THE EFFECT OF MULTINATIONALITY
ON THE PERFORMANCE OF INITIAL PUBLIC OFFERINGS
Submission #483
There is a large literature on the effects of multinationality on firm performance, focusing on large, established firms. There is also a large literature on the performance of initial public offerings (IPOs). However, the effect of multinationality on IPO performance has not been studied. We address this question using data from the London main market for the period 1991-95. We find that multinationality does not have a significant effect on short run IPO performance. However, its effects on long run performance are significant and positive. Further, we find that the factors affecting multinational IPO performance are systematically different from those affecting domestic IPO performance, in both the short and long runs.
Keywords: Initial public offerings, multinational activity
INTRODUCTION
Diversification strategies are commonly developed in order to stabilize cash flows and therefore to reduce risk as perceived by investors and creditors (Madura and Whyte, 1990). This diversification can take two forms. Pure product line diversification involves producing a wider range of products in the same geographical setting, while pure geographical diversification involves producing the same line of products but developing a wider geographical reach (Buhner, 1987; Thompson, 1985, 1988). Multinational operations are typically the end result of geographical diversification (Buckley and Casson, 1981; Hood and Young, 1979; Mudambi and Ricketts, 1998).
The level of multinational activity has been increasing for firms in most parts of the world (UNCTAD, 1997). This has been observed for firms in virtually all countries of the world. Further, it is not limited to firms in industrialized countries – large firms in many third world countries are increasingly turning to multinational activity (Dunning and Hamdani, 1997). While risk reduction is certainly a factor, historically, multinational activity has also been aimed at generating excess returns.
A second important business trend is the increasing levels of stock market listing and capitalization in both developed and developing countries. In countries ranging from Indonesia to Switzerland, the last decade and a half has seen a prodigious rate of increase (Jenkinson and Ljungqvist, 1996). The market for initial public offerings (IPOs) has been extensively studied. The motivations for going public are many, and include gaining access to capital markets, increasing the liquidity of the firm’s assets through expanded trading opportunities and establishing the value of the firm.
The literature on IPO performance is virtually unanimous and has documented two anomalies in virtually all stock markets of the world (see Jenkinson and Ljungqvist (1996) and Khurshed and Mudambi (1998) for recent reviews of this literature). In the short run, prices tend to be characterized by underpricing, so that the shares of companies going public are offered to investors at prices considerably below the levels at which they subsequently trade on the stock market. However, in the long run, relative to other quoted companies, IPO firms tend to underperform market indices for similarly sized firms. A number of theories have been put forward to explain these anomalies (Tinic, 1988; Jenkinson and Ljungqvist, 1996). In addition, several studies have suggested that long run underperformance disappears if IPO firms are compared to portfolios of firms that are similar in terms of risk and capitalization (Spiess and Affleck-Graves, 1995; Jegadeesh, 1997; Eckbo et al, 1998). It should be noted that most of these studies focus on seasoned equity offerings. However, the question of how IPO performance is affected by multinationality has not been addressed.
In the literature examining the impact of international and multinational activity on firm performance, the results generally support the notion that multinational activity is successful in improving the firm’s rate of return and survival prospects. Grant (1987) finds that multinationality improves profitability for the largest UK firms and that this is consistent across a number of different measures of profit. In the same vein, Grant et al (1988) report that for large firms, multinational diversification is spurred by domestic profitability and that it in turn increases profitability. On the basis of a sample of large US and European multinational enterprises (MNEs), Geringer et al (1989) report that both geographical and product line diversification strategies improve MNE performance.
Mitchell et al (1993) report that the effects of a multinational diversification strategy exhibit considerable path dependence. International expansion is found to be necessary for survival when domestic markets are entered by foreign rivals. However, only firms with international experience were found to be able to take advantage of this strategy and further, firms that retrenched from such a strategy failed to survive. However, this finding with regard to reversals in level of multinational activity is in contrast to the findings of Tsetsekos and Gombola (1992), who report that foreign plant closings have insignificant effects on firm value.
Hitt et al (1997) find that geographical and product line diversification are related. They report that the effects of multinational activity depend on the level of product line diversification. Thus, multinationality has a strong positive effect on returns for firms that are highly diversified in terms of their product line. However, multinationality has a negative effect on returns for firms that are undiversified in terms of their product line.
In an important study, Morck and Yeung (1991) point out that multinationality may be a summary measure for a number of other factors that impact on firm performance. Thus, they report that multinationals are more R&D-intensive and have unique intangible assets. In the absence of these assets, multinationality per se does not seem to improve firm performance or increase firm value.
