1

1
Booster shots An examination of the strength of
analyst coverage following IPOs

• Chris James
• Jason Karceski
• University of Central FloridaOctober 2004

2

• IPO firms purchase analyst coverage with
  underpricing.
• The role of analysts has intensified.
• What exactly do IPO firms purchase?
• More coverage / more likely to be covered
• Rajan and Servaes (1997 JF)
• Chen and Ritter (2000 JF)
• Aggarwal, Krigman and Womack (2002 JFE)

3

• What exactly do IPO firms purchase?
• More prestigious analyst (all star)
• Cliff and Denis (forthcoming JF) If all star
  analyst, initial return is 9 higher.
• Stronger (more favorable) coverage(Almost all
  initiating coverage is strong.)
• Michaely and Womack (1999 RFS) Leads issue
  positively biased recommendations.
• Bradley, Jordan and Ritter (2003 JF) 76 of
  firms get coverage at end of quiet period. 96
  of recommendations are favorable. 70 of
  covered firms have more than one analyst.

4
IPO firms care about analyst coverage.

• Analyst coverage is rewarded by the market.
• Buy recommendation by lead ? 2.7 AR
• Buy rec by unaffiliated underwriter ? 4.4 AR
  Michaely and Womack (1999)
• If coverage at end of quiet period ? 4.1 AR,
  otherwise 0.1 AR
• If covered by three or more analysts, AR is 4.7
  higher. Bradley, Jordan and Ritter (2003)

5
IPO firm is happy if it doesnt switch
underwriters from IPO to SEO.

• Timeliness and research quality are main factors
  in decision to switch. Krigman, Shaw and
  Womack (2001 JFE)
• Firms with more underpricing are less likely to
  switch. Aggarwal, Krigman and Womack
  (2002)
• Less likely to switch if lead analyst has a buy
  recommendation at the one-year IPO anniversary.
  Cliff and Denis (2004)

6
Analyst coverage Quantity vs. qualityMain
research questions

• Are IPO firms also purchasing stronger coverage
  with underpricing?
• If so (not), how long does the stronger coverage
  last?
• Does the market care about the strength of
  coverage, controlling for quantity?
• Does the relationship between target prices and
  aftermarket returns change during the bubble?

7
Measures of analyst coverage strength

• Target price / current stock price (TP/CP)
  target price ratio Higher ? stronger
• Buy / sell recommendations 1strong buy
  2buy 3accumulate 4hold 5sell
  Lower ? stronger

8
Target pricesserious valuations or marketing?

• Target prices are more finely tuned than
  recommendations.
• Average target price/current price of all stocks
  is 1.28 from 1997-2000 Brav and Lehavy (2003)
• Stock prices react to target price revisions
  Brav and Lehavy (2003)
• In our sample, stock prices react to target
  prices.
• ? the market believes there is some information
  in target prices.
• The only goal of research is a target
  price.Sallie Krawcheck, CEO Smith Barney

9

• The only goal of a research report is a
  target price. Just because Henry Blodget put a
  400 target on Amazon does not mean target prices
  are bad. They are the end result of research.
• Sallie Krawcheck, CEO Smith Barney, Fortune
  2003

10
Analyst categories / types

• Lead analystslead or co-lead banks
• Co-manager analystsany co-managing bank as
  identified by SDC
• Unaffiliated analystsany other firm (may or may
  not be in the IPO syndicate)

11
Deal categories / typesBroken deals vs.
successful deals

• Initial return (IR)the first-day return, from
  offer price to close of first day
• Return to coverage (RTC) the stock return from
  the offer price to the end of the day before the
  analyst report date
• If IR 0 and/or return to coverage 0, we call
  these broken / busted deals.
• Practitioners refer to broken deals as P

12
What we find.

• Affiliated analysts give stronger recs, but only
  give stronger TP/CP in busted deals.
• All analysts give stronger coverage (recs and
  TP/CP) for broken deals.
• But lead analysts have the largest discrepancy
  between deal types (TP/CP of 1.82 vs. 1.46). ?
  Firms do not purchase strength of coverage with
  underpricing.

13
What we find.

• Relation between TP/CP and IR/RTC is more
  negative for broken deals, especially for lead
  analysts.? Lead analysts give booster shots to
  broken deals.

14
What we find.

• These booster shots (high TP/CP in broken deals)
  are short lived.
• They are gone by the one-year IPO anniversary.
• TP/CP is dramatically reduced by the second
  analyst report, and gone by the third analyst
  report.
• The market cares about coverage strength (TP/CP
  and recs), even after controlling for quantity of
  coverage.
• If TP/CP increases by 0.5, analyst report AR is
  higher by 2.8.

15
Comparing our results to the literature

• Cliff and Denis (2004)
• More money left on the table buys all star
  coverage.
• The coverage comes first, then the first-day
  return.
• James and Karceski (2004)
• Less money left on the table gets a higher target
  price ratio.
• The target price comes second.
• A form of insurance.

16
Data

• IPOs from SDC/Ritter IPO database.
• Identify analyst type using SDC.
• Target prices and recommendations from First
  Call.
• Stock prices and returns from CRSP.
• 1,355 IPOs from November 1996 to August 2000.
• 1,189 IPO firms have at least one target price
  within the first year.

17
Table 1 Summary statistics

• Firms with coverage
• Are larger
• Older
• Have more syndicate members
• Have more highly ranked lead underwriters
• 
  All deals Broken deals
• High ranked underwriter 96
  94
• Lower ranked underwriter 68
  53
• Broken deals without a top-ranked underwriter are
  less likely to be covered.
• Have higher aftermarket returns (initial returns
  and quiet period returns)
• Bubble period IPOs are more likely to have
  coverage, but time period does not drive all of
  the difference.
• Number of syndicate members affects the timing
  but not the likelihood of coverage.

