@manuscripts/transform 4.3.53 → 4.4.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/cjs/jats/__tests__/jats-roundtrip.test.js +3 -3
- package/dist/cjs/jats/exporter/jats-exporter.js +3 -0
- package/dist/cjs/jats/importer/jats-dom-parser.js +41 -1
- package/dist/cjs/schema/index.js +9 -0
- package/dist/cjs/schema/nodes/headshot_element.js +38 -0
- package/dist/cjs/schema/nodes/headshot_grid.js +37 -0
- package/dist/cjs/schema/nodes/headshot_image.js +51 -0
- package/dist/cjs/transformer/node-names.js +2 -0
- package/dist/cjs/transformer/node-title.js +6 -0
- package/dist/cjs/version.js +1 -1
- package/dist/es/jats/__tests__/jats-roundtrip.test.js +3 -3
- package/dist/es/jats/exporter/jats-exporter.js +3 -0
- package/dist/es/jats/importer/jats-dom-parser.js +41 -1
- package/dist/es/schema/index.js +9 -0
- package/dist/es/schema/nodes/headshot_element.js +34 -0
- package/dist/es/schema/nodes/headshot_grid.js +33 -0
- package/dist/es/schema/nodes/headshot_image.js +48 -0
- package/dist/es/transformer/node-names.js +2 -0
- package/dist/es/transformer/node-title.js +6 -0
- package/dist/es/version.js +1 -1
- package/dist/types/jats/__tests__/utils.d.ts +1 -1
- package/dist/types/schema/index.d.ts +3 -0
- package/dist/types/schema/nodes/headshot_element.d.ts +25 -0
- package/dist/types/schema/nodes/headshot_grid.d.ts +24 -0
- package/dist/types/schema/nodes/headshot_image.d.ts +26 -0
- package/dist/types/schema/types.d.ts +1 -1
- package/dist/types/version.d.ts +1 -1
- package/package.json +22 -21
- package/src/errors.ts +28 -0
- package/src/getVersion.ts +21 -0
- package/src/index.ts +28 -0
- package/src/jats/__tests__/__fixtures__/debug-example.xml +136 -0
- package/src/jats/__tests__/__fixtures__/jats-abstract-no-body.xml +163 -0
- package/src/jats/__tests__/__fixtures__/jats-citations.xml +472 -0
- package/src/jats/__tests__/__fixtures__/jats-comments.xml +46 -0
- package/src/jats/__tests__/__fixtures__/jats-document.xml +134 -0
- package/src/jats/__tests__/__fixtures__/jats-example-doc.xml +1235 -0
- package/src/jats/__tests__/__fixtures__/jats-example-front-only.xml +26 -0
- package/src/jats/__tests__/__fixtures__/jats-example-full.xml +209 -0
- package/src/jats/__tests__/__fixtures__/jats-example-no-back.xml +156 -0
- package/src/jats/__tests__/__fixtures__/jats-example-no-body.xml +91 -0
- package/src/jats/__tests__/__fixtures__/jats-example.xml +221 -0
- package/src/jats/__tests__/__fixtures__/jats-fn-group.xml +60 -0
- package/src/jats/__tests__/__fixtures__/jats-import-no-refs.xml +1077 -0
- package/src/jats/__tests__/__fixtures__/jats-import.xml +1578 -0
- package/src/jats/__tests__/__fixtures__/jats-roundtrip.xml +133 -0
- package/src/jats/__tests__/__fixtures__/jats-tables-example.xml +2043 -0
- package/src/jats/__tests__/__fixtures__/math-fragment.xml +12 -0
- package/src/jats/__tests__/__snapshots__/citeproc.test.ts.snap +57 -0
- package/src/jats/__tests__/__snapshots__/create-article-node.test.ts.snap +65 -0
- package/src/jats/__tests__/__snapshots__/jats-exporter.test.ts.snap +7 -0
- package/src/jats/__tests__/__snapshots__/jats-importer.test.ts.snap +59796 -0
- package/src/jats/__tests__/__snapshots__/jats-roundtrip.test.ts.snap +18 -0
- package/src/jats/__tests__/citations.ts +25 -0
- package/src/jats/__tests__/citeproc.test.ts +195 -0
- package/src/jats/__tests__/create-article-node.test.ts +29 -0
- package/src/jats/__tests__/data/section-categories.ts +212 -0
- package/src/jats/__tests__/files.ts +24 -0
- package/src/jats/__tests__/jats-exporter.test.ts +223 -0
- package/src/jats/__tests__/jats-importer.test.ts +567 -0
- package/src/jats/__tests__/jats-roundtrip.test.ts +66 -0
- package/src/jats/__tests__/utils.ts +90 -0
- package/src/jats/exporter/citeproc.ts +265 -0
- package/src/jats/exporter/jats-exporter.ts +2093 -0
- package/src/jats/exporter/jats-versions.ts +50 -0
- package/src/jats/exporter/labels.ts +101 -0
- package/src/jats/importer/create-article-node.ts +52 -0
- package/src/jats/importer/jats-comments.ts +111 -0
- package/src/jats/importer/jats-dom-parser.ts +1306 -0
- package/src/jats/importer/jats-parser-utils.ts +234 -0
- package/src/jats/importer/jats-transformations.ts +591 -0
- package/src/jats/importer/parse-jats-article.ts +106 -0
- package/src/jats/index.ts +21 -0
- package/src/jats/types.ts +16 -0
- package/src/lib/__tests__/footnotes.test.ts +36 -0
- package/src/lib/citeproc.ts +30 -0
- package/src/lib/credit-roles.ts +91 -0
- package/src/lib/deafults.ts +17 -0
- package/src/lib/footnotes.ts +85 -0
- package/src/lib/html.ts +56 -0
- package/src/lib/section-categories.ts +27 -0
- package/src/lib/utils.ts +145 -0
- package/src/lib/xml.ts +28 -0
- package/src/schema/__tests__/docs.ts +1911 -0
- package/src/schema/__tests__/groups.test.ts +61 -0
- package/src/schema/__tests__/migration.test.ts +23 -0
- package/src/schema/groups.ts +36 -0
- package/src/schema/index.ts +295 -0
- package/src/schema/marks.ts +254 -0
- package/src/schema/migration/migrate.ts +77 -0
- package/src/schema/migration/migration-script.ts +26 -0
- package/src/schema/migration/migration-scripts/1.2.5.ts +44 -0
- package/src/schema/migration/migration-scripts/2.3.22.ts +52 -0
- package/src/schema/migration/migration-scripts/3.0.12.ts +33 -0
- package/src/schema/migration/migration-scripts/3.0.21.ts +50 -0
- package/src/schema/migration/migration-scripts/3.0.30.ts +65 -0
- package/src/schema/migration/migration-scripts/3.0.31.ts +40 -0
- package/src/schema/migration/migration-scripts/3.0.41.ts +37 -0
- package/src/schema/migration/migration-scripts/3.0.55.ts +42 -0
- package/src/schema/migration/migration-scripts/3.0.56.ts +114 -0
- package/src/schema/migration/migration-scripts/4.2.13.ts +37 -0
- package/src/schema/migration/migration-scripts/4.2.15.ts +48 -0
- package/src/schema/migration/migration-scripts/4.3.23.ts +55 -0
- package/src/schema/migration/migration-scripts/4.3.34.ts +132 -0
- package/src/schema/migration/migration-scripts/4.3.35.ts +41 -0
- package/src/schema/migration/migration-scripts/index.ts +49 -0
- package/src/schema/nodes/abstracts.ts +29 -0
- package/src/schema/nodes/affiliation.ts +88 -0
- package/src/schema/nodes/affiliations.ts +30 -0
- package/src/schema/nodes/alt_text.ts +36 -0
- package/src/schema/nodes/alt_title.ts +40 -0
- package/src/schema/nodes/alt_titles_section.ts +58 -0
- package/src/schema/nodes/attachment.ts +46 -0
- package/src/schema/nodes/attachments.ts +45 -0
- package/src/schema/nodes/attribution.ts +38 -0
- package/src/schema/nodes/author_notes.ts +44 -0
- package/src/schema/nodes/award.ts +52 -0
- package/src/schema/nodes/awards.ts +45 -0
- package/src/schema/nodes/backmatter.ts +31 -0
- package/src/schema/nodes/bibliography_element.ts +58 -0
- package/src/schema/nodes/bibliography_item.ts +108 -0
- package/src/schema/nodes/bibliography_section.ts +60 -0
- package/src/schema/nodes/blockquote_element.ts +79 -0
- package/src/schema/nodes/body.ts +29 -0
- package/src/schema/nodes/box_element.ts +67 -0
- package/src/schema/nodes/caption.ts +67 -0
- package/src/schema/nodes/caption_title.ts +67 -0
- package/src/schema/nodes/citation.ts +70 -0
- package/src/schema/nodes/comment.ts +50 -0
- package/src/schema/nodes/comments.ts +28 -0
- package/src/schema/nodes/contributor.ts +87 -0
- package/src/schema/nodes/contributors.ts +40 -0
- package/src/schema/nodes/core_section.ts +27 -0
- package/src/schema/nodes/corresp.ts +49 -0
- package/src/schema/nodes/cross_reference.ts +64 -0
- package/src/schema/nodes/doc.ts +21 -0
- package/src/schema/nodes/embed.ts +55 -0
- package/src/schema/nodes/equation.ts +65 -0
- package/src/schema/nodes/equation_element.ts +61 -0
- package/src/schema/nodes/figure.ts +65 -0
- package/src/schema/nodes/figure_element.ts +73 -0
- package/src/schema/nodes/footnote.ts +95 -0
- package/src/schema/nodes/footnotes_element.ts +69 -0
- package/src/schema/nodes/footnotes_section.ts +61 -0
- package/src/schema/nodes/general_table_footnote.ts +42 -0
- package/src/schema/nodes/graphical_abstract_section.ts +62 -0
- package/src/schema/nodes/hard_break.ts +36 -0
- package/src/schema/nodes/headshot_element.ts +48 -0
- package/src/schema/nodes/headshot_grid.ts +46 -0
- package/src/schema/nodes/headshot_image.ts +65 -0
- package/src/schema/nodes/hero_image.ts +46 -0
- package/src/schema/nodes/highlight_marker.ts +68 -0
- package/src/schema/nodes/image_element.ts +49 -0
- package/src/schema/nodes/inline_equation.ts +68 -0
- package/src/schema/nodes/inline_footnote.ts +66 -0
- package/src/schema/nodes/keyword.ts +61 -0
- package/src/schema/nodes/keyword_group.ts +63 -0
- package/src/schema/nodes/keywords.ts +59 -0
- package/src/schema/nodes/keywords_element.ts +64 -0
- package/src/schema/nodes/link.ts +79 -0
- package/src/schema/nodes/list.ts +134 -0
- package/src/schema/nodes/listing.ts +91 -0
- package/src/schema/nodes/listing_element.ts +61 -0
- package/src/schema/nodes/long_desc.ts +36 -0
- package/src/schema/nodes/manuscript.ts +86 -0
- package/src/schema/nodes/missing_figure.ts +59 -0
- package/src/schema/nodes/paragraph.ts +77 -0
- package/src/schema/nodes/placeholder.ts +62 -0
- package/src/schema/nodes/placeholder_element.ts +60 -0
- package/src/schema/nodes/pullquote_element.ts +81 -0
- package/src/schema/nodes/quote_image.ts +64 -0
- package/src/schema/nodes/section.ts +69 -0
- package/src/schema/nodes/section_label.ts +39 -0
- package/src/schema/nodes/section_title.ts +58 -0
- package/src/schema/nodes/subtitle.ts +40 -0
- package/src/schema/nodes/subtitles.ts +51 -0
- package/src/schema/nodes/supplement.ts +77 -0
- package/src/schema/nodes/supplements.ts +60 -0
- package/src/schema/nodes/table.ts +144 -0
- package/src/schema/nodes/table_col.ts +76 -0
- package/src/schema/nodes/table_element.ts +67 -0
- package/src/schema/nodes/table_element_footer.ts +42 -0
- package/src/schema/nodes/text.ts +31 -0
- package/src/schema/nodes/title.ts +43 -0
- package/src/schema/nodes/trans_abstract.ts +58 -0
- package/src/schema/nodes/trans_graphical_abstract.ts +67 -0
- package/src/schema/types.ts +243 -0
- package/src/tests.ts +33 -0
- package/src/transformer/id.ts +23 -0
- package/src/transformer/index.ts +21 -0
- package/src/transformer/node-names.ts +75 -0
- package/src/transformer/node-title.ts +116 -0
- package/src/transformer/node-types.ts +39 -0
- package/src/transformer/node-validator.ts +82 -0
- package/src/types.ts +25 -0
- package/src/version.ts +1 -0
|
@@ -0,0 +1,1235 @@
|
|
|
1
|
+
<?xml version="1.0" encoding="utf-8"?>
|
|
2
|
+
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.1 20151215//EN" "https://jats.nlm.nih.gov/archiving/1.1d1/JATS-archivearticle1.dtd">
|
|
3
|
+
<article article-type="research-article" dtd-version="1.1" xml:lang="en" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML">
|
|
4
|
+
<front>
|
|
5
|
+
<journal-meta>
|
|
6
|
+
<journal-id journal-id-type="publisher-id">NLM-Export</journal-id>
|
|
7
|
+
<journal-id journal-id-type="nlm-ta">NLM-Export</journal-id>
|
|
8
|
+
<journal-title-group>
|
|
9
|
+
<journal-title>NLM-Export</journal-title>
|
|
10
|
+
<abbrev-journal-title abbrev-type="pubmed">NLM-Export</abbrev-journal-title>
|
|
11
|
+
</journal-title-group>
|
|
12
|
+
<issn>1878-108X</issn>
|
|
13
|
+
<publisher><publisher-name>Inera</publisher-name></publisher>
|
|
14
|
+
</journal-meta>
|
|
15
|
+
<article-meta>
|
|
16
|
+
<article-id pub-id-type="publisher-id">9999999</article-id>
|
|
17
|
+
<title-group>
|
|
18
|
+
<article-title>WHEN ELECTIONS FAIL TO RESOLVE UNCERTAINTY: THE CASE OF THE 2016 U.S. <bold>PRESIDENTIAL ELECTION</bold></article-title>
|
|
19
|
+
</title-group>
|
|
20
|
+
<contrib-group>
|
|
21
|
+
<contrib contrib-type="author"><name><surname>Cox</surname><given-names>Justin</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib>
|
|
22
|
+
<contrib contrib-type="author"><name><surname>Griffith</surname><given-names>Todd</given-names></name><xref ref-type="aff" rid="aff2"></xref></contrib>
|
|
23
|
+
<aff id="aff1">University of Mississippi</aff>
|
|
24
|
+
<aff id="aff2">Utah State University</aff>
|
|
25
|
+
</contrib-group>
|
|
26
|
+
<pub-date pub-type="ppub"><year></year></pub-date>
|
|
27
|
+
<volume></volume>
|
|
28
|
+
<fpage>1</fpage>
|
|
29
|
+
<permissions>
|
|
30
|
+
</permissions>
|
|
31
|
+
<abstract>
|
|
32
|
+
<p>In this article, we examine whether certain political election outcomes create, rather than resolve, uncertainty in financial markets. We posit that the uncertainty associated with unanticipated election outcomes is not resolved before or on the election dates. To test this claim, we use the surprise outcome of the 2016 U.S. presidential election and two previous U.S. presidential elections as benchmarks. In contrast to prior elections, we find that the 2016 U.S. presidential election outcome did not resolve market uncertainty. Specifically, we show significant increases in transactions costs, adverse selection costs, and volatility in the days following the election date. We contribute to the literature by suggesting that unexpected elections can engender, rather than resolve, market uncertainty.</p>
|
|
33
|
+
</abstract>
|
|
34
|
+
<kwd-group kwd-group-type="author"><title>JEL Classification: </title><kwd>D80</kwd><kwd>G00</kwd><kwd>G12</kwd><kwd>G14</kwd></kwd-group>
|
|
35
|
+
<kwd-group kwd-group-type="author"><title>KEYWORDS: </title><kwd>suicide attempters</kwd><kwd>schizophrenia</kwd><kwd><italic>DβH</italic></kwd><kwd>attention</kwd><kwd>polymorphism</kwd></kwd-group>
|
|
36
|
+
<history>
|
|
37
|
+
<date date-type="received" iso-8601-date="1998-01-05">
|
|
38
|
+
<day>05</day>
|
|
39
|
+
<month>01</month>
|
|
40
|
+
<year>1998</year>
|
|
41
|
+
</date>
|
|
42
|
+
<date date-type="rev-request" iso-8601-date="1998-03-14">
|
|
43
|
+
<day>14</day>
|
|
44
|
+
<month>03</month>
|
|
45
|
+
<year>1998</year>
|
|
46
|
+
</date>
|
|
47
|
+
<date date-type="rev-recd" iso-8601-date="1998-05-24">
|
|
48
|
+
<day>24</day>
|
|
49
|
+
<month>05</month>
|
|
50
|
+
<year>1998</year>
|
|
51
|
+
</date>
|
|
52
|
+
<date date-type="accepted" iso-8601-date="1998-06-06">
|
|
53
|
+
<day>06</day>
|
|
54
|
+
<month>06</month>
|
|
55
|
+
<year>1998</year>
|
|
56
|
+
</date>
|
|
57
|
+
</history>
|
|
58
|
+
</article-meta>
|
|
59
|
+
</front>
|
|
60
|
+
<body>
|
|
61
|
+
<sec id="secI" sec-type="intro">
|
|
62
|
+
<label>I</label><title>Introduction</title>
|
|
63
|
+
<p>Do political elections resolve financial market uncertainty? This study examines precise measures of stock liquidity surrounding the 2016 U.S. presidential election. First, we investigate whether market liquidity changes in anticipation of the election outcome. Second, we analyze whether the election outcome resolved uncertainty, resulting in better market liquidity and less adverse selection risk. Third, we examine the effect of election uncertainty on systematic and idiosyncratic volatility. Fourth, we examine whether election uncertainty induced greater informational asymmetries across various stock characteristics and for stocks in politically sensitive industries.</p>
|
|
64
|
+
<p>We contend that not all election outcomes are created equal and that the 2016 U.S. presidential election engendered, rather than resolved, financial market uncertainty. This became apparent as media outlets reported Donald Trump’s victory late Tuesday night, November 8, 2016, and Dow futures dropped almost 900 points and the S&P 500 lost nearly 5% that same evening.<xref ref-type="fn" rid="fn1"><sup>1</sup></xref><sup>,</sup><xref ref-type="fn" rid="fn2"><sup>2</sup></xref> We use the 2012 U.S. presidential election as a benchmark, as the outcome of that election appears to have been known in the weeks leading up to the election date, which is clearly not the case for the 2016 election.<xref ref-type="fn" rid="fn3"><sup>3</sup></xref> <xref ref-type="fig" rid="figI">Figure I</xref> provides a time-series depiction of the daily Iowa Election Markets (IEM) winner-takes-all probabilities for both the Democratic and Republican candidates during the 2012 and 2016 election campaigns.<xref ref-type="fn" rid="fn4"><sup>4</sup></xref> Panel A shows that the Democratic candidate led the Republican candidate throughout the 2016 election cycle, with the gap widening in the final three months before the election date, but failed to win the election. Panel B shows that the Democratic candidate led the Republican candidate for the entire final year of the 2012 election cycle, and eventually was voted into the presidential office. Thus, voter expectations during the 2012 election campaign were more consistent with the eventual result relative to the 2016 election.</p>
|
|
65
|
+
<fig id="figI" position="float" fig-type="half-right"><label>Figure I</label>
|
|
66
|
+
<caption>
|
|
67
|
+
<p><bold>Iowa Election Markets Winner Takes All.</bold> The figure displays the daily results of the Iowa Election Markets winner takes all (WTA) for both the 2012 and 2016 U.S. presidential elections (available at <underline><ext-link ext-link-type="uri" xlink:href="https://iemweb.biz.uiowa.edu/closed/">https://iemweb.biz.uiowa.edu/closed/</ext-link>)</underline>. [Color figure can be viewed at wileyonlinelibrary.com.]</p>
|
|
68
|
+
<p>A. 2016 U.S. Presidential Election</p>
|
|
69
|
+
<p>B. 2012 U.S. Presidential Election</p>
|
|
70
|
+
</caption>
|
|
71
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image001"></graphic>
|
|
72
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image002"></graphic>
|
|
73
|
+
</fig>
|
|
74
|
+