The literature on multinational diversification has focused on the performance of large, established firms. The literature on IPOs has not examined the effects of multinational diversification on IPO performance. In this paper, we are interested in addressing this gap in the literature. In particular, we are interested in the following question: what is the effect of multinational diversification on IPO performance? We are interested in performance both in the short run and in the long run. It is of particular interest to address the question of whether multinationality has the same effect on small newly public firms as it does on large, established firms.
We find that multinationality does not have a significant effect on IPO performance in the short run. However, in the long run, the effect is extremely significant and positive. When focusing only on IPO firms with multinational activity, an increasing degree of multinationality is associated with a significantly improved long run performance. Further, the factors affecting short run and long run performance are different for IPO firms with multinational activity and those with purely domestic activity.
The paper is organized as follows. In the following section, we describe our methodology and data. Subsequently, we discuss the estimation and our results. Finally, we offer some concluding comments.
METHODOLOGY AND DATA
In order to address the question at hand, two steps need to be undertaken. In the first step, we need to define and estimate IPO returns and in the second we need to relate these to multinational activity. In the second step, we also need to normalize the relationship for a number of other factors that affect IPO price performance.
Estimating IPO returns
Short run IPO returns are estimated using the market adjusted abnormal returns at the close of the first day of trading (MAAR0). In computing MAAR0, we follow the methodology of Aggarwal et al (1993). These returns are computed relative to a market index. Since the IPO firms are generally quite small, we use a small companies’ index as the appropriate benchmark. The index we use is the Hoare Govett Small Companies (HGSC) index.
Thus, the total return for firm ‘i’ at the end of the first day’s trading is Ri1 = (Pi1/Pi0), where Pi0 is the IPO offer price and Pi1 is the firm’s stock price at the end of the first day’s trading. Similarly, the return on the market index for the same period is Rm1 = (Ii1/Ii0), where Ii0 is the value of the market index at the beginning of the offer day of firm ‘i’ and Ii1 is the value of the index at the close of trading on the offer day. Using these two returns, MAAR0 for firm ‘i’ is:
MAAR0i = {[(1+Ri1)/(1+Rm1)] - 1} (1)
The market adjusted long run buy-and-hold abnormal returns in the after-market are calculated for a period of 36 months following the first month of trading (MAAR36). This computation uses the stock price on the last trading day of each month. Dividend payments and rights and scrip issues are incorporated. Allowing for the initial underpricing and the possibility of price support in the first few days of trading, the first month of trading is excluded in the computation of long run returns. In computing long run returns, we use the computation methodology from Ritter (1991):
MAAR36 = å t=2, t=37 [ln(Pit/Pit- 1) - ln(Imt/Imt- 1)] (2)
Multinational activity and normalizing factors
The IPO firm’s level of multinational
activity is determined by the geographical scope of its subsidiaries. The geographical
areas considered are: the UK, mainland Europe, North America, South America,
Asia, Australia and Africa. The index is constructed as follows – each firm
is automatically assigned a score of 1, since all are found to have UK operations.
Its score increases by 1 for every other area in which it has operations. Thus,
if it has operations in one area outside the UK, it receives a score of 2,while
if it has operations in two areas outside the UK, it receives a score of 3,
and so on.
The maximum value of the index is 7.
Obviously, IPO performance is affected by a number of factors other than its multinationality. Hence we use a number of normalizing factors to control for these effects. As reported by Hitt et al (1997), product line diversification is an important normalizing factor. Hence we introduce product line diversification by considering the number of two-digit industry codes in which the firm reports operations. Following Ritter (1984), we control for so-called ‘hot-issue periods’ by introducing time dummies.
Characteristics of the firm before
and at the IPO are likely to have an influence on post-IPO performance. Hence,
we introduce the firm’s pre-IPO reported profits and the cost of flotation.
The cost variable is based on the notion that firms judged to be weaker by investment
bankers have to incur higher costs in going public (Ritter, 1991). In the case
of long run performance, we also use short run performance as an explanatory
variable, following Rajan and Servaes (1997). It has been suggested that size,
age and underwriter reputation are likely to have a positive effect on performance
(Carter et al, 1998; Mudambi, 1998). We introduce two different measures
of size. We capture the size of the firm by its reported assets and the size
of the IPO by the amount of funds raised. We introduce age by using the duration
of its pre-IPO operations. We proxy the reputation of the underwriters of the
IPO by using a measure of underwriter size.
Our methodology here is similar to Carter and Manaster (1990).
We capture agency effects (Mello and Parsons, 1992) by using the percentage of the equity of the firm that is issued at the IPO. The rationale here is that the greater of proportion of the firm that is sold to the public (and consequently the smaller proportion that is held by its owners), the lower is the incentive for the owners to put a maximal level of effort into the firm. Finally, we take account of industry-specific effects by introducing industry dummies. These dummies are based on the industry categories used by the London Stock Exchange. Full details of all variables used are provided in the Appendix.