18
Hypotheses

• Information momentumAggarwal, Krigman and Womack
  (2002)
• Strategic underpricing and analyst coverage to
  maximize the share price at the end of the lockup
  period.
• Lead analysts give stronger support to firms with
  higher aftermarket performance.
• Booster shotsMichaely and Womack (1999)
• Lead analysts agree to give strong support if the
  firm underperforms in the aftermarket.
• Highest TP/CP for broken deals, asymmetric
  relation between TP/CP and aftermarket
  performance.

19
Hypotheses

• Mean reversion / anchored TP level
• Lead analysts pre-commit to a target price level
  and recommendation at the time of the offer.
• As aftermarket performance falls, TP/CP
  increases, but no asymmetry is expected.
• Anchored target price ratio
• Lead analysts pre-commit to a target price ratio
  and recommendation at the time of the offer.
• TP/CP is invariant to aftermarket performance.

20
Information momentum
Booster shot
Anchored TP/CP
21
Table 2 Strength of Coverage by Performance and
Analyst Affiliation

• Unaffiliated analysts give coverage later.
• Leads have more favorable recs than other
  analysts, but all have about the same TP/CPs.
• Leads also have more favorable TP/CPs, but only
  for broken deals.
• Each analyst type has higher TP/CP for broken
  deals (even unaffiliatedresearch guarantees?)
• Leads have highest TP/CP for broken deals
• ? 0.36 vs. 0.24 vs. 0.21
• Results are not driven by timing differences
  between affiliated and unaffiliated analysts.
• Supports booster shot and mean reversion
  hypotheses.

22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
Table 3 Coverage strength vs. return to coverage

• Lead analyst target price ratios are most
  sensitive (lower RTC ? higher TP/CP).
• Broken deals have much higher TP/CP sensitivity
  than successful deals (hockey stick).
• Outside 30 days from the offer date, lead
  analysts have the most TP/CP asymmetry.
• Asymmetry is more consistent with booster shot
  hypothesis.
• Prob(strong buy) is also sensitive to return to
  coverage, but no asymmetry.

27
Why we think Table 3s results are not
mechanically driven

• The concern is that the denominator of
  TP/CP is mechanically related to IR and RTC.
• We see the same asymmetry when we look at
  TP/Sales vs. RTC (-19.7 vs. 5.97).
• The asymmetry is different by analyst type.
• In Table 8, nonlead analysts change target prices
  more than the return from prior report (for
  broken deals). This gives a positive slope
  between ? TP/CP and prior return.

28
Why we think Table 3s results are not
mechanically driven

• After making log transformations, the asymmetry
  remains strong (-0.58 for broken deals, -0.07 for
  successful deals).

29
Table 4 Piecewise linear regressions of TP/CP

• If the aftermarket return is huge, the analyst
  will not set a target price ratio of less than
  one.
• So consider three regionsbroken deals,
  below-average successful deals, above-average
  successful deals.
• When RTC is really high, TP/CP is constant
  (around 1.39).
• We still see a change at RTC 0.

30
Table 5 Add control variables to Table 4

• TP/CP is higher when
• Underwriter rank is lower
• Firm is smaller
• IPO is during the bubble period
• Adding controls slightly amplifies the asymmetry.
• The asymmetry for lead analysts is larger than
  for nonlead analysts.

31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
Table 6 Coverage strength by analyst report date

• Lead analysts support for broken deals fades
  over time and rather quickly.
• Mostly gone by the second report (about 6 months
  after offer date)
• Completely gone by third report and at the one
  year offer date anniversary

38
Table 9 Market reaction to coverage strength

• When a new initiating analyst report comes out,
  the market reacts favorably (AR 2).
• Not much difference between lead and nonlead
  reports, except when only nonlead analysts
  initiate in the first 30 days (AR of 3.79 vs.
  1.58).
• Controlling for of analysts, strength matters.
  Higher TP/CP and better rec ? higher AR
• For TP/CP, 1.8 vs. 1.3 ? higher AR by 2.8
• For rec, 1 vs. 2 ? higher AR by 3.1

39
Analyst coverage Quantity vs. qualityMain
research questions

• Q Are IPO firms also purchasing stronger
  coverage with underpricing
• A No. Analysts give higher target price ratios
  to firms that perform poorly in the aftermarket.
  The sensitivity of TP/CP to aftermarket
  performance is asymmetric, especially for lead
  analysts. When a firm goes public, part of what
  it is buying is a booster shot if things go
  poorly in the aftermarket.

40
Analyst coverage Quantity vs. qualityMain
research questions

• Q How long does the stronger coverage last?
• A Most of it is gone by the second analyst
  report (about 150 days after the initiating
  report). All of it is gone by the third report /
  the one-year offer date anniversary.

41
Analyst coverage Quantity vs. qualityMain
research questions

• Q Does the market care about the strength of
  coverage, controlling for quantity?
• A Yes. A higher target price ratio by 0.5 is
  associated with a 2.8 higher abnormal return. A
  stronger recommendation by 1.0 is associated with
  a 3.1 higher abnormal return.

42
Analyst coverage Quantity vs. qualityMain
research questions

• Q Does the relationship between target prices
  and aftermarket returns change during the bubble?
• A No. There is no evidence that booster shots
  became stronger in the bubble period.