SME Financing Data Initiative
sme-fdi.gc.ca
Common menu bar links
First
Previous
Initial Public Offerings: Status, Flaws and Dysfunctions
4. Implications for Regulation and Intervention
In this section, we try to explain the implications of the observations made in the earlier sections in terms of policy and regulation. The dysfunctions in the small issues market (relative rarity, high cost, poor medium-term return) are all the more paradoxical given that governments have tried very hard to stimulate the venture capital supply, for which the primary issue is a natural route. This stimulation has been such that the amounts raised every year through technology company issues are far less than are amassed by the venture capital industry. In fact, venture capital supports very few primary issues. There do not appear to be any plans in the stock brokerage industry to change issuing approaches, and the role of institutional investors is ambiguous. These different factors need careful consideration.
In 1994, MacIntosh referred to a number of institutional and regulatory barriers to the financing of innovative businesses in Canada. He said then (p. 7) that "[Canadian] small firms appear to 'Go Public' or sell their shares to public investors, far less often than their US counterparts." This study shows that the situation is largely unchanged, although the CPC program has appreciably increased market listings for small companies. Yet, over the 1990s, major changes were introduced in the institutional and regulatory environment.
4.1 The lowering of market listing standards
4.1.1 Process of change
Over the years, and often in tandem with programs set up to promote the capitalization of small businesses, the minimum standards for stock market listing have gradually come down in Canada. The CPCs represent an extreme, where listing can be done without a track or operating record and capitalization of a few thousand dollars. CPCs enable very small firms to get listed.Footnote 47 The Quebec Stock Saving Plan had already caused an appreciable relaxation of basic listing requirements for the Montreal Stock Exchange. The TSX allows listings for technology and industrial firms with net tangible assets of $1 million (Tier 1) or even with no tangible assets (Tier 3). By comparison, the minimum requirement for access to the least demanding tier of the West African Regional Stock Exchange is $500,000. In other words, the minimum standards for exchange listings in Canada are similar or even lower than those for markets in developing countries.
This study shows that small business issues are costly, are significantly underpriced and, on average, provide poor medium-term returns. In a parallel study (Carpentier et al., 2003), we show that companies that grow out of the CPC program are still small businesses in the usual sense years after the qualifying transaction, with sales figures of around $10 million. In analysing the QSSP (Suret, 1994), we highlighted the disappearance rate and very poor performance of small issues made under this program. Of 215 issues from 1979 to 1990 in the "development" component of this program, 34 have remained closed, 15 were apparently trading in late December 1993 at levels above their issue price adjusted for index fluctuations, and 9 had been redeemed in similar circumstances. The success rate was thus 15/215 or about 7%. This rate is very similar to the one reported in this study for small Canadian issues ($1 million or less) or locally regulated US ones.
4.1.2 Issue performance and lowering of standards
The QSSP – Development, CPC and, in the US, SCOR programs and the lowering of minimum stock market listing standards have the common aim of making it easier for companies to gain faster and cheaper access to public financing. However, these various programs are apparently not very successful in developing profitable businesses that would provide investors with a decent rate of return. In economic terms, and despite the importance of small businesses, there is reason to wonder whether these incentives lead to a less than optimal allocation of funds. When investors lose about 50% of their outlay in 5 years, the answer seems obvious. The various Canadian and US studies show a generally bleak future for small firms with stock exchange listings. The European experience, where the new market's index lost 84% of its value between December 1999 and December 2002, shows that this phenomenon is not limited to North America. A mere stock listing will not ensure a company's growth. An analysis of the minimum conditions for potentially successful issues is beyond the compass of this study, but one should be undertaken. It seems clear that the lowering of standards and the market listing situation are not effective ways of promoting the development of growth companies. The fact that only 5% of IPOs for amounts of less than $1 million do not [sic-TR] manage to achieve enough growth after 5 or 10 years for the company to move beyond SME status should prompt authorities to limit, rather than facilitate, the market listing of small businesses.
4.1.3 Monitoring by analysts
Minimum listing standards to achieve success should be reviewed in light of the following factors.
- The idea of adequate available capitalization. The float is the outcome of issue size and the distribution of shares among investors. Capitalization standards cannot be lowered unless the shares going to institutions are limited. There should be a study of listing standards based on available rather than total capital.
- Coverage by analysts. The interest of financial analysts in a stock is a major success factor for issues in the medium term. Chen et al. (2002) show that the level of coverage of shares by analysts affects their value, the effect often varying with types of analysts. Those with the big national stock brokerage houses have more effect on share values than the ones working for regional firms. Analysts with the least clout are from companies that are not involved in brokerage. Sayrak and Dhiensiri (2003) show that, in the US, coverage of a share by analysts generally starts when there is a positive yield of around 3%. These findings are consistent with the assumption that financial analysts provide investors with valid information and help to close the information gap between companies and the market. Sayrak and Dhiensiri note that the effect of stock coverage is greater for NASDAQ shares than for NYSE shares. This reflects the fact that information asymmetry is probably greater for NASDAQ issues. Analytical activity thus becomes more important in their case.
