Science topic

Economic Forecasting - Science topic

Explore the latest questions and answers in Economic Forecasting, and find Economic Forecasting experts.
Questions related to Economic Forecasting
  • asked a question related to Economic Forecasting
Question
4 answers
Hello,
If we look to forecast the trajectory of a time series and we detect breakpoints (especially for small sample size), how should we manage: delete them, impute them...
Regards
Relevant answer
Answer
Given that you have a relatively small sample (less than 30?) but working with a univariate time series model, the simplest approach is to model it as an interrupted time series model as suggested by David Booth above. This is the same as adding a binary (0, 1) exogenous variable to the ARIMA model as suggested by Guy Melard. This assumes that 1) you know where the break occurs; and 2) the break is in the intercept (constant) term.
  • asked a question related to Economic Forecasting
Question
6 answers
Hi,
I am currently working on a research project which aims to develop a forecasting model for prices of agricultural commodities. I am mainly using ARIMA method for this purpose. If my ARIMA Model is such that there exists ARCH effect/Heteroskedasticity in that. Then can I use the same ARIMA model for Price forecasting. Can anybody advise me on this?
Thank you
Relevant answer
Answer
It would be better to control the Heteroskedasticity using a GARCH-type model to forecast.
  • asked a question related to Economic Forecasting
Question
39 answers
If not forecasting, would you replace the forecast with a foresight prediction method?
We are entering the new year 2022 and it is at the turn of the year that many macroeconomic and microeconomic forecasts for individual markets and industries appear. On the other hand, companies listed on the stock exchange are happy to boast of their potential to create value in the first quarter of the year, present attractive forecasts of increasing sales, acquiring new customers and profits in the markets ... to tempt new investors into shares. Is it not sometimes as Peter Drucker stated years ago?
„The best way to predict the future is to create it.”
“Time is the scarcest resource. Unless it is managed, nothing else can be managed.”
What do you think about forecasting and planning the future?
I've added some inspirational leads
Relevant answer
Answer
Economic forecasting is the process of attempting to predict the future condition of the economy using a combination of widely followed indicators. Government officials and business managers use economic forecasts to determine fiscal and monetary policies and plan future operating activities, respectively.
  • asked a question related to Economic Forecasting
Question
33 answers
Some banks conducting an analysis of the creditworthiness of an enterprise applying for a long-term investment or mortgage loan take into account the issues of climate change, if these changes may affect the business profitability of a specific lending business venture. For example, in the case of a hotel application for a long-term investment loan operating on the sea, the level of which can rise and flood the hotel area.
Another example is a hotel located in the mountains, where winter sports tourists come. Climate change predictions may indicate that 10 years of snow will no longer be the place where this hotel in the mountains provides its services. Therefore, the bank may not grant credit due to the forecasted secondary effects of progressive climate changes and, above all, the rising average temperature.
On the other hand, companies are developing which produce components for new power plants producing electricity as part of renewable energy sources, produce electric car equipment components, e.g. electric motors, batteries, etc. More and more innovative startups are being produced as part of cooperation with large enterprises and renewable energy plants Wind turbine type subassemblies, charging devices for electric cars, etc. Other companies manufacture packaging from recycled materials, recycled or from biodegradable materials.
Other companies are developing innovative solutions for automatic sorting of rubbish. If eco-friendly products become popular and the state creates good institutional, legal and financial conditions for the development of such projects, then the process of implementing sustainable green economy based on the green economy concept will be implemented more quickly and business probes will become more and more profitable. Financial institutions, including banks, will gradually take into consideration eco-friendly processes and business activities of clients in concluded financial transactions.
Do you agree with me on the above matter?
In the context of the above issues, I am asking you the following question:
Are there banks and / or companies that take into account forecasted climate changes in their business decisions?