<p>To the extent that election uncertainty affects market conditions, we focus on an extended event window that accounts for the various stages of the political process. Pantzalis, Strangeland, and Turtle (2000) argue that market participants incorporate both news and political information into prices in anticipation of election outcomes in three stages. In the first stage, market participants weigh potential policy changes before the election event. In the second stage, market participants use available sources to continually update their expectations about the future political outcome. In the final stage as election outcomes become certain, market participants update their previous probability distributions of not only the policies to be implemented but also the long-term economic response to these policies. <xref ref-type="bibr" rid="bib11">Brown, Harlow, and Tinic (1988)</xref> argue that as election uncertainty resolves, market uncertainty reduces, resulting in positive price changes.</p>
|
|
75
|
+
<p>Although several studies offer evidence related to the timing of this resolution of uncertainty, no consensus exists. A host of studies suggest that uncertainty is resolved before the voting date. Pantzalis, Strangeland, and Turtle (2000) contend that uncertainty may resolve well before the actual election date, showing that much of the market uncertainty is resolved in the weeks leading up to the election. Similarly, Santa-Clara and Valkanov (2003) find no evidence that returns are driven by uncertainty surrounding elections but instead find that market prices incorporate information about a new presidency gradually during the period before the election, with minimal political uncertainty left for the election day. Pasquareillo and Zafeiridou (2014) explore nearly 20 U.S. presidential elections, finding that market liquidity worsens before the election event, only to improve in the subsequent months. This suggests that even if uncertainty is not resolved before the election date, the election itself serves as the resolution mechanism of any residual uncertainty.</p>
|
|
76
|
+
<p>Other studies document that election outcomes may create uncertainty in financial markets. <xref ref-type="bibr" rid="bib13">Gemmill (1992)</xref> shows that inefficiencies associated with opinion polls for the British parliamentary election of 1987 resulted in a speculative bubble in the options market. <xref ref-type="bibr" rid="bib29">Nippani and Medlin (2002)</xref>, <xref ref-type="bibr" rid="bib28">Nippani and Arize (2005)</xref>, and <xref ref-type="bibr" rid="bib16">He et al. (2009)</xref> all show that the uncertainty associated with the 2000 U.S. presidential election served as a catalyst for uncertainty in financial markets. <xref ref-type="bibr" rid="bib24">Li and Born (2006)</xref> use a 40-year sample period covering multiple U.S. presidential elections, finding that when election uncertainty increases via candidate preference polling numbers, both stock market volatility and returns increase. <xref ref-type="bibr" rid="bib6">Białkowski, Gottschalk, and Wisniewski (2008)</xref> use international data across 27 Organization for Economic Co-operation and Development (OECD) countries and provide evidence that stock market volatility nearly doubles during election week, which they attribute to the surprise outcomes of national elections. Finally, <xref ref-type="bibr" rid="bib14">Goodell and Vähämaa (2013)</xref> show that the presidential election process creates rather than alleviates market concerns, forcing investors to reform their expectations about future economic policy.</p>
|
|
77
|
+
<p>The uncertainty associated with an election outcome can affect market liquidity through three channels: information asymmetry, ambiguity, and disagreement. The information asymmetry channel argues that political events widen the information gap between informed and uninformed investors, which increases the implementation shortfall in financial markets. This gap is likely to widen just before the political outcome when uncertainty is at its highest, increasing the adverse selection risk between informed and uninformed participants. The resulting higher adverse selection risk may increase the costs of market making, resulting in higher spreads and greater volatility (see <xref ref-type="bibr" rid="bib17">Ho and Stoll 1981</xref>; <xref ref-type="bibr" rid="bib22">Kyle 1985</xref>; <xref ref-type="bibr" rid="bib2">Amihud and Mendelson 1986</xref>). Once the election outcome is certain, informational asymmetries reduce, and market liquidity improves. The ambiguity channel assumes that uncertainty, regardless of the political component, reduces the consensus belief about liquidity information in the market.<xref ref-type="fn" rid="fn5"><sup>5</sup></xref> Thus, as fundamental information becomes inaccurate because of increased uncertainty, market participants may avoid trading altogether to evade adverse selection risk. As participants withdraw from the market, both trading volume and market liquidity decline. However, <xref ref-type="bibr" rid="bib32">Pasquariello and Zafeiridou (2014)</xref> argue that once the election outcome is determined, the ambiguity associated with the quality of information reduces, resulting in improved liquidity conditions. Finally, the disagreement channel asserts that a lack of consensus about the fundamental value of an asset enhances the incentive to trade and, therefore, improves market liquidity. This follows from the premise that increases in the divergence of opinions among market participants results in more trading during periods of high uncertainty (i.e., before the election). Once the election outcome is known and uncertainty reduces, trading volume and liquidity decline.<xref ref-type="fn" rid="fn6"><sup>6</sup></xref></p>
|
|
78
|
+
<p>We contend that the unexpected outcome of the 2016 election likely increased information asymmetries among market participants.<xref ref-type="fn" rid="fn7"><sup>7</sup></xref> Consistent with the information asymmetry argument, we predict that the surprise outcome of the 2016 election reduced financial market liquidity, defined as the ability to trade a large quantity of securities quickly, at a low cost, and without substantially moving prices (see <xref ref-type="bibr" rid="bib33">Pástor and Stambaugh 2003</xref>). Specifically, we expect to observe higher quoted and effective spreads (<xref ref-type="bibr" rid="bib17">Ho and Stoll, 1981</xref>) and greater price impacts (<xref ref-type="bibr" rid="bib12">Easley and O’Hara 1987</xref>) following the 2016 election. We also expect stock prices to fluctuate more as investors alter their expectations around unanticipated policy changes (see <xref ref-type="bibr" rid="bib34">Pástor and Veronesi 2012</xref>).<xref ref-type="fn" rid="fn8"><sup>8</sup></xref><sup>,</sup><xref ref-type="fn" rid="fn9"><sup>9</sup></xref></p>
|
|
79
|
+
<p>The results from our empirical analysis indicate that the outcome of the 2016 election engendered more uncertainty than the 2012 election, resulting in deteriorating liquidity.<xref ref-type="fn" rid="fn10"><sup>10</sup></xref> We find that transactions costs, measured by quoted and effective spreads, increased considerably following the 2016 election, relative to the 2012 election. Furthermore, our results indicate that the higher transactions costs associated with the 2016 election persisted for nearly two months following the outcome. We find similar results when analyzing the effects of the 2016 election on measures of adverse selection costs. In fact, both the price impact component of the bid–ask spread and <xref ref-type="bibr" rid="bib1">Amihud’s (2002)</xref> illiquidity increased significantly following the 2016 election, relative to the 2012 election. Additionally, we find that total, systematic, and idiosyncratic volatility increased substantially following the 2016 election. Although systematic volatility subsided four weeks after the 2016 election, idiosyncratic volatility continued to increase.</p>
|
|
80
|
+
<p>In our next set of tests, we attempt to identify the economic channels through which election uncertainty creates market illiquidity. We find differential effects across stock characteristics around the 2016 election. More infrequently traded and small-cap securities experienced the greatest illiquidity shock following the 2016 election, relative to the 2012 election. We also document that stocks within politically sensitive industries (e.g. healthcare, pharmaceutical, banking, trading, natural gas, defense, and tobacco) experienced more adverse liquidity effects after the 2016 election. The latter finding is consistent with studies showing that firms in politically sensitive industries are more exposed to higher political uncertainty (see Boutchkova, Doshi, and Durnev 2011; <xref ref-type="bibr" rid="bib20">Julio and Yook 2012</xref>; <xref ref-type="bibr" rid="bib19">Hong and Kostovetsky 2012</xref>).</p>
|
|
81
|
+
<p>Our article contributes to the literature linking political uncertainty and financial market outcomes in the following ways.<xref ref-type="fn" rid="fn11"><sup>11</sup></xref> First, our analysis differs from previous studies (Pantzalis, Strangeland, and Turtle 2000; Santa-Clara and Valkanov 2003; Pasqauriello and Zafeiridou 2014) in that we contend that not all election outcomes are known in advance and that certain elections may create, rather than reduce, uncertainty. Using the 2004 and 2012 elections as benchmarks, we detail that the uncertainty associated with the 2016 election resulted in greater illiquidity and worsened trading conditions for market participants. Second, in contrast to <xref ref-type="bibr" rid="bib32">Pasquariello and Zafeiridou (2014)</xref>, who examine the effects of election uncertainty on market liquidity using monthly data, we use more precise measurements of stock liquidity based on intraday New York Stock Exchange (NYSE) Trade and Quote (TAQ) files. Finally, we argue that our empirical design and analysis provide inferences from pinpointing the economic channel through which election uncertainty affects market liquidity. Specifically, our findings support the information asymmetry channel, but not the disagreement channel, as illiquidity after the 2016 election persisted for several weeks and stocks more exposed to information asymmetries (i.e. infrequently traded small-cap stocks in politically sensitive industries) experienced the most significant decrease in liquidity. Overall, our findings provide evidence that uncertainty is not necessarily resolved before the political event, or on the political event day itself, and that unanticipated political outcomes can be detrimental to financial market liquidity.</p>
|
|
82
|
+
</sec>
|
|
83
|
+
<sec id="secII" sec-type="other1">
|
|
84
|
+
<label>II</label><title>Measures of Liquidity and Data Description</title>
|
|
85
|
+
<sec>
|
|
86
|
+
<title>Measuring Liquidity</title>
|
|
87
|
+
<p>We follow <xref ref-type="bibr" rid="bib18">Holden and Jacobsen (2014)</xref> and analyze standard measures of market liquidity. First, we evaluate the quoted bid–ask spread, which is the difference between the prices quoted for an immediate sale and an immediate buy for one share of stock. For a given time interval <italic>t</italic>, the percent quoted spread is measured as</p><disp-formula id="eq1"><mml:math id="m1">
|
|
88
|
+
<mml:mrow>
|
|
89
|
+
<mml:mi>P</mml:mi><mml:mi>e</mml:mi><mml:mi>r</mml:mi><mml:mi>c</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:mi>t</mml:mi><mml:mo> </mml:mo><mml:mi>Q</mml:mi><mml:mi>u</mml:mi><mml:mi>o</mml:mi><mml:mi>t</mml:mi><mml:mi>e</mml:mi><mml:mi>d</mml:mi><mml:mo> </mml:mo><mml:mi>S</mml:mi><mml:mi>p</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>a</mml:mi><mml:msub>
|
|
90
|
+
<mml:mi>d</mml:mi>
|
|
91
|
+
<mml:mi>t</mml:mi>
|
|
92
|
+
</mml:msub>
|
|
93
|
+
<mml:mo>=</mml:mo><mml:mo> </mml:mo><mml:mfrac>
|
|
94
|
+
<mml:mrow>
|
|
95
|
+
<mml:mi>A</mml:mi><mml:mi>s</mml:mi><mml:msub>
|
|
96
|
+
<mml:mi>k</mml:mi>
|
|
97
|
+
<mml:mi>t</mml:mi>
|
|
98
|
+
</mml:msub>
|
|
99
|
+
<mml:mo>−</mml:mo><mml:mi>B</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
100
|
+
<mml:mi>d</mml:mi>
|
|
101
|
+
<mml:mi>t</mml:mi>
|
|
102
|
+
</mml:msub>
|
|
103
|
+
</mml:mrow>
|
|
104
|
+
<mml:mrow>
|
|
105
|
+
<mml:msub>
|
|
106
|
+
<mml:mi>M</mml:mi>
|
|
107
|
+
<mml:mi>t</mml:mi>
|
|
108
|
+
</mml:msub>
|
|
109
|
+
</mml:mrow>
|
|
110
|
+
</mml:mfrac>
|
|
111
|
+
</mml:mrow>
|
|
112
|
+
</mml:math>,<label>(1)</label></disp-formula><p>where <italic>Ask<sub>t</sub></italic> is the National Best Ask and <italic>Bid<sub>t</sub></italic> is the National Best Bid assigned to time <italic>t</italic> and <inline-formula><mml:math id="m2">
|
|
113
|
+
<mml:mrow>
|
|
114
|
+
<mml:msub>
|
|
115
|
+
<mml:mi>M</mml:mi>
|
|
116
|
+
<mml:mi>t</mml:mi>
|
|
117
|
+
</mml:msub>
|
|
118
|
+
</mml:mrow>
|
|
119
|
+
</mml:math></inline-formula> is the midpoint, which is the average of <italic>Ask<sub>t</sub></italic> and <italic>Bid<sub>t</sub></italic>.</p>
|
|
120
|
+
<p>To more accurately measure the round-trip trading costs when trades occur at prices inside the posted bid and ask quotes, we estimate the effective spread. For a given security, the percent effective spread on the <italic>k</italic>th trade is defined as</p><disp-formula id="eq2"><mml:math id="m3">
|
|
121
|
+
<mml:mrow>
|
|
122
|
+
<mml:mi>P</mml:mi><mml:mi>e</mml:mi><mml:mi>r</mml:mi><mml:mi>c</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:mi>t</mml:mi><mml:mo> </mml:mo><mml:mi>E</mml:mi><mml:mi>f</mml:mi><mml:mi>f</mml:mi><mml:mi>e</mml:mi><mml:mi>c</mml:mi><mml:mi>t</mml:mi><mml:mi>i</mml:mi><mml:mi>v</mml:mi><mml:mi>e</mml:mi><mml:mo> </mml:mo><mml:mi>S</mml:mi><mml:mi>p</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>a</mml:mi><mml:msub>
|
|
123
|
+
<mml:mi>d</mml:mi>
|
|
124
|
+
<mml:mi>k</mml:mi>
|
|
125
|
+
</mml:msub>
|
|
126
|
+
<mml:mo>=</mml:mo><mml:mo> </mml:mo><mml:mfrac>
|
|
127
|
+
<mml:mrow>
|
|
128
|
+
<mml:mn>2</mml:mn><mml:msub>
|
|
129
|
+
<mml:mi>D</mml:mi>
|
|
130
|
+
<mml:mi>k</mml:mi>
|
|
131
|
+
</mml:msub>
|
|
132
|
+
<mml:mfenced>
|
|
133
|
+
<mml:mrow>
|
|
134
|
+
<mml:msub>
|
|
135
|
+
<mml:mi>P</mml:mi>
|
|
136
|
+
<mml:mi>k</mml:mi>
|
|
137
|
+
</mml:msub>
|
|
138
|
+
<mml:mo>−</mml:mo><mml:msub>
|
|
139
|
+
<mml:mi>M</mml:mi>
|
|
140
|
+
<mml:mi>k</mml:mi>
|
|
141
|
+
</mml:msub>
|
|
142
|
+
</mml:mrow>
|
|
143
|
+
</mml:mfenced></mml:mrow>
|
|
144
|
+
<mml:mrow>
|
|
145
|
+
<mml:msub>
|
|
146
|
+
<mml:mi>M</mml:mi>
|
|
147
|
+
<mml:mi>k</mml:mi>
|
|
148
|
+
</mml:msub>
|
|
149
|
+
</mml:mrow>
|
|
150
|
+
</mml:mfrac>
|
|
151
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo></mml:mrow>
|
|
152
|
+
</mml:math><label>(2)</label></disp-formula><p>where <italic>D<sub>k</sub></italic> is an indicator variable that equals +1 if the <italic>k</italic>th trade is a buy and −1 if the <italic>k</italic>th trade is a sell, <inline-formula><mml:math id="m4">
|
|
153
|
+
<mml:mi>P</mml:mi>
|
|
154
|
+
</mml:math></inline-formula><italic><sub>k</sub></italic> is the price of the <italic>k</italic>th trade, and <inline-formula><mml:math id="m5">
|
|
155
|
+
<mml:mi>M</mml:mi>
|
|
156
|
+
</mml:math></inline-formula><italic><sub>k</sub></italic> is the midpoint of the National Best Bid and Offer (NBBO) quotes assigned to the <italic>k</italic>th trade. We use the <xref ref-type="bibr" rid="bib23">Lee and Ready (1991)</xref> algorithm to determine whether a given trade is a buy or sell. Under this convention, a trade is a buy when <inline-formula><mml:math id="m6">
|
|
157
|
+
<mml:mrow>
|
|
158
|
+
<mml:msub>
|
|
159
|
+
<mml:mi>P</mml:mi>
|
|
160
|
+
<mml:mi>k</mml:mi>
|
|
161
|
+
</mml:msub>
|
|
162
|
+
</mml:mrow>
|
|
163
|
+
</mml:math></inline-formula> > <inline-formula><mml:math id="m7">
|
|
164
|
+
<mml:mrow>
|
|
165
|
+
<mml:msub>
|
|
166
|
+
<mml:mi>M</mml:mi>
|
|
167
|
+
<mml:mi>k</mml:mi>
|
|
168
|
+
</mml:msub>
|
|
169
|
+
</mml:mrow>
|
|
170
|
+
</mml:math></inline-formula>, a sell when <inline-formula><mml:math id="m8">
|
|
171
|
+
<mml:mrow>
|
|
172
|
+
<mml:msub>
|
|
173
|
+
<mml:mi>P</mml:mi>
|
|
174
|
+
<mml:mi>k</mml:mi>
|
|
175
|
+
</mml:msub>
|
|
176
|
+
</mml:mrow>
|
|
177
|
+
</mml:math></inline-formula> < <inline-formula><mml:math id="m9">
|
|
178
|
+
<mml:mrow>
|
|
179
|
+
<mml:msub>
|
|
180
|
+
<mml:mi>M</mml:mi>
|
|
181
|
+
<mml:mi>k</mml:mi>
|
|
182
|
+
</mml:msub>
|
|
183
|
+
</mml:mrow>
|
|
184
|
+
</mml:math></inline-formula>, and the tick test is used when <inline-formula><mml:math id="m10">
|
|
185
|
+
<mml:mrow>
|
|
186
|
+
<mml:msub>
|
|
187
|
+
<mml:mi>P</mml:mi>
|
|
188
|
+
<mml:mi>k</mml:mi>
|
|
189
|
+
</mml:msub>
|
|
190
|
+
</mml:mrow>
|
|
191
|
+
</mml:math></inline-formula> = <inline-formula><mml:math id="m11">
|
|
192
|
+
<mml:mrow>
|
|
193
|
+
<mml:msub>
|
|
194
|
+
<mml:mi>M</mml:mi>
|
|
195
|
+
<mml:mi>k</mml:mi>
|
|
196
|
+
</mml:msub>
|
|
197
|
+
</mml:mrow>
|
|
198
|
+
</mml:math></inline-formula>. The tick test determines a trade is a buy (sell) if the most recent prior trade at a different price was at a lower (higher) price.</p>
|
|
199
|
+
<p>To measure the adverse selection component of the bid–ask spread, we calculate price impact. For a given security, the percent price impact on the <italic>k</italic>th trade are defined as</p><disp-formula id="eq3"><mml:math id="m12">
|
|
200
|
+
<mml:mrow>
|
|
201
|
+
<mml:mi>P</mml:mi><mml:mi>e</mml:mi><mml:mi>r</mml:mi><mml:mi>c</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:mi>t</mml:mi><mml:mo> </mml:mo><mml:mi>P</mml:mi><mml:mi>r</mml:mi><mml:mi>i</mml:mi><mml:mi>c</mml:mi><mml:mi>e</mml:mi><mml:mo> </mml:mo><mml:mi>I</mml:mi><mml:mi>m</mml:mi><mml:mi>p</mml:mi><mml:mi>a</mml:mi><mml:mi>c</mml:mi><mml:msub>
|
|
202
|
+
<mml:mi>t</mml:mi>
|
|
203
|
+
<mml:mi>k</mml:mi>
|
|
204
|
+
</mml:msub>
|
|
205
|
+
<mml:mo>=</mml:mo><mml:mo> </mml:mo><mml:mfrac>
|
|
206
|
+
<mml:mrow>
|
|
207
|
+
<mml:mn>2</mml:mn><mml:msub>
|
|
208
|
+
<mml:mi>D</mml:mi>
|
|
209
|
+
<mml:mi>k</mml:mi>
|
|
210
|
+
</mml:msub>
|
|
211
|
+
<mml:mfenced>
|
|
212
|
+
<mml:mrow>
|
|
213
|
+
<mml:msub>
|
|
214
|
+
<mml:mi>M</mml:mi>
|
|
215
|
+
<mml:mrow>
|
|
216
|
+