Estimating the effects of multinational activity
In order to estimate the effect of multinational activity on IPO performance, we estimate the following relationship:
MAARi(j) = f(multinational activity, normalizing factors) (3)
We estimate the effects on short run performance when j=0 and on long run performance when j=36. Examining the effects of the normalizing factors enables to relate our findings to those reported in the literature. If we are able to replicate earlier findings, this provides strong evidence that the estimated results are not sample specific, but have more general validity.
Data
The sample used in this study comprises 240 IPOs of non-investment trust companies floated on the UK Official List from January 1991 through June 1995. There were a total of 252 non-investment trust IPOs during this period. Thus, the sample coverage is over 95% of new issues of placements and offers for sale during the period. The identification of the target population of issues is undertaken using the London Share Price Database (LSPD). The 12 excluded IPOs had incomplete data entries in this database, as well as in other publications. Details of sample coverage are provided in Table 1.
Table 1: IPOs on the London Main Market: Year-by-year Sample Coverage
Year |
Total IPOs |
Sample |
Sample Coverage (%) |
1991 |
15 |
12 |
80.00 |
1992 |
26 |
25 |
96.15 |
1993 |
66 |
64 |
96.96 |
1994 |
119 |
114 |
95.79 |
1995 |
26 |
25 |
96.15 |
Total |
252 |
240 |
- |
Mean |
51.8 |
48.00 |
95.23 |
Information concerning the particulars of each offering, including multinationality and normalizing factor data, are obtained from the following sources: (a) The Extel Book of Takeovers, New Issues and Offers; (b) KPMG New Issue Statistics; (c) Extel Company Research; (d) the London Stock Exchange; and (e) The Hamilton Scott Smaller Companies Guide. The multiplicity of the sources was also useful in checking for any discrepancies in the data.
The age of the issuing firm at the time of the offering was obtained from the London Stock Exchange Yearbooks. Age was defined to be the number of days between the date of registration and the first day of trading. Initial trading prices and initial returns were determined using Datastream’s UK Equity Database and monthly returns over three years were obtained from the LSPD.
Summary statistics with regard to all variables used in the study are provided in Table 2. These statistics are reported separately for all IPO firms, multinational firms and domestic firms. Multinational IPO firms are those firms which report subsidiary operations in at least one region other than the UK, while domestic IPO firms are those which report operations wholly within the UK. Multinationals appear to have higher unconditional returns in the short run, but in the long run the difference is very small. It is interesting to note that the multinational firms in the sample are actually smaller, on average, than domestic firms in terms of assets. However, they seem to raise a slightly higher level of IPO funds and their cost of flotation is slightly lower. The domestic firms are also older on average. Both types of firm issue roughly the same percentage of their equity at the IPO. Thus, the effects of multinationality that appear in this study cannot be attributed to size, operational experience or agency effects.
Table 2: Summary Statistics
Variable |
All Firms n=194 |
Multinationals n=92 |
Domestic Firms n=102 |
|||
Mean |
S.D. |
Mean |
S.D. |
Mean |
S.D. |
|
MAAR36 |
- 0.1826 |
0.7383 |
- 0.1784 |
0.8056 |
- 0.1864 |
0.6759 |
MAAR0 |
0.0949 |
0.1768 |
