Brian Harding
Formerly of James Capel Stockbrokers
"In recent years UK companies in the quoted engineering, automotive and aerospace sectors have spent more on acquisitions than on R&D and capital investment combined."
In view of the benefits associated with increased R&D investment
intensity highlighted in this and previous Scoreboards, it is important to gain
some insight into the success or failure of acquisition-led growth in general, and
to compare this with the total shareholder returns (TSR) generated by high investment
intensity companies. In carrying out this comparison, it is helpful to make some
international comparison with respect to the propensity to grow by acquisition.
This has been done via a direct comparison with a large group of quoted, US companies
in the same sectors. This article addresses these two related issues.
The analysis that follows is based on a sample of 46 quoted UK companies in the
engineering, automotive and aerospace sectors and, for the international comparison,
on 184 quoted US companies in the same sectors. Both samples account for a very
high proportion of total sector market capitalisation. The research on the relationship
between TSR and levels of investment, and the work on the TSR effects of large acquisitions
relates primarily to the ten year period 1988 to 1997 when all 46 UK companies were
operating independently throughout the entire period. There is some extension to
cover later years, although several of the original companies disappeared from the
sample as a result of M&A activity. The comparison between the UK and the US
has been based on the period 1997 to 2000 because of earlier discontinuities in
the US data. The underlying source documents for the US were the IRI Industrial
R&D Scoreboards and US Mergers and Acquisitions published by Thomson Financial.
The UK data was derived from the R&D Scoreboards, company accounts, Datastream
and Acquisitions Monthly.
The relationship between total shareholder returns and investment in research and
development and capital spending
As the 2001 Capex Scoreboard *1 showed, there is a high degree of variability
between companies with respect to TSR, particularly over longer periods of time.
Over the period 1988-97 total shareholder returns vary from +839% (Bodycote), to
–71% (ASW). Full details for all companies in the broad engineering sector are found
in the Capex Scoreboard*1.
It is, however, short run share price volatility which poses the greatest difficulty
in procuring satisfactory measurements of the success or failure of particular investment
strategies. This difficulty can be circumvented to some extent by using ranking
techniques. Rather than working in terms of share prices per se, companies are ranked
by total shareholder return and by investment intensity, (research and development
plus capital spending as a percentage of sales). The correlations between these
two rankings are then calculated. This approach has the advantage that it effectively
removes systematic share price variation arising from changes in investor fashion
(eg, a switch from engineering to ‘dot-com’ stocks), broad changes in investor sentiment
(eg, a switch from equities to cash), and sectoral rotation (switching into or out
of defensive or cyclical stocks). Ranking therefore focuses on outperformance or
underperformance within sector in terms of TSR.
On this basis it is possible to discern highly significant positive correlations
between within-sector rankings for TSR and investment intensity. The correlation
coefficient for the ten-year period 1988-1997 for example, is Rs = 0.55. The chances
of observing a coefficient of this magnitude if no real relationship existed are
about 2000 to 1 against. Similarly high Rs values can be observed for almost every
period of four years or more, (and most periods of three years or more). To put
these findings another way, most high investment intensity companies will be found
in the top half of the TSR ranking table most of the time if the measurements are
taken over a period of three years or more.
* Brian Harding has a fellowship at Warwick Business School and was previously with
James Capel for a period of some 20 years. Before this, he held senior management
positions in UK engineering companies.
*1 Brian Harding in 2001 Capex Scoreboard, DTI & Company Reporting Ltd
The relationship between total shareholder returns and major acquisitions
The argument for acquisitions advanced most frequently by acquiring companies is
that of synergy. It is argued that the enlarged group will achieve improved manufacturing
efficiencies via plant closure or consolidation or economies of scale. Alternatively,
or additionally, better marketing efficiency is sought through the completion of
product ranges, the introduction of existing products into new markets where the
acquired company has an established presence, or the introduction of the acquired
company’s products into markets controlled by the acquirer. This is undoubtedly
a popular argument amongst companies. We have identified a total of 540 acquisitions
made by the 46 companies in our data set over the period 1988-97, rising to 667
over the period 1998-2000. It should be added that integrating an acquired company
involves much management time and attention leaving less for issues of organic growth
through investment.
The aggregate expenditure on acquisitions, by year, for the 46 UK companies involved
in the study is set out in table 1 below.
It is fairly obvious that this data is intensely cyclical in nature with very little
acquisition activity occurring in periods of economic decline when prices would
tend to be lower. Maximum expenditure tends to occur when the economy is buoyant
and the price of making an acquisition is relatively high. This phenomenon relates
to the frequent desire of predator companies to use their own paper in the acquisition
process and/or, to the desire of sellers to deal only when prices are relatively
buoyant.
