Markets by Trading view

Alternative Data and Nowcasting are trending amongst central banks, Dow Jones explains why


It is a statement of the obvious to say that market conditions are at historically volatile levels. As the markets initially responded to the outbreak of the pandemic last March, the S&P 500 stock index dropped over 30% in a month. The Dow Jones crashed by almost 3,000 points on one day alone. The yields on US government bond plummeted as investors rushed for the security of Treasury-backed cash. In times like this, Alternative Data and Nowcasting become even more valuable.

Such dramatic movements inevitably posed huge problems for central banks attempting to adjust their monetary policy to rapidly changing conditions. Official data traditionally used by monetary committees for decision-making – such as GDP figures and consumer sentiment indexes – are always backward looking. They are published only at monthly or quarterly intervals.

In the midst of a crisis, how can central banks work out what is happening in the market in real-time? How can they obtain the necessary data to make quick and important decisions?

Increasingly, major central banks are experimenting with ‘nowcasting’ – using news data to improve their economic forecasts and obtain a more immediate insight into market conditions. To discuss this we spoke to Simon Rodda, Market Specialist Director at Dow Jones.

Simon began by discussing the original foray into nowcasting by the Norwegian Central Bank (Norges).

“The Norges had a project running with the The Norwegian Business School’s Centre of Applied Macroeconomics and Commodity Prices. The academics were put in both camps, building indicators of consumer sentiment. This was built around the analysis of local newspapers. They looked at how they could get a read of consumer sentiment from the articles and the kind of language that was used. They then matched this against the official data to see whether or not it was actually front-running – if the sentiment they found was later picked up in the official quarterly data. And it produced some encouraging signals.”

The academics of the Norges and The Norwegian Business School’s Centre of Applied Macroeconomics and Commodity Prices approached Dow Jones and worked with Simon and his team to build a similar model for the US economy. This work had been going on in the background for “four or five years” before “we hit the pandemic” and it suddenly became of pressing significance.

“With the pandemic, we have seen the biggest changes in economic outlook for not just decades but centuries. In terms of volatility, it is off the scale. And so central banks have been scrambling to adjust monetary policy to cope with this. Waiting for the three-monthly cycle and GDP data is just not good enough. Whereas they had already started to experiment around with nowcasting, this has now become of paramount importance.”

Quite simply, the urgency of the situation meant central banks required immediate data on which to make major decisions. As markets moved at historically fast rates, they needed instant insights into market sentiment to try and get to grip with the situation.

For Simon, nowcasting also offers another quality that traditional sources of data do not. “News includes not just information”, Simon argues, “but sentiment”.

“Part of the writing process is the selection of what you write about. There is an inference in the topic selection, and in the language itself. Through textual analysis, we can pick up on the language journalists are using and look for positive and negative words and phrases associated with certain economic topics.

“That is what the academics and analysts within the central banks are doing here. They are trying to build up their sentiment signal over time, and to detect any changes that may occur in reaction to events.

“Obviously, the events that have unfolded over the past fifteen months have been extreme. And so having a very quick read on shifts indicating an improvement or decline in sentiment is very important to central banks looking to adjust monetary policy.”

Simon also pointed out that nowcasting can help central banks “prepare the ground for monetary policy”. As we have seen lately with the “taper tantrums” in the bond markets caused by the possibility of higher interest rates, markets react to potential monetary policies as much as those that have already been implemented.

It is important for central banks “to gauge what the reaction might be to a policy shift – as we have seen with the Bank of England asking banks to prepare for negative rates”. Nowcasting is increasingly “part of this policy management as banks bring different kinds of tools into the policy toolkit”.

Nowcasting can also help financial institutions such as asset managers and investment banks make judgements on issues that cannot be decided using traditional data. This is particularly the case with regards to the ESG issues they are now grappling with.

“How do you measure one company against another company on ESG?” Simon asked. “You cannot simply look at the P/E ratio and say ‘well, this one is better than that one’. For binary measurements like this you can use a machine or programmes to determine which is better.

“But what data point do you use for trying to decide which company is better on diversity and inclusion? You will need to use the kind of data nowcasting looks at in order to reach a judgement.”

Of course, when analysing a text, “there is no one truth”. With nowcasting a huge amount rests on “inference” and “personal judgement”. But as Simon argued, nowcasting can be a useful tool by creating an “aggregate of hundreds or thousands of articles written on any one day. Some of them may just be purely factual, some may be a combination of facts and opinion. Some may not be expressing an adverse opinion but using adjectives or phrases that infer a certain development was good or bad”. The picture this paints can be a useful “sentiment indicator” informing the decisions of financial institutions.

Various central banks have already brought nowcasting into their decision-making processes, and more will no doubt follow. The Banca D’Italia is using nowcasting techniques to forecast GDP by tracking weekly variables, whilst the Asian Development Bank is currently working on a project to incorporate it into its planned response for future crises.

What is sure is that nowcasting offers immediate – if imperfect – market insights that can be a useful tool during times of extreme economic turbulence like our own.  

Author: Harry Clynch

#Nowcasting #DowJones #CentralBanks #AlternativeData #Norges #Covid #ESG #SentimentIndicator

One Response

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts


Write your email to verify subscription


Sign up for our free newsletter and receive the latest banking and fintech stories, straight to your inbox - every week