Please reply
I invite you to the discussion
Thank you very much
Best wishes
Relevant answer
Answer
yes i agree
  • asked a question related to Economic Forecasting
Question
8 answers
Will the development of computerized business analytics of large collections of economic information collected in Big Data database systems improve the forecasting of future economic processes?
Please reply
I invite you to the discussion
Thank you very much
Dear Colleagues and Friends from RG
The key aspects and determinants of applications of data processing technologies in Big Data database systems are described in the following publications:
I invite you to discussion and cooperation.
Best wishes
Relevant answer
Answer
More than 2 years have passed since I asked the above question. During these two years, has there been a significant progress in analytics based on Big Data Analytics technology towards using this analytics to forecast complex climate, natural, social and economic processes?
Regards,
Dariusz Prokopowicz
  • asked a question related to Economic Forecasting
Question
7 answers
Currently, it is difficult to define this type of analytic problem. The key issue is forecasting future global problems. It is necessary to collect additional analytical data over the next years and perhaps in about 100 years in huge Big Data database systems supported by another generation of artificial intelligence, it will be possible to forecast what can happen to the planet Earth in the next 1000 years.
In view of the above, the current question is: Will I be able to precisely forecast in the 21st century what will be the future of planet Earth in the next 1000 years?
Please, answer, comments. I invite you to the discussion.
Relevant answer
Answer
Forecasting climatic, geological, natural and other processes that may occur in the next several dozen years is burdened with a very large scope of research error. Despite the progress made in the field of predictive analytics, the impact of the development of civilization on the climate and the biosphere of the planet Earth is still large and growing. Therefore, forecasting the development of climatic, geological and natural processes that may appear in the next several hundred years may border on the proverbial "fortune-telling on tea grounds".
Best regards,
Dariusz Prokopowicz
  • asked a question related to Economic Forecasting
Question
6 answers
Will future Big Data database systems supported by artificial intelligence be used in precise forecasting in order to verify futurological projections?
Currently, it is difficult to define this type of analytic problem. The key issue is forecasting future global problems. It is necessary to collect additional analytical data over the next years, and perhaps in the 21st century, in huge Big Data database systems supported by another generation of artificial intelligence, it will be possible to predict what may happen to the planet Earth in the future.
In view of the above, the current question is: Will future Big Data database systems supported by artificial intelligence be used in precise forecasting for the verification of futurological projections?
Please, answer, comments. I invite you to the discussion.
Relevant answer
Answer
The combination of database and analytical technologies Big Data Analytics and other technologies typical for Industry 4.0, such as artificial intelligence, machine learning, etc., can be used to improve the processes of forecasting analysis in the situation of obtaining and processing large amounts of information and data in real time, e.g. .large amounts of data sourced from many different websites. In the field of this type of analytics, the formula of sentiment analysis is more and more often used, carried out on the basis of the verification of information contained on many websites, many entries, comments, posts posted on online discussion forums and social networks.
Best regards,
Dariusz Prokopowicz
  • asked a question related to Economic Forecasting
Question
20 answers
Apparently, on the financial markets and in macroeconomic determinants of the economic situation in particular sectors and entire economies of developed countries, there are symptoms that suggest a high probability of economic slowdown from 2020 in individual countries and, consequently, in the entire global economy.
Therefore, I am asking you: Do you know the forecasts of the global economic development that would suggest a high probability of deceleration (or possibly acceleration) of economic growth from 2020 in individual countries and, consequently, in the entire global economy?
What are the symptoms of potential changes in the financial markets and / or the scope of macroeconomic determinants of the economic situation in particular sectors and entire economies?
If you know the results of prognostic research in this area, please send links to websites or scientific publications in which this type of prognostic issues are taken.
I wish you the best in New Year 2019.
Best wishes
Relevant answer
Answer
9 May MMXXI
Please read attached article, THE ART OF GREED...
Cordially...
ASJ
  • asked a question related to Economic Forecasting
Question
23 answers
Is there any experience in the use of big data analytics in this area?
