of reasons involving issuers and investors. High primary issue costs pay off if the transaction later enables companies to secure affordable financing with subsequent issues. This will not happen if share prices fall significantly after the issue. Few investors will maintain an interest in IPOs if the corporate survival rate is low and performance mediocre. In terms of public policy, however, there is another factor to consider, and that is the survival and growth of newly listed companies. Their stock exchange listings are not free for either the issuers or the investors buying the primary issues. The lowering of standards likely has a cost for the whole market, since it reduces overall liquidity. There is reason to wonder whether the benefits in terms of corporate growth warrant these different costs.

The remainder of this part begins with a discussion of the survival and growth of companies having gone through an IPO, then goes on to analyse their stock market performance before getting into the factors conditioning this performance and foreign experiences.

3.1 Survival and growth

Table 13 provides a breakdown^{Footnote 34} of IPOs by Canadian companies in 1991-1995 based on their status in 2002. The following categories were defined after systematic research using all accessible sources (SEDAR, Financial Post, stock exchanges, the Internet): delisted shares, bought-back or merged shares, and shares still being bought and sold.

Delisted shares: these are shares that have been suspended or delisted, generally due to financial problems, or shares bought and sold over the counter.

Bought-back/merged shares: shares that have disappeared from the list when bought back by another company or the issuing company or in a merger or reverse takeover. As a rule, small issue shares are bought back at very low prices.

Still active shares were divided into three groups to reflect the book value of equity capital in 2001:^{Footnote 35}

• Foundering shares have negative shareholders' equity. Their continued existence is in question.
• Poorly performing shares have positive shareholders' equity, though it is below gross proceeds at issue. The book return over the period is therefore negative.
• Companies with net assets worth more than the proceeds of the issue have a positive book return.^{Footnote 36} However, this evaluation criterion is very generous, since it does not consider assets prior to the initial public offering or the rate of return required by the shareholders.

Finally, we calculated the number of companies still being bought and sold where net assets exceeded $10 million, among the group whose gross proceeds had been under $5 million. In our view, these companies have gone beyond the bounds of small business to become medium-sized concerns.^{Footnote 37}

Table 13 shows the very high rates of disappearance through delisting or acquisition. This phenomenon is seen in all issue categories, but the significance of buy-backs varies with issue size. Small issues are generally picked up for very minor amounts, whereas big issues are acquired largely for the purposes of consolidation, which has affected a number of sectors, including pulp and paper. In addition to the delistings, if we classify as failures issues where net assets are negative or below the gross proceeds, the rate rises to 52.94% for small issues. Some 28.10% of companies do survive and increase their yield above the gross proceeds at issue. Only nine companies in this category, however, have net assets exceeding $10 million and can therefore be viewed as successful. The success rate of small issues is thus 5.9%. The corresponding values for the $1 to $5 million group are 28.42% (failure) and 17.89% (success). In this case, the redemption rate is higher (37.89%). At 40.74%, the failure rate remains very high for companies with $5 to $10 million share issues. It is very low, however, for issues exceeding $100 million (0.6%).

Table 13: Distribution of Canadian companies with IPOs in 1991-1995

GP	Delisted	Acquired	Bought/Sold			Total
			SE<0	0<SE<GP	GP<SE
The table is based on their status in 2002, the book value of their shareholders' equity (SE) and gross proceeds (GP) at the time of issue. Excluded are privatizations, issues under the CPC program and issues by companies headquartered abroad. Relative frequency for the different size categories is shown as a percentage in brackets. Sources: Financial Post Report of New Issues, Cancorp Financials, www.sedar.com, www.tsx.com and additional research on the Internet.
Under $1 M	41 (26.80)	29 (18.95)	28 (18.30)	12 (7.84)	43 (28.10)	153
$1 to $5 M	15 (15.79)	36 (37.89)	8 (8.42)	4 (4.21)	32 (33.68)	95
$5 to $10 M	8 (29.63)	11 (40.74)	2 (7.41)	1 (3.70)	5 (18.52)	27
$10 to $50 M	9 (8.04)	54 (48.21)	4 (3.57)	7 (6.25)	38 (33.93)	112
$50 M or over	0 (0)	17 (48.6)	1 (0.3)	1 (0.3)	16 (45.8)	35

As can be seen, very small issues have an extremely small chance of succeeding. The economic rationale behind programs to promote stock listings for very junior companies should therefore be revisited.

Generally speaking, researchers have shown that newly issued shares suffer from abnormally poor performance over the first 3 to 5 years following the offering when this performance is adjusted against a benchmark made up of comparable securities.