<mml:mi>k</mml:mi><mml:mo>+</mml:mo><mml:mn>5</mml:mn></mml:mrow>
|
|
217
|
+
</mml:msub>
|
|
218
|
+
<mml:mo>−</mml:mo><mml:msub>
|
|
219
|
+
<mml:mi>M</mml:mi>
|
|
220
|
+
<mml:mi>k</mml:mi>
|
|
221
|
+
</mml:msub>
|
|
222
|
+
</mml:mrow>
|
|
223
|
+
</mml:mfenced></mml:mrow>
|
|
224
|
+
<mml:mrow>
|
|
225
|
+
<mml:msub>
|
|
226
|
+
<mml:mi>M</mml:mi>
|
|
227
|
+
<mml:mi>k</mml:mi>
|
|
228
|
+
</mml:msub>
|
|
229
|
+
</mml:mrow>
|
|
230
|
+
</mml:mfrac>
|
|
231
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo></mml:mrow>
|
|
232
|
+
</mml:math><label>(3)</label></disp-formula><p>where <inline-formula><mml:math id="m13">
|
|
233
|
+
<mml:mrow>
|
|
234
|
+
<mml:msub>
|
|
235
|
+
<mml:mi>M</mml:mi>
|
|
236
|
+
<mml:mi>k</mml:mi>
|
|
237
|
+
</mml:msub>
|
|
238
|
+
</mml:mrow>
|
|
239
|
+
</mml:math></inline-formula> is the midpoint of the NBBO quotes assigned to the <italic>k</italic>th trade, <inline-formula><mml:math id="m14">
|
|
240
|
+
<mml:mrow>
|
|
241
|
+
<mml:msub>
|
|
242
|
+
<mml:mi>M</mml:mi>
|
|
243
|
+
<mml:mrow>
|
|
244
|
+
<mml:mi>k</mml:mi><mml:mo>+</mml:mo><mml:mn>5</mml:mn></mml:mrow>
|
|
245
|
+
</mml:msub>
|
|
246
|
+
<mml:mo> </mml:mo></mml:mrow>
|
|
247
|
+
</mml:math></inline-formula>is the midpoint five minutes after the <italic>k</italic>th trade, and <inline-formula><mml:math id="m15">
|
|
248
|
+
<mml:mrow>
|
|
249
|
+
<mml:msub>
|
|
250
|
+
<mml:mi>D</mml:mi>
|
|
251
|
+
<mml:mi>k</mml:mi>
|
|
252
|
+
</mml:msub>
|
|
253
|
+
</mml:mrow>
|
|
254
|
+
</mml:math></inline-formula> is an indicator variable that equals +1 if the <italic>k</italic>th trade is a buy and −1 if the <italic>k</italic>th trade is a sell.<xref ref-type="fn" rid="fn12"><sup>12</sup></xref> A low-frequency proxy of price impact often used in the literature is <xref ref-type="bibr" rid="bib1">Amihud (2002)</xref> illiquidity, which is defined as the absolute daily return divided by dollar volume, scaled by 10<sup>6</sup>, where larger values indicate greater illiquidity.</p>
|
|
255
|
+
<p>Finally, we estimate three measures of stock volatility: total, systematic, and idiosyncratic. Total volatility is the 30-day rolling standard deviation of daily stock returns. Systematic volatility is the 30-day rolling beta coefficient estimated from the capital asset pricing model (CAPM) using Center for Research in Security Prices (CRSP) value-weighted market returns. Idiosyncratic volatility is the 30-day rolling standard deviation of the residuals obtained from the CAPM.</p>
|
|
256
|
+
</sec>
|
|
257
|
+
<sec>
|
|
258
|
+
<title>Data Description</title>
|
|
259
|
+
<p>To examine the effects of both the 2012 and 2016 U.S. presidential elections on financial market liquidity, we obtain NYSE TAQ files for the 121 days (90 days before and 31 days after) surrounding the election dates. These data include all trades and quotes for all issues traded on NYSE, NASDAQ, and regional exchanges. We calculate the NBBO in accordance with <xref ref-type="bibr" rid="bib18">Holden and Jacobsen (2014)</xref>, who account for most of the official NBBO quotes from the Securities Industries Processors.<xref ref-type="fn" rid="fn13"><sup>13</sup></xref></p>
|
|
260
|
+
<p>We make the following adjustments to the initial sample. We retain quotes between 9:00 a.m. and 4:00 p.m. to observe all trades during regular trading hours 9:30 a.m. and 4:00 p.m. For each exchange, or market maker, we delete instances where the bid price of one exchange is greater than or equal to the ask price of the same exchange (i.e., crossed quotes). We remove quotes with non-normal conditions and when the bid/ask price or bid/ask size is equal to zero or missing. We merge the TAQ data with closing prices and shares outstanding from CRSP and retain only common stocks (i.e., share codes 10 and 11). We also require the stock to have a closing price greater than $5 with nonzero trading volume.<xref ref-type="fn" rid="fn14"><sup>14</sup></xref> To reduce the influence of outliers and potential data errors, we winsorize the market liquidity measures at the 5th and 95th percentile levels of their distributions.</p>
|
|
261
|
+
<p><xref ref-type="table" rid="tbl1">Table 1</xref> displays summary statistics for the market liquidity measures, along with stock and trade characteristics for the sample before the 2012 and 2016 elections. The event window for the summary statistics is [−90,−10], which allows us to control for liquidity changes at various stages of the political process and potential illiquidity run-up in the 10 days before the elections. Our sample consists of 3,042 unique stocks around the 2012 election and 3,106 unique stocks around the 2016 election.<xref ref-type="fn" rid="fn15"><sup>15</sup></xref> In Panel A, we report that the average percentage quoted and effective spreads for the sample of stocks in the days before the 2012 election are 52.55 and 36.08 basis points (bps), with standard deviations of 71.31 and 52.07 bps, respectively. We also show that the average price impact is 11.78 bps, whereas <xref ref-type="bibr" rid="bib1">Amihud (2002)</xref> illiquidity is 0.4258. The average trading price and market capitalization for the sample of 2012 stocks are $29.66 and $5,059 million, respectively.</p>
|
|
262
|
+
<table-wrap id="tbl1" position="float">
|
|
263
|
+
<label>TABLE 1</label><caption><title>Descriptive Statistics.</title>
|
|
264
|
+
</caption>
|
|
265
|
+
<table frame="hsides" rules="groups">
|
|
266
|
+
<col width="12.55%"/>
|
|
267
|
+
<col width="7%"/>
|
|
268
|
+
<col width="8.01%"/>
|
|
269
|
+
<col width="6.95%"/>
|
|
270
|
+
<col width="8.82%"/>
|
|
271
|
+
<col width="8.42%"/>
|
|
272
|
+
<col width="9.44%"/>
|
|
273
|
+
<col width="11.26%"/>
|
|
274
|
+
<col width="9.73%"/>
|
|
275
|
+
<col width="8.06%"/>
|
|
276
|
+
<col width="9.76%"/>
|
|
277
|
+
<thead>
|
|
278
|
+
<tr>
|
|
279
|
+
<th valign="bottom" align="left" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"></th>
|
|
280
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Quoted Spread</italic></th>
|
|
281
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Effective Spread</italic></th>
|
|
282
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Price Impact</italic></th>
|
|
283
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Amihud Illiquidity</italic></th>
|
|
284
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Total Volatility</italic></th>
|
|
285
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Systematic Volatility</italic></th>
|
|
286
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Idiosyncratic Volatility</italic></th>
|
|
287
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Trading Volume</italic></th>
|
|
288
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>Price</italic></th>
|
|
289
|
+
<th valign="bottom" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><italic>MCAP</italic><break/>($millions)</th>
|
|
290
|
+
</tr>
|
|
291
|
+
</thead>
|
|
292
|
+
<tbody>
|
|
293
|
+
<tr>
|
|
294
|
+
<td valign="middle" colspan="11" align="left" style="border-top: solid 0.50pt" scope="col">Panel A. 2012 U.S. Presidential Election (3,042 common stocks)</td>
|
|
295
|
+
</tr>
|
|
296
|
+
<tr>
|
|
297
|
+
<td valign="middle" align="left" scope="row">Mean</td>
|
|
298
|
+
<td valign="middle" align="center">0.5255</td>
|
|
299
|
+
<td valign="middle" align="center">0.3608</td>
|
|
300
|
+
<td valign="middle" align="center">0.1178</td>
|
|
301
|
+
<td valign="middle" align="center">0.4258</td>
|
|
302
|
+
<td valign="middle" align="center">0.0232</td>
|
|
303
|
+
<td valign="middle" align="center">1.0778</td>
|
|
304
|
+
<td valign="middle" align="center">0.0201</td>
|
|
305
|
+
<td valign="middle" align="center">1,284,758</td>
|
|
306
|
+
<td valign="middle" align="center">$29.66</td>
|
|
307
|
+
<td valign="middle" align="center">$5,059</td>
|
|
308
|
+
</tr>
|
|
309
|
+
<tr>
|
|
310
|
+
<td valign="middle" align="left" scope="row">Std. Dev.</td>
|
|
311
|
+
<td valign="middle" align="center">0.7131</td>
|
|
312
|
+
<td valign="middle" align="center">0.5207</td>
|
|
313
|
+
<td valign="middle" align="center">0.1197</td>
|
|
314
|
+
<td valign="middle" align="center">0.8455</td>
|
|
315
|
+
<td valign="middle" align="center">0.0140</td>
|
|
316
|
+
<td valign="middle" align="center">0.6787</td>
|
|
317
|
+
<td valign="middle" align="center">0.0139</td>
|
|
318
|
+
<td valign="middle" align="center">4,800,735</td>
|
|
319
|
+
<td valign="middle" align="center">$56.26</td>
|
|
320
|
+
<td valign="middle" align="center">$21,535</td>
|
|
321
|
+
</tr>
|
|
322
|
+
<tr>
|
|
323
|
+
<td valign="middle" align="left" scope="row">25th percentile</td>
|
|
324
|
+
<td valign="middle" align="center">0.0955</td>
|
|
325
|
+
<td valign="middle" align="center">0.0575</td>
|
|
326
|
+
<td valign="middle" align="center">0.0311</td>
|
|
327
|
+
<td valign="middle" align="center">0.0057</td>
|
|
328
|
+
<td valign="middle" align="center">0.0153</td>
|
|
329
|
+
<td valign="middle" align="center">0.6463</td>
|
|
330
|
+
<td valign="middle" align="center">0.0122</td>
|
|
331
|
+
<td valign="middle" align="center">49,313</td>
|
|
332
|
+
<td valign="middle" align="center">$10.02</td>
|
|
333
|
+
<td valign="middle" align="center">$225</td>
|
|
334
|
+
</tr>
|
|
335
|
+
<tr>
|
|
336
|
+
<td valign="middle" align="left" scope="row">Median</td>
|
|
337
|
+
<td valign="middle" align="center">0.2097</td>
|
|
338
|
+
<td valign="middle" align="center">0.1268</td>
|
|
339
|
+
<td valign="middle" align="center">0.0665</td>
|
|
340
|
+
<td valign="middle" align="center">0.0368</td>
|
|
341
|
+
<td valign="middle" align="center">0.0206</td>
|
|
342
|
+
<td valign="middle" align="center">1.0501</td>
|
|
343
|
+
<td valign="middle" align="center">0.0171</td>
|
|
344
|
+
<td valign="middle" align="center">245,033</td>
|
|
345
|
+
<td valign="middle" align="center">$18.67</td>
|
|
346
|
+
<td valign="middle" align="center">$735</td>
|
|
347
|
+
</tr>
|
|
348
|
+
<tr>
|
|
349
|
+
<td valign="middle" align="left" scope="row">75 percentile</td>
|
|
350
|
+
<td valign="middle" align="center">0.5553</td>
|
|
351
|
+
<td valign="middle" align="center">0.3505</td>
|
|
352
|
+
<td valign="middle" align="center">0.1675</td>
|
|
353
|
+
<td valign="middle" align="center">0.3028</td>
|
|
354
|
+
<td valign="middle" align="center">0.0277</td>
|
|
355
|
+
<td valign="middle" align="center">1.4754</td>
|
|
356
|
+
<td valign="middle" align="center">0.0242</td>
|
|
357
|
+
<td valign="middle" align="center">860,998</td>
|
|
358
|
+
<td valign="middle" align="center">$36.41</td>
|
|
359
|
+
<td valign="middle" align="center">$2,610</td>
|
|
360
|
+
</tr>
|
|
361
|
+
<tr>
|
|
362
|
+
<td valign="middle" colspan="11" align="left" scope="col">Panel B. 2016 U.S. Presidential Election (3,106 common stocks)</td>
|
|
363
|
+
</tr>
|
|
364
|
+
<tr>
|
|
365
|
+
<td valign="middle" align="left" scope="row">Mean</td>
|
|
366
|
+
<td valign="middle" align="center">0.4694</td>
|
|
367
|
+
<td valign="middle" align="center">0.3953</td>
|
|
368
|
+
<td valign="middle" align="center">0.1541</td>
|
|
369
|
+
<td valign="middle" align="center">0.2674</td>
|
|
370
|
+
<td valign="middle" align="center">0.0229</td>
|
|
371
|
+
<td valign="middle" align="center">1.2043</td>
|
|
372
|
+
<td valign="middle" align="center">0.0197</td>
|
|
373
|
+
<td valign="middle" align="center">1,131,824</td>
|
|
374
|
+
<td valign="middle" align="center">$39.97</td>
|
|
375
|
+
<td valign="middle" align="center">$6,935</td>
|
|
376
|
+
</tr>
|
|
377
|
+
<tr>
|
|
378
|
+
<td valign="middle" align="left" scope="row">Std. Dev.</td>
|
|
379
|
+
<td valign="middle" align="center">0.6017</td>
|
|
380
|
+
<td valign="middle" align="center">0.5762</td>
|
|
381
|
+
<td valign="middle" align="center">0.1639</td>
|
|
382
|
+
<td valign="middle" align="center">0.5198</td>
|
|
383
|
+
<td valign="middle" align="center">0.0219</td>
|
|
384
|
+
<td valign="middle" align="center">0.8005</td>
|
|
385
|
+
<td valign="middle" align="center">0.0219</td>
|
|
386
|
+
<td valign="middle" align="center">3,225,708</td>
|
|
387
|
+
<td valign="middle" align="center">$86.59</td>
|
|
388
|
+
<td valign="middle" align="center">$27,028</td>
|
|
389
|
+
</tr>
|
|
390
|
+
<tr>
|
|
391
|
+
<td valign="middle" align="left" scope="row">25th percentile</td>
|
|
392
|
+
<td valign="middle" align="center">0.0844</td>
|
|
393
|
+
<td valign="middle" align="center">0.0552</td>
|
|
394
|
+
<td valign="middle" align="center">0.0465</td>
|
|
395
|
+
<td valign="middle" align="center">0.0038</td>
|
|
396
|
+
<td valign="middle" align="center">0.0139</td>
|
|
397
|
+
<td valign="middle" align="center">0.7759</td>
|
|
398
|
+
<td valign="middle" align="center">0.0106</td>
|
|
399
|
+
<td valign="middle" align="center">69,435</td>
|
|
400
|
+
<td valign="middle" align="center">$11.91</td>
|
|
401
|
+
<td valign="middle" align="center">$272</td>
|
|
402
|
+
</tr>
|
|
403
|
+
<tr>
|
|
404
|
+
<td valign="middle" align="left" scope="row">Median</td>
|
|
405
|
+
<td valign="middle" align="center">0.1951</td>
|
|
406
|
+
<td valign="middle" align="center">0.1286</td>
|
|
407
|
+
<td valign="middle" align="center">0.0945</td>
|
|
408
|
+
<td valign="middle" align="center">0.0221</td>
|
|
409
|
+
<td valign="middle" align="center">0.0188</td>
|
|
410
|
+
<td valign="middle" align="center">1.1691</td>
|
|
411
|
+
<td valign="middle" align="center">0.0154</td>
|
|
412
|
+
<td valign="middle" align="center">282,248</td>
|
|
413
|
+
<td valign="middle" align="center">$23.42</td>
|
|
414
|
+
<td valign="middle" align="center">$957</td>
|
|
415
|
+
</tr>
|
|
416
|
+
<tr>
|
|
417
|
+
<td valign="middle" align="left" style="border-bottom: solid 0.50pt" scope="row">75 percentile</td>
|
|
418
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.5500</td>
|
|
419
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.4304</td>
|
|
420
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.1951</td>
|
|
421
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.1869</td>
|
|
422
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.0263</td>
|
|
423
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">1.5961</td>
|
|
424
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">0.0232</td>
|
|
425
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">961,105</td>
|
|
426
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">$46.57</td>
|
|
427
|
+
<td valign="middle" align="center" style="border-bottom: solid 0.50pt">$3,306</td>
|
|
428
|
+
</tr>
|
|
429
|
+
</tbody>
|
|
430
|
+
</table><table-wrap-foot>
|
|
431
|
+
<p>Note: This table reports average daily summary statistics for common stocks with a share price greater than $5 and nonzero trading volume in U.S. equities before the 2012 and 2016 U.S. presidential elections. The event window for the summary statistics is [−90,−10]. All variables are first averaged by stock and then averaged across stocks. <italic>Quoted Spread</italic> is defined as <inline-formula><mml:math id="m16">
|
|
432
|
+
<mml:mrow>
|
|
433
|
+
<mml:mfenced close="]" open="[">
|
|
434
|
+
<mml:mrow>
|
|
435
|
+
<mml:mfenced>
|
|
436
|
+
<mml:mrow>
|
|
437
|
+
<mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:msub>
|
|
438
|
+
<mml:mi>k</mml:mi>
|
|
439
|
+
<mml:mi>t</mml:mi>
|
|
440
|
+
</mml:msub>
|
|
441
|
+
<mml:mo>−</mml:mo><mml:mi>b</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
442
|
+
<mml:mi>d</mml:mi>
|
|
443
|
+
<mml:mi>t</mml:mi>
|
|
444
|
+
</mml:msub>
|
|
445
|
+
</mml:mrow>
|
|
446
|
+
</mml:mfenced><mml:mo>/</mml:mo><mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
447
|
+
<mml:mi>t</mml:mi>
|
|
448
|
+
<mml:mi>t</mml:mi>
|
|
449
|
+
</mml:msub>
|
|
450
|
+
</mml:mrow>
|
|
451
|
+
</mml:mfenced><mml:mo>×</mml:mo><mml:mn>100</mml:mn></mml:mrow>
|
|
452
|
+
</mml:math></inline-formula>. <italic>Effective Spread</italic> is measured as <inline-formula><mml:math id="m17">
|
|
453
|
+
<mml:mrow>
|
|
454
|
+
<mml:mfenced close="]" open="[">
|
|
455
|
+
<mml:mrow>
|
|
456
|
+
<mml:mn>2</mml:mn><mml:msub>
|
|
457
|
+
<mml:mi>D</mml:mi>
|
|
458
|
+
<mml:mi>k</mml:mi>
|
|
459
|
+
</mml:msub>
|
|
460
|
+
<mml:mfenced>
|
|
461
|
+
<mml:mrow>
|
|
462
|
+
<mml:mi>p</mml:mi><mml:mi>r</mml:mi><mml:mi>i</mml:mi><mml:mi>c</mml:mi><mml:msub>
|
|
463
|
+
<mml:mi>e</mml:mi>
|
|
464
|
+
<mml:mi>k</mml:mi>
|
|
465
|
+
</mml:msub>
|
|
466
|
+
<mml:mo>−</mml:mo><mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
467
|
+
<mml:mi>t</mml:mi>
|
|
468
|
+
<mml:mi>k</mml:mi>
|
|
469
|
+
</mml:msub>
|
|
470
|
+
<mml:mo>/</mml:mo><mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
471
|
+
<mml:mi>t</mml:mi>
|
|
472
|
+
<mml:mi>k</mml:mi>
|
|
473
|
+
</mml:msub>
|
|
474
|
+
</mml:mrow>
|
|
475
|
+
</mml:mfenced></mml:mrow>
|
|
476
|
+
</mml:mfenced><mml:mo>×</mml:mo><mml:mn>100</mml:mn></mml:mrow>
|
|
477
|
+
</mml:math></inline-formula>, where <inline-formula><mml:math id="m18">
|
|
478
|
+
<mml:mrow>
|
|
479
|
+
<mml:msub>
|
|
480
|
+
<mml:mi>D</mml:mi>
|
|
481
|
+
<mml:mi>k</mml:mi>
|
|
482
|
+
</mml:msub>
|
|
483
|
+
</mml:mrow>
|
|
484
|
+
</mml:math></inline-formula> is an indicator variable that equals +1 if the <italic>k</italic>th trade is a buy and −1 if the <italic>k</italic>th trade is a sell (trade direction is defined as in <xref ref-type="bibr" rid="bib23">Lee and Ready, 1991</xref>). <italic>Price Impact</italic> is defined as <inline-formula><mml:math id="m19">
|
|
485
|
+
<mml:mrow>
|
|
486
|
+
<mml:mfenced close="]" open="[">
|
|
487
|
+
<mml:mrow>
|
|
488
|
+
<mml:mn>2</mml:mn><mml:msub>
|
|
489
|
+
<mml:mi>D</mml:mi>
|
|
490
|
+
<mml:mi>k</mml:mi>
|
|
491
|
+
</mml:msub>
|
|
492
|
+
<mml:mfenced>
|
|
493
|
+
<mml:mrow>
|
|
494
|
+
<mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
495
|
+
<mml:mi>t</mml:mi>
|
|
496
|
+
<mml:mrow>
|
|
497
|
+
<mml:mi>k</mml:mi><mml:mo>+</mml:mo><mml:mn>5</mml:mn></mml:mrow>
|
|
498
|
+
</mml:msub>
|
|
499
|
+