In the US, the key role of analysts in primary issues is acknowledged by NASDAQ (2001): "An effective investor relation (IR) program can reinforce a company's strengths, strategy and key executives – and it is vital to helping a company achieve its goal of maximizing shareholder value… The foundation of most IR programs is built on contact with sell-side analysts. They are the key link to the buy-side, provide research, recommending stocks and overseeing transactions." Das et al. (2002) are of the view that the results of their study of IPO coverage "are consistent with a market of powerful analysts, whose coverage on firms determines investor's demand and interest in IPO stocks." These authors show that, after a year, 73% of IPOs in 1985-1995 were being followed by analysts, and this figure rises to 83% after three years. Speed and intensity of coverage are directly linked to issue size and the prestige of the capital underwriters. There is also a positive relationship between coverage and medium- and long-term share performance.
Interest on the part of analysts therefore seems to be a major condition for IPO success. Yet even in the US, where issues and issuing companies are much bigger, a large percentage of primary issues is not monitored on a regular basis. It is therefore more than likely that the vast majority of IPOs in Canada go unnoticed by analysts.
This would need to be verified in Canada, but L'Her and Suret (1996) report that in 1995, only 203 Canadian stocks were being actively covered by more than five financial analysts. The number of stocks being tracked was 458. It is therefore highly unlikely that new issues are followed closely by Canadian analysts.
Lack of coverage by analysts, insufficient available capital and a poor profit picture leave little hope for new Canadian issues. The policies that lowered the standards and stimulated new issues can certainly be termed dangerous. They do not lead to an optimal distribution of savings, they cause transfers of wealth to middlemen, and they culminate in only rare successes. Given the conditions under which small issues are made in Canada, such offerings are more of a problem than a solution to financing difficulties.
4.2 Venture capital and primary issues
4.2.1 Paradox
Public financing for technology companies is very modest in Canada, compared with the US. This is paradoxical considering the intense activity reported by venture capital firms. In 1999, for example, the Canadian Venture Capital Association reported over 1,200 investments in technology and managed funds in excess of $12 billion. In Canada, venture capital has become a much more important source of financing for growing companies than public financing, as shown in Table 16. Over the period 1992-2000, there was more than twice the amount of backing by venture capital than capital from initial public offerings by technology concerns.
Some results suggest that venture capital firms have little involvement in financing through primary issues. Cumming and McIntosh (2000), using CVCA data, found 33 primary issues connected with venture capital firms in Canada in 1992-1995, or 4.5% of the 662 issues made in that period. This percentage is consistent with our estimates, which put the percentage of issues involving venture capital at about 10% for 1991-1999 (excluding CPC issues).
However, issues initiated by venture capital companies have focussed on technology firms. They were involved in 33 of the 86 technology issues in the period 1992-1995, or 38%. Venture capital companies seem to be having a modest effect on primary issue activity overall, but they are responsible for a significant percentage of technology issues. There is a major disproportion between the overall activity of venture capital companies in Canada and the average of eight technology issues they can be credited with every year. Moreover, the exit methods of venture capital investors differ in two major ways: buy-backs by issuers are far more numerous (as a percentage) in Canada than in the US, and delistings are commoner in the US. There is a similarity, however, in the proportions of primary issues (see Table 15).
| Year | Annual IPO Financing by Technology Companies | Annual Total Funds Invested by Venture Capitalists |
|---|---|---|
| ($ billions) | ||
|
* 1997 was highly influenced by a few issues, especially by subsidiaries like Bell Canada International and AT&T Canada for $640 million and Microcell and Telesystem for $307.6 million. Source of data on annual amounts invested by venture capital investors, 1992-1997:48 SECOR, 1998, p. 13, (diagram 5), and for 1998-2000: Web site of the Canadian Venture Capital Association at http://www.cvca.ca/resources/statistics/. Annual amount of public financing for technology companies: Financial Post, Report of New Issues. |
||
| 1992 | 0.17 | 0.3 |
| 1993 | 0.62 | 0.42 |
| 1994 | 0.27 | 0.46 |
| 1995 | 0.37 | 0.67 |
| 1996 | 0.76 | 1.09 |
| 1997 | 2.19* | 1.82 |
| 1998 | 0.54 | 1.66 |
| 1999 | 0.42 | 2.72 |
| 2000 | 1.85 | 6.63 |
| Total | 7.20 | 15.77 |
4.2.2 Pre-IPO financing
Pension funds are not big investors in growth companies in Canada. In 1999, only 5.6% of new venture capital generated in Canada came from pension funds investing in venture funds. In 2000, this figure rose to 11%, only to dip back to 3% in 2001. In the US, by contrast, 40% of all new capital commitments came from pension funds in 2000.Footnote 49 SECOR (1998, p. 11) echoes this observation for the 1990s as a whole, stating that although pension funds, insurance companies and foundations are the main venture capital sources in the United States, these institutional investors have actually abandoned the venture capital market in Canada over the past 10 years.