0.1118 |
0.1984 |
0.0797 |
0.1542 |
YR92 |
0.1082 |
0.3115 |
0.1304 |
0.3386 |
0.0882 |
0.2850 |
YR93 |
0.2887 |
0.4543 |
0.3478 |
0.4789 |
0.2353 |
0.4263 |
YR94 |
0.4845 |
0.5011 |
0.4348 |
0.4984 |
0.5294 |
0.5016 |
YR95 |
0.1134 |
0.3179 |
0.0761 |
0.2666 |
0.1471 |
0.3559 |
GSCOPE |
1.8866 |
1.2374 |
2.8696 |
1.1787 |
1 |
0.000 |
DIVERPRD |
1.6186 |
0.8750 |
1.6739 |
0.8658 |
1.5686 |
0.8845 |
PROFLOAT |
2759.765 |
8642.685 |
3081.370 |
9699.818 |
2469.689 |
7601.696 |
COST |
0.0591 |
0.0384 |
0.0557 |
0.0336 |
0.0622 |
0.0423 |
FUNDS |
38766.06 |
76985.46 |
42581.08 |
81672.56 |
35325.07 |
72733.81 |
ASSFLOAT |
17099.05 |
51589.83 |
13566.03 |
53886.07 |
20285.70 |
49477.46 |
DURATION |
2600.582 |
3903.758 |
2182.837 |
3054.832 |
2977.372 |
4517.778 |
MSHARE |
0.0454 |
0.0655 |
0.0428 |
0.0565 |
0.0478 |
0.0729 |
EQUISSUE |
0.4599 |
0.1814 |
0.4564 |
0.1689 |
0.4630 |
0.1928 |
ID1 |
0.0258 |
0.1589 |
0.0326 |
0.1786 |
0.0196 |
0.1393 |
ID2 |
0.0567 |
0.2319 |
0.0761 |
0.2666 |
0.0392 |
0.1951 |
ID3 |
0.1443 |
0.3523 |
0.1957 |
0.3989 |
0.0980 |
0.2988 |
ID4 |
0.1495 |
0.3575 |
0.1304 |
0.3386 |
0.1667 |
0.3745 |
ID5 |
0.0361 |
0.1870 |
0.0326 |
0.1786 |
0.0392 |
0.1951 |
ID6 |
0.1753 |
0.3812 |
0.1522 |
0.3612 |
0.1961 |
0.3990 |
ID7 |
0.0567 |
0.2319 |
0.0326 |
0.1786 |
0.0784 |
0.2702 |
ESTIMATION AND RESULTS
Base-line estimates were obtained by estimating equation (3) for all IPO firms using ordinary least squares (OLS). Since the firms varied considerably in terms of size, heteroskedasticity was a very real concern. Running the Breusch-Pagan test confirms this fear, as the null hypothesis of homoskedasticity is strongly rejected. We deal with this problem by using White’s heteroskedasticity-consistent variance-covariance matrix in estimating the ‘t’ statistics. We note that using the White matrix, the problem appears to be resolved (see Tables 3 and 4).
We estimate equation (3) separately for short run returns (j=0) and for long run returns (j=36). Estimation of short run returns in presented in Table 3 and estimation of long run returns is presented in Table 4. We compare the base-line estimates for all firms, with estimates that are obtained when multinational IPO firms are separated from domestic IPO firms.
Short run returns
Examining the results presented in Table 3, we note that the overall fit of the base-line estimates for short run returns is not very good. Only two variables are significant at the 10% level, one of which is an industry dummy (ID5). Duration of pre-IPO operations (DURATION) has a very weak positive influence. However, when we disaggregate the sample into multinational and domestic firms, we find slightly stronger effects. However, the degree of multinationality (GSCOPE) does not appear to have a significant effect on short run returns for the full sample or for multinational firms. However, for domestic firms, GSCOPE always takes the value 1, and the constant term, which is highly significant, may pick up some of this effect.
Table 3: Estimating
Short Run Returns
OLS estimates of abnormal
returns at the end of the first day’s trading
Regressand: MAAR0
Regressor |
Coefficient (‘t’ statistic)@ |
||
All Firms |
Multinational Firms |
Domestic Firms |
|
Constant |
4.996 (0.86) |
- 2.140 (0.19) |
15.902 (2.87)*** |
YR92 |
6.114 (1.06) |
15.860 (1.87)* |
- 1.777 (0.58) |
YR93 |
15.126 (1.97)* |
25.304 (2.11)** |
7.696 (1.37) |
YR94 |
4.779 (0.92) |
12.150 (1.61) |
- 1.477 (0.50) |
YR95 |
6.514 (1.13) |
13.139 (1.56) |
- |
GSCOPE |
1.108 (1.15) |
1.982 (1.15) |
- |
DIVERPRD |
- 1.233 (1.07) |
- 1.563 (0.75) |
- 0.887 (0.67) |