<mml:mo>−</mml:mo><mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
500
|
+
<mml:mi>t</mml:mi>
|
|
501
|
+
<mml:mi>k</mml:mi>
|
|
502
|
+
</mml:msub>
|
|
503
|
+
</mml:mrow>
|
|
504
|
+
</mml:mfenced><mml:mo>/</mml:mo><mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
505
|
+
<mml:mi>t</mml:mi>
|
|
506
|
+
<mml:mi>k</mml:mi>
|
|
507
|
+
</mml:msub>
|
|
508
|
+
</mml:mrow>
|
|
509
|
+
</mml:mfenced><mml:mo>×</mml:mo><mml:mn>100</mml:mn></mml:mrow>
|
|
510
|
+
</mml:math></inline-formula>, where <inline-formula><mml:math id="m20">
|
|
511
|
+
<mml:mrow>
|
|
512
|
+
<mml:mi>m</mml:mi><mml:mi>i</mml:mi><mml:mi>d</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
513
|
+
<mml:mi>t</mml:mi>
|
|
514
|
+
<mml:mrow>
|
|
515
|
+
<mml:mi>k</mml:mi><mml:mo>+</mml:mo><mml:mn>5</mml:mn></mml:mrow>
|
|
516
|
+
</mml:msub>
|
|
517
|
+
</mml:mrow>
|
|
518
|
+
</mml:math></inline-formula> is the midpoint five minutes following the <italic>k</italic>th trade. <italic>Amihud Illiquidity</italic> is defined as the absolute value of the daily return divided by dollar volume scaled by 10<sup>6</sup>. <italic>Total Volatility</italic> is measured as the 30-day rolling standard deviation of returns. <italic>Systematic Volatility</italic> is the 30-day rolling beta coefficient estimated from the capital asset pricing model (CAPM) using Center for Research in Security Prices (CRSP) value-weighted market returns. <italic>Idiosyncratic Volatility</italic> is the 30-day rolling standard deviation of the residuals obtained from the CAPM. <italic>Trading Volume</italic> is the average daily share volume. <italic>Price</italic> is the average daily CRSP closing price. <italic>MCAP</italic> is a security’s market capitalization defined as price times shares outstanding. We also report the average closing price and the average daily trading volume (number of shares).</p>
|
|
519
|
+
</table-wrap-foot></table-wrap>
|
|
520
|
+
<p>Panel B of <xref ref-type="table" rid="tbl1">Table 1</xref> reports descriptive statistics for the sample of stocks before the 2016 election. Although differences exist between the two election periods, most of the liquidity and trading measures are comparable. In fact, the median percentage effective spread in 2016 is nearly identical to that of 2012. The mean percentage quoted and effective spreads for the sample of stocks in the days before the 2016 election are 46.94 and 39.53 bps, with standard deviations of 60.17 and 57.62 bps, respectively. The average daily five-minute price impact for stocks in the 2016 election sample is 0.1541, with a standard deviation of 0.1639. The average price and market capitalization are slightly higher in the 2016 sample, $39.97 and $6,935 million, respectively.</p>
|
|
521
|
+
</sec>
|
|
522
|
+
</sec>
|
|
523
|
+
<sec id="secIII" sec-type="methods">
|
|
524
|
+
<label>III</label><title>Empirical Methodology</title>
|
|
525
|
+
<p>In our empirical analysis, we use the highly unexpected victory of Trump in the 2016 U.S. presidential election to assess the effects of political uncertainty on market liquidity. Again, this event is chosen as its outcome provides a source of political uncertainty, rather than the resolution of uncertainty. We use the 2012 U.S. presidential election as our primary benchmark, although we show that our results are robust to an alternative election. We estimate the following general fixed-effects regression equation for common stock observations in the 121 days [−90,30] surrounding both the 2012 and 2016 elections:</p><disp-formula id="eq4"><mml:math id="m21">
|
|
526
|
+
<mml:mrow>
|
|
527
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
528
|
+
<mml:mi>q</mml:mi>
|
|
529
|
+
<mml:mrow>
|
|
530
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
531
|
+
</mml:msub>
|
|
532
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
533
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
534
|
+
<mml:mrow>
|
|
535
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
536
|
+
<mml:mrow>
|
|
537
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
538
|
+
</mml:munderover >
|
|
539
|
+
<mml:mi>β</mml:mi><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
540
|
+
<mml:mi>d</mml:mi>
|
|
541
|
+
<mml:mi>t</mml:mi>
|
|
542
|
+
</mml:msub>
|
|
543
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
544
|
+
<mml:mi>φ</mml:mi>
|
|
545
|
+
<mml:mi>i</mml:mi>
|
|
546
|
+
</mml:msub>
|
|
547
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
548
|
+
<mml:mi>ε</mml:mi>
|
|
549
|
+
<mml:mrow>
|
|
550
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
551
|
+
</mml:msub>
|
|
552
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo></mml:mrow>
|
|
553
|
+
</mml:math><label>(4)</label></disp-formula><p>where the dependent variable is one of seven liquidity metrics: <italic>Quoted Spread, Effective Spread, Price Impact, Amihud Illiquidity, Total Volatility, Systematic Volatility,</italic> or <italic>Idiosyncratic Volatility</italic>. We include 10 pre-trend indicator variables for the 10 trading days leading up to the election, 31 post-trend indicator variables for the 31 trading days following the election (election day included), and stock fixed effects.<xref ref-type="fn" rid="fn16"><sup>16</sup></xref> All indicator variables are interpreted in comparison to the 80-day pre-election period. We estimate <xref ref-type="disp-formula" rid="eq4">equation (4)</xref> for each election separately and cluster standard errors at the stock level. Rather than report coefficient estimates in a table setting, we provide the plot of the beta coefficient estimates with 95% confidence intervals across the 40-day event window, covering the 10 days before and 30 days after the election dates.</p>
|
|
554
|
+
<p>To assess the statistical impact of the 2016 election on market liquidity, relative to the 2012 benchmark election, we estimate the following difference-in-difference regression equation:</p>
|
|
555
|
+
<disp-formula id="eq5"><mml:math id="m22">
|
|
556
|
+
<mml:mrow>
|
|
557
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
558
|
+
<mml:mi>q</mml:mi>
|
|
559
|
+
<mml:mrow>
|
|
560
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
561
|
+
</mml:msub>
|
|
562
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
563
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
564
|
+
<mml:mrow>
|
|
565
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
566
|
+
<mml:mrow>
|
|
567
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
568
|
+
</mml:munderover >
|
|
569
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
570
|
+
<mml:mi>d</mml:mi>
|
|
571
|
+
<mml:mi>t</mml:mi>
|
|
572
|
+
</mml:msub>
|
|
573
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
574
|
+
<mml:mi>p</mml:mi>
|
|
575
|
+
<mml:mi>j</mml:mi>
|
|
576
|
+
</mml:msub>
|
|
577
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
578
|
+
<mml:mrow>
|
|
579
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
580
|
+
<mml:mi>p</mml:mi>
|
|
581
|
+
<mml:mi>j</mml:mi>
|
|
582
|
+
</mml:msub>
|
|
583
|
+
</mml:mrow>
|
|
584
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
585
|
+
<mml:mi>φ</mml:mi>
|
|
586
|
+
<mml:mi>i</mml:mi>
|
|
587
|
+
</mml:msub>
|
|
588
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
589
|
+
<mml:mi>δ</mml:mi>
|
|
590
|
+
<mml:mi>t</mml:mi>
|
|
591
|
+
</mml:msub>
|
|
592
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
593
|
+
<mml:mi>ε</mml:mi>
|
|
594
|
+
<mml:mrow>
|
|
595
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
596
|
+
</mml:msub>
|
|
597
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
598
|
+
</mml:math><label>(5)</label></disp-formula>
|
|
599
|
+
<p>where the dependent variable is one of seven liquidity measures: <italic>Quoted Spread, Effective Spread, Price Impact, Amihud Illiquidity, Total Volatility, Systematic Volatility,</italic> or <italic>Idiosyncratic Volatility</italic>. We include 10 pre-trend interaction variables for the 10 days leading up to each election, 31 post-trend interaction variables for the 31 days following the election outcomes (election days included), time trends, level effects, and stock fixed effects. The interaction terms serve as difference-in-difference estimators between the pre- and post-election windows and the 2012 and 2016 elections. The <italic>Trump</italic> indicator variable equals 1 if the observation is during the 2016 election cycle and 0 otherwise, which controls for differences in liquidity between the two election periods. The event fixed effects, <inline-formula><mml:math id="m23">
|
|
600
|
+
<mml:mrow>
|
|
601
|
+
<mml:msub>
|
|
602
|
+
<mml:mi>δ</mml:mi>
|
|
603
|
+
<mml:mi>t</mml:mi>
|
|
604
|
+
</mml:msub>
|
|
605
|
+
</mml:mrow>
|
|
606
|
+
</mml:math></inline-formula>, control for variation in liquidity surrounding the election outcomes. The stock fixed effects, <inline-formula><mml:math id="m24">
|
|
607
|
+
<mml:mrow>
|
|
608
|
+
<mml:msub>
|
|
609
|
+
<mml:mi>φ</mml:mi>
|
|
610
|
+
<mml:mi>i</mml:mi>
|
|
611
|
+
</mml:msub>
|
|
612
|
+
</mml:mrow>
|
|
613
|
+
</mml:math></inline-formula>, control for stock-specific factors that influence liquidity. As before, we forgo traditional tables and plot the beta coefficients with 95% confidence intervals. Standard errors are clustered at the stock level to account for possible serial correlation.<xref ref-type="fn" rid="fn17"><sup>17</sup></xref></p>
|
|
614
|
+
</sec>
|
|
615
|
+
<sec id="secIV" sec-type="results">
|
|
616
|
+
<label>IV</label><title>Empirical Results</title>
|
|
617
|
+
<sec>
|
|
618
|
+
<title>Election Uncertainty and Market Liquidity</title>
|
|
619
|
+
<p>We begin by examining the effect of election uncertainty on stock transactions costs. We estimate <xref ref-type="disp-formula" rid="eq4">equations (4)</xref> and <xref ref-type="disp-formula" rid="eq5">(5)</xref> using either quoted spreads or effective spreads as the dependent variable. <xref ref-type="fig" rid="figII">Figure II</xref> plots the beta coefficient estimates, where circles are the mean estimates for the 2012 election, triangles are the mean estimates for the 2016 election, and squares are the difference-in-difference estimates. In Panel A, we find a large difference in quoted spreads following the 2016 election outcome. In economic terms, we find that the average quoted spread is 13.2 bps higher the second day following the 2016 election, relative to the 80-day pre-event period. This represents a 28% increase above the average pre-election quoted spread level. The average quoted spread remains elevated throughout the 31-day post-2016 election period. The plots for the 2012 election indicate moderate positive differences in quoted spreads in the post-event period, relative to the 80-day pre-event period. However, the differences become negative, starting two weeks after the 2012 election day. In Panel B, we provide the difference-in-difference plot estimates, where statistical significance is inferred if the “notches” associated with the box do not touch the horizontal axis. The difference-in-difference plot estimates indicate that the increase in quoted spreads around the election date are significantly higher for the 2016 election relative to the 2012 election.</p>
|
|
620
|
+
<fig id="figII" position="float" fig-type="half-right"><label>Figure II</label><caption><p><bold>Election Uncertainty and Transactions Costs.</bold> This figure reports the results of estimating the following fixed-effects regression equations for common stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m25">
|
|
621
|
+
<mml:mrow>
|
|
622
|
+
<mml:mi>S</mml:mi><mml:mi>p</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>a</mml:mi><mml:msub>
|
|
623
|
+
<mml:mi>d</mml:mi>
|
|
624
|
+
<mml:mrow>
|
|
625
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
626
|
+
</mml:msub>
|
|
627
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
628
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
629
|
+
<mml:mrow>
|
|
630
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
631
|
+
<mml:mrow>
|
|
632
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
633
|
+
</mml:munderover >
|
|
634
|
+
<mml:mi>β</mml:mi><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
635
|
+
<mml:mi>d</mml:mi>
|
|
636
|
+
<mml:mi>t</mml:mi>
|
|
637
|
+
</mml:msub>
|
|
638
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
639
|
+
<mml:mi>φ</mml:mi>
|
|
640
|
+
<mml:mi>i</mml:mi>
|
|
641
|
+
</mml:msub>
|
|
642
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
643
|
+
<mml:mi>ε</mml:mi>
|
|
644
|
+
<mml:mrow>
|
|
645
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
646
|
+
</mml:msub>
|
|
647
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
648
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m26">
|
|
649
|
+
<mml:mrow>
|
|
650
|
+
<mml:mtext>and</mml:mtext></mml:mrow>
|
|
651
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m27">
|
|
652
|
+
<mml:mrow>
|
|
653
|
+
<mml:mi>S</mml:mi><mml:mi>p</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>a</mml:mi><mml:msub>
|
|
654
|
+
<mml:mi>d</mml:mi>
|
|
655
|
+
<mml:mrow>
|
|
656
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
657
|
+
</mml:msub>
|
|
658
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
659
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
660
|
+
<mml:mrow>
|
|
661
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
662
|
+
<mml:mrow>
|
|
663
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
664
|
+
</mml:munderover >
|
|
665
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
666
|
+
<mml:mi>d</mml:mi>
|
|
667
|
+
<mml:mi>t</mml:mi>
|
|
668
|
+
</mml:msub>
|
|
669
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
670
|
+
<mml:mi>p</mml:mi>
|
|
671
|
+
<mml:mi>j</mml:mi>
|
|
672
|
+
</mml:msub>
|
|
673
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
674
|
+
<mml:mrow>
|
|
675
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
676
|
+
<mml:mi>p</mml:mi>
|
|
677
|
+
<mml:mi>j</mml:mi>
|
|
678
|
+
</mml:msub>
|
|
679
|
+
</mml:mrow>
|
|
680
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
681
|
+
<mml:mi>φ</mml:mi>
|
|
682
|
+
<mml:mi>i</mml:mi>
|
|
683
|
+
</mml:msub>
|
|
684
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
685
|
+
<mml:mi>δ</mml:mi>
|
|
686
|
+
<mml:mi>t</mml:mi>
|
|
687
|
+
</mml:msub>
|
|
688
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
689
|
+
<mml:mi>ε</mml:mi>
|
|
690
|
+
<mml:mrow>
|
|
691
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
692
|
+
</mml:msub>
|
|
693
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
694
|
+
</mml:math></inline-formula></p><p>where the dependent variable is either <italic>Quoted Spread</italic> or <italic>Effective Spread</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We estimate the first equation for each election separately and plot the beta coefficients with 95% confidence bands in Panels A and C. We include 10 pre-trend indicator variables for the 10 days leading up to each election, 31 post-trend indicator variables for the 31 days following the election outcomes (election days included), and stock fixed effects. We then estimate the second equation for the two elections jointly where we control for time trends, level effects, and stock fixed effects and plot the beta coefficients with 95% confidence intervals in Panels B and D. Standard errors in both equations are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Quoted Spread</italic> (Differences)</p><p>B. <italic>Quoted Spread</italic> (Diff-in-Diff)</p><p>C. <italic>Effective Spread</italic> (Differences)</p><p>D. <italic>Effective Spread</italic> (Diff-in-Diff)</p></caption>
|
|
695
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image003"></graphic>
|
|
696
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image004"></graphic>
|
|
697
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image005"></graphic>
|
|
698
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image006"></graphic></fig>
|
|
699
|
+
<p>In Panel C of <xref ref-type="fig" rid="figII">Figure II</xref>, we provide plot estimates for the difference in effective spreads around each election. We find an increase in transactions costs in the 10 days before the 2016 election; however, transactions costs peak the day following the election date. In fact, the average effective spread is 7.3 bps higher the second day after the 2016 election relative to the 80-day pre-event period. Economically, this represents 18.5% of the pre-election effective spread level. The average effective spread remains significantly higher in the 31-day post-event period relative to the pre-event period. The results in Panel D confirm the statistical difference-in-difference between the two elections. In sum, the evidence provided in <xref ref-type="fig" rid="figII">Figure II</xref> indicates that election uncertainty causes an increase in transactions costs.</p>
|
|
700
|
+
<p>Next, we analyze how election uncertainty affects information asymmetries in financial markets. We estimate <xref ref-type="disp-formula" rid="eq4">equations (4)</xref> and <xref ref-type="disp-formula" rid="eq5">(5)</xref> using price impact and <xref ref-type="bibr" rid="bib1">Amihud (2002)</xref> illiquidity in as the dependent variables. <xref ref-type="fig" rid="figIII">Figure III</xref> plots the beta coefficients for the 2012 election (circles), 2016 election (triangles), and difference-in-difference (squares). In Panel A, we report the pre- and post-event differences in price impact across the 2012 and 2016 elections. Consistent with the premise the election uncertainty increases informational asymmetries and results in greater adverse selection risk, we find large positive differences for the 2016 election. In economic terms, the average price impact the fifth day after the 2016 election is 5 bps higher than the 80-day pre-event average, which represents roughly a 32% increase above pre-event levels. Although the larger positive differences following the 2016 election subside in the following weeks, they remain significantly higher than the 80-day pre-event benchmark. In Panel B, we provide the plot estimates for our difference-in-difference analysis, showing that the price impact differences between the pre- and post-event periods are significantly higher in the 2016 election relative to the 2012 election. Our results also show that except for the three or four trading days in the two weeks following the election, the difference-in-difference estimates are significantly higher in the 31-day post-event period for the 2016 election.</p>