It may be that, as a matter of tradition, Canadian pension funds are choosing not to favour this category of assets, or that regulations are curbing this type of investment. It is also likely that these institutions lack qualified staff to gauge the profitability of this type of project. Perhaps the funds are concerned about insufficient liquidity, as the preferred exit method of venture capital investors is the buy-back of the company by its executives. Finally, the pension funds may simply be shut out of the Canadian venture capital market by an overabundance of venture capital from government sources in particular. As SECOR shows (1998, diagram 1), the government's involvement in supplying venture capital has risen steadily since the early 1990s. Cumming and MacIntosh (2002) show that certain policies may be directly responsible for defects in the way venture capital works in Canada. They highlight the generally negative effect of labour-sponsored funds on the capital supply, owing to a crowding-out effect on private funds.
4.2.3 Venture capital fragmentation
Venture capital in Canada is broadly influenced by labour-sponsored funds, the portfolios of which are far from typical of venture capital investments. Some of these funds have achieved significant size, but they have adopted a regionalized structure that fragments their resources. Funds other than labour-sponsored funds are generally small. The profile of respondents to the annual survey of the Canadian Venture Capital Association indicates that they had an average of 27 investments in 1999. Cumming (2003) analyses 214 Canadian funds and reckons that each labour-sponsored fund has an average of 32 investments. This means that Canadian funds are basically small. Successive rounds of financing may prove difficult and funds with limited resources may be reluctant to back rounds of financing preceding IPOs. This situation may offer an explanation for hasty company listings and the rarity of "venture-backed" IPOs.
Considering that this major support to the venture capital supply is not bringing a significant number of viable firms to the stock market or financing the money-raising rounds preliminary to IPOs, intervention policies in this area, including tax credits to labour-sponsored funds, need to be re-examined.
4.3 Institutional investors
Basically, existing primary issues procedures seem to enable brokers to allocate a major proportion of issues to institutional investors. In the US, institutional investors get a disproportionate share of "good" issues. The advantage afforded to institutional investors is viewed as essential and often defended by brokers as a way of keeping these buyers, which absorb large percentages of new issues. Boehmer et al. (2002) estimate this proportion as 77% of available shares.
Admittedly, the sale of a major proportion of these issues to institutional investors helps the broker, which unloads the issue quickly. This practice would be necessary if the non-institutional investors were reluctant to pick up shares in IPOs. But the average price hike in early negotiating sessions contradicts this assumption. There is also a great deal of empirical evidence that IPOs are oversubscribed and individual investors have problems obtaining securities at the offering price. Moreover, the concentration of shares in the hands of institutionals could have the following negative effects:
- A significant loss of liquidity for small cap securities;
- Lack of interest by analysts and brokers due to low transaction levels;
- Major price fluctuations as blocks of shares are sold.
4.4 Brokerage
A number of researchers, including Loughran et al. (1994) and Chowdhry and Sherman (1996), argue that the mechanisms used for valuing and distributing the shares of a primary issue influence the stock market listing process. The most widespread practice in the US and Canada is called book building, in which the lead broker gathers information about the buying plans of potential (institutional and individual) investorsFootnote 50 during "road-shows" and establishes a demand curve based on a number of criteria.
Cornelli and Goldreich (2001) stress the preference for institutional investorsFootnote 51 that are prepared to buy and hold the shares over the long term. The issue price is set so as to create the impression that there is surplus demand and the distribution of the shares is left to the underwriter's discretion.
The second mechanism, the fixed price offer, used in the UK, Singapore and Finland, is based on setting a fixed issue price even before demand for the share is determined. If there is surplus demand, shares are rationed on a prorated basis or by drawing lots. In other countries, rationing reflects the size of the orders received. To reduce the likelihood of a failed primary issue, a low issue price is often set, which might also explain the high average initial returns (Loughran et al., 1994).
In Israel, IPOs use the "dutch auction" system, where the issue price is set after the buy orders are collected to reflect supply and demand and the shares are allocated to all successful bidders. In the US, "open" IPOs use the same system on the Internet and claim that it is more efficient than book building. In France, a sales approach that combines book building and auctions, which could also be termed a book building hybrid, is used to set the issue price and distribute the shares. The release involves making a quantity of shares publicly available at a minimum price with the near certainty, especially in bullish periods, of obtaining a higher rate that is close to market price.
