Why Have Auctions Been Losing Market Shares to Bookbuilding in IPO Markets?

Ji-Chai Lin

Louisiana State University

Yi-Tsung Lee

National Chengchi University

Yu-Jane Liu

National Chengchi University

Comments welcome

Good things about IPO auctions

All bidders are on equal footing

No favoritism

Less average underpricing (e.g., Derrien and Womack (2002), Kandel, Sarig, and Wohl (1999), and Pettway and Kaneko (1996)).

Some puzzling factsabout IPO auctions

Sherman (2002) Losing market share to bookbuilding Only Taiwan and Israel Many countries that experimented with IPO

auctions in the 1990s or 1980s abandoned them within a few years.

Ljungqvist, Jenkinson, and Wilhelm (2003) Auctions are rarely used in IPO markets Increasing popularity of bookbuilding

Why are IPO auctions not popular?

Auctions are popular in many settings in selling items with uncertain valuation.

Why not in IPO markets? Could we get some answers from auctions

data? Would key players in IPO markets be better

off under bookbuilding?

Three Issues

1) In bookbuilding, allocation policies favor institutional investors.

Would auctions be a fairer mechanism in

IPO allocations?

2) Sherman (2002) postulates that bookbuilding has an edge in pricing over auctions.

What empirical evidence can be offered to suggest that auctions may not perform as well as bookbuilding?

Second Issue

3) Benveniste and Spindt (1989) posit that institutional investors could increase their expected profits under bookbuilding from a tradeoff between a higher allocation of shares and a lower level of underpricing.

Do observed institutional allocations and underpricing required by institutional investors in auctions make the tradeoff feasible?

Third Issue

The objective

Analyze the bidding data from 89 Taiwan IPO auctions to provide answers to the three empirical questions.

Shed some light on why auctions have been losing market shares to bookbuilding.

Taiwan Auctioned IPOs

Competitive bidding for half of the IPO shares

Public subscription at a fixed price for the remaining half of the shares.

The public subscription is limited to only one round lot (i.e., 1000 shares) per person.

No bidder shall be allowed to win more than six percent of the shares designated for auction.

Winning bidders pay what they bid.

Table 1. The Distribution of the 89 IPO Auctions Held in Taiwan by Year

Year Frequency

1996

1997

1998

1999

2000

2001

2002

7

23

23

19

11

4

2

Total 89

Table 2. Summary Statistics of IPO Auctions

Table 2. ( Continued )

Institutional Allocations

Inst_Alloc=The percentage of shares won by institutions.

Average Inst_Alloc=19%(much lower than the norm of around 70% in countries using bookbuilding)

Ten auctions with Inst_Alloc=0 Auctions with high Inst_Alloc

if Inst_Alloc > 20%

Auctions with high Inst_Alloc

set a higher base price put more shares for IPO auction attract more bidders

Institutional Investors

Represent less than 5% of all bidders. Submit about 22% of shares in all bids. Seem to know in which auctions and at

which prices they should bid to win shares.

Table 3. Characteristics of Institutional Investors and Retail Investors in IPO Auctions

Three Evaluation Metrics

H o w w e l l d o a u c t i o n s r e v e a l t h e v a l u e o f I P O s ?

T h e h o l d i n g p e r i o d r e t u r n :

, ,( ) /w w wi t i t i iR P P P

T h e m a r k e t - a d j u s t e d r e t u r n :

, , ,w w

i t i t m tM A R R R

T h e a l p h a i n t h e f o l l o w i n g c r o s s - s e c t i o n a l e x c e s s r e t u r n r e g r e s s i o n :

, , , , ,( )wi t f t m t f t i tR R R R e ,

Table 4. Event Study with MAR

Table 5. The Initial Returns of the 89 IPO Auctions

Evaluation Metric

Mean (%) t-valuea

Std (%)

Min (%)

Median (%)

Max (%)

Panel A: The whole sample, N=89 R 2.39 0.74 30.47 -57.56 -1.58 89.31 MAR 2.79 1.02 25.69 -53.91 1.05 90.22 2.80 0.32

The pricing performance of Taiwan IPO auctions

On average, winning bidders earn about two percent from the auction day to the 20th day of exchange listing.

All three metrics suggest that the initial returns are insignificantly different from zero.

Our finding appears to be the first one showing that an IPO pricing mechanism leads to insignificant underpricing.

Evaluation Metric

Mean (%) t-valuea

Std (%)

Min (%)

Median (%)

Max (%)

Panel B: The zero institutional allocation group, N=10 R 0.99 0.12 25.18 -38.39 -5.36 38.79 MAR -4.31 -0.71 19.24 -24.62 -10.20 37.21 -4.65 -1.27 Panel C: The low institutional allocation group, N=50 R -2.16 -0.53 28.80 -57.56 -2.71 77.94 MAR -2.28 -0.68 23.47 -53.91 0.03 53.57 -2.20 -0.67 Panel D: The high institutional allocation group, N=29 R 10.71 1.70 33.97 -33.81 1.37 89.31 MAR 13.98 2.67** 28.25 -21.66 5.09 90.22 14.38 2.55**

Auction performance conditioned on Inst_Alloc

Results

For auctions with Inst_Alloc=0, retail investors lose an average of 4.3%.

For auctions with high Inst_Alloc, the average initial return is about 14%.

