Wednesday, 15 February 2023

Nowcasting

 


From Wikipedia, the free encyclopedia

Nowcasting in economics is the prediction of the present, the very near future, and the very recent past state of an economic indicator. The term is a portmanteau of "now" and "forecasting" and originates in meteorology. It has recently become popular in economics as typical measures used to assess the state of an economy (e.g., gross domestic product (GDP)), are only determined after a long delay and are subject to revision.[1] Nowcasting models have been applied most notably in Central Banks, who use the estimates to monitor the state of the economy in real-time as a proxy for official measures.[2][3]

Principle[edit]

While weather forecasters know weather conditions today and only have to predict future weather, economists have to forecast the present and even the recent past. Many official measures are not timely due to the difficulty in collecting information. Historically, nowcasting techniques have been based on simplified heuristic approaches but now rely on complex econometric techniques. Using these statistical models to produce predictions eliminates the need for informal judgement.[4]

Nowcast models can exploit information from a large quantity of data series at different frequencies and with different publication lags.[5] Signals about the direction of change in GDP can be extracted from this large and heterogeneous set of information sources (such as jobless figures, industrial orders, trade balances) before the official estimate of GDP is published. In nowcasting, this data is used to compute sequences of current quarter GDP estimates in relation to the real time flow of data releases.

Development[edit]

Selected academic research papers show how this technique has developed.[6][7][8][9][10][11][12][13]

Banbura, Giannone and Reichlin (2011)[14] and Marta Banbura, Domenico Giannone, Michele Modugno & Lucrezia Reichlin (2013)[15] provide surveys of the basic methods and more recent refinements.

Nowcasting methods based on social media content (such as Twitter) have been developed to estimate hidden sentiment such as the 'mood' of a population[16] or the presence of a flu epidemic.[17]

A simple-to-implement, regression-based approach to nowcasting involves mixed-data sampling or MIDAS regressions.[18] The MIDAS regressions can also be combined with machine learning approaches.[19]

Econometric models can improve accuracy.[20] Such models can be built using bayesian vector autoregressions, dynamic factors, bridge equations using time series methods, or some combination with other methods.[21]

Implementation[edit]

Economic nowcasting is largely developed by and used in central banks to support monetary policy.

Many of the Reserve Banks of the US Federal Reserve System publish macroeconomic nowcasts. The Federal Reserve Bank of Atlanta publishes GDPNow to track GDP.[3][21] Similarly, the Federal Reserve Bank of New York publishes a dynamic factor model nowcast.[2] Neither are official forecasts of the Federal Reserve regional bank, system, or the FOMC; nor do they incorporate human judgment.

Nowcasting can also be used to estimate inflation[22] or the business cycle.[23]

References[edit]

