How Big Data Can Help In A Recession
With the chances of Britain voting to exit the EU looking increasingly likely, and even the most optimistic of leave campaigners anticipating a recession in such an eventuality, companies should be bracing for a prolonged period of yet more economic turbulence.
We walked into the last major recent recession of 2008 with our eyes closed. In the third quarter of 2008, less than 30% of forecasters in the the Survey of Professional Forecasters predicted a that GDP would fall in the fourth quarter of the year. It fell 8%, one of the biggest drops on record. It total, the crisis would go on to cost the global economy around $22 billion. This time around, however, big data should prove a new secret weapon that better enables governments to steer the economy in the right direction, and businesses to weather the storm.
The impression that economists were asleep at the wheel is well earned, and has sparked a real skepticism when around the viability of economic forecasting. Data analytics has added new dimensions to this forecasting though, and this can help governments to better pinpoint the causes of the recession and the problem areas, and enable them to enact policies that will best counteract them.
Among such initiatives to have been developed in recent years is BlackRock’s attempt to mine senior management’s conversations for signifiers to gauge the economic outlook. Every quarter, the financial services firm mines 3,000 earnings conference calls and compares the topics and language used by CEOs and CFOs, looking for any minor changes in their language that could suggest a change in thinking around a topic. In their blog, they claim that ‘Rather than relying on a few anecdotal bits of evidence, big data allows us to measure exactly how much more frequently words like ‘recession’ have crept back into use. We can then decide if we should be worried, too.’
In the same way that analyzing senior executives can be a good indicator of what the business world is thinking, researchers are also looking to social media sites to gauge wider economic sentiment. Zillow, an online real-estate service, for example, collects information about home sales and mortgages to better understand the housing market, and sentiment around recruitment is also a good indicator of the labor market. All of this provides insights before official statistics are released, meaning that a response to a crisis can be implemented even quicker.
Not only can data analytics identify where a recession is hitting, it can also gauge the potential impact of government’s response. In Ben Bernanke’s paper, ‘Measuring the effects of monetary policy: a factor-augmented vector autoregressive approach’, a Factor-Augmented Vector Autoregressive (FAVAR) model was used for Big Data forecasting and structural analysis to accurately identify the monetary policy transmission mechanism and establish how exactly monetary policy would impact the economy. They found that the proposed FAVAR model greatly outperformed the Structural VAR model because it used content that was far more useful for the assessment of the monetary policy transmission mechanism.
Data does not only help governments better manage a downturn, businesses themselves can also adopt analytics to best prepare. There are a small number of recession-proof industries. With industries like healthcare and discount retail often resisting the negative impact of a downturn, as well as — somewhat ironically — condoms. For other businesses, they can learn rapidly how the recession has impacted their customers’ purchasing habits, and adjust their marketing and product strategies accordingly before major damage is done. Using predictive analytics models also allows companies to cut investment to parts of their business in such a way that, firstly, will actually have an impact that’s beneficial in the short term and, secondly, that will not hamper growth in the long term as the economy climbs out of the downturn.
Making decisions based on experience of prior recessions is foolish. Every recession is different. The change in demographics as baby boomers retire and millennials enter the job market make it impossible for governments to really know what will happen to the job market. Evolving consumer habits as a result in the explosion in technological advancements over the past few years means that this will also be hard to predict, and it could be that companies that barely survived previous recessions do better this time. Analytics can anticipate the direction that it will take and allow for responses to be formulated quickly enough to have an impact. The next recession will likely be even harder to predict, and we ignore data at our peril.