This situation is apparently common in many countries, and Ritter (1998) provides a summary (Table 4). Underperformance was around –32% to -34% between 1980 and 2000 in the US, except for the 1990-1994 sub-period, when it was only -7% (Ritter and Welch, 2002). Loughran and Ritter (1995) obtained similar results for 1970 and 1990 in the US, as did Levis (1993) for the UK, as well as Cai and Wei (1997) for Japan in 1971-1982. Jog (1997) examined a sample of 254 Canadian primary issues listed on the Toronto Stock Exchange in the period 1971-1992 and also noted abnormally poor performance. Recent studies, however, have been critical of the methods used so far for analyzing this phenomenon and have sometimes shown contrary results. Paudyal et al. (1998) examined a sample of 77 Malaysian IPOs in 1984-1995 and arrived at adjusted abnormal returns for the three years following the offerings that were positive and insignificant (12.85%). Buser and Chan (1987) reported adjusted positive returns of 11.2% in the two years following the offerings based on a sample of primary issues on the NASDAQ in 1981-1985. Kim et al. (1995) reported overperformance based on a sample of 169 primary issues in Korea. This finding could have something to do with the fact that the Korean issuing companies had been in business longer than, for example, the US issuing companies. Their average age at the time of issue was 19.63 years for the sample used by Kim et al. (1995), but only 6.46 years for the 1,526 companies in Ritter's sample (1991). Entering the stock market later, the Korean companies would have a better chance of survival. A second possible explanation has to do with a high initial underpricing in Korea and a major rise in prices during the first month following the offering. By excluding the first month from the adjusted returns calculation, the writers note that the performance of primary issues in Korea does not differ statistically from the benchmark sample. This explanation relates to measurement problems.

3.2   Measurement problems

Measuring the long-term performance of primary issues raises major problems the analysis of which is beyond the purview of this study.^{Footnote 38} Numerous methodological options are available to researchers, involving benchmark performances (weighted or unweighted indexes, control group of companies), methods used to calculate cumulative abnormal returns (based on events or years), which are the differences between observed and benchmark returns, and, especially, the methods used to simulate investment strategies. Methods that use passive strategies (BHAR, for "buy and hold average return") are very different from cumulative returns approaches. In the first case, the simulated strategy is to buy and hold part of every new issue for 3 to 5 years without rebalancing the portfolio. The cumulative returns approach assumes that the portfolio is reweighted from time to time. Since no method has proven to be clearly better than the others and, as noted by Brav et al. (2000), the methodology used to measure long-term performance directly affects the size and strength of the statistical test, it is recommended that a number of different methods be used simultaneously. This is what we do below.

3.3 Medium- and long-term performance of issues

Figures 3 and 4 show cumulative abnormal returns calculated on the basis of the issue price for all Canadian primary issues except CPC issues for which stock market data were obtained. The excess yields are calculated for an equal-weighted portfolio where an equal amount is invested in each issue, and then on a weighted basis where the investment reflects the issue's relative size. Figure 4 shows that calculation methods have little effect on results and so they are discussed in general terms.

The high initial return in the graphs (Day 1) represents the high initial return discussed in the section above. By purchasing shares directly from the broker, the investor sees a big initial return, but the value of the investment drops rapidly and, on an equal-weighted basis, an initial investment in the issuing companies results in a loss to the investor of 19.96% and 26.5% after, respectively, 36 or 60 months, compared with an investment in companies that have not issued securities for 5 years. The investor also loses by investing in issuing companies in the first year of trading, with the peak loss occurring in the 60th month. When value weighted, an initial investment in the issuing companies results in a loss to the investor of 12.32% and 20.61% after, respectively, 36 and 60 months, compared with an investment in non-issuing companies. Value weighting appreciably reduces Canadian primary issues' medium- and long-term under performance, although it does not do away with it. This suggests that under performance is more significant for small issues than for bigger ones.

3.4 Issue success factors

3.4.1 Size

When the sample is broken down on the basis of the gross proceeds of issues, it can be seen that issues with gross proceeds below $10 million give rise to relatively bigger losses than the ones with gross proceeds above $10 million. In the 48th and 60th months of issue, the BHARs from small issues are -42.58% and -44.86% respectively, whereas the figures for big issues are -7.9% and -3.1%. In the 5 years following their primary issue, small cap shares lose half of their value, taking market fluctuations into account. This is due to the fact that a large percentage of shares in small businesses listed on the exchange disappear. A study of the current status of issues would require in-depth analysis. Many securities are viewed as still trading even though prices are minuscule and transactions rare. As seen in section 3.1, the survival rate is positively linked to company size.

