Using Real Time Data to Forecast Impacts of Financial Crises

With the United States still working to recover from the 2008 financial crisis, the world watches closely to see how Greece responds to its two bailouts, while simultaneously turning an anxious eye towards China’s crashing stock market. With these worrying financial crises brewing on the horizon, it becomes even more important to gain a greater understanding of the long-term impacts of such events.

A new study explores these impacts by developing a new model with specific criteria tailored to measuring impacts of financial crises. Christina and David Romer use the Organization for Economic Co-operation and Development (OECD) Economic Outlook as their primary source of data. The OECD Economic Outlook is published semiannually, allowing the Romers to examine almost-real-time data.

This use of more modern data makes the Romer study more applicable to modern crises, as previous studies have focused predominantly on the postwar era.

However, this perspective is only concerned with advanced countries, limiting the generalizability of conclusions drawn from these data.

Using these data, the authors look for developments such as increases in financial institutions’ costs of borrowing at a safe rate, decreases in institutional willingness to lend, and other (real or perceived) disruptions in normal borrower-lender relationships.

After identifying these instances, the Romers separate them into categories of “credit disruption,” “minor crisis,” “moderate crisis,” “major crisis,” and “extreme crisis.” Previous research has been confined to analyzing financial crisis in a binomial framework, comparing crisis to non-crisis. In contrast, Christina and David Romer incorporate a 0-15 range for categorizing levels of financial crisis. Using multiple categories allows the authors to later have a more complete understanding of impacts. To enhance their analysis, they also retrieve information about financial distress from the annual reports of central banks, reports from the IMF’s Article IV consultations, and the Wall Street Journal.

To measure long-term or ongoing impacts, the authors use two components: industrial production and real GDP. While both components are seemingly straightforward, the authors willingly admit that GDP may be less reliable since it is less transparent. Furthermore, the authors acknowledge that their sourcing does not allow them to parse out other influential external factors, such as a natural disaster that disrupts financial markets.

After a crisis, the authors find that industrial production falls significantly but has only a minimal long-term impact. The authors report that the effect dissipates quickly and that, within two years of the catalyst event, the decline in production is zero. Similarly, the authors report that the shock of a crisis to GDP begins to dissipate after six months. It should be noted that these figures reflect a study population that excludes Japan. The study excludes Japan because Japan’s crisis weight becomes overestimated in the calculations and would otherwise accredit too great a change to both industrial production and GDP.

The authors outline several other main impacts of a financial crisis, such as an industry fear of lending, where few healthy lenders are “willing to step into the void” following a recession. Furthermore, the authors highlight the negative spillover effects of a recession. The great potential impact of spillover effects is particularly relevant now, as Greece is part of the integrated Eurozone economy, and China is the second-largest economy in the world. Christina and David Romer offer a sound basis for estimating the long- and medium-term impacts of these brewing crises, which could potentially have great effects locally and across the globe.

 

Article Source: New Evidence on the Impact of Financial Crises in Advanced Countries. Romer, Christina D. Romer, David H. The National Bureau of Economic Research, Working Paper, November 2014

Featured Photo: cc/(London Permaculture)