Seduced by data

Sep. 25, 2017

The discovery of Swedish data changed Paolo Sodini’s career path. He went from focusing on theoretical asset pricing to being part of establishing the then new research field of Household Finance.

When Paolo Sodini first came to Sweden in 2004 he had his mind set on continuing his research in the field of asset pricing theory. But when he took up a position as Assistant Professor at the Stockholm School of Economics he quickly stumbled upon a set of data that would change his focus.

“It was at a meeting at Statistiska Centralbyrån (Statistics Sweden) where I first realized the amount and level of detail of the data available in Sweden. It felt like I was suddenly sitting on this treasure”, Paolo says.

Testing theories in real life

While studying at his hometown of Gorizia in northern Italy, Paolos plan for the future was to find a way to leave Italy and go out and discover the world. A career in academia provided both interesting work and a ticket out of Italy. He did his Master at London School of Economics and then went on to Massachusetts Institute of Technology in the US for a PhD, all the time focusing on the more theoretical and mathematical side of economics. 

“My work took a U-turn when I found the Swedish data. I realized that I had the opportunity to test many of the theories I had studied and see whether they held up in real life. It was an exciting moment.”

Paolo and his colleagues Laurent Calvet and John Campbell worked on cleaning and preparing the data for their first paper for two and a half years. People were wondering what they were up to. Their first paper “Down or Out: Assessing The Welfare Costs of Household Investment Mistakes” was published in the Journal of Political Economy. It studied for the first time, the level of diversification in household portfolios, and established a link between household level of financial sophistication, portfolio diversification and the level of financial risk households take in their liquid wealth. The paper subverted the previous belief that household suffer large losses from holding poorly diversified portfolios. They don’t, and largely because those households that hold concentrated portfolios, typically financially unsophisticated, do not invest much wealth in financial markets, thereby limiting their losses. It seems like people are somehow aware of their limitations.

This was before the term “Household finance” was defined and since then, the area has contributed with a lot of useful information for policy makers. The group’s second paper “Fight or Flight? Portfolio Rebalancing by Individual Investors”, published in the Quarterly Journal of Economics, discussed how households rebalance their portfolio in response to stock market movements. It shows that households rebalance their portfolios, for example by buying stocks when the market tanks, and that more sophisticated households display less inertia in doing so.

Since then, Paolo and his coauthors have been using the Swedish data to understand household financial risk taking, income hedging, the shape of the wealth distribution, and the effects of homeownership on consumption and savings. Since 2007, as Sweden cut the wealth tax, the collection of detailed wealth data also stopped.

“The whole world is going into big data, and meanwhile Sweden is going into reverse. We used to be best at it, now we are the worst. I think the next policy revolution is going to be that we cannot take monetary policy decisions without micro-data. A lot of policy decisions now are taken without the right amount of information. How do we know if we are on the brink of collapse in the real estate market if we don’t have the data? It will be revolutionary for research as well as policy if we start collecting the data again.”

Pioneering work through the Data Center

Some of the research questions that Paolo tries to answer are; how financial education should be designed, whether households act rationally or in opposite of their own interest, and the implications on society of bad household economic decisions. Coming up with relevant questions has been one of Paolos strengths during his career.

Apart from teaching, researching and various other assignments, Paolo is also the director of the Swedish House of Finance Data Center. The Data Center was formed in 2011 when NasdaqOMX donated FinBas, a database of financial data from the Nordic Stock Exchanges, MTF’s and OTC markets, to Swedish House of Finance. The database came without the possibility to update to new data feeds and extract data so the team at the center has been busy working to resuscitate Finbas and make it available to the academic community. In addition to FinBas, other sets of data have been added along the way and the data center is working to provide a foundation for research in the Nordic region. 

“Data is important to me, and to the entire research and policy community. Without it, we cannot find the answers we are looking for in our research and for the policy decisions that affect so many in our society.”