3.4.2 Issuing periods

In 1991, 1992, 1995 and 1998, the market was "hot." In 1993, 1994, 1996 and 1997, it was "cold," although a few issues by subsidiaries raised significant capital in Canada. This breakdown shows that underperformance is not a generalized phenomenon. At the 36th month after issue, the BHARs are -24.06% for "hot" issues and -8.06% for "cold" issues. At the 60th month, the BHARs are -40.38% for "hot" issues and -11.47% for "cold" ones. Under performance therefore appears to be worse when issues occur during periods of intense issuing activity.

3.4.3 Technology sector

When the sample is broken down by industry,^{Footnote 39} it becomes clear that underperformance varies considerably from sector to sector. For example, primary issues in the financial services sector see a positive adjusted return in the five years following the offering. The special characteristics of financial companies compared with industrial and commercial concerns might explain this positive performance. This group includes demutualizations, which play a major role when value weighting is used.

Issues in the mining and oil and gas sectors underperform substantially five years after issue. The ex ante uncertainty typical of these issues might account for this long-term under performance.

Primary technology issues perform poorly in the first year, after marked initial underpricing (19.77%), but under perform thereafter through the fifth year of issue. This observation holds true only when performance is calculated based on the issue price. Performance measured on the basis of market price is sharply negative.

In short, Canadian primary issues yielded lower returns over the 5 years studied than a sample of similar non-issuing companies. The implications here are significant, especially for small issues. To begin with, there is the problem of sustaining investor interest in this type of investment. Secondly, there is the high cost of subsequent issues. It would seem that a major objective of primary issues, lowering financing costs, is not being achieved.

3.5 Foreign experiences

In the US, the Small Corporate Offering Registration (SCOR) program and certain forms of direct investment exempted from registering with the SEC (Regulations A and D and intrastate regulations) enable small start-ups to raise funds more easily and cheaply^{Footnote 40} than through conventional primary issues. The amounts raised are smaller, ranging from $1 to $20 million, like the vast majority of Canadian issues. Disclosure requirements are less stringent and oversight is local. These issues are not covered by the SEC under the terms of the NSMIA (1996). They can be sold by the company itself or a commission sales representative and make use of mass communications, including the Internet. The SCOR program,^{Footnote 41} developed by the North American Securities Administrators Association (NASAA) in conjunction with the American Bar Association, exists in 45 US states. Originally, it was intended to register, at the state level, share offerings exempt under SEC Rule 504 for sales of securities up to $1 million. Issues are limited to $1 million in each 12-month period with a minimum unit price of $5 ($2 in some states). Companies have to fill out a U-7 form that is simpler than the SEC's.^{Footnote 42} In most states, financial statements have to be audited if the issue exceeds $500,000 and minimum income conditions are sometimes placed on investors. To make things easier for issuing companies in a number of states, SCOR programs have been implemented on a regional basis.^{Footnote 43}

At the outset, securities issued but exempted from SEC registration were not traded. To solve this problem, in 1995, the SEC authorised the Pacific Stock Exchange to publicly list shares subject to SCOR regulations and Regulation A.^{Footnote 44} The vast majority of companies, however, use this type of issue as a private investment, and investors hold the shares until the company is sold, is listed on an exchange, or buys back its own shares.

SB-1 and SB-2 issues stand out from other types of issues intended for small businesses in that they are like initial public offerings that have to be registered with the SEC, but are reserved for small businesses. They are issues worth $10 million or less (SB-1) and $25 million or less (SB-2). Companies have to register with the SEC, but the procedures have been simplified. It remains to be seen whether these pared-down procedures lead to the creation of viable companies.

Table 13 shows the annual distribution of small company issues in the US in 1984-2001. The data are from Stewart Gordon (Dallas SCOR Report) and the SEC.^{Footnote 45} Overall, there were 1,983  issues yielding total gross proceeds of US $3.7 billion in the period covered (starting sample). Very rare in the 1980s, these issues became much more common in 1993 (166 issues, for an increase of 77% over 1992) and thereafter, peaking in 2000 (285).

Based on this initial sample, we were able to track companies that subsequently registered their shares on a national stock exchange. Inasmuch as this affords tangible evidence of corporate growth, in that the company satisfies the minimum listing criteria and needs financing in order to continue growing, the listing of a small business that has released an IPO is a sign of success. Based on this observation, the overall issue success rate is only 5.40%, however, or 107 IPOs (final sample). The annual distribution of successful issues is similar to that of the starting sample. Table 17 shows a large percentage of issues under SB-2 and Regulation A in the final sample: 80 of the 107 companies succeeded (47 issued under Regulation A and 33 under SB-2 regulations). However, no company that arranged an issue under intrastate legislation or the SB-1 in 1984-2001 later achieved listing status.