  1. ^ Hueng, C. James (2020-08-25), "Alternative Economic Indicators", W.E. Upjohn Institute, pp. 1–4, doi:10.17848/9780880996778.ch1, ISBN 978-0-88099-677-8 {{citation}}: Missing or empty |title= (help)
  2. ^ Jump up to:a b "Nowcasting Report - FEDERAL RESERVE BANK of NEW YORK". www.newyorkfed.org. Retrieved 2020-09-24.
  3. ^ Jump up to:a b "GDPNow". www.frbatlanta.org. Retrieved 2020-09-24.
  4. ^ Giannone, Domenico; Reichlin, Lucrezia; Small, David (May 2008). "Nowcasting: The real-time informational content of macroeconomic data". Journal of Monetary Economics. 55 (4): 665–676. CiteSeerX 10.1.1.597.705. doi:10.1016/j.jmoneco.2008.05.010. Retrieved 12 June 2015.
  5. ^ Bańbura, Marta; Modugno, Michele (2012-11-12). "Maximum Likelihood Estimation of Factor Models on Datasets with Arbitrary Pattern of Missing Data". Journal of Applied Econometrics. 29 (1): 133–160. doi:10.1002/jae.2306. hdl:10419/153623. ISSN 0883-7252. S2CID 14231301.
  6. ^ Camacho, Maximo; Perez-Quiros, Gabriel (2010). "Introducing the euro-sting: Short-term indicator of euro area growth". Journal of Applied Econometrics. 25 (4): 663–694. doi:10.1002/jae.1174. Retrieved 12 June2015.
  7. ^ Matheson, Troy D. (January 2010). "An analysis of the informational content of New Zealand data releases: The importance of business opinion surveys". Economic Modelling. 27 (1): 304–314. doi:10.1016/j.econmod.2009.09.010. Retrieved 12 June 2015.
  8. ^ Evans, Martin D. D. (September 2005). "Where Are We Now? Real-Time Estimates of the Macroeconomy". International Journal of Central Banking. 1 (2). Retrieved 12 June 2015.
  9. ^ Rünstler, G.; Barhoumi, K.; Benk, S.; Cristadoro, R.; Den Reijer, A.; Jakaitiene, A.; Jelonek, P.; Rua, A.; Ruth, K.; Van Nieuwenhuyze, C. (2009). "Short-term forecasting of GDP using large datasets: a pseudo real-time forecast evaluation exercise". Journal of Forecasting. 28 (7): 595–611. doi:10.1002/for.1105.
  10. ^ Angelini, Elena; Banbura, Marta; Rünstler, Gerhard (2010). "Estimating and forecasting the euro area monthly national accounts from a dynamic factor model". OECD Journal: Journal of Business Cycle Measurement and Analysis. 1: 7. Retrieved 12 June 2015.
  11. ^ Domenico, Giannone; Reichlin, Lucrezia; Simonelli, Saverio (23 November 2009). "Is the UK still in recession? We don't think so". Vox. Retrieved 12 June 2015.
  12. ^ Kajal, Lahiri; Monokroussos, George (2013). "Nowcasting US GDP: The role of ISM business surveys". International Journal of Forecasting. 29 (4): 644–658. CiteSeerX 10.1.1.228.3175. doi:10.1016/j.ijforecast.2012.02.010. S2CID 12028550.
  13. ^ Antolin-Diaz, Juan; Drechsel, Thomas; Petrella, Ivan (2014). "Following the Trend: Tracking GDP when Long-Run Growth is Uncertain". CEPR Discussion Papers 10272. Retrieved 12 June 2015.
  14. ^ Banbura, Marta; Giannone, Domenico; Reichlin, Lucrezia (2010). "Nowcasting". In Clements, Michael P.; Hendry, David F. (eds.). Oxford Handbook on Economic Forecasting. Oxford University Press.
  15. ^ Banbura, Marta; Giannone, Domenico; Modugno, Michele; Reichlin, Lucrezia (2013). "Chapter 4. Nowcasting and the Real-Time Dataflow". In Elliot, G.; Timmerman, A. (eds.). Handbook on Economic Forecasting. Handbook of Economic Forecasting. Vol. 2. Elsevier. pp. 195–237. doi:10.1016/B978-0-444-53683-9.00004-9. hdl:10419/153997. ISBN 9780444536839. S2CID 14278918.
  16. ^ Lansdall‐Welfare, Thomas; Lampos, Vasileios; Cristianini, Nello (August 2012). "Nowcasting the mood of the nation". Significance. 9 (4): 26–28. doi:10.1111/j.1740-9713.2012.00588.x. Archived from the original on 20 August 2012.
  17. ^ Lampos, Vasileios; Cristianini, Nello (2012). "Nowcasting Events from the Social Web with Statistical Learning" (PDF). ACM Transactions on Intelligent Systems and Technology. 3 (4): 1–22. doi:10.1145/2337542.2337557. S2CID 8297993.
  18. ^ Andreou, Elena; Ghysels, Eric; Kourtellos, Andros (2011-07-08). "Forecasting with Mixed-Frequency Data". Oxford Handbooks Online: 225–246. doi:10.1093/oxfordhb/9780195398649.013.0009. ISBN 978-0195398649.
  19. ^ Babii, Andrii; Ghysels, Eric; Striaukas, Jonas (2020). "Machine learning time series regressions with an application to nowcasting".
  20. ^ Tessier, Thomas H.; Armstrong, J. Scott (2015). "Decomposition of time-series by level and change". Journal of Business Research. 68 (8): 1755–1758. doi:10.1016/j.jbusres.2015.03.035.
  21. ^ Jump up to:a b Higgins, Patrick (July 2014). "GDPNow: A Model for GDP "Nowcasting"" (PDF). Federal Reserve Bank of Atlanta Working Paper Series.
  22. ^ Ahn, Hie Joo; Fulton, Chad (2020). "Index of Common Inflation Expectations". FEDS Notes. 2020 (2551). doi:10.17016/2380-7172.2551. ISSN 2380-7172. S2CID 225316591 – via Board of Governors of the Federal Reserve System.
  23. ^ Aruoba, S. Boragan; Diebold, Francis; Scotti, Chiara (2008). "Real-Time Measurement of Business Conditions". Cambridge, MA. doi:10.3386/w14349.