|
|
701
|
+
<fig id="figIII" position="float" fig-type="half-right"><label>Figure III</label><caption><p><bold>Election Uncertainty and Adverse Selection.</bold> This figure reports the results of estimating the following fixed-effects regression equations for stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m28">
|
|
702
|
+
<mml:mrow>
|
|
703
|
+
<mml:mi>A</mml:mi><mml:mi>d</mml:mi><mml:mi>v</mml:mi><mml:mi>e</mml:mi><mml:mi>r</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi><mml:mo> </mml:mo><mml:mi>S</mml:mi><mml:mi>e</mml:mi><mml:mi>l</mml:mi><mml:mi>e</mml:mi><mml:mi>c</mml:mi><mml:mi>t</mml:mi><mml:mi>i</mml:mi><mml:mi>o</mml:mi><mml:msub>
|
|
704
|
+
<mml:mi>n</mml:mi>
|
|
705
|
+
<mml:mrow>
|
|
706
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
707
|
+
</mml:msub>
|
|
708
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
709
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
710
|
+
<mml:mrow>
|
|
711
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
712
|
+
<mml:mrow>
|
|
713
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
714
|
+
</mml:munderover >
|
|
715
|
+
<mml:mi>β</mml:mi><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
716
|
+
<mml:mi>d</mml:mi>
|
|
717
|
+
<mml:mi>t</mml:mi>
|
|
718
|
+
</mml:msub>
|
|
719
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
720
|
+
<mml:mi>φ</mml:mi>
|
|
721
|
+
<mml:mi>i</mml:mi>
|
|
722
|
+
</mml:msub>
|
|
723
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
724
|
+
<mml:mi>ε</mml:mi>
|
|
725
|
+
<mml:mrow>
|
|
726
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
727
|
+
</mml:msub>
|
|
728
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
729
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m29">
|
|
730
|
+
<mml:mrow>
|
|
731
|
+
<mml:mtext>and</mml:mtext></mml:mrow>
|
|
732
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m30">
|
|
733
|
+
<mml:mrow>
|
|
734
|
+
<mml:mi>A</mml:mi><mml:mi>d</mml:mi><mml:mi>v</mml:mi><mml:mi>e</mml:mi><mml:mi>r</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi><mml:mo> </mml:mo><mml:mi>S</mml:mi><mml:mi>e</mml:mi><mml:mi>l</mml:mi><mml:mi>e</mml:mi><mml:mi>c</mml:mi><mml:mi>t</mml:mi><mml:mi>i</mml:mi><mml:mi>o</mml:mi><mml:msub>
|
|
735
|
+
<mml:mi>n</mml:mi>
|
|
736
|
+
<mml:mrow>
|
|
737
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
738
|
+
</mml:msub>
|
|
739
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
740
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
741
|
+
<mml:mrow>
|
|
742
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
743
|
+
<mml:mrow>
|
|
744
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
745
|
+
</mml:munderover >
|
|
746
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
747
|
+
<mml:mi>d</mml:mi>
|
|
748
|
+
<mml:mi>t</mml:mi>
|
|
749
|
+
</mml:msub>
|
|
750
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
751
|
+
<mml:mi>p</mml:mi>
|
|
752
|
+
<mml:mi>j</mml:mi>
|
|
753
|
+
</mml:msub>
|
|
754
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
755
|
+
<mml:mrow>
|
|
756
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
757
|
+
<mml:mi>p</mml:mi>
|
|
758
|
+
<mml:mi>j</mml:mi>
|
|
759
|
+
</mml:msub>
|
|
760
|
+
</mml:mrow>
|
|
761
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
762
|
+
<mml:mi>φ</mml:mi>
|
|
763
|
+
<mml:mi>i</mml:mi>
|
|
764
|
+
</mml:msub>
|
|
765
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
766
|
+
<mml:mi>δ</mml:mi>
|
|
767
|
+
<mml:mi>t</mml:mi>
|
|
768
|
+
</mml:msub>
|
|
769
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
770
|
+
<mml:mi>ε</mml:mi>
|
|
771
|
+
<mml:mrow>
|
|
772
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
773
|
+
</mml:msub>
|
|
774
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
775
|
+
</mml:math></inline-formula></p><p>where the dependent variable is either the <italic>Price Impact</italic> component of the bid–ask spread or <xref ref-type="bibr" rid="bib1">Amihud’s (2002)</xref> illiquidity measure (<italic>Amihud Illiquidity</italic>). The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We estimate the first equation for each election separately and plot the beta coefficients with 95% confidence bands in Panels A and C. We include 10 pre-trend indicator variables for the 10 days leading up to each election, 31 post-trend indicator variables for the 31 days following the election outcomes (election days included), and stock fixed effects. We then estimate the second equation for the two elections jointly where we control for time trends, level effects, and stock fixed effects and plot the beta coefficients with 95% confidence intervals in Panels B and D. Standard errors in both equations are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Price Impact</italic> (Differences)</p><p>B. <italic>Price Impact</italic> (Diff-in-Diff)</p><p>C. <italic>Amihud Illiquidity</italic> (Differences)</p><p>D. <italic>Amihud Illiquidity</italic> (Diff-in-Diff)</p></caption>
|
|
776
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image007"></graphic>
|
|
777
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image008"></graphic>
|
|
778
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image009"></graphic>
|
|
779
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image010"></graphic></fig>
|
|
780
|
+
<p>To test for robustness, we analyze whether the low-frequency proxy of price impact, <xref ref-type="bibr" rid="bib1">Amihud (2002)</xref> illiquidity, increases following the 2016 election. In Panel C of <xref ref-type="fig" rid="figIII">Figure III</xref>, we find that illiquidity spikes following both the 2012 and 2016 election days. However, the post-event differences become negative for the 2012 election, whereas they remain positive for the 2016 election. In Panel D, we show that the post-event increases in illiquidity are significantly higher for the 2016 election than for the 2012 election. Thus, the results in <xref ref-type="fig" rid="figIII">Figure III</xref> provide compelling evidence that election uncertainty associated with the 2016 election increases information asymmetries among market participants, resulting in higher adverse selection costs.</p>
|
|
781
|
+
<p>In our next set of tests, we examine whether election uncertainty and the aftermath of the 2016 election increase market volatility. We posit that election uncertainty increases informational asymmetries, creating a larger equity risk premium and greater volatility (<xref ref-type="bibr" rid="bib34">Pástor and Veronesi 2012</xref>, <xref ref-type="bibr" rid="bib35">2013</xref>). We replace the liquidity variable in <xref ref-type="disp-formula" rid="eq4">equations (4)</xref> and <xref ref-type="disp-formula" rid="eq5">(5)</xref> with one of three measures of volatility: total volatility, systematic volatility, or idiosyncratic volatility. We then plot the beta coefficients with 95% confidence intervals surrounding the elections. In Panel A of <xref ref-type="fig" rid="figIV">Figure IV</xref>, we report the pre- and post-event differences in total volatility across the 2012 and 2016 elections. We find a significant increase in total volatility surrounding the 2016 election, which peaks roughly four weeks after the outcome. In fact, average daily total volatility increases by 0.0047 at the highest point ex post, which represents a 20.5% increase above the 80-day pre-event average.</p>
|
|
782
|
+
<fig id="figIV" position="float" fig-type="half-right"><label>Figure IV</label><caption><p><bold>Election Uncertainty and Volatility.</bold> This figure reports the results of estimating the following fixed-effects regression equations for stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m31">
|
|
783
|
+
<mml:mrow>
|
|
784
|
+
<mml:mi>V</mml:mi><mml:mi>o</mml:mi><mml:mi>l</mml:mi><mml:mi>a</mml:mi><mml:mi>t</mml:mi><mml:mi>i</mml:mi><mml:mi>l</mml:mi><mml:mi>i</mml:mi><mml:mi>t</mml:mi><mml:msub>
|
|
785
|
+
<mml:mi>y</mml:mi>
|
|
786
|
+
<mml:mrow>
|
|
787
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
788
|
+
</mml:msub>
|
|
789
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
790
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
791
|
+
<mml:mrow>
|
|
792
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
793
|
+
<mml:mrow>
|
|
794
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
795
|
+
</mml:munderover >
|
|
796
|
+
<mml:mi>β</mml:mi><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
797
|
+
<mml:mi>d</mml:mi>
|
|
798
|
+
<mml:mi>t</mml:mi>
|
|
799
|
+
</mml:msub>
|
|
800
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
801
|
+
<mml:mi>φ</mml:mi>
|
|
802
|
+
<mml:mi>i</mml:mi>
|
|
803
|
+
</mml:msub>
|
|
804
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
805
|
+
<mml:mi>ε</mml:mi>
|
|
806
|
+
<mml:mrow>
|
|
807
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi></mml:mrow>
|
|
808
|
+
</mml:msub>
|
|
809
|
+
<mml:mo> </mml:mo><mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
810
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m32">
|
|
811
|
+
<mml:mrow>
|
|
812
|
+
<mml:mtext>and</mml:mtext></mml:mrow>
|
|
813
|
+
</mml:math></inline-formula></p><p><inline-formula><mml:math id="m33">
|
|
814
|
+
<mml:mrow>
|
|
815
|
+
<mml:mi>V</mml:mi><mml:mi>o</mml:mi><mml:mi>l</mml:mi><mml:mi>a</mml:mi><mml:mi>t</mml:mi><mml:mi>i</mml:mi><mml:mi>l</mml:mi><mml:mi>i</mml:mi><mml:mi>t</mml:mi><mml:msub>
|
|
816
|
+
<mml:mi>y</mml:mi>
|
|
817
|
+
<mml:mrow>
|
|
818
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
819
|
+
</mml:msub>
|
|
820
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
821
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
822
|
+
<mml:mrow>
|
|
823
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
824
|
+
<mml:mrow>
|
|
825
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
826
|
+
</mml:munderover >
|
|
827
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
828
|
+
<mml:mi>d</mml:mi>
|
|
829
|
+
<mml:mi>t</mml:mi>
|
|
830
|
+
</mml:msub>
|
|
831
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
832
|
+
<mml:mi>p</mml:mi>
|
|
833
|
+
<mml:mi>j</mml:mi>
|
|
834
|
+
</mml:msub>
|
|
835
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
836
|
+
<mml:mrow>
|
|
837
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
838
|
+
<mml:mi>p</mml:mi>
|
|
839
|
+
<mml:mi>j</mml:mi>
|
|
840
|
+
</mml:msub>
|
|
841
|
+
</mml:mrow>
|
|
842
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
843
|
+
<mml:mi>φ</mml:mi>
|
|
844
|
+
<mml:mi>i</mml:mi>
|
|
845
|
+
</mml:msub>
|
|
846
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
847
|
+
<mml:mi>δ</mml:mi>
|
|
848
|
+
<mml:mi>t</mml:mi>
|
|
849
|
+
</mml:msub>
|
|
850
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
851
|
+
<mml:mi>ε</mml:mi>
|
|
852
|
+
<mml:mrow>
|
|
853
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
854
|
+
</mml:msub>
|
|
855
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
856
|
+
</mml:math></inline-formula></p><p>where the dependent variable is either <italic>Total Volatility</italic>, <italic>Systematic Volatility</italic>, or <italic>Idiosyncratic Volatility</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We estimate the first equation for each election separately and plot the beta coefficients with 95% confidence bands in Panels A, C, and E. We include 10 pre-trend indicator variables for the 10 days leading up to each election, 31 post-trend indicator variables for the 31 days following the election outcomes (election days included), and stock fixed effects. We then estimate the second equation for the two elections jointly where we control for time trends, level effects, and stock fixed effects and plot the beta coefficients with 95% confidence intervals in Panels B, D, and F. Standard errors in both equations are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Total Volatility</italic> (Differences)</p><p>B. <italic>Total Volatility</italic> (Diff-in-Diff)</p><p>C. <italic>Systematic Volatility</italic> (Differences)</p><p>D. <italic>Systematic Volatility</italic> (Diff-in-Diff)</p><p>E. <italic>Idiosyncratic Volatility</italic> (Differences)</p><p>F. <italic>Idiosyncratic Volatility</italic> (Diff-in-Diff)</p></caption>
|
|
857
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image011"></graphic>
|
|
858
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image012"></graphic>
|
|
859
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image013"></graphic>
|
|
860
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image014"></graphic>
|
|
861
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image015"></graphic>
|
|
862
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image016"></graphic></fig>
|
|
863
|
+
<p>In contrast to the 2016 election, the difference estimates for the 2012 election are either negative or not significantly different from zero, suggesting either a decrease or no change in total volatility around the election outcome. The plot estimates in Panel B of <xref ref-type="fig" rid="figIV">Figure IV</xref> show that the differences between the two elections are significant at the .05 level. We argue that the elevated differences in total volatility displayed in Panels A and B are likely attributed to the uncertainty surrounding potential policy changes. To the extent that the 2012 election resulted in the re-election of the incumbent, investor expectations did not fluctuate as no major shift in policy changes were expected. Conversely, the surprise outcome associated with the 2016 election likely forced investors to revise their expectations about new governmental policies as a new political party was placed into power.</p>
|
|
864
|
+
<p>To identify whether the increase in total risk is market wide or firm specific, we separate volatility into systematic and idiosyncratic. The plot estimates provided in Panel C of <xref ref-type="fig" rid="figIV">Figure IV</xref> show that systematic volatility increases significantly around the 2016 election. In economic terms, beta risk increases by an average of 0.1679 at the highest point following the 2016 election, which indicates a 13.9% increase above the 80-day pre-event average. The shock to market volatility around the 2016 election disappears three weeks after the outcome. We do not find a significant increase in systematic volatility surrounding the 2012 election. The results in Panels E and F show a significant increase in idiosyncratic volatility around the 2016 election relative to the 2012 election. We find that firm-specific risk continues to increase up to a month after the 2016 election. The results in <xref ref-type="fig" rid="figIV">Figure IV</xref> have important economic implications, as investors appear exposed to multiple layers of risk during times of political unrest. Because systematic risk is not diversifiable for investors, we argue that despite proper asset allocation and diversification strategies of individual and professional traders, the election uncertainty associated with the 2016 election exposed all investors to greater risk.</p>
|
|
865
|
+
</sec>
|
|
866
|
+
<sec>
|
|
867
|
+
<title>Election Uncertainty and Market Liquidity—Stock Characteristics</title>
|
|
868
|
+
<p>In this next section, we examine whether certain stocks are more, or less, exposed to election uncertainty. We believe this provides a direct test of the information asymmetry hypothesis that asserts that the information gap between informed and uninformed investors widens during times of political unrest. For instance, both smaller capitalized stocks and those with lower trading volume are likely to possess greater firm-specific informational asymmetry and adverse selection risk.<xref ref-type="fn" rid="fn18"><sup>18</sup></xref> Thus, we examine the effects of the two presidential elections on market liquidity across varying stock-level characteristics.</p>
|
|
869
|
+
<p>We estimate <xref ref-type="disp-formula" rid="eq5">equation (5)</xref> for various liquidity measures and for stocks in the top and bottom quintiles of average daily trading volume during the pre-election period. In Panel A of <xref ref-type="fig" rid="figV">Figure V</xref>, we analyze whether the positive difference-in-differences in round-trip transaction costs are higher among low-volume stocks. In fact, we find no significant change in either effective spread or price impact measures for high-volume stocks. Consistent with the premise that election uncertainty has a greater effect on stocks more susceptible to information asymmetries, we find that differences in round-trip costs are higher for less frequently traded stocks. In Panel B, we show that increases in adverse selection risks are also higher among less actively traded securities. In Panels C and D, we find that the increase in average volatility around the 2016 election, relative to the 2012 election, is generally higher for low-volume stocks.</p>
|
|
870
|
+
<fig id="figV" position="float" fig-type="half-right"><label>Figure V</label><caption><p><bold>Election Uncertainty and Liquidity—By Stock Trading Volume.</bold> This figure reports the results of estimating specifications of the following fixed-effects regression equation for stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m34">