PROFLOAT |
- 0.446´ 10- 4 (0.37) |
0.243´ 10- 3 (1.67)* |
0.17´ 10- 3 (0.96) |
COST |
5.469 (0.19) |
2.936 (0.05) |
- 5.524 (0.17) |
FUNDS |
0.174´ 10- 4 (1.58) |
- 0.271´ 10- 5 (0.12) |
0.47´ 10- 4 (2.31)** |
ASSFLOAT |
- 0.155´ 10- 4 (1.01) |
- 0.77´ 10- 4 (2.44)** |
- 0.22´ 10- 4 (0.61) |
DURATION |
0.399´ 10- 3 (1.79)* |
- 0.196´ 10- 3 (0.59) |
- 0.53´ 10- 3 (1.90)* |
MSHARE |
- 30.937 (1.36) |
- 84.884 (2.22)** |
- 7.478 (0.33) |
EQUISSUE |
- 6.752 (1.28) |
- 4.487 (0.43) |
- 12.615 (1.89)* |
ID1 |
0.643 (0.11) |
4.575 (0.64) |
- 21.876(2.90)*** |
ID2 |
0.502 (0.14) |
- 3.459 (0.54) |
1.362 (0.33) |
ID3 |
1.446 (0.45) |
- 1.742 (0.33) |
- 0.030 (0.01) |
ID4 |
0.959 (0.36) |
0.578 (0.12) |
- 0.258 (0.09) |
ID5 |
- 4.971 (1.86)* |
- 4.358 (0.83) |
- 12.772 (3.37)*** |
ID6 |
5.777 (1.21) |
1.768 (0.23) |
- |
ID7 |
0.525 (0.15) |
- 8.991 (1.46) |
- |
Diagnostics |
|||
R2 |
0.0069 |
0.0935 |
0.0520 |
F stat (d.f.) |
2.07 (20, 173) |
2.61 (20, 71) |
2.69 (16, 85) |
Log-L |
- 820.7777 |
- 397.6088 |
- 417.0904 |
Restricted Log-L |
- 832.0638 |
- 404.9106 |
- 423.3004 |
Likelihood Ratio |
22.5722 |
14.6036 |
12.4200 |
AIC |
8.678 |
9.100 |
8.512 |
Breusch-Pagan test: c 2(d.f.); (‘p’ value) |
20.3435 (20) (0.4366) |
17.1007 (20) (0.6464) |
18.5756 (18) (0.4184) |
Sample Size |
194 |
92 |
102 |
@ White’s heteroskedasticity-consistent
standard errors used to compute ‘t’ statistics.
* Coefficient significant
at the 10% level. ** Coefficient
significant at the 5% level ***
Coefficient significant at the 1% level
Table 4: Estimating
Long Run Returns
OLS estimates of abnormal
returns at the end of 36 months’ trading
Regressand: MAAR36
Regressor |
Coefficient (‘t’ statistic)@ |
||
All Firms |
Multinational Firms |
Domestic Firms |
|
Constant |
- 0.712 (2.53) |
- 0.876 (2.15)** |
0.659 (1.92)* |
YR92 |
0.949 (3.75)*** |
0.826 (2.06)** |
- 0.0078 (0.04) |
YR93 |
0.794 (3.42)*** |
0.693 (1.90)* |
- 0.184 (0.80) |
YR94 |
0.895 (3.82)*** |
0.749 (2.43)** |
- 0.031 (0.16) |
YR95 |
0.981 (3.63)*** |
0.970 (2.50)** |
- |
GSCOPE |
0.097 (2.44)** |
0.206 (3.09)*** |
- |
DIVERPRD |
- 0.071 (0.91) |
- 0.118 (1.00) |
0.006 (0.07) |
PROFLOAT |
- 0.021 (3.23)*** |
- 3.3´ 10- 5 (4.39)*** |
- 3.7´ 10- 6 (0.68) |
MAAR0 |
- 0.006 (2.30)** |
- 0.0076 (2.04)** |
- 4.97´ 10- 4 (0.14) |
COST |
- 2.53 (1.83)* |
0.265 (0.12) |
- 4.157 (2.80)*** |
FUNDS |
2.06´ 10- 4 (0.35) |
- 1.70´ 10- 6 (1.47) |
1.48´ 10- 6 (1.54) |
ASSFLOAT |
1.543´ 10- 3 (1.84)* |
- 4.39´ 10- 7 (0.37) |
1.8´ 10- 7 (0.12) |
DURATION |
7.2´ 10- 6 (0.69) |
3.4´ 10- 5 (1.46) |
- 1.32´ 10- 5 (1.14) |
MSHARE |
0.026 (0.03) |
3.541 (2.40)** |
- 0.652 (0.97) |
EQUISSUE |
- 0.635 (2.36)** |
- 1.156 (2.35)** |
- 0.830 (2.60)** |
ID1 |
0.789 (2.93)*** |
0.296 (1.13) |
0.653 (2.67)*** |
ID2 |
0.014 (0.06) |
- 0.141 (0.51) |
- 0.033 (0.11) |
ID3 |
- 0.229 (1.48) |
0.010 (0.05) |
- 0.877 (3.69)*** |
ID4 |
0.052 (0.39) |
0.132 (0.53) |
- 0.195 (1.38) |
ID5 |
- 0.097 (0.47) |
- 0.149 (0.36) |
- 0.447 (2.83)*** |
ID6 |
0.361 (2.31)** |
0.385 (1.67) |
- |
ID7 |
0.516 (2.51)** |
1.052 (5.57)*** |
- |
Diagnostics |
|||
R2 |
0.0846 |
0.1770 |
0.0515 |
F stat (d.f.) |
2.85 (21, 172) |
3.93 (21, 70) |
2.32 (17, 84) |
Log-L |
- 196.1679 |
- 89.1196 |
- 92.1803 |
Restricted Log-L |
- 215.9118 |
- 110.1503 |
- 104.2765 |
Likelihood Ratio |
39.4878 |
42.0614 |
24.1924 |
AIC |
2.249 |
2.416 |
2.160 |
Breusch-Pagan test: c 2 (d.f.); (‘p’ value) |
26.0689 (21) (0.2038) |
22.3244 (21) (0.3810) |
22.3314 (17) (0.1723) |
Sample Size |
194 |
92 |
102 |
@ White’s heteroskedasticity-consistent
standard errors used to compute ‘t’ statistics.