External links[edit]

on No comments:
Labels: , , , ,

Monday, 6 February 2023

How can carbon tagging digital payments help to tackle climate change?

ENERGY & CLIMATE CHANGE • 2 FEB 2023 JOHN BARRETT/ JOHN GATHERGOOD, DAVID LEAKE, ANNA TRENDL, ALEX WARD/ Economics Observatory

The majority of the goods and services that we purchase each day generate greenhouse gas emissions. Linking the digital data on these transactions to their carbon footprint could help households and businesses to make more informed decisions, and enable better targeted policy interventions.

On our current trajectory, carbon emissions will increase by over 10% by 2030. This contrasts with the 43% reduction needed to limit temperature rises to 1.5°C (United Nations, 2022). Radical changes are urgently needed across society to reduce emissions and tackle climate change.

The carbon measurement problem

One challenge for speeding up decarbonisation is effective measurement. While country-level analysis is generally seen as robust, there is not currently a consistent method that provides a meaningful carbon assessment for businesses or households.

Without this information, we are blinkered to the size of our personal carbon footprints. This is likely to make both firms and individuals indecisive on the next steps to take, and it means that they lack accountability.

We therefore need more granular measures of carbon footprints, which we can aggregate up from individual businesses or households to the total footprint of the economy. Without such a system in place, it is impossible to understand the collective effort needed to make rapid reductions in emissions.

How could tagging carbon emissions help?

A promising route forward on these measurement issues may already be available. The daily economic activity of businesses and households is captured digitally, in real-time, through electronic payment data that flows between banks, suppliers and retailers.

By embedding a carbon attribute into these payments – a measure of how much carbon the underlying activity represented by the payment generates – it is possible to introduce an automated, detailed and live measure of carbon accounts. This can then, in turn, inform and drive change.

The concept of ‘carbon tagging’ draws inspiration from widely used food labelling, which informs consumers about the calorific content of food. This nutritional information prompts consumers to evaluate their choices, firms to change their ingredients and governments to take action to improve diets.

Such systems provide objective data that can create change. Carbon tagging would do the same in the equally urgent and universal challenge of emissions reduction.

The benefits from carbon tagging digital payments would extend beyond just consumer labelling. Such a scheme would introduce much needed transparency across our economies, providing a key data resource to underpin general policy interventions, such as new carbon taxation. It would also help to inform specific initiatives that could be focused on the highest emitting sectors or activities.

Does the digital technology exist to make this possible?

All key data, technology and reporting practices for carbon tagging already exist. Using data from a UK-based high street bank, it is possible to demonstrate how this process might work in practice.