Attention was focused on the changes in TSR ranking which followed major acquisition
events to see whether or not companies which completed major acquisitions outperformed
those which did not. A ‘major’ acquisition is arbitrarily defined as one where the
consideration is equal to at least 15% of the annual sales of the acquirer at the
time of acquisition. On the basis of this definition a total of 18 companies completed
25 ‘major’ acquisitions between 1988 and 1997. For each of these acquisitions total
shareholder return was tracked on a monthly basis for four years after each takeover.
This revealed that a major acquisition was followed by under-performance against
the sector average on 16 occasions out of 25 (64%). Additionally, and just after
the end of our arbitrary four year period, a further three companies (Wagon, Invensys
and Baynes) suffered a collapse in share price and total shareholder return; on
this basis 76% of all major acquisitions were followed by underperformance. However
the period is defined, share price underperformance after a major acquisition is
significantly more likely than outperformance. It is important to note that some
major acquisitions appeared to be successful with 7 of the 25 associated with relative
TSR outperformance of over 20% after 4 years. However, three of these 7 were made
by Wagon, Invensys and Baynes whose share prices collapsed just after the period
studied. The point is that significant outperformance is just a much less likely
outcome with two to four times as many acquisitions associated with underperformance
against average shareholder returns for the sector.
Research in the US by Sirower*2, which was based on an examination of
all major acquisitions (excluding regulated sectors such as banks, utilities) completed
between 1979 and 1990, found that about 66% of all deals were followed by reduced
shareholder value. The similarity between our figures for the UK engineering, aerospace
and automotive sectors and the more broadly based US data is striking. It is suggested
that the reasons for this relatively high failure rate include overpayment for the
target company, (particularly likely in the case of contested takeovers), overestimation
of the synergy benefits, and underestimation of the time taken to realise these
benefits. Such under-performance could, of course, result from a variety of factors
and it might be unwise to attribute it solely to ill-judged acquisition strategies.
Nevertheless, the contrast between high investment intensity companies with their
sustained and frequent appearance at the top of the TSR rankings, (reference 1),
and the frequent under-performance which follows major acquisitions is very marked.
*2 Sirower: The Synergy Trap, 1997
A comparison between the UK and the US with respect to acquisition-led growth in
the engineering, automotive and aerospace sectors This comparison relates to the
four-year period 1997 to 2000 and is based on the behaviour of two groups of quoted
companies, 184 in America and 46 in the UK, all of which operated in the engineering,
automotive and aerospace sectors. Aggregate expenditure on acquisitions is displayed
in tables 2 and 3 below:
Table 4 contains a comparison of acquisition intensities, (aggregate expenditure
on acquisitions as a percentage of aggregate sales), for the UK and the US.
A number of interpretive comments have to be made about the data in table 4:
- Acquisition spend in the US is not lower because the sector was more fully consolidated
before 1997. In 1999, about 89% of the 184 US companies included in the sample contributed
less than 1% to total sector sales. The corresponding figure for the UK is 56%.
If anything, the UK is in a more advanced state of consolidation than the US so
that the US has greater potential for acquisitions and might have been expected
to show a higher acquisition intensity. The majority of UK companies in this broad
sector are international in scope so UK domestic market considerations should not
predominate.
- The UK engineering, automotive and aerospace industries spent about five times
as much on acquisitions per £ of sales as the US over the period 1997 to 2000. If
the particularly large acquisitions noted beneath tables 2 and 3 are arbitrarily
removed (the turnover figures remaining unchanged), the UK still spends over four
times as much per £ of sales as the US.
- There is an argument for excluding Ford and GM from the American data on the grounds
that these companies are so large, (Ford $163bn of sales, GM $173bn of sales in
1999) that they can no longer make meaningful acquisitions. Also, there are no broadly
comparable companies in the UK. Exclusion of these two companies does, of course,
raise US acquisition intensity – but only to an average of 3.28% over the period
1997 to 2000. This still leaves UK acquisition intensity about 3.5 times higher
than that seen in the US.
- Even the entirely arbitrary removal of Ford and GM on the grounds of size and
lack of comparability (removing both acquisition expenditure and turnover), plus
the removal of all other major acquisitions noted under tables 2 and 3, leaves the
residual acquisition intensity in the UK close to three times that seen in the US.
While table 4 establishes the difference in acquisition intensities between the
US and the UK, it does not relate the expenditure on acquisitions to organic investment
in R&D and Capex. This is done in table 5 below.
Table 5 demonstrates that UK companies invested only 60% as much in R&D plus
Capex as they spent on acquisitions. US companies, on the other hand invested between
four and five times as much on R&D plus Capex as they spent on acquisitions.
