Quantifying Economic Behaviour Using Big Data Tobias Preis Warwick - PowerPoint PPT Presentation

Quantifying Economic Behaviour Using Big Data Tobias Preis Warwick Business School Tobias.Preis@wbs.ac.uk www.tobiaspreis.de

A map of the world built only from GPS locations of Flickr photos ¡ The big data explosion

1 The advantage of looking forward

more Google searches for “2009” more Google searches for “2011” Based on Preis, Moat, Stanley and Bishop (2012) 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Future Orientation Index 2010

Richer countries look forward B Preis, Moat, Stanley & GDP / Capita [10 4 USD] Bishop (2012) 4 3 2 Featured by: 1 0.0 0.5 1.0 1.5 2.0 Future-Orientation Index

Photo: Perpetual Tourist 2 Predicting stock markets

Financial markets: big data Photo: Perpetual Tourist

The Internet: big data

Moat et al. (2013); Hypothetical strategy Preis et al. (2013) number of Google searches for keyword week t

Moat et al. (2013); Hypothetical strategy Preis et al. (2013) number of Google searches for keyword week t-3 t-2 t-1 t

Moat et al. (2013); Hypothetical strategy Preis et al. (2013) number of Google searches Search volume for keyword decreased: BUY stock in week t+1 week t-3 t-2 t-1 t

Moat et al. (2013); Hypothetical strategy Preis et al. (2013) number of Google searches Search volume for keyword decreased: BUY stock in week t+1 Search volume increased: SELL stock week in week t+1 t-3 t-2 t-1 t

Featured by: Example: “culture” % pro f t 40 16 0 5 − 40 2005 2007 2009 2011 “culture” trading strategy Preis, Moat buy and hold strategy & Stanley (2013) mean ± 1 sd of random strategies

Featured by: Example: “debt” % pro f t 326 300 200 100 16 0 2005 2007 2009 2011 “debt” trading strategy Preis, Moat buy and hold strategy & Stanley (2013) mean ± 1 sd of random strategies

How di ff erent keywords perform return (random strategy sds) Random strategy mean + 2 sds 2 Random strategy mean + 1 sd 1 0 -1 “culture” “debt” Preis, Moat & Stanley (2013)

How di ff erent keywords perform return (random strategy sds) Random strategy mean + 2 sds 2 Random strategy mean + 1 sd 1 0 -1 “credit” “culture” “debt” “train” “stocks” “garden” Preis, Moat & Stanley (2013)

How di ff erent keywords perform return (random strategy sds) Random strategy mean + 2 sds 2 Random strategy mean + 1 sd 1 0 -1 Financial relevance Returns signi fj cantly correlated with indicator # occurrences in FT of fj nancial relevance # hits on Google Preis, Moat & Stanley (2013)

Wikipedia: Dow Jones companies Moat, Curme, Wikipedia Views 0.6 Avakian, Kenett, DJIA Companies Wikipedia Edits Stanley & Preis DJIA Companies (2013) 0.4 Density Featured by: Random Strategy 0.2 0.0 − 2 0 2 Return [Std. Dev. of Random Strategies] Views strategies pro fj table

Wikipedia: Financial topics Moat, Curme, 1.00 Avakian, Kenett, Stanley & Preis Wikipedia Edits Financial Topics 0.75 (2013) Wikipedia Views Density Financial Topics Featured by: 0.50 Random Strategy 0.25 0.00 − 2 0 2 Return [Std. Dev. of Random Strategies] Views strategies pro fj table

Wikipedia: Actors and fj lmmakers? Moat, Curme, Avakian, Kenett, 0.4 Stanley & Preis Random Wikipedia Views Strategy Actors & Filmmakers (2013) 0.3 Density Featured by: 0.2 0.1 0.0 − 2 0 2 Return [Std. Dev. of Random Strategies] Strategies NOT pro fj table

What is searched for before falls? debt apple housing crisis housing orange tree debt Curme, Preis, Stanley & Moat (2014)

What is searched for before falls? Cumulative Returns (%) -100 0 100 200 Random Strategy Politics I Business 55 groups of search terms Business and politics most related Curme, Preis, Stanley & Moat (2014)

3 Sensing disasters

Flickr and Hurricane Sandy A 0.10 0.08 Flickr : photos taken 0.06 Landfall of Hurricane Sandy 0.04 Flickr Photos with Hurricane Related Tags 0.02 0.00 20 Oct 25 Oct 30 Oct 05 Nov 10 Nov 15 Nov 20 Nov Preis and Moat � Time (under review); ¡ B Landfall of Hurricane Sandy 1020 1000 Averaged Pressure in Air pressure US State New Jersey 980 960 20 Oct 25 Oct 30 Oct 05 Nov 10 Nov 15 Nov 20 Nov Preis, Moat, Bishop, Treleaven and Stanley (2013)

How can open data help you? Tobias.Preis@wbs.ac.uk Data from the Internet may help us measure and even predict human behaviour