Political uncertainty and stock market volatility: new evidence from the 2014 Scottish Independence Referendum
| Date | 01 May 2019 |
| DOI | http://doi.org/10.1111/sjpe.12186 |
| Published date | 01 May 2019 |
| Author | Julia Darby,Graeme Roy |
POLITICAL UNCERTAINTY AND
STOCK MARKET VOLATILITY: NEW
EVIDENCE FROM THE 2014 SCOTTISH
INDEPENDENCE REFERENDUM
Julia Darby* and Graeme Roy*
ABSTRACT
We investigate the impact of heightened political uncertainty in the run-up to,
and after, the 2014 Scottish independence referendum. The conditional volatili-
ties of stock returns of our Scottish index and the FTSE all share index are
characterised by the same GARCH parameters for a sample ending in late 2013,
but this no longer holds when estimation extends closer to the referendum. The
relative volatility of Scottish companies’ stock returns peaked when polls indi-
cated the referendum result was ‘too close to call,’ fell back on the result, but
rose again in the run-up to publication of proposals for further devolution.
II
NTRODUCTION
This paper examines information from financial markets in the run-up to, and
immediately after, the 2014 Scottish Independence Referendum. To the best
of our knowledge, it is the first study to focus on the impact of this referen-
dum on stock market volatility. The specific questions we address are: to what
extent were companies headquartered in Scotland exposed to heightened
uncertainty in the run-up to the Independence Referendum? Did the volatility
of stock returns of these companies increase relative to a UK wide bench-
mark? Was any referendum induced uncertainty resolved after the result was
known? And finally, how robust are our findings?
Heightened stock market volatility matters in part because it is likely to dis-
courage new share issues and initial public offerings and can impacts on len-
ders’ risk assessments and therefore borrowing costs. Recent research, for
example Baker et al. (2013), has suggested that political uncertainty can
increase unemployment and reduce investment, particularly if sustained.
We demonstrate, for the early part of our sample, that the conditional volatil-
ities of the Scottish stock returns and of returns in the FTSE all share index can
be characterised by the same GARCH parameters, but that this is no longer the
case once estimation extends beyond mid December 2013. Our estimates show
*University of Strathclyde
Scottish Journal of Political Economy, DOI: 10.1111/sjpe.12186, Vol. 66, No. 2, May 2019
©2018 Scottish Economic Society.
314
that the relative stock market volatility of Scottish companies’ stock returns
peaked in early September 2014, when the polls suggested the referendum result
was too close to call, fell back after the result was known, but built up again
prior to the publication of a report –the Smith Commission –into Scotland’s
future in the United Kingdom in November 2014. We find these findings are
robust to the inclusion/exclusion of the Royal Bank of Scotland and to the
exclusion of the whole of the Scottish financial and oil and gas sectors.
The remainder of the paper proceeds as follows: section II briefly discusses
the related literature and section III sets out a timeline of key events in the
2014 Scottish independence referendum. In section IV, we explain how we iden-
tify the publicly listed companies in Scotland and describe our data. Section V
outlines our econometric methodology. Section VI presents the key results.
Robustness checks are carried out in section VII and section VIII concludes.
II RELATED LITERATURE
There are few empirical studies of the impact of referendums on stock market
volatility. Nonetheless numerous studies have examined stock market volatil-
ity around the time of elections: these include Goodell and V€
ah€
amaa (2013)
and Li and Born (2006) who focused on USA presidential elections; Smales
(2014) for Australian elections; and Białkowski et al. (2008) who studied evi-
dence for 27 OECD countries. Each of these studies concluded that elections
are accompanied by a significant hike in stock market volatility, which is
more pronounced in close contests. They also point to cases in which the
excessive volatility could not be resolved quickly, e.g. when the political orien-
tation of the government changes or when an election was inconclusive. These
results are consistent with the theoretical predictions of P
astor and Veronesi
(2012, 2013). Vuchelen (2003) stressed that within a coalition-based political
system, election results may eliminate some possibilities but that more decisive
positions only become clear once a new coalition has been announced. The
same is likely to be true of a referendum outcome when the timing and nature
of the next steps are unclear.
Arin et al. (2013) examined the effects of a number of political variables on
stock returns and their volatilities in a panel of 17 parliamentary democracies
spanning the post-war period to 1995. They presented a strong empirical case
that the effects of political variables on stock return volatilities are consider-
ably stronger than the effects of these variables on stock returns.
Far fewer studies have explored how political uncertainty influences financial
markets outside the election cycle. An exception, with some similarities to our
own, is Beaulieu et al. (2006). They focus on the 1995 Quebec referendum
which could have led to the separation from the Canadian federation. As in
the Scottish case, opinion polls did not point to a clear winner, so uncertainty
associated with the referendum could not be resolved prior to the vote. Even
after the Quebec result, when 50.6% voted ‘no’, it was not clear that the refer-
endum resolved uncertainty with respect to Quebec’s future. And in the Scot-
tish case, it was made clear even before the vote that neither a ‘yes’ or ‘no’
POLITICAL UNCERTAINTY AND STOCK MARKET VOLATILITY 315
Scottish Journal of Political Economy
©2018 Scottish Economic Society
To continue reading
Request your trial