If GM and Ford are arbitrarily removed from the US data, the remaining US companies
still spent 2.4 times as much as on R&D and capital investment as they did on
acquisitions. This ratio thus remains well in excess of the UK figure of 0.6.
Conclusion
We have argued that, in terms of within sector total shareholder return, growth
by acquisition is frequently inferior to that achieved by ‘organic’ investment in
R&D and capital spending. Despite this, vast sums continue to be spent on acquisition-led
growth and there is some evidence which suggests that, in the late 1990s, it was
becoming increasingly important relative to R&D and capital spending.
The argument is not against acquisitions per se. A process of acquisition and consolidation
is probably the only way of dealing with any industry which is facing a long term
contraction in demand or substantial overcapacity. The hostile takeover, or mere
threat of such a takeover, has an important place in the general disciplinary function
of the Financial Market. Nor are we criticising individual acquisitions. We are
merely pointing out that, as a group, those companies which spend heavily on research
and development and new plant outperform, on average, those which undertake large
acquisitions. The disparity between the frequent out-performance of high investment
intensity companies, and the under-performance which too often follows major acquisitions,
is too great to be dismissed as mere chance.
Commentary
The position in the US appears to be fundamentally different from that in the UK
in that major acquisitions are, somewhat surprisingly, less frequent.
In the US, expenditure on acquisitions over the period 1997-2000 was less than one
third of that seen in the UK on a per £ of sales basis. This lower incidence of
acquisition activity would not appear to be related to lack of opportunity since
the US sectors remain less consolidated than their UK counterparts. Additionally,
there is a fundamentally different relationship between research and development
plus capital spending and acquisition spending in the US vis-a-vis the UK. In 1999
and 2000 American companies in the above sectors spent only 22% as much on acquisitions
as they invested in R&D plus Capex whereas UK companies spent 160% of organic
investment on acquisitions. Even the arbitrary removal of BAe with its huge acquisition
of the Marconi defence interests from the samples leaves this ratio at 110% and,
hence, with acquisition spend still larger than organic investment; if, in addition,
GM and Ford are removed from the US data, the US ratio of R&D plus Capex to
acquisitions is still three times the UK figure.
Although some acquisitions are undoubtedly successful, the question remains as to
why UK CEOs should continue heavily to favour strategies of investing in major acquisitions.
The answers could include the desire for quick results, a desire to compensate for
previous underinvestment in R&D and Capex and, perhaps, the prospect of enhanced
rewards and security thought to be associated with a larger and expanding company
*3,4. There could also be a tacit coincidence of interest between investment
bankers (with their interest in M&A fees), plc directors (with concern for their
personal position and income) and fund managers (who are likely to choose an immediate
and substantial premium on the shares they hold in the target company rather than
voting against a takeover on grounds of insufficient industrial logic).
The key factual conclusions are therefore:
- A study following a large group of UK companies over 10 years shows that high
levels of R&D and capital investment (essential components of organic growth)
are positively correlated with increased relative shareholder returns over periods
of 3 to 4 or more years.
- Major acquisitions (in the broad engineering industry and also across industries)
are associated with a reduction in relative shareholder returns over the succeeding
four years in some twothirds of cases.
- US companies in the broad engineering sector invest more heavily in R&D and
Capex than do their UK counterparts.
- The result is that the broad engineering sector in the US spent on acquisitions
just over 20% of the sum invested in R&D and capital equipment, while the equivalent
sector in the UK spent on acquisitions a sum which was over 150% of their R&D
and capital investment. These percentages remain very substantially different even
if the largest UK acquisition and the largest US companies are arbitrarily excluded.
This evidence on long-term shareholder returns shows that, in formulating their
strategy for future growth and success, companies need to consider carefully the
merits of organic growth based on high R&D and capital investment against the
generally less successful outcomes associated with major acquisitions.
*3 Charkham ‘Keeping Good Company’ OUP 1992(p311-5)
*4 Marsh ‘Short Termism on Trial’ Booklet for Institutional Fund Managers Association.
1990
Acknowledgement
Brian Harding wishes to acknowledge the valuable assistance given by Mark Whittington
of the Warwick University Business school.
Appendix
The basic methodology used was as follows. We calculate, for each acquisition event,
the cumulative post acquisition total shareholder return and subtract from this
the average cumulative total shareholder return achieved by the entire group of
companies over the same period. A negative result indicates under-performance against
the group average, a positive result indicates out-performance. This data was computed
on a monthly basis for four years after the completion of each acquisition. This
approach possesses the advantage that it effectively removes systematic share price
variation arising from changes in investor fashion, investor sentiment and sectoral
rotation. It therefore focuses on outperformance or underperformance within sector
in terms of total shareholder return. The results of this analysis are summarised
in the table.