Drawing on earlier research (Trendl et al, 2022), we construct a dataset of carbon multipliers – the amount of carbon produced per £1 spend on a good or service – and apply these to financial transactions from an anonymous population of 3,582,448 customers and 773,503 firms, using data on all their transactions during October 2021.

The analysis shows that emissions – when measured through digital payments – are highly concentrated in just a small number of payment types. Specifically, 15 digital payments account for almost 60% of all emissions for both individuals and households, and small and medium-sized businesses (see Figures 1 and 2).

The transactions with the highest carbon footprint are regular direct debit payments for gas and electricity, and card payments for petrol, gas and airfares. Unsurprisingly, these are carbon-intensive products and dominate household and business carbon footprints. The carbon footprints arising from purchases of food, clothing and other retail products are modest in comparison.


The technology already exists to add carbon attributes to the digital payments data we see in our bank accounts and on our personal and business finance apps. Researchers have compiled 'carbon multipliers’: key parameters that represent the emissions generated by a given unit of spending on a particular product. These rely on the continual development of multi-regional input-output analyses, which trace the flow of inputs of raw materials, and the flow of outputs of intermediate and final products through the production process (Owen et al, 2018).

For example, £1 spent on fuel has a carbon multiplier close to three – which means that 3kg of carbon are generated for each £1 spent on fuel (Kilian, 2022). Using these calculations, we can readily transform spending into corresponding carbon emissions. We should see not just the financial cost of the goods and services we buy every day measured in pounds and pence, but also the carbon cost measured in kilos and tonnes.

What are the benefits of digital carbon tagging?

For businesses, this new information could radically simplify the calculation of carbon intensities of supply chains, business strategies and, ultimately, end products. This, in turn, would empower business leaders to make informed decisions to reduce their emissions footprints.

For consumers, it would create a new, salient dimension of consumer choice. Individuals would be able to opt to change their spending patterns to reduce their own carbon footprints.

There are also important benefits for governments, as this would enable emissions to be measured at multiple levels of potential policy interventions in a consistent way. This precision would empower policy-makers to target more effective policies and learn from interventions that are working.

It would also ensure that the sum of individual parts adds up to the whole. This is currently not the case, as contemporary reporting of carbon footprints by firms tends to be inconsistent – covering different geographies, time periods and using different carbon multipliers (Giesekam et al, 2019).

What comes next?

We see three main steps to creating a new, mandatory carbon emissions attribute for those digital payments that carry the highest carbon impact.

First, an independent body needs to be established that defines the carbon multipliers, sets the standards for reporting and oversees implementation in the payments system.

This would also involve defining the scope of carbon tagging. A first generation of labelling for the most carbon-intensive forms of production and consumption should be made a priority.

Further, new policies would need to require businesses operating in sectors with high carbon intensity to carry additional responsibilities that support the tracking of emissions. With new software upgrades to payment technologies, these requirements can be automated to help to minimise the implementation burden.

For example, it would be possible to create live data feeds of carbon footprints from payments data, which could be provided in real-time to business and households, allowing them to track their carbon footprints against targets in the same way that firms might track sales or revenues.

The second key step involves engagement with payment networks to embed carbon information in digital payments alongside other existing information. The specifics of this can vary based on the particular payment type being made and other factors, such as its purpose (Bank of England, 2020).

Finally, guidance and support needs to be provided to circulate the new carbon information. This could be via reporting guidelines for bank statements, business accountancy software or data sharing for government entities and regulators.

There are already useful examples of these types of practices. For example, mandatory information disclosures are required of banks and other payments services providers when generating bank statements, and when reporting key information to tax authorities.

There is a unique opportunity to leverage our digital payment infrastructure to help society to decarbonise. This would represent a major modernisation of our approach to reaching net-zero targets.

It would ensure consistency with country-level carbon accounting while providing much needed information to government, businesses and consumers to help them make more informed choices. It would also strengthen a key requirement for our eventual success: personal accountability.

on No comments:
Labels: , ,
Subscribe to: Posts (Atom)