Table 17 shows the initial underpricing of IPOs by mechanism and country. Auction sales seem to result in far lower underpricing than the other methods of setting prices. Conversely, the advent of book building in a market where it did not exist can produce, as in Hong Kong's case, an increase in brokerage fees that were previously very low (2.5%) (Butler and Huang, 2002). To the extent that they are done by Internet, primary issues help to create auctions that might allow for wider distribution of shares and a more objective valuation process. Primary issues by auction are still unusual. They have been used in Israel (Kandel, Sarig and Wohl, 1999) and, more recently, in the US, especially by Hambrecht. This initiative has not garnered much enthusiasm among the competition, though the technology developed by Hambrecht for auctioning primary issues is attracting more and more interest.Footnote 52 In particular, it is used for auctioning fixed-income securities, in amounts of around $5 billion. At any rate, the book building technique is becoming generalized worldwide except in the rare countries where it is prohibited.
In order to establish issuing systems less geared to brokers' needs and interests, it is first necessary to achieve genuine competition among the firms in this field. The Canadian market is doing the opposite.
| Mechanism | Period | Average Number of Shares | Average Initial Return |
|---|---|---|---|
|
Sources: France – Derrien and Womack (2000, Table 1); Japan – Loughran et al. (1994), Pettway and Kaneko (1996), Hamao et al. (2000), and Kaneko and Pettway (2001); Taiwan – Lin & Sheu (1997)Footnote 53 and Liaw et al. (2000, Table VII). |
|||
| France: | |||
| Fixed price | 1992-1998 | 24 | 8.9% |
| Auction | 1992-1998 | 99 | 9.7% |
| Book building | 1992-1998 | 135 | 16.9% |
| Japan: | |||
| Fixed price | 1970-1988 | 441 | 32.5% |
| Auction | 1989-1997 | 733 | 14.1% |
| Book building | 1997-2000 | 368 | 43.7% |
| Taiwan: | |||
| Fixed price | 1986-1995 | 241 | 34.6% |
| Auction | 1995-1998 | 52 | 7.8% |
4.5 Conclusion
Our study highlights a number of major problems in the operation of the primary issues market. In our view, the main one is the low survival rate of small businesses involved in initial public offerings. We estimate that 5.9% of small issues lead to a business that can claim big business status within 5 to 10 years. The policy of stimulating small firms' stock exchange listings should be re-evaluated. The emphasis should be on mechanisms that would allow IPOs to be deferred until businesses have had a reasonable chance to ensure they can survive. Stress should be placed on developing a supply of capital prior to the issue.
The arguments in favour of lightening the regulatory burden for small issues should be reviewed in the light of these findings as well. The success rates of primary issues are similar to those of venture capital. However, this form of financing is handled by experts, which is not the case with primary issues that are bought in part by institutional investors. Lastly, small primary issues in all likelihood have a negative effect on the market. They perform poorly and hamper market liquidity.
Primary issues remain few in number compared with the situation in the US, which may seem paradoxical given the efforts made by the different governments to increase the venture capital supply. It may be that the fragmentation of this supply is one the factors accounting for the relative rarity of IPOs involving venture capital firms. The role of government in the area of venture capital supply should be revisisted. However, the main points of comparison should be issue size and medium-term success. The vast majority of Canadian issues are very small, and a very small percentage of these can be seen as successful.
Footnote 47 Policy 2.4 states that an issuer resulting from a restructuring (after the qualifying transaction) has to meet the minimum listing requirements related to its industry sector and Tier 1 or Tier 2 as defined by Policy 2.1 minimum requirements, that is, $5 million or $1 million in net tangible assets respectively.
Footnote 48 Note that in Canada, labour-sponsored funds are included in the definition of venture capital enven though some of these funds' investments do not behave like venture capital.
Footnote 49 According to the findings of the latest Round Table on the Potential of Canadian E-Business:
http://benefice-net.branchez-vous.com/nouvelles/02-03/06-176501.html (in French only).
Footnote 50 These buying plans mauy take the form of a quantity of shares at a maximum or minimum price.
Footnote 51 Jog (1997), conducting a survey of a number of issuers, says they agree on the fact that institutional investors not only play a major role in the initial sale of their shares but hold on to the shares they buy: private investors will sell their shares shortly after buying them, which strengthens the importance issuing companies attach to the potential influence of the brokers.
Footnote 52 For an analysis of reactions, see: http://en.wikipedia.org/wiki/Initial_public_offering#auction.
Footnote 53 Quoted in Loughran et al. (1994), updated September 2002.