* Coefficient significant
at the 10% level. ** Coefficient
significant at the 5% level. ***
Coefficient significant at the 1% level
The fit of the equation is best for multinational firms. For these firms, we find a weak positive effect of pre-IPO profits (PROFLOAT), but a strong negative effect associated with firm size (ASSFLOAT). We also find that a higher underwriter reputation (MSHARE) is associated with lower short run returns (less short run underpricing).
The fit of the equation for domestic firms is also better than the fit of the base-line equation, suggesting that the base-line equation is composed of two separate underlying relationships. Here we find a weak negative effect associated with DURATION. However, we find a strong positive effect associated with the size of the IPO as measured by the funds raised (FUNDS). Weak agency effects are observed through the negative effect of the percentage of equity issued (EQUISSUE). Two of the industry dummies (ID1 and ID5) have strong negative effects.
Long run returns
Turning to the estimates in Table 4, we note that the fit of all estimations of equation (3) is very good for long run returns. Again the base-line estimates for all firms act as a composition of the underlying estimates for the multinational and domestic firms. However, since the all fits are so much better, we see many statistically significant effects.
Beginning with the base-line estimates for all firms, we see strong effects associated with all year dummies, indicating that returns were relatively poorer in the reference year (1991). Multinationality (GSCOPE) has a significant positive effect on long run returns in the base-line estimates. Product line diversification (DIVERPRD), however, does not appear significant. Pre-IPO profits (PROFLOAT) now have a very strong negative effect, as does the initial level of underpricing (MAAR0). Firm size as measured by assets (ASSFLOAT) now has only a weak positive effect and the per unit cost of flotation (COST) has a weak negative effect. An increasing percentage of the firm’s equity offered (EQUISSUE) is associated with a lower level of long run returns. Three of the industry dummies emerge as significant and positive (ID1, ID6 and ID7).
Turning to the estimates for the multinational firms, we see that the fit is surprisingly good (relative to typical estimates reported in this literature). Further, they drive many of the results that appear in the base-line estimates. These include the estimates of the parameters associated with the year dummies, the strong negative effect of pre-IPO profits (PROFLOAT) and the initial level of underpricing (MAAR0). The percentage of the equity of the firm offered (EQUISSUE) also has a strong and significant negative effect.
The level of multinationality (GSCOPE) is also strongly positive. The effect of underwriter reputation (MSHARE) now emerges as strongly significant and positive. This effect is not seen either for the base-line estimates or for domestic firms. Only one industry dummy (ID7) appears significant.
The fit for domestic firms is poorer, but more typical of the earlier literature. Only two effects emerge strongly here (and carry over to drive similar results in the base-line estimates). The first is that the per unit cost of flotation (COST) has a strong negative effect. The second is that the percentage of equity issued (EQUISSUE) also has a strong negative effect. Three industry dummies appear significant (ID1, ID3 and ID5).
Discussion
We note that the estimates for short run returns are far poorer than the estimates for the long run returns. The only factors that seem to matter are size and underwriter reputation. Firm size has a negative effect on short run returns for multinational firms, whereas flotation size has a positive effect for domestic firms. The reputation of the underwriter has a negative effect on short run returns for multinational firms. This could be because better underwriters set the issue price to extract a greater amount of surplus from investors. Similar results are reported by Carter et al (1998).
However, we find that multinationality plays no significant role in determining short run returns. Further, the effects of duration or the age of the firm are ambiguous. While duration seems to have a weak positive effect for all firms, it does not appear significant for multinationals and appears to be negatively related to returns for domestic firms. Thus, we really cannot reach any meaningful conclusions here.
Our findings with regard to long run returns are far more interesting. Here we find that multinationality has a strong and significant positive effect. Thus, in the base-line estimates, multinationality is significant and positive, indicating that multinational firms, ceteris paribus, have better long run returns than domestic firms. Further, even amongst multinationals, the index of multinationality has a strong positive effect. Thus, even amongst firms that have multinational operations, an increasing level of multinationality is associated with an improved long run performance.
