Knowledge sharing is the way forward
The boundaries between the traditional disciplines and interests have become more fluid. To the benefit of all parties, because according to Anders Rahbek there is much to gain by collaborating on the task of modelling multi-dimensional, dynamic systems, both nationally and internationally.
By Eline Mørch Jensen
Professor Anders Rahbek
Anders Rahbek is professor of econometrics at the Department of Economics, University of Copenhagen (KU). He holds an MSc in Econometrics from the London School of Economics and an MSc in Mathematics-Economics as well as a PhD in mathematical statistics and econometrics from KU.
When asked what the term econometrics covers exactly, he explains using the following definition: Econometrics is the application of statistical and mathematical-statistical methods in working with economic and financial empirical problems.
- To work in a field such as econometrics it is necessary to be able to handle areas within economics as well as applied and theoretical (mathematical) statistics. And it is also in these areas, at the intersection of various scientific fields and special interests, where we can capitalize on the synergies that arise in the research project "Dynamic models" as well as in our work in general, Anders Rahbek explains and elaborates:
- There is a very positive effect from our interdisciplinary collaboration, an exciting and surprising synergy. As researchers in general we are often working with many of the same types of methodological problems or at least overlapping areas of interest, something that has become obvious at several of the meetings we have already held on the project. So it is encouraging that we have the opportunity to create a common understanding, a common platform if you will, for the overlapping areas so that new research may occur.
Anders Rahbek stresses that he sees the cooperation in "Dynamic models" as a great opportunity for the expansion of new types of cooperation. A process that is now underway and reflects the fluid boundaries between past classic differentiated disciplines:
- The food chain is very clear in the sense that, for instance, the development of dynamic econometric models, based on the existing economic and financial theory, contributes to a better understanding of what is happening in the economy. One might say that the development of dynamic (econometric) models enables a dialogue between new empirical data and theory.
With the aim to become better at predicting trends?
- You might think that, but analysing a dynamic system is not necessarily directed at actual predictions. Understanding and analysing what has happened provides an insight in itself, insight that may help create new economic theory and can be used as input for economic policy. It does not necessarily make us better at predicting exactly what will happen. In economics the unexpected happens, simply because the economic reality is constantly changing, but the hope is of course that we will also become more skilled at assessing or, if you will, anticipating possible future economic scenarios.
One can´t look at data from the depression of the 1930s and transfer them to today's financial crisis?
- That is in practice what we do because we analyse historical data, time series. To a large extent this is how we work. We seek to gain understanding from what has happened. But although many of the mechanisms to some degree are the same as then, the conditions are constantly changing, as I said. In part because of widening scope of possibility on the financial markets of today, Anders Rahbek explains and adds:
- Somewhat simplified you might say that it is possible to isolate economic or financial factors which are time constant, i.e. factors which were also present, for instance, during the previous crisis and corresponding time-varying, new factors or mechanisms, and thus be more able to understand the possible scenarios of economic development.
- If you for instance wish to buy a house and need to decide whether the mortgage you need to take out should have a short or a long-term rate, then being able to say something about how the interest rates will develop is important. The two rates will change in very unstable and unpredictable ways, but there is a certain balance between them, a mutual stability, so to speak. This stable equilibrium is in principle a predictable or time constant factor.
- Similarly we search for equilibrium states and relationships that can be used as a direct input in the financial policy so that we may say something about the cohesiveness and make predictions for parts of the economic reality.
Do you go even further back in time - to even older data?
- With some of the data that we have in macroeconomics we can go back to the 1900s, and in a few areas, such as wheat prices, we have data that goes back to the Middle Ages. Typically, however, we work with newer data. In financial econometrics the analysis of stock prices, interest rates, options and similar financial data are often based on completely new and also very high-dimensional data sets in order to, among other things, be better at lowering the risks. The developed econometric techniques are used by banks, central banks and major finance companies, Anders Rahbek explains.
He adds that to a certain extent one has to leave the classical statistical theory when working with newer dynamic econometric models:
- We use simulation-based inference instead, for instance the bootstrap method, where you, in popular terms, pull yourself up by the boot-strap. Or in other words recycle the data we have, explains Anders Rahbek.
- And we do not have that much data as economics is not an experimental field. You have some data over a certain period of time, to which you can add some noise so as to recreate new data. So what you do is add noise to your system in a controlled manner which allows you to take into account the non-modelled uncertainty.
According to Anders Rahbek the amount of data used in fields such as financial econometrics is very large, and much of the data is defective and has to be cleaned up in order to be of any use; making this type of data collection is very costly.
Is this why you have to draw on foreign numbers instead of Danish ones?
- Yes, but hopefully the mechanisms are essentially the same. Because a large part of the economy is not an experimental science it is the hope of all economic theory building that some recognizable features will prove themselves. And ultimately this is not Danish research, but very much international research. It is our clear goal to publish internationally, says Anders Rahbek, adding:
- The Danish support to "Dynamic models" is given in order to raise our profile internationally. Moreover, the research work - as always in basic research - takes place in collaboration with foreign colleagues. The project is based on, as our work in general, the development and exchange internationally, on bringing people together to create the greatest possible understanding through knowledge sharing.