|
|
871
|
+
<mml:mrow>
|
|
872
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
873
|
+
<mml:mi>q</mml:mi>
|
|
874
|
+
<mml:mrow>
|
|
875
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
876
|
+
</mml:msub>
|
|
877
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
878
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
879
|
+
<mml:mrow>
|
|
880
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
881
|
+
<mml:mrow>
|
|
882
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
883
|
+
</mml:munderover >
|
|
884
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
885
|
+
<mml:mi>d</mml:mi>
|
|
886
|
+
<mml:mi>t</mml:mi>
|
|
887
|
+
</mml:msub>
|
|
888
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
889
|
+
<mml:mi>p</mml:mi>
|
|
890
|
+
<mml:mi>j</mml:mi>
|
|
891
|
+
</mml:msub>
|
|
892
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
893
|
+
<mml:mrow>
|
|
894
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
895
|
+
<mml:mi>p</mml:mi>
|
|
896
|
+
<mml:mi>j</mml:mi>
|
|
897
|
+
</mml:msub>
|
|
898
|
+
</mml:mrow>
|
|
899
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
900
|
+
<mml:mi>φ</mml:mi>
|
|
901
|
+
<mml:mi>i</mml:mi>
|
|
902
|
+
</mml:msub>
|
|
903
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
904
|
+
<mml:mi>δ</mml:mi>
|
|
905
|
+
<mml:mi>t</mml:mi>
|
|
906
|
+
</mml:msub>
|
|
907
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
908
|
+
<mml:mi>ε</mml:mi>
|
|
909
|
+
<mml:mrow>
|
|
910
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
911
|
+
</mml:msub>
|
|
912
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
913
|
+
</mml:math></inline-formula></p><p>where the dependent variable is one of four liquidity measures: <italic>Effective Spread</italic>, <italic>Price Impact</italic>, <italic>Systematic Volatility</italic>, or <italic>Idiosyncratic Volatility</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We separate the Trump sample into pre-election trading volume quintiles and estimate the equation above separately for the top and bottom quintiles. We include 10 pre-trend interaction variables for the 10 days leading up to each election, 31 post-trend interaction variables for the 31 days following the election outcomes (election days included), time trends, level effects, and stock fixed effects. We plot the beta coefficients with 95% confidence intervals below. Standard errors are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Effective Spread</italic> (Diff-in-Diff)</p><p>B. <italic>Price Impact</italic> (Diff-in-Diff)</p><p>C. <italic>Systematic Volatility</italic> (Diff-in-Diff)</p><p>Panel D. <italic>Idiosyncratic Volatility</italic> (Diff-in-Diff)</p></caption>
|
|
914
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image017"></graphic>
|
|
915
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image018"></graphic>
|
|
916
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image019"></graphic>
|
|
917
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image020"></graphic></fig>
|
|
918
|
+
<p>Next, we estimate <xref ref-type="disp-formula" rid="eq5">equation (5)</xref> for stocks in the top and bottom quintiles of average daily market capitalization ex ante. We posit that small-cap stocks likely experience greater fluctuations in liquidity and volatility during times of election uncertainty. In Panel A of <xref ref-type="fig" rid="figVI">Figure VI</xref>, we plot the difference-in-difference estimates in effective spreads for large- and small-cap stocks. We find that the increase in transactions costs around the 2016 election, relative to the 2012 election, is isolated to small-cap securities. We report similar findings in Panel B for price impact. Though to a lesser extent, the results in Panels C and D provide corroborating evidence that small-cap stocks are more exposed to election uncertainty, resulting in greater volatility.</p>
|
|
919
|
+
<fig id="figVI" position="float" fig-type="half-right"><label>Figure VI</label><caption><p><bold>Election Uncertainty and Liquidity—By Stock Market Capitalization.</bold> This figure reports the results of estimating specifications of the following fixed-effects regression equation for stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m35">
|
|
920
|
+
<mml:mrow>
|
|
921
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
922
|
+
<mml:mi>q</mml:mi>
|
|
923
|
+
<mml:mrow>
|
|
924
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
925
|
+
</mml:msub>
|
|
926
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
927
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
928
|
+
<mml:mrow>
|
|
929
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
930
|
+
<mml:mrow>
|
|
931
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
932
|
+
</mml:munderover >
|
|
933
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
934
|
+
<mml:mi>d</mml:mi>
|
|
935
|
+
<mml:mi>t</mml:mi>
|
|
936
|
+
</mml:msub>
|
|
937
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
938
|
+
<mml:mi>p</mml:mi>
|
|
939
|
+
<mml:mi>j</mml:mi>
|
|
940
|
+
</mml:msub>
|
|
941
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
942
|
+
<mml:mrow>
|
|
943
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
944
|
+
<mml:mi>p</mml:mi>
|
|
945
|
+
<mml:mi>j</mml:mi>
|
|
946
|
+
</mml:msub>
|
|
947
|
+
</mml:mrow>
|
|
948
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
949
|
+
<mml:mi>φ</mml:mi>
|
|
950
|
+
<mml:mi>i</mml:mi>
|
|
951
|
+
</mml:msub>
|
|
952
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
953
|
+
<mml:mi>δ</mml:mi>
|
|
954
|
+
<mml:mi>t</mml:mi>
|
|
955
|
+
</mml:msub>
|
|
956
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
957
|
+
<mml:mi>ε</mml:mi>
|
|
958
|
+
<mml:mrow>
|
|
959
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
960
|
+
</mml:msub>
|
|
961
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
962
|
+
</mml:math></inline-formula></p><p>where the dependent variable is one of four liquidity measures: <italic>Effective Spread</italic>, <italic>Price Impact</italic>, <italic>Systematic Volatility</italic>, or <italic>Idiosyncratic Volatility</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We separate the Trump sample into pre-election market capitalization quintiles and estimate the above equation separately for the top and bottom quintiles. We include 10 pre-trend interaction variables for the 10 days leading up to each election, 31 post-trend interaction variables for the 31 days following the election outcomes (election days included), time trends, level effects, and stock fixed effects. We plot the beta coefficients with 95% confidence intervals below. Standard errors are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Effective Spread</italic> (Diff-in-Diff)</p><p>B. <italic>Price Impact</italic> (Diff-in-Diff)</p><p>C. <italic>Systematic Volatility</italic> (Diff-in-Diff)</p><p>Panel D. <italic>Idiosyncratic Volatility</italic> (Diff-in-Diff)</p></caption>
|
|
963
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image021"></graphic>
|
|
964
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image022"></graphic>
|
|
965
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image023"></graphic>
|
|
966
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image024"></graphic></fig>
|
|
967
|
+
<p>In summary, we find that the surprise victory of Trump in 2016 reduced market liquidity for less frequently traded small-cap stocks, consistent with the notion that stock-level characteristics associated with greater adverse selection risk are an important consideration in determining how political uncertainty can affect financial market outcomes.</p>
|
|
968
|
+
</sec>
|
|
969
|
+
<sec>
|
|
970
|
+
<title>Election Uncertainty and Market Liquidity—Politically Sensitive Industries</title>
|
|
971
|
+
<p>Several studies suggest that political uncertainty can have more severe effects on stocks operating in politically sensitive industries given their increased exposure to potential policy changes (Boutchkova, Doshi, and Durnev 2011; <xref ref-type="bibr" rid="bib20">Julio and Yook 2012</xref>; <xref ref-type="bibr" rid="bib19">Hong and Kostovetsky 2012</xref>). We next examine the impact on stocks operating in politically sensitive industries. According to a report published one month before the 2016 election, several industries (i.e., healthcare, financials, energy, and infrastructure) were identified as being the most sensitive to potential policy changes.<xref ref-type="fn" rid="fn19"><sup>19</sup></xref> For example, the financial sector was considered “highly sensitive” given that the two leading candidates had distinct perspectives on bank regulations and support for the Dodd–Frank Act, and the healthcare industry was considered more sensitive because of the candidates’ opposing views on the Affordable Care Act. In our analysis, we classify a stock as politically sensitive if it falls into the healthcare (Standard Industrial Classification [SIC] 8000–8099), pharmaceutical (SIC 2830–2836), banking (SIC 6000–6199), trading (SIC 6200–6299 and 6700–6799), natural gas (SIC 1300–1389, 2900–2912, and 2990–2999), defense (SIC 3760–3769, 3795, and 3480–3489), or tobacco (SIC 2100–2199) industry.</p>
|
|
972
|
+
<p>We estimate <xref ref-type="disp-formula" rid="eq5">equation (5)</xref> for various liquidity measures and for politically sensitive and “nonsensitive” stocks. The results are reported in <xref ref-type="fig" rid="figVII">Figure VII</xref>, where we plot the beta coefficients with 95% confidence intervals. In Panel A, we plot difference-in-difference estimates for average effective spreads and find that stocks operating in politically sensitive industries are more exposed to greater election uncertainty. Specifically, we find that differences in round-trip costs are generally higher among stocks in politically sensitive industries. In Panel B, we show that the difference-in-difference estimates in adverse selection costs, measured as percentage price impact, are higher among stocks in politically sensitive industries. In Panels C and D, we show that difference-in-difference estimates in systematic and idiosyncratic volatility are higher among firms in politically sensitive industries. In fact, the difference in firm-specific risk increases over the post-election period.</p>
|
|
973
|
+
<fig id="figVII" position="float" fig-type="half-right"><label>Figure VII</label><caption><p><bold>Election Uncertainty and Liquidity—Politically Sensitive Industries.</bold> The figure reports the results of estimating specifications of the following fixed-effects regression equation for stock observations in the 121 days [−90,30] surrounding the 2012 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m36">
|
|
974
|
+
<mml:mrow>
|
|
975
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
976
|
+
<mml:mi>q</mml:mi>
|
|
977
|
+
<mml:mrow>
|
|
978
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
979
|
+
</mml:msub>
|
|
980
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
981
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
982
|
+
<mml:mrow>
|
|
983
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
984
|
+
<mml:mrow>
|
|
985
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
986
|
+
</mml:munderover >
|
|
987
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
988
|
+
<mml:mi>d</mml:mi>
|
|
989
|
+
<mml:mi>t</mml:mi>
|
|
990
|
+
</mml:msub>
|
|
991
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
992
|
+
<mml:mi>p</mml:mi>
|
|
993
|
+
<mml:mi>j</mml:mi>
|
|
994
|
+
</mml:msub>
|
|
995
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
996
|
+
<mml:mrow>
|
|
997
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
998
|
+
<mml:mi>p</mml:mi>
|
|
999
|
+
<mml:mi>j</mml:mi>
|
|
1000
|
+
</mml:msub>
|
|
1001
|
+
</mml:mrow>
|
|
1002
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
1003
|
+
<mml:mi>φ</mml:mi>
|
|
1004
|
+
<mml:mi>i</mml:mi>
|
|
1005
|
+
</mml:msub>
|
|
1006
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
1007
|
+
<mml:mi>δ</mml:mi>
|
|
1008
|
+
<mml:mi>t</mml:mi>
|
|
1009
|
+
</mml:msub>
|
|
1010
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
1011
|
+
<mml:mi>ε</mml:mi>
|
|
1012
|
+
<mml:mrow>
|
|
1013
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
1014
|
+
</mml:msub>
|
|
1015
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
1016
|
+
</mml:math></inline-formula></p><p>where the dependent variable is one of four liquidity measures: <italic>Effective Spread</italic>, <italic>Price Impact</italic>, <italic>Systematic Volatility</italic>, or <italic>Idiosyncratic Volatility</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We separate the sample by industry and define those in the healthcare (Standard Industrial Classification [SIC] 8000–8099), pharmaceutical (SIC 2830–2836), banking (SIC 6000–6199), trading (SIC 6200–6299 and 6700–6799), natural gas (SIC 1300–1389, 2900–2912, and 2990–2999), defense (SIC 3760–3769, 3795, and 3480–3489), and tobacco (SIC 2100–2199) industries as politically sensitive. We estimate the equation above separately for politically sensitive and insensitive stocks. We include 10 pre-trend interaction variables for the 10 days leading up to each election, 31 post-trend interaction variables for the 31 days following the election outcomes (election days included), time trends, level effects, and stock fixed effects. We plot the beta coefficients with 95% confidence intervals below. Standard errors are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Effective Spread</italic> (Diff-in-Diff)</p><p>B. <italic>Price Impact</italic> (Diff-in-Diff)</p><p>C. <italic>Systematic Volatility</italic> (Diff-in-Diff)</p><p>D. <italic>Idiosyncratic Volatility</italic> (Diff-in-Diff)</p></caption>
|
|
1017
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image025"></graphic>
|
|
1018
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image026"></graphic>
|
|
1019
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image027"></graphic>
|
|
1020
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image028"></graphic></fig>
|
|
1021
|
+
</sec>
|
|
1022
|
+
</sec>
|
|
1023
|
+
<sec id="secV" sec-type="other2">
|
|
1024
|
+
<label>V</label><title>Robustness</title>
|
|
1025
|
+
<p>To this point, our results indicate that relative to the 2012 U.S. presidential election, the 2016 election induced, rather than resolved, market uncertainty, resulting in deteriorating market liquidity. We also account for an alternative benchmark election. Our intraday data allow us to use both the 2004 and 2008 U.S. presidential elections as comparable events. However, the confounding effects associated with the financial crisis, short-selling ban, and government bailouts surrounding the 2008 U.S. presidential election hinder our ability to make any proper statistical inferences. Therefore, we exclude the 2008 election from our analysis and include the 2004 U.S. presidential election as an additional benchmark. The 2004 U.S. presidential election is a unique benchmark given that the race was tightly contested yet the winning candidate was an incumbent, presumably resulting in less uncertainty surrounding future policy changes.</p>
|
|
1026
|
+
<p>In <xref ref-type="fig" rid="figVIII">Figure VIII</xref>, we plot the difference-in-difference estimates comparing the pre- and post-event differences in effective spreads, price impact, and volatility across both the 2004 and 2016 U.S. presidential elections. Consistent with our earlier findings using the 2012 U.S. presidential election as a benchmark, we find that the election uncertainty associated with the 2016 election resulted in greater illiquidity relative to the changes experienced in the 2004 election. In fact, the difference-in-difference estimates provided in Panel A indicate that the higher transaction costs remained elevated throughout the entire post-event period. We provide similar findings in Panels B–D for our measures of price impact and volatility. Hence, we suggest that the surprise outcome of the 2016 U.S. presidential election worsened market liquidity more than the two U.S. presidential election benchmarks. Furthermore, our results indicate that the effects of the 2016 U.S. presidential election were not transitory and, in fact, appear to have increased in the subsequent 30 trading days.</p>
|
|
1027
|
+
<fig id="figVIII" position="float" fig-type="half-right"><label>Figure VIII</label><caption><p><bold>Election Uncertainty and Liquidity—2004 U.S. Presidential Election as a Benchmark.</bold> This figure reports the results of estimating specifications of the following fixed-effects regression equation for stock observations in the 121 days [−90,30] surrounding the 2004 and 2016 U.S. presidential elections:</p><p><inline-formula><mml:math id="m37">
|
|
1028
|
+
<mml:mrow>
|
|
1029
|
+
<mml:mi>L</mml:mi><mml:mi>i</mml:mi><mml:msub>
|
|
1030
|
+
<mml:mi>q</mml:mi>
|
|
1031
|
+
<mml:mrow>
|
|
1032
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
1033
|
+
</mml:msub>
|
|
1034
|
+
<mml:mo>=</mml:mo><mml:mi>α</mml:mi><mml:mo>+</mml:mo><mml:munderover>
|
|
1035
|
+
<mml:mstyle mathsize="140%" displaystyle="true"><mml:mo>∑</mml:mo></mml:mstyle>
|
|
1036
|
+
<mml:mrow>
|
|
1037
|
+
<mml:mi>t</mml:mi><mml:mo>=</mml:mo><mml:mo>−</mml:mo><mml:mn>10</mml:mn></mml:mrow>
|
|
1038
|
+
<mml:mrow>
|
|
1039
|
+
<mml:mn>30</mml:mn></mml:mrow>
|
|
1040
|
+
</mml:munderover >
|
|
1041
|
+
<mml:mi>β</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>n</mml:mi><mml:msub>
|
|
1042
|
+
<mml:mi>d</mml:mi>
|
|
1043
|
+
<mml:mi>t</mml:mi>
|
|
1044
|
+
</mml:msub>
|
|
1045
|
+
<mml:mo>×</mml:mo><mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
1046
|
+
<mml:mi>p</mml:mi>
|
|
1047
|
+
<mml:mi>j</mml:mi>
|
|
1048
|
+
</mml:msub>
|
|
1049
|
+
<mml:mo stretchy="false">)</mml:mo><mml:mo>+</mml:mo><mml:mi>τ</mml:mi><mml:mfenced>