The level of short run return seems to be significant and negatively associated with long run performance. This effect appears in the base-line estimates and is traced to the multinational firms on disaggregation. The same is true for pre-IPO profits. These effects may be ascribed to overoptimism, i.e., the larger the short run returns generated by the IPO, the greater the overoptimism, so that the long run returns are relatively lower. This is consistent with the results reported by Rajan and Servaes (1997).
The influence of underwriters appears in different forms, depending upon whether the firm is multinational or domestic. For the multinationals, underwriter reputation appears to be associated in a direct and significantly positive way with long run performance. However, for domestic firms, the influence of underwriters appears in an indirect way, through the finding that higher per unit flotation costs are associated with significantly worse long run performance. Thus, firms judged to be poor by the investment bankers, in fact do perform relatively worse.
A significant finding of this study is the extremely strong evidence of agency effects in IPOs. Higher percentages of equity issued at IPO are consistently associated with worse long run performance. Thus, the more the IPO is used as an ‘exit’ strategy for the entrepreneur, the poorer the long run prospects for the firm.
Product line diversification does not appear significant. This is in contrast to the findings of Hitt et al (1997). We find that the introduction of interactions between geographical and product line diversification considerably worsens the fit of the equation and also does not produce statistically significant results. Firm age also does not appear to be a significant factor. This is in contrast to the findings reported by Carter et al (1998) for US data. Size, both in terms of the size of the firm and the size of the offering, does not seem to matter. While, we find a weak positive effect associated with firm size in the base-line estimates, this effect disappears in the disaggregated estimates.
CONCLUDING REMARKS
There is a large literature examining the effect of multinationality on large, established firms. There is also a large literature examining the performance of newly public firms (typically, small and medium sized firms), coming to market through an IPO. However, the effect of multinationality on IPO firms has not been addressed. Our objective in this paper is to study the effect of multinationality on IPO performance in the short and long run.
We approach the question using a comprehensive dataset from the London main market. We find that multinationality has little explanatory power in the context of short run performance. Indeed, the only significant factors in explaining short run performance are some firm and offer size factors. However, we are able to explain long run performance quite well. Here multinationality appears to have a strong positive effect, a result that agrees with many findings for large, established firms. Further, a higher level of multinationality has a positive effect on performance even for the sub-sample of multinational firms.
We also find strong and consistent evidence of agency effects whereby the greater the percentage of the firm sold at the IPO (and consequently the lower the percentage retained by the entrepreneur), the worse the long run performance of the firm. The influence of the underwriters is also significant and we find that these effects are different for multinational and domestic IPO firms.
Our bottom line finding is that multinational and domestic IPO firms differ systematically in terms of the factors that affect their short and long run performance. We note that the sub-samples of multinational and domestic firms are similar along the dimensions of size, age and percentage of equity issued. Thus, we are confident that the effects that we capture can be ascribed to multinationality and not to an underlying latent factor. In the spirit of Morck and Yeung (1991), we can conclude that even at the stage of IPO, multinational firms possess unique intangible assets that create superior long run returns.
APPENDIX
Variables Used in the Analysis
Variable |
Definition |
MAAR36 |
Market adjusted long run buy-and-hold abnormal returns for 36 months following the first month of trading, incl. dividend payments and rights and scrip issues, using the stock price on the last trading day of each month. Excludes the first month of trading |
MAAR0 |
Market adjusted abnormal returns at the close of the first day of trading relative to the Hoare Govett Small Companies (HGSC) index. |
YR‘i’ |
Year dummy for IPO occurring in year ‘i’ |
GSCOPE |
Number of regions in which the IPO firm has subsidiaries. Regions are: The UK, other Europe, North America, South America, Asia, Australasia and Africa. |
DIVERPRD |
Number of two-digit industries in which the IPO firm has lines of business |
PROFLOAT |
Average pre-tax profits or losses for the last 3 years before the IPO (£ ‘000) |
COST |
Total direct costs, as a percentage of funds raised, incurred for the listing. Includes underwriting fees, legal expenses, accountancy and audit fees |
FUNDS |
The amount of funds raised in the IPO (£ ‘000) |
ASSFLOAT |
Net assets of the firm in the year before listing (£ ‘000) |
DURATION |
Difference between the date of registration of the company and the date of listing (days) |
MSHARE |
Market share (by value) of the lead underwriter expressed as a percentage of the underwriting market in the year of the issue |
EQUISSUE |
Percentage of the equity of the firm issued at the IPO |
ID1 |
Industry dummy – Extraction of minerals, mineral products and ores and chemicals |
ID2 |
Industry dummy – Metal goods, engineering and vehicle industries |
ID3 |
Industry dummy – Other manufacturing industries |
ID4 |
Industry dummy – Construction |
ID5 |
Industry dummy – Distribution including hotels and catering |
ID6 |
Industry dummy – Transport and communication |
ID7 |
Industry dummy – Banking, finance, insurance, business services and leasing |
REFERENCES
Aggarwal, R., Leal, R. and Hernandez, L. 1993. The after-market performance of initial public offerings in Latin America. Financial Management, 22: 42-53.