The sectoral breakdown of the final sample (Table 14) shows that companies in the financial, technology and service sectors were predominant, at 60.7%,^{Footnote 46} while the basic materials and durable goods sectors were underrepresented, at 10.3%.

Table 14: Starting sample of issues by small US companies, including direct issues, in 1984-2001, and final sample of issues subsequently listed

Type of issue	Starting Sample		Final Sample		Success Rate
	Number	Gross Proceeds (US$ M)	Number	Gross Proceeds (US$ M)
Data sources: SEC at http://www.sec.gov/ and SCOR Report at http://www.scor-report.com. The success rate is the ratio of the number of successful (listed) issues under regulations to the total number of issues made under the same regulations.
Direct issues
SCOR	770	537.1	23	15.8	2.99%
Reg. A	566	1 479.5	47	123.9	8.30%
Reg. D-504	53	32.4	4	3.1	7.55%
Intrastate	174	437.0	0	0	0.00%
Primary issues
SB-1	14	21.8	0	0	0.00%
SB-2	406	1 236.9	33	137.7	8.12%
Total	1983	3 744.8	107	280.5	5.40%

US initiatives geared to achieving listings for small businesses do not appear to be successful. The rate of actual listing is very low and most companies financed in this way do not manage to grow.

3.6 Conclusion and policy implications

The performance of primary issues in Canada is disappointing. In this respect, it is no different from issues in other developed markets. However, the medium-term return on small issues is about -50% over the five years following the offering and the five-year survival rate is around 60% for issues from the early 1990s. The US has chosen to relax its regulations to facilitate access to financing for small businesses. Based on the fragmentary data we have been able to gather and review, these programs do not appear successful.

It may be that this disappointing market performance reflects poor operational performance after issue that is concealed from the investors by accounting tricks when preparing the prospectus. At least four US studies have highlighted this phenomenon and the investors' inability to detect these tricks. If this is so, it would imply a need for increased oversight by securities commissions. For now, we have no evidence of this problem in Canada. It seems more plausible to associate the mediocre performance of Canadian IPOs with the fact that they are conducted by small companies that often have very short track records. These companies have not been able to develop and build a competitive edge, which is the key factor for achieving wealth and increasing value.

^{Footnote 34} A detailed analysis of the growth of newly listed companies calls for a laborious data search that is beyond the scope of this study. This summary analysis is based solely on the status of companies listed at the end of the fiscal year ended in 2001.

^{Footnote 35} This is the book value of assets as of the latest financial statements ending between September 30, 2001 and August 31, 2002 and available from the Cancorp Financials database.

^{Footnote 36} Or slightly negative where the shareholders' equity was significant before the issue.

^{Footnote 37} The European Community carried out an extensive project to review and consult on the definition of SME. The final recommendation of June 25, 2002 draws the line between small and medium-sized businesses at 10 million Euros (C$16 million) in sales or total assets. With $10 million in equity, the company has usually crossed this line.

^{Footnote 38} See Barber and Lyon (1997) and Kothari and Warner (1997) for a study in this regard and Kooli and Suret (2003b) for a summary.

^{Footnote 39} The sectoral distribution is based on the distribution under the System for Electronic Document Analysis and Retrieval (SEDAR), available at www.sedar.com.

^{Footnote 40} The cost of a Direct Public Offering (DPO), done without brokers, is about 10% of the cost of an IPO, according to T. Stewart Gordon, editor of SCOR Report, the journal of alternative financing: see http://www.dfdpo.com/article8.htm.

^{Footnote 41} The Full Disclosure Act, New Issues Act, Truth in Securities Act or Prospectus Act. For more details about this program, log onto http://www.nasaa.org/.

^{Footnote 42} The form asks 50 questions about the company's background, the main risks facing investors, the amount needed and how the company plans to spend it.

^{Footnote 43} A region is a group of states: there are currently four regions.

^{Footnote 44} Provided they meet the minimum registration conditions. However, the SEC, under the NSMIA (1996), incorporated the Pacific Stock Exchange into the so-called national markets on January 13, 1998, and the securities it lists are now covered within the meaning of the NSMIA.

^{Footnote 45} http://www.scor-report.com and http://www.sec.gov/.

^{Footnote 46} Based on the sectoral classification suggested by Stewart Gordon.