|
|
1050
|
+
<mml:mrow>
|
|
1051
|
+
<mml:mi>T</mml:mi><mml:mi>r</mml:mi><mml:mi>u</mml:mi><mml:mi>m</mml:mi><mml:msub>
|
|
1052
|
+
<mml:mi>p</mml:mi>
|
|
1053
|
+
<mml:mi>j</mml:mi>
|
|
1054
|
+
</mml:msub>
|
|
1055
|
+
</mml:mrow>
|
|
1056
|
+
</mml:mfenced><mml:mo>+</mml:mo><mml:msub>
|
|
1057
|
+
<mml:mi>φ</mml:mi>
|
|
1058
|
+
<mml:mi>i</mml:mi>
|
|
1059
|
+
</mml:msub>
|
|
1060
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
1061
|
+
<mml:mi>δ</mml:mi>
|
|
1062
|
+
<mml:mi>t</mml:mi>
|
|
1063
|
+
</mml:msub>
|
|
1064
|
+
<mml:mo>+</mml:mo><mml:msub>
|
|
1065
|
+
<mml:mi>ε</mml:mi>
|
|
1066
|
+
<mml:mrow>
|
|
1067
|
+
<mml:mi>i</mml:mi><mml:mo>,</mml:mo><mml:mi>t</mml:mi><mml:mo>,</mml:mo><mml:mi>j</mml:mi></mml:mrow>
|
|
1068
|
+
</mml:msub>
|
|
1069
|
+
<mml:mo>,</mml:mo><mml:mo> </mml:mo><mml:mo> </mml:mo></mml:mrow>
|
|
1070
|
+
</mml:math></inline-formula></p><p>where the dependent variable is one of four liquidity measures: <italic>Effective Spread</italic>, <italic>Price Impact</italic>, <italic>Systematic Volatility</italic>, or <italic>Idiosyncratic Volatility</italic>. The variables are defined in <xref ref-type="table" rid="tbl1">Table 1</xref>. We include 10 pre-trend interaction variables for the 10 days leading up to each election, 31 post-trend interaction variables for the 31 days following the election outcomes (election days included), time trends, level effects, and stock fixed effects. We plot the beta coefficients with 95% confidence intervals below. Standard errors are clustered at the stock level. [Color figure can be viewed at wileyonlinelibrary.com.]</p><p>A. <italic>Effective Spread</italic> (Diff-in-Diff)</p><p>B. <italic>Price Impact</italic> (Diff-in-Diff)</p><p>C. <italic>Systematic Volatility</italic> (Diff-in-Diff)</p><p>D. <italic>Idiosyncratic Volatility</italic> (Diff-in-Diff)</p></caption>
|
|
1071
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image029"></graphic>
|
|
1072
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image030"></graphic>
|
|
1073
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image031"></graphic>
|
|
1074
|
+
<graphic xlink:href="Wiley-JFR-Final Manuscript-MT-Image032"></graphic></fig>
|
|
1075
|
+
</sec>
|
|
1076
|
+
<sec id="secVI" sec-type="conclusions">
|
|
1077
|
+
<label>VI</label><title>Conclusion</title>
|
|
1078
|
+
<p>This study contributes to the growing literature investigating the relation between politics and financial markets. We examine whether the recent surprise outcome of the 2016 U.S. presidential election induced, or resolved, uncertainty in financial markets. Using two U.S. presidential elections (2004 and 2012) as comparable benchmarks, we provide evidence that the uncertainty from the 2016 U.S. presidential election increased information asymmetries among market participants, which reduced liquidity. Specifically, we find that post-event liquidity following the 2016 U.S. presidential election worsened with respect to transaction costs, adverse selection costs, and volatility, relative to previous elections.</p>
|
|
1079
|
+
<p>From an economic perspective, our results suggest that election uncertainty can have significant consequences for market participants. First, an increase in transactions costs affects investors’ net returns and may deter future market participation. Second, an increase in price impact suggests a widening of the information gap between uninformed and informed traders. Third, shocks to both idiosyncratic and systematic volatility expose investors to multiple layers of risk, regardless of their diversification strategies. Overall, our article contributes to the literature in presenting evidence that not all election outcomes are created equal and that election uncertainty can have severe and persistent consequences for financial market participants. We argue that our findings are useful to investors attempting to diversify and hedge against potential shocks to political and policy uncertainty.</p>
|
|
1080
|
+
</sec>
|
|
1081
|
+
</body>
|
|
1082
|
+
<back>
|
|
1083
|
+
<app-group>
|
|
1084
|
+
<app id="app"><label>Appendix</label>
|
|
1085
|
+
<table-wrap id="tblA.1" position="float">
|
|
1086
|
+
<label>Table A1</label><caption><title><?AuthorQuery id="Q12" queryProcess="eXtyles Citation Match Check" queryText="The table "Table A1" is not cited in the text. Please add an in-text citation or delete the table."?>Difference between exporting vs. non-exporting firms</title>
|
|
1087
|
+
</caption>
|
|
1088
|
+
<table frame="hsides" rules="groups">
|
|
1089
|
+
<col width="49.98%" />
|
|
1090
|
+
<col width="14.26%" />
|
|
1091
|
+
<col width="17.19%" />
|
|
1092
|
+
<col width="18.57%" />
|
|
1093
|
+
<thead>
|
|
1094
|
+
<tr>
|
|
1095
|
+
<th valign="top" align="left" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt" />
|
|
1096
|
+
<th valign="top" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><bold>Exporter</bold></th>
|
|
1097
|
+
<th valign="top" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><bold>EU exporter</bold></th>
|
|
1098
|
+
<th valign="top" align="center" scope="col" style="border-top: solid 0.50pt; border-bottom: solid 0.50pt"><bold>HIC exporter</bold></th>
|
|
1099
|
+
</tr>
|
|
1100
|
+
</thead>
|
|
1101
|
+
<tbody>
|
|
1102
|
+
<tr>
|
|
1103
|
+
<td valign="top" align="left" style="border-top: solid 0.50pt" scope="row">Average number of employed (log)</td>
|
|
1104
|
+
<td valign="top" align="center" style="border-top: solid 0.50pt">1.4904***<break />(.0107)</td>
|
|
1105
|
+
<td valign="top" align="center" style="border-top: solid 0.50pt">.6878***<break />(.0225)</td>
|
|
1106
|
+
<td valign="top" align="center" style="border-top: solid 0.50pt">.6473***<break />(.0222)</td>
|
|
1107
|
+
</tr>
|
|
1108
|
+
<tr>
|
|
1109
|
+
<td valign="top" align="left" scope="row">Average number of female employed (log)</td>
|
|
1110
|
+
<td valign="top" align="center">1.0341***<break />(.0104)</td>
|
|
1111
|
+
<td valign="top" align="center">.6266***<break />(.0220)</td>
|
|
1112
|
+
<td valign="top" align="center">.5870***<break />(.0219)</td>
|
|
1113
|
+
</tr>
|
|
1114
|
+
<tr>
|
|
1115
|
+
<td valign="top" align="left" scope="row">Female employment share</td>
|
|
1116
|
+
<td valign="top" align="center">-.1245***<break />(.0023)</td>
|
|
1117
|
+
<td valign="top" align="center">.0271***<break />(.0036)</td>
|
|
1118
|
+
<td valign="top" align="center">.0355***<break />(.0036)</td>
|
|
1119
|
+
</tr>
|
|
1120
|
+
<tr>
|
|
1121
|
+
<td valign="top" align="left" scope="row">Average wages (log)</td>
|
|
1122
|
+
<td valign="top" align="center">.9177***<break />(.0089)</td>
|
|
1123
|
+
<td valign="top" align="center">.2975***<break />(.0138)</td>
|
|
1124
|
+
<td valign="top" align="center">.2755***<break />(.0136)</td>
|
|
1125
|
+
</tr>
|
|
1126
|
+
<tr>
|
|
1127
|
+
<td valign="top" align="left" scope="row">Average female wages (log)</td>
|
|
1128
|
+
<td valign="top" align="center">.8062***<break />(.0089)</td>
|
|
1129
|
+
<td valign="top" align="center">.2322***<break />(.0142)</td>
|
|
1130
|
+
<td valign="top" align="center">.2103***<break />(.0141)</td>
|
|
1131
|
+
</tr>
|
|
1132
|
+
<tr>
|
|
1133
|
+
<td valign="top" align="left" scope="row">Female-to-male wage ratio</td>
|
|
1134
|
+
<td valign="top" align="center">-.0572***<break />(.0122)</td>
|
|
1135
|
+
<td valign="top" align="center">-.0316<break />(.0194)</td>
|
|
1136
|
+
<td valign="top" align="center">-.0488**<break />(.0192)</td>
|
|
1137
|
+
</tr>
|
|
1138
|
+
<tr>
|
|
1139
|
+
<td valign="top" align="left" style="border-bottom: solid 0.50pt" scope="row">Average turnover (log)</td>
|
|
1140
|
+
<td valign="top" align="center" style="border-bottom: solid 0.50pt">2.5000***<break />(.0171)</td>
|
|
1141
|
+
<td valign="top" align="center" style="border-bottom: solid 0.50pt">.9636***<break />(.0292)</td>
|
|
1142
|
+
<td valign="top" align="center" style="border-bottom: solid 0.50pt">.9103***<break />(.0288)</td>
|
|
1143
|
+
</tr>
|
|
1144
|
+
</tbody>
|
|
1145
|
+
</table><table-wrap-foot>
|
|
1146
|
+
<p>Source: GEOSTAT Statistics Survey of Enterprises for 2006-2017.</p>
|
|
1147
|
+
<p>Notes: HIC = high income country. Exporters are firms that exported in at least 1 year of the sample. EU / HIC exporters are firms that exported to the EU / high income country in at least 1 year of the sample. Column 2: difference in means in log workers, log wages, log turnover between exporters and non-exporters, controlling for year and 1-digit industry. Column 3: difference in means between firms that export to at least one EU destination and non-EU exporters, controlling for year and 1-digit industry. Column 4: difference in means between firms that export to at least one HIC destination and non-HIC exporters, controlling for year and 1-digit industry. ***Significant at the 1% level, **Significant at the 5% level, *Significant at the 10% level.</p>
|
|
1148
|
+
</table-wrap-foot></table-wrap>
|
|
1149
|
+
</app>
|
|
1150
|
+
</app-group>
|
|
1151
|
+
<ref-list>
|
|
1152
|
+
<title>References</title>
|
|
1153
|
+
<ref id="bib1"><mixed-citation publication-type="journal"><?AuthorQuery id="Q1" queryText="eXtyles test comment"?><person-group person-group-type="author"><string-name><surname>Amihud</surname>, <given-names>Y.</given-names></string-name></person-group>, <year>2002</year>, <article-title>Illiquidity and stock returns: Cross-section and time-series effects</article-title>, <source>Journal of Financial Markets</source> <volume>5</volume>, <fpage>31</fpage>–<lpage>56</lpage>.</mixed-citation></ref>
|
|
1154
|
+
<ref id="bib2"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Amihud</surname>, <given-names>Y.</given-names></string-name>, and <string-name><given-names>H.</given-names> <surname>Mendelson</surname></string-name></person-group>, <year>1986</year>, <article-title>Asset pricing and the bid-ask spread</article-title>, <source>Journal of Financial Economics</source> <volume>17</volume>, <fpage>223</fpage>–<lpage>49</lpage>.</mixed-citation></ref>
|
|
1155
|
+
<ref id="bib3"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Baker</surname>, <given-names>S. R.</given-names></string-name>, <string-name><given-names>N.</given-names> <surname>Bloom</surname></string-name>, and <string-name><given-names>S. J.</given-names> <surname>Davis</surname></string-name></person-group>, <year>2016</year>, <article-title>Measuring economic policy uncertainty</article-title>, <source>Quarterly Journal of Economics</source> <volume>131</volume>, <fpage>1593</fpage>–<lpage>1636</lpage>.</mixed-citation></ref>
|
|
1156
|
+
<ref id="bib4"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Barry</surname>, <given-names>C. B.</given-names></string-name>, and <string-name><given-names>S. J.</given-names> <surname>Brown</surname></string-name></person-group>, <year>1984</year>, <article-title>Differential information and the small firm effect</article-title>, <source>Journal of Financial Economics</source> <volume>13</volume>, <fpage>283</fpage>–<lpage>94</lpage>.</mixed-citation></ref>
|
|
1157
|
+
<ref id="bib5"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Bhattacharya</surname>, <given-names>U.</given-names></string-name>, <string-name><given-names>P. H.</given-names> <surname>Hsu</surname></string-name>, <string-name><given-names>X.</given-names> <surname>Tian</surname></string-name>, and <string-name><given-names>Y.</given-names> <surname>Xu</surname></string-name></person-group>, <year>2017</year>, <article-title>What affects innovation more: Policy or policy uncertainty?</article-title> <source>Journal of Financial and Quantitative Analysis</source> <volume>52</volume>, <fpage>1869</fpage>–<lpage>1901</lpage>.</mixed-citation></ref>
|
|
1158
|
+
<ref id="bib6"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Białkowski</surname>, <given-names>J.</given-names></string-name>, <string-name><given-names>K.</given-names> <surname>Gottschalk</surname></string-name>, and <string-name><given-names>T. P.</given-names> <surname>Wisniewski</surname></string-name></person-group>, <year>2008</year>, <article-title>Stock market volatility around national elections</article-title>, <source>Journal of Banking & Finance</source> <volume>32</volume>, <fpage>1941</fpage>–<lpage>53</lpage>.</mixed-citation></ref>
|
|
1159
|
+
<ref id="bib7"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Bonaparte</surname>, <given-names>Y.</given-names></string-name>, <string-name><given-names>A.</given-names> <surname>Kumar</surname></string-name>, and <string-name><given-names>J. K.</given-names> <surname>Page</surname></string-name></person-group>, <year>2017</year>, <article-title>Political climate, optimism, and investment decisions</article-title>, <source>Journal of Financial Markets</source> <volume>34</volume>, <fpage>69</fpage>–<lpage>94</lpage>.</mixed-citation></ref>
|
|
1160
|
+
<ref id="bib8"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Boutchkova</surname>, <given-names>M.</given-names></string-name>, <string-name><given-names>H.</given-names> <surname>Doshi</surname></string-name>, <string-name><given-names>A.</given-names> <surname>Durnev</surname></string-name>, and <string-name><given-names>A.</given-names> <surname>Molchanov</surname></string-name></person-group>, <year>2011</year>. <article-title>Precarious politics and return volatility.</article-title> <source>Review of Financial Studies</source> <volume>25</volume>, <fpage>1111</fpage>-<lpage>1154</lpage>.</mixed-citation></ref>
|
|
1161
|
+
<ref id="bib9"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Boutchkova</surname>, <given-names>M.</given-names></string-name>, <string-name><given-names>H.</given-names> <surname>Doshi</surname></string-name>, <string-name><given-names>A.</given-names> <surname>Durnev</surname></string-name>, and <string-name><given-names>A.</given-names> <surname>Molchanov</surname></string-name></person-group>, <year>2012</year>, <article-title>Precarious politics and return volatility</article-title>, <source>Review of Financial Studies</source> <volume>25</volume>, <fpage>1111</fpage>–<lpage>54</lpage>.</mixed-citation></ref>
|
|
1162
|
+
<ref id="bib10"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Brogaard</surname>, <given-names>J.</given-names></string-name>, and <string-name><given-names>A.</given-names> <surname>Detzel</surname></string-name></person-group>, <year>2015</year>, <article-title>The asset-pricing implications of government economic policy uncertainty</article-title>, <source>Management Science</source> <volume>61</volume>, <fpage>3</fpage>–<lpage>18</lpage>.</mixed-citation></ref>
|
|
1163
|
+
<ref id="bib11"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Brown</surname>, <given-names>K. C.</given-names></string-name>, <string-name><given-names>W. V.</given-names> <surname>Harlow</surname></string-name>, and <string-name><given-names>S. M.</given-names> <surname>Tinic</surname></string-name></person-group>, <year>1988</year>, <article-title>Risk aversion, uncertain information, and market efficiency</article-title>, <source>Journal of Financial Economics</source> <volume>22</volume>, <fpage>355</fpage>–<lpage>85</lpage>.</mixed-citation></ref>
|
|
1164
|
+
<ref id="bib12"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Easley</surname>, <given-names>D.</given-names></string-name>, and <string-name><given-names>M.</given-names> <surname>O’Hara</surname></string-name></person-group>, <year>1987</year>, <article-title>Price, trade size, and information in securities markets</article-title>, <source>Journal of Financial Economics</source> <volume>19</volume>, <fpage>69</fpage>–<lpage>90</lpage>.</mixed-citation></ref>
|
|
1165
|
+
<ref id="bib13"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Gemmill</surname>, <given-names>G.</given-names></string-name></person-group>, <year>1992</year>, <article-title>Political risk and market efficiency: Tests based in British stock and options markets in the 1987 election</article-title>, <source>Journal of Banking & Finance</source> <volume>16</volume>, <fpage>211</fpage>–<lpage>31</lpage>.</mixed-citation></ref>
|
|
1166
|
+
<ref id="bib14"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Goodell</surname>, <given-names>J. W.</given-names></string-name>, and <string-name><given-names>S.</given-names> <surname>Vähämaa</surname></string-name></person-group>, <year>2013</year>, <article-title>US presidential elections and implied volatility: The role of political uncertainty</article-title>, <source>Journal of Banking & Finance</source> <volume>37</volume>, <fpage>1108</fpage>–<lpage>17</lpage>.</mixed-citation></ref>
|
|
1167
|
+
<ref id="bib15"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Gungoraydinoglu</surname>, <given-names>A.</given-names></string-name>, <string-name><given-names>G.</given-names> <surname>Çolak</surname></string-name>, and <string-name><given-names>Ö.</given-names> <surname>Öztekin</surname></string-name></person-group>, <year>2017</year>, <article-title>Political environment, financial intermediation costs, and financing patterns</article-title>, <source>Journal of Corporate Finance</source> <volume>44</volume>, <fpage>167</fpage>–<lpage>92</lpage>.</mixed-citation></ref>
|
|
1168
|
+
<ref id="bib16"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>He</surname>, <given-names>Y.</given-names></string-name>, <string-name><given-names>H.</given-names> <surname>Lin</surname></string-name>, <string-name><given-names>C.</given-names> <surname>Wu</surname></string-name>, and <string-name><given-names>U. B.</given-names> <surname>Dufrene</surname></string-name></person-group>, <year>2009</year>, <article-title>The 2000 presidential election and the information cost of sensitive versus non-sensitive S&P 500 stocks</article-title>, <source>Journal of Financial Markets</source> <volume>12</volume>, <fpage>54</fpage>–<lpage>86</lpage>.</mixed-citation></ref>