Buckley, P. and Casson, M. 1981. The optimal timing of foreign direct investment. Economic Journal, 91: 75-87.
Carter, R.B. and Manaster, 1990. Initial public offerings and underwriter reputation. Journal of Finance, 45(4): 1045-1067.
Carter, R.B., Dark, F.H. and Singh, A.K. 1998. Underwriter reputation, initial returns, and the long-run performance of IPO stocks. Journal of Finance, 53(1): 285-311.
Buhner, R. 1987. Assessing international diversification of West German corporations. Strategic Management Journal, 8(1): 25-37.
Dunning, J. and Hamdani, K. 1997. The new globalism and developing countries. Tokyo: United Nations University.
Eckbo, E., Masulis, R.W. and Norli, O. 1998. The conditional long run performance of security offerings: Is there a new issues puzzle? Working Paper, Dartmouth College.
Geringer, J.M., Beamish, P.W. and daCosta, R.C. 1989. Diversification strategy and internationalization. Strategic Management Journal, 10(2): 109-119.
Grant, R.M. 1987. Multinationality, and performance among British manufacturing companies. Journal of International Business Studies, 18(3): 79-89.
Grant, R.M., Jammine, A.P. and Thomas, H. 1988. Diversity, diversification and profitability among British manufacturing companies. Academy of Management Journal, 31(4): 771-801.
Khurshed, A. and Mudambi, R. 1998. The short run price performance of investment trust IPOs on the UK main market. Discussion Papers in Economics and Management, Series A, 11, No.388, University of Reading.
Hitt, M.A., Hoskisson, R.E. and Kim, H.1997. International diversification: Effects on innovation and firm performance in product-diversified firms. Academy of Management Journal, 40(4): 767-798.
Hood, N. and Young, S. 1979. The economics of multinational enterprise. London: Longman.
Jegadeesh, N. 1997. Long-run performance of seasoned equity offerings: Benchmark errors and biases in expectations. Working paper, University of Illinois.
Jenkinson, T.J. and Ljungqvist, A.P. 1996. Going public: The theory and evidence on how companies raise equity finance. Oxford: The Clarendon Press.
Madura, J. and Whyte, A.M. 1990. Diversification benefits of direct foreign investment. Management International Review, 30(1): 73-85.
Mello, A.S. and Parsons, J.E. 1992. Measuring the Agency Cost of Debt. Journal of Finance, 47(5): 1887-1904.
Mitchell, W., Shaver, J.M. and Yeung, B. 1993. Performance following changes of international presence in diversified industries. Journal of International Business Studies, 24(4): 647-669.
Morck, R. and Yeung, B. 1991. Why investors value multinationality. Journal of Business, 64(20): 165-187.
Mudambi, R. 1998. The Role of Duration in MNE Investment Strategies. Journal of International Business Studies, 29(2): 239-262.
Mudambi, R. and Ricketts, M.J. 1998. Economic organisation and the multinational firm. In R. Mudambi and M.J.Ricketts, editors, The organisation of the firm: International business perspectives. London: Routledge.
Rajan, R. and Servaes, H. 1997. Analyst following of initial public offerings. Journal of Finance, 52(2): 507-529.
Ritter, J.R. 984. The "hot issue" market of 1980. Journal of Business, 57, 215-241.
Ritter, J.R. 1991. The long-run performance of initial public offerings. Journal of Finance, 46, 3-27.
Spiess, D.K. and Affleck-Graves, J. 1995. Underperformance in long-run stock returns following seasoned equity offerings. Journal of Financial Economics, 38, 243-267.
Thompson, R.S. 1985. Risk reduction and international diversification: An analysis of large UK multinational companies. Applied Economics, 17(3): 529-541.
Thompson, R.S. 1988. Product differentiation in newspaper industry: A hedonic price approach. Applied Economics, 20(3): 367-376.
Tinic, S. 1988. Anatomy of the IPOs of common stock. Journal of Finance, 43(4): 789-822.
Tsetsekos, G.P. and Gombola, M.J. 1992. Foreign and domestic divestments: Evidence of valuation effects of plant closings. Journal of International Business Studies, 23(2): 203-223.
UNCTAD. 1997. World investment report: Transnational corporations, market structure and competition policy. Geneva: United Nations.