Table 6. Institutional Allocations in Winners and Losers

Mean (%) t-value a

Std (%)

Min (%)

Max (%)

Panel A: iMAR 20%, N=19

Inst_Alloc 26.72 19.67 0 61.95 MAR 40.95 19.16 21.71 90.22

Panel B: 0 iMAR <20%, N=30

Inst_Alloc 19.52 1.22 20.27 0 84.78 MAR 5.91 5.18 0.18 16.57

Panel C: 0> iMAR -20%, N=27

Inst_Alloc 16.37 2.02** 15.02 0 47.67 MAR -11.07 6.38 -19.93 -0.12

Panel D: iMAR <-20%, N=13

Inst_Alloc 11.60 2.56** 9.50 0 35.25 MAR -31.38 10.47 -53.90 -21.09

A winner’s curse

Retail investors win significantly higher proportions of shares in auctions that yield negative returns.

The higher the returns institutions could earn in auctions, the more shares they obtain.

Are auctions a fair allocation mechanism?

The answer is no.

Retail investors are more likely to suffer a winner’s curse in IPO auctions (Rock(1986)).

They are not better off under auctions than under bookbuilding.

Table 8. The Relation Between Inst_Alloc and Underpricing

0 1 2 3 4 ( 1, 30)_ i i i i i m iInst Alloc a a MAR a Size a BaseP a NHt R u

0 1 _ ( )i f i m f iR R Inst Alloc R R e

3sls Results

Inst_Alloc R_i - R_f

Constant -0.257 -0.142***

Inst_Alloc 0.904***

MAR 0.325***

Size 0.040**

BaseP -2.134**

NHt*R_m(-1,-30) 0.165**

R_m – R_f 1.127***

The relation is robust.

0 1 1 ( 1, 30) 2 ( 1, 30)

3 4

_ ( )

%

i f i m f i m i m

i i i

R R Inst Alloc R R Ht R NHt R

Share Elast e

0 1 2 3 4 ( 1, 30)_ i i i i i m iInst Alloc a a MAR a Size a BaseP a NHt R u

Underpricing and Inst_Allco

Underpricing is positively related to Inst_Alloc.

This structural relation is opposite to that of bookbuilding IPOs (Ljungqvist and Wilhelm (2002)).

Implication: As Inst_Alloc increases, bookbuilding could outperform auctions.

Table 9. Inst_Alloc, Bids, and Initial Returns

0 1 2 3 4 ( 1, 30)

6

_

( )

i inst i i i m

inst retail i

Inst Alloc a a MAR a Size a BaseP a NHt R

a P P u

0 1 _ ( )inst f i m f iR R Inst Alloc R R e

0 1 2 ( 1, 30)_inst retail i i m iP P c c Inst Alloc c NHt R

Table 9. ( Continued )

Inst_Alloc R_inst – R_f P_inst – P_retail

Constant -0.151 -0.140** -0.013***

Inst_Alloc 0.853*** 0.057***

P_inst – P_retail 4.936***

MAR 0.328***

Size 0.030**

BaseP -0.829

NHt*R_m(-1,-30) 0.488*** -0.059***

R_m – R_f 1.097***

The required underpricing by institutions

Underpricing_inst = -0.14 + 0.853 Inst_Alloc

Institutions tend to earn higher returns in auctions in which they win more shares.

When Inst_Alloc < 16.4%, underpricing <0, suggesting that not all institutions are informed.

Allocation-Weighted Returns

(R_inst –R_f) w_inst=0.105+1.032(R_m -R_f) w_inst

(2.03)

(R_retl –R_f) w_retl=0.011+1.051(R_m -R_f) w_retail

(0.43)

Institutions vs. Retail investors

The allocation-weighted abnormal return for retail investors is closed to zero.

The allocation-weighted abnormal return for institutions is 10.5%.

Consistent with Rock’s (1986) winner’s curse theory in which uninformed investors earn the R_f.

Could institutions be better off under bookbuilding?

Institutions are informed and have advantages in auctions.

The average Inst_Alloc of 19% is much lower than the norm of 70% in countries using bookbuilding.

Under bookbuilding, their expected profits could be increased with a tradeoff of a higher allocation and a lowered required underpricing.

For example, Inst_Alloc could be doubled to 38%, and the required underpricing be lowered from 10.5% to 6%.

Conclusion

There could be many reasons for why auctions have been losing market shares to bookbuilding.

This paper contributes three.1) Retail investors are likely to suffer a winner’s curse in auctions.2) For auctions with high Inst_Allco, underpricing could be large, which could be reduced under bookbuilding. 3) Institutions could be better off under bookbuilding.

The dark side of bookbuilding

If underwriters are given discretion in share allocations, the discretion will not automatically be used in the best interests of the issuing firm.

For bookbuilding to outperform auctions, underwriters must act in the best interest of the issuing firm

Figure 1. The Average Daily Stock Price of the 89 IPOs

0

10

20

30

40

50

60

70

80

90

100

1 10 19 28 37 46

Day of Exchange Listing

NT$

Average Winning Bid Price

Average Daily Stock Price

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1 4 7

10

13

16

19

22

25

28

31

34

37

40

43

46

49

52

55

58

61

64

67

70

73

76

79

82

85

88

Auctions Sorted by MAR within Each Allocation Group

Und

erpr

icin

g (M

AR

)

Figure 2. The Distributions of Underpricing in the Zero (auctions1-10), Low (auctions 11-60), and High (auctions 61-89) Institutional Allocation Groups.