|
|
1169
|
+
<ref id="bib17"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Ho</surname>, <given-names>T.</given-names></string-name>, and <string-name><given-names>H. R.</given-names> <surname>Stoll</surname></string-name></person-group>, <year>1981</year>, <article-title>Optimal dealer pricing under transactions and return uncertainty</article-title>, <source>Journal of Financial Economics</source> <volume>9</volume>, <fpage>47</fpage>–<lpage>73</lpage>.</mixed-citation></ref>
|
|
1170
|
+
<ref id="bib18"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Holden</surname>, <given-names>C. W.</given-names></string-name>, and <string-name><given-names>S.</given-names> <surname>Jacobsen</surname></string-name></person-group>, <year>2014</year>, <article-title>Liquidity measurement problems in fast, competitive markets: Expensive and cheap solutions</article-title>, <source>Journal of Finance</source> <volume>69</volume>, <fpage>1747</fpage>–<lpage>85</lpage>.</mixed-citation></ref>
|
|
1171
|
+
<ref id="bib19"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Hong</surname>, <given-names>H.</given-names></string-name>, and <string-name><given-names>L.</given-names> <surname>Kostovetsky</surname></string-name></person-group>, <year>2012</year>, <article-title>Red and blue investing: Values and finance</article-title>, <source>Journal of Financial Economics</source> <volume>103</volume>, <fpage>1</fpage>–<lpage>19</lpage>.</mixed-citation></ref>
|
|
1172
|
+
<ref id="bib20"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Julio</surname>, <given-names>B.</given-names></string-name>, and <string-name><given-names>Y.</given-names> <surname>Yook</surname></string-name></person-group>, <year>2012</year>, <article-title>Political uncertainty and corporate investment cycles</article-title>, <source>Journal of Finance</source> <volume>67</volume>, <fpage>45</fpage>–<lpage>83</lpage>.</mixed-citation></ref>
|
|
1173
|
+
<ref id="bib21"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Kelly</surname>, <given-names>B.</given-names></string-name>, <string-name><given-names>L.</given-names> <surname>Pástor</surname></string-name>, and <string-name><given-names>P.</given-names> <surname>Veronesi</surname></string-name></person-group>, <year>2016</year>, <article-title>The price of political uncertainty: Theory and evidence from the option market</article-title>, <source>Journal of Finance</source> <volume>71</volume>, <fpage>2417</fpage>–<lpage>79</lpage>.</mixed-citation></ref>
|
|
1174
|
+
<ref id="bib22"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Kyle</surname>, <given-names>A. S.</given-names></string-name></person-group>, <year>1985</year>, <article-title>Continuous auctions and insider trading</article-title>, <source>Econometrica</source> <volume>53</volume>, <fpage>1315</fpage>–<lpage>35</lpage>.</mixed-citation></ref>
|
|
1175
|
+
<ref id="bib23"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Lee</surname>, <given-names>C.</given-names></string-name>, and <string-name><given-names>M. J.</given-names> <surname>Ready</surname></string-name></person-group>, <year>1991</year>, <article-title>Inferring trade direction from intraday data</article-title>, <source>Journal of Finance</source> <volume>46</volume>, <fpage>733</fpage>–<lpage>46</lpage>.</mixed-citation></ref>
|
|
1176
|
+
<ref id="bib24"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Li</surname>, <given-names>J.</given-names></string-name>, and <string-name><given-names>J. A.</given-names> <surname>Born</surname></string-name></person-group>, <year>2006</year>, <article-title>Presidential election uncertainty and common stock returns in the United States</article-title>, <source>Journal of Financial Research</source> <volume>29</volume>, <fpage>609</fpage>–<lpage>22</lpage>.</mixed-citation></ref>
|
|
1177
|
+
<ref id="bib25"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Liu</surname>, <given-names>L. X.</given-names></string-name>, <string-name><given-names>H.</given-names> <surname>Shu</surname></string-name>, and <string-name><given-names>K. J.</given-names> <surname>Wei</surname></string-name></person-group>, <year>2017</year>, <article-title>The impacts of political uncertainty on asset prices: Evidence from the Bo scandal in China</article-title>, <source>Journal of Financial Economics</source> <volume>125</volume>, <fpage>286</fpage>–<lpage>310</lpage></mixed-citation></ref>
|
|
1178
|
+
<ref id="bib26"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Marshall</surname>, <given-names>B. R.</given-names></string-name>, <string-name><given-names>H. T.</given-names> <surname>Nguyen</surname></string-name>, <string-name><given-names>N. H.</given-names> <surname>Nguyen</surname></string-name>, and <string-name><given-names>N.</given-names> <surname>Visaltanachoti</surname></string-name></person-group>, <year>2017</year>, <article-title>Politics and liquidity</article-title>, <source>Journal of Financial Markets</source> <volume>38</volume>, <fpage>1</fpage>–<lpage>13</lpage>.</mixed-citation></ref>
|
|
1179
|
+
<ref id="bib27"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Mei</surname>, <given-names>J.</given-names></string-name>, and <string-name><given-names>L.</given-names> <surname>Guo</surname></string-name></person-group>, <year>2004</year>, <article-title>Political uncertainty, financial crisis and market volatility</article-title>, <source>European Financial Management</source> <volume>10</volume>, <fpage>639</fpage>–<lpage>57</lpage>.</mixed-citation></ref>
|
|
1180
|
+
<ref id="bib28"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Nippani</surname>, <given-names>S.</given-names></string-name>, and <string-name><given-names>A. C.</given-names> <surname>Arize</surname></string-name></person-group>, <year>2005</year>, <article-title>US presidential election impact on Canadian and Mexican stock markets</article-title>, <source>Journal of Economics and Finance</source> <volume>29</volume>, <fpage>271</fpage>–<lpage>79</lpage>.</mixed-citation></ref>
|
|
1181
|
+
<ref id="bib29"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Nippani</surname>, <given-names>S.</given-names></string-name>, and <string-name><given-names>W. B.</given-names> <surname>Medlin</surname></string-name></person-group>, <year>2002</year>, <article-title>The 2000 presidential election and the stock market</article-title>, <source>Journal of Economics and Finance</source> <volume>26</volume>, <fpage>162</fpage>–<lpage>69</lpage>.</mixed-citation></ref>
|
|
1182
|
+
<ref id="bib30"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Ozsoylev</surname>, <given-names>H.</given-names></string-name>, and <string-name><given-names>J.</given-names> <surname>Werner</surname></string-name></person-group>, <year>2011</year>, <article-title>Liquidity and asset prices in rational expectations equilibrium with ambiguous information</article-title>, <source>Economic Theory</source> <volume>48</volume>, <fpage>469</fpage>–<lpage>91</lpage>.</mixed-citation></ref>
|
|
1183
|
+
<ref id="bib31"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Pantzalis</surname>, <given-names>C.</given-names></string-name>, <string-name><given-names>D. A.</given-names> <surname>Stangeland</surname></string-name>, and <string-name><given-names>H. J.</given-names> <surname>Turtle</surname></string-name></person-group>, <year>2000</year>, <article-title>Political elections and the resolution of uncertainty: The international evidence</article-title>, <source>Journal of Banking & Finance</source> <volume>24</volume>, <fpage>1575</fpage>–<lpage>1604</lpage>.</mixed-citation></ref>
|
|
1184
|
+
<ref id="bib32"><mixed-citation specific-use="unstructured-citation" publication-type="other">Pasquariello, P., and C. Zafeiridou, 2014, Political uncertainty and financial market liquidity, Working Paper, University of Michigan.</mixed-citation></ref>
|
|
1185
|
+
<ref id="bib33"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Pástor</surname>, <given-names>L.</given-names></string-name>, and <string-name><given-names>R.</given-names> <surname>Stambaugh</surname></string-name></person-group>, <year>2003</year>, <article-title>Liquidity risk and expected stock returns</article-title>, <source>Journal of Political Economy</source> <volume>111</volume>, <fpage>642</fpage>–<lpage>85</lpage>.</mixed-citation></ref>
|
|
1186
|
+
<ref id="bib34"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Pástor</surname>, <given-names>L.</given-names></string-name>, and <string-name><given-names>P.</given-names> <surname>Veronesi</surname></string-name></person-group>, <year>2012</year>, <article-title>Uncertainty about government policy and stock prices</article-title>, <source>Journal of Finance</source> <volume>67</volume>, <fpage>1219</fpage>–<lpage>64</lpage>.</mixed-citation></ref>
|
|
1187
|
+
<ref id="bib35"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Pástor</surname>, <given-names>L.</given-names></string-name>, and <string-name><given-names>P.</given-names> <surname>Veronesi</surname></string-name></person-group>, <year>2013</year>, <article-title>Political uncertainty and risk premia</article-title>, <source>Journal of Financial Economics</source> <volume>110</volume>, <fpage>520</fpage>–<lpage>45</lpage>.</mixed-citation></ref>
|
|
1188
|
+
<ref id="bib36"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Pérignon</surname>, <given-names>C.</given-names></string-name>, and <string-name><given-names>B.</given-names> <surname>Vallée</surname></string-name></person-group>, <year>2017</year>, <article-title>The political economy of financial innovation: Evidence from local governments</article-title>, <source>Review of Financial Studies</source> <volume>30</volume>, <fpage>1903</fpage>–<lpage>34</lpage>.</mixed-citation></ref>
|
|
1189
|
+
<ref id="bib37"><mixed-citation publication-type="journal"><person-group person-group-type="author"><string-name><surname>Santa‐Clara</surname>, <given-names>P.</given-names></string-name>, and <string-name><given-names>R.</given-names> <surname>Valkanov</surname></string-name></person-group>, <year>2003</year>, <article-title>The presidential puzzle: Political cycles and the stock market</article-title>, <source>Journal of Finance</source> <volume>58</volume>, <fpage>1841</fpage>–<lpage>72</lpage>.</mixed-citation></ref>
|
|
1190
|
+
<ref id="bib38"><mixed-citation specific-use="unstructured-citation" publication-type="other">Voth, H. J., 2002, Stock price volatility and political uncertainty: Evidence from the interwar period, Working Paper, University of Zurich.</mixed-citation></ref>
|
|
1191
|
+
<ref id="bib39"><mixed-citation specific-use="unstructured-citation" publication-type="other">test reference which contains unparsed text</mixed-citation></ref>
|
|
1192
|
+
</ref-list>
|
|
1193
|
+
<fn-group>
|
|
1194
|
+
<fn id="fn1">
|
|
1195
|
+
<p>See Matt Egan, “Wall Street Welcomes Trump with a Bang” CNN Money (November 9, 2016), available at <underline><ext-link ext-link-type="uri" xlink:href="http://money.cnn.com/2016/11/09/investing/dow-jones-trump-wins-election/index.html">http://money.cnn.com/2016/11/09/investing/dow-jones-trump-wins-election/index.html</ext-link></underline></p>
|
|
1196
|
+
</fn><fn id="fn2">
|
|
1197
|
+
<p>Overseas markets also experienced dramatic swings following initial selloffs around the Trump election (see Jethro Mullen and Matt Egan, “Global Markets Drop as U.S. Election Results Shock Investors” CNN Money (November 9, 2016), available at <underline><ext-link ext-link-type="uri" xlink:href="http://money.cnn.com/2016/11/08/investing/global-markets-stocks-trump-clinton-us-presidential-election/index.html">http://money.cnn.com/2016/11/08/investing/global-markets-stocks-trump-clinton-us-presidential-election/index.html</ext-link></underline></p>
|
|
1198
|
+
</fn><fn id="fn3">
|
|
1199
|
+
<p>As an alternative benchmark, we use the 2004 U.S. presidential election.</p>
|
|
1200
|
+
</fn><fn id="fn4">
|
|
1201
|
+
<p>Full results are available at <underline><ext-link ext-link-type="uri" xlink:href="https://iemweb.biz.uiowa.edu/closed/">https://iemweb.biz.uiowa.edu/closed/</ext-link></underline></p>
|
|
1202
|
+
</fn><fn id="fn5">
|
|
1203
|
+
<p><xref ref-type="bibr" rid="bib30">Ozsoylev and Werner (2011)</xref> argue that ambiguous information results in potential market illiquidity because it reduces market arbitrage opportunities.</p>
|
|
1204
|
+
</fn><fn id="fn6">
|
|
1205
|
+
<p><xref ref-type="bibr" rid="bib32">Pasquariello and Zafeiridou (2014)</xref> argue that this may not hold when informed participants are reluctant to reveal information via transacting, resulting in lower trading activity before the election outcome.</p>
|
|
1206
|
+
</fn><fn id="fn7">
|
|
1207
|
+
<p>According to FiveThirtyEight’s election forecasts, the probability of the Republican candidate, Donald Trump, winning was 28.16% on election day. These election forecasts are available at <underline><ext-link ext-link-type="uri" xlink:href="https://projects.fivethirtyeight.com/2016-election-forecast/">https://projects.fivethirtyeight.com/2016-election-forecast/</ext-link></underline></p>
|
|
1208
|
+
</fn><fn id="fn8">
|
|
1209
|
+
<p><xref ref-type="bibr" rid="bib38">Voth (2002)</xref>, <xref ref-type="bibr" rid="bib27">Mei and Guo (2004)</xref>, <xref ref-type="bibr" rid="bib24">Li and Born (2006)</xref>, <xref ref-type="bibr" rid="bib6">Białkowski, Gottschalk, and Wisniewski (2008)</xref>, <xref ref-type="bibr" rid="bib9">Boutchkova et al. (2012)</xref>, and <xref ref-type="bibr" rid="bib3">Baker, Bloom, and Davis (2016)</xref> all provide evidence that increased political uncertainty is associated with increased return volatility in the equity markets. <xref ref-type="bibr" rid="bib21">Kelly, Pástor, and Veronesi (2016)</xref> show that political uncertainty is priced into the option markets as it pertains to implied volatility.</p>
|
|
1210
|
+
</fn><fn id="fn9">
|
|
1211
|
+
<p><xref ref-type="bibr" rid="bib35">Pástor and Veronesi (2013)</xref> provide evidence that both the magnitude of a political shock and the state of the economy help determine the risk premium required by investors.</p>
|
|
1212
|
+
</fn><fn id="fn10">
|
|
1213
|
+
<p>We also compare the 2004 and 2016 U.S. presidential elections in a difference-in-difference setting and show that liquidity deteriorates for the 30 trading days following the 2016 election.</p>
|
|
1214
|
+
</fn><fn id="fn11">
|
|
1215
|
+
<p>This study is related to the growing literature surrounding the role of political uncertainty in financial markets surrounding political event outcomes. <xref ref-type="bibr" rid="bib9">Boutchkova et al. (2012)</xref>, <xref ref-type="bibr" rid="bib34">Pástor and Veronesi (2012</xref>, <xref ref-type="bibr" rid="bib35">2013</xref>), <xref ref-type="bibr" rid="bib10">Brogaard and Detzel (2015)</xref>, <xref ref-type="bibr" rid="bib15">Gungoraydinoglu, Çolak, and Öztekin (2017)</xref>, <xref ref-type="bibr" rid="bib7">Bonaparte, Kumar, and Page (2017)</xref>, <xref ref-type="bibr" rid="bib5">Bhattacharya et al. (2017)</xref>, <xref ref-type="bibr" rid="bib26">Marshall et al. (2017)</xref>, <xref ref-type="bibr" rid="bib25">Liu, Shu, and Wei (2017)</xref>, and <xref ref-type="bibr" rid="bib36">Perignon and Vallée (2017)</xref> all examine the effects of political <xref ref-type="bibr" rid="bib39">test-author</xref>uncertainty of financial markets as it relates to prices, volatility, risk premia, politically sensitive industries, financial intermediation, innovation activity, investment decisions, and corporate governance.</p>
|
|
1216
|
+
</fn><fn id="fn12">
|
|
1217
|
+
<p>We also estimate price impact based on shorter windows such as one minute and 30 seconds, and our results are robust.</p>
|
|
1218
|
+
</fn><fn id="fn13">
|
|
1219
|
+
<p><xref ref-type="bibr" rid="bib18">Holden and Jacobsen (2014)</xref> show that the NBBO file is incomplete by itself, but the complete official NBBO can be constructed using the NBBO file and the quote file.</p>
|
|
1220
|
+
</fn><fn id="fn14">
|
|
1221
|
+
<p>Because our analysis uses liquidity metrics such as bid–ask spreads that are directly influenced by the minimum tick size, we remove stocks that are included in the U.S. Securities and Exchange Commission (SEC) 2016 Tick Size Pilot Program, and our results are robust.</p>
|
|
1222
|
+
</fn><fn id="fn15">
|
|
1223
|
+
<p>We rerun the entire analysis enforcing a balanced panel and our results are identical.</p>
|
|
1224
|
+
</fn><fn id="fn16">
|
|
1225
|
+
<p>To test for robustness, we include firm-specific control variables such as price, volume, market capitalization, and turnover in each of our regression equations, and our results are robust.</p>
|
|
1226
|
+
</fn><fn id="fn17">
|
|
1227
|
+
<p>We also bootstrap the standard errors and our results are robust.</p>
|
|
1228
|
+
</fn><fn id="fn18">
|
|
1229
|
+
<p><xref ref-type="bibr" rid="bib1">Amihud (2002)</xref> documents that smaller, illiquid stocks are associated with greater adverse selection costs. Similarly, <xref ref-type="bibr" rid="bib4">Barry and Brown (1984)</xref> document that investors command higher returns on small stocks as compensation for having less available information.</p>
|
|
1230
|
+
</fn><fn id="fn19">
|
|
1231
|
+
<p>“Trump vs. Clinton: The Destiny of These 4 Major Stock Industries Is at Stake,” IHODL (August 31, 2016), available at: <underline><ext-link ext-link-type="uri" xlink:href="https://en.insider.pro/economics/2016-08-31/trump-vs-clinton-destiny-these-4-major-stock-industries-stake/">https://en.insider.pro/economics/2016-08-31/trump-vs-clinton-destiny-these-4-major-stock-industries-stake/</ext-link></underline></p>
|
|
1232
|
+
</fn>
|
|
1233
|
+
</fn-group>
|
|
1234
|
+
</back>
|
|
1235
|
+
</article>
|