3.6.1. The target 52-week high as a reference point
In this subsection, robustness results are provided for the target 52-week high as a reference point by
first employing other price measures related to the target 52-weeks: the target 52-week minimum and
the target 52-week mean. Moreover, the target monthly returns prior to the takeover announcement
date are utilised the expectation that the bidder may also extrapolate the past performance of the target.
Results on regressions of the offer premium on the reference point are robustness after checking for
different price measures related to the target 52-weeks. The target 52-week minimum and the target
52-week mean were taken into consideration besides the target 52-week high in terms of the reference
point price, as reported in Table 3.8. Huddart et al. (2009) provided an overall review with regard to
the reference price using both 52-week highs and 52-week lows. In a similar way, the target 52-week
mean is also considered as the reference price. It is generally believed that the target 52-week prices
tend to attract a great deal of investors’ attention, making more conservative investors likely to focus
on the firm’s mean value over 52 weeks prior to the announcement date since it reflects the firm’s
performance in a certain period of time rather than a snapshot price. The results show that the target
52-week high remains strong after checking for those proposed reference point price measures.
In addition, robustness checks were carried out by including the target past returns prior to the
Chapter 3. Reference Point Theory on Cross-Border and Domestic M&A Deals: U.K. Evidence
133
3.9). According to Rau and Vermaelen’s extrapolation hypothesis (1998), market investors tend to
forecast a firm’s prospects based on its past returns, creating a strong performance momentum around
a bid announcement for those with best past performance. Therefore, targets’ past returns of each
month were taken into consideration, reflecting the firm’s returns over the past months, which are
more likely be a proxy for the market perception of the firm’s valuation than a single item of price
information (i.e. the target 52-week high). The extrapolation effect on the offer premium was studied
by including the target past returns up to 12 months prior to the takeover announcement date. In
particular, specification (2) controls for the target monthly returns for the recent three months prior
to the takeover announcement date, specification (3) controls for the target monthly returns for the
recent six months prior to the takeover announcement date, and specification (4) controls for the
target monthly returns up to a year prior to the takeover announcement date. It was found that the
offer premium is positively related to the target 52-week high. Once again, our results show that the
target 52-week high is a proper measure for the reference point in an M&A deal.
3.6.2. The reference point effect on the offer premium by market conditions
In addition, the reference point effect on the offer premium by the market wide valuation was
examined. Investors’ have a rationale that bidders may under- or over-react to a target’s valuation
when the market valuation fluctuates. The sample periods are classified into High-, Low- and Neutral-
valuation markets according to Bouwman et al.'s method (2009). Results of Table 3.10 show that the
reference point effect is strong when the market-wide valuation is either high or low. It can be
interpreted that a high market valuation is likely to push the target’s current price closer to the 52-
week high, making bidders to believe that they can outperform the target’s recent performance. While
Chapter 3. Reference Point Theory on Cross-Border and Domestic M&A Deals: U.K. Evidence
134
for the target’s performance given that no other factors drive the firm’s value beyond or below its
fundamental value. Moreover, it was found that cross-border acquisitions tend to pay higher offer
premiums when the market valuation is high, implying that the high market wide valuation increases
information asymmetry and thus reinforces the target’s bargaining position over cross-border bidders.
3.6.3. The reference point effect on the offer premium across different subsamples
The reference point effect on the offer premium was studied by different nations with the expectation
that bidders whose nation was farthest away compared with that of the U.K. firms would tend to have
greater information barriers and be more likely to rely on the target 52-week high. In addition, the
reference point effect on the offer premiums in different M&A subsamples was analysed: the payment
method for the M&A deals, whether or not the deal is diversified and whether the deal involves a
negotiation between the boards of the two firms (i.e. tender offers and mergers).
A target’s bargaining position is likely to be reinforced when the distance involved between the two
firms’ home countries is large. The M&A sample was divided into two subsamples, the one includes
acquisitions comprising U.K. bidders, the other comprising acquisitions involving U.S. bidders,
farthest away compared with other cross-border bidders (see Table 3.11). It can be seen that the offer
premium increases by approximately 6% in the U.S. bidders compared with 4% in the U.K. domestic
bidders. Greater distance between the two nations impedes the transaction, making the bidder more
Chapter 3. Reference Point Theory on Cross-Border and Domestic M&A Deals: U.K. Evidence
135
of Erel et al. (2012) and Uysal et al. (2008), finding that geographical distance plays a role in
information dissemination.45
The reference point effect on offer premiums across different subsamples by deal information was
also tested. The reference point effect on the offer premium being tested by the method of payment
(see Table 3.12), by whether or not the deal is diversified reported in (see Table 3.13), and the type
of merger reported in (see Table 3.14). Taken together, the results show that the target 52-week high
as a proxy for the reference point effect is very pronounced in different subsamples, and has a
substantial impact on the offer premium in different M&A subsamples.
3.6.4. The reference point effect on bidder announcement returns
Table 3.15 tabulates results on regressions of the offer premium on the bidder announcement returns
and with the target 52-week high as an instrumental variable of the offer premium. Instead of using
the market-adjust model to calculate the firm’s abnormal returns around the announcement date, the
market model is also used. The market model parameters were estimated based on an event window
from 261 to 28 trading days prior to the announcement date, i.e. [-261,-28]. The results obtained from
the market model are consistent with those by the market-adjusted model. More specifically, a 10%
increase in the reference point driven offer premium decreased the bidder announcement returns by
1.83% in the full M&A sample. While dividing it into two subsamples, the reference point effect leads
to more negative bidder announcement returns for domestic bidders than for cross-border bidders.
45 It can be argued that the information asymmetry are relatively low due to culture similarity between the two countries.
However, the sample included in this chapter is small, which is likely to result in bias in the results. Further investigation by including more observations needs to be done to explain whether the culture or the distance between the countries matter the bargaining power. We leave it to the future research, since the focus of this chapter is to test the target reference point effect on the firm’s valuation.
Chapter 3. Reference Point Theory on Cross-Border and Domestic M&A Deals: U.K. Evidence
136
In addition, bidder abnormal returns in a 3-day window around the takeover announcement date were
examined using both market-adjusted model and market model. The reference point effect on bidder
3-day announcement returns calculated with the market-adjusted model is reported in Table 3.16, and
results with the market model are reported in Table 3.17, which are consistent with the main results
reported in Table 3.7.