Relevant answer
Answer
You should be careful. There is a methodological issue (and I know I am in a minority on this) that calibration of an ABM by "fitting" has the potential to undermine the value of the approach. This is because most ABM have lots of parameters and if they were statistical models, we would already know that too many parameters and not enough data = rubbish. The trouble is, unlike for statistics, we don't have a "formal" way of deciding how many parameters we are "allowed" not to know the values of (but fit) relative to the "amount" of data. I suspect that for ABM we will have to rely on some sort of operational procedure (analogous to sensitivity analysis) to understand this issue. But I do not want to overstate the case. Some fitting could sometimes be legitimate depending on the data and the model - but at the moment we are not really sure when or why IMO, we just do it. (But also, ask yourself, is it a bad design principle to specify a model where you cannot see how to collect the data "for real" and only fit? Sometimes the problem is only practical I agree but sometimes it is definitely bad design.) I have tried to develop these arguments more rigorously here: http://methods.sagepub.com/foundations/agent-based-models. I have also shown that calibration/validation without fitting does not have to be an impossible goal: (though I know this article has many weaknesses.)
  • asked a question related to Economic Forecasting
Question
6 answers
The original series is nonstationary as it has a clear increasing trend and its ACF plot gradually dampens. To make the series stationary, what optimum order of differencing (d) is needed?
Furthermore, if the ACF and PACF plots of the differenced series do not cut off after a definite value of lags but have peaks at certain intermittent lags. How to choose the optimum values of 'p' and 'q' in such a case?
Relevant answer
Answer
You can use auto.arima function from 'forecast' package for R.
Alternatively, if you have many observations, you can try out-of-sample comparison of alternative models with different values of d.
To compare alternative models, you can use the instructions described here:
  • asked a question related to Economic Forecasting
Question
9 answers
What is the growing share of protectionism that limits cross-border trade between major economies in the projected slowdown in global economic growth?
Do you agree with my opinion that in many developing countries the influence of foreign direct, capital and financial investments is significant.
However, the analysis of this process in individual countries results in a significantly different scope and nature of the impact of foreign investment capital.
According to the doctrine of classical economics, all countries should benefit from opening up the economy to foreign investments and the development of trade, including the export and import of economic goods.
However, are all countries always benefiting from this process economic benefits and the process develops faster?
It's not always like that. If all countries benefited from the growth of trade, protectionism, such as the establishment of anti-dumping duties to reduce cross-border trade, would be unnecessary.
What is the impact of foreign investment capital in the globalization era on the economic development of developing countries?
What is the growing share of protectionism that limits cross-border trade between major economies in the projected slowdown in global economic growth?
Are the currently limited protectionist practices cross-border trade is the main factor in the forecasted slowdown in global economy growth?
Please reply
Best wishes
Relevant answer
Answer
Protectionist practice is one of the reason in the slowdown in global economic growth amid Covid-19, however, benefits of so called Globalization & Liberalization have failed to address the Equality, access to equal opportunities & removing poverty in many emerging nations.
  • asked a question related to Economic Forecasting
Question
6 answers
I am performing multi-period optimization (e.g. 2020-2035) of a MSW management system within the European context and I am struggling to find reliable data/methods/approaches for MSW composition forecasting. Most of the references I found link the generation of specific waste materials (e.g.plastic packaging, food waste, etc.) to GDP evolution. In other words, all materials contained in MSW evolve at the same rate as the GDP. This approach does not account for prevention measures for specific wastes (e.g. plastic ban) or the potential increase on the generation of other waste materials (e.g. an increased use of glass packaging). Do you have any idea/reference on this issue? (I am thinking, for example, on the use of consumption patterns to derive waste composition). I appreciate any suggestion provided.
Relevant answer
Answer
Dear Robert,
I suggest that you have a look at MLS Journal, there is a special issue on this topic.
Sincerely,
Antonio
  • asked a question related to Economic Forecasting
Question
4 answers
I have been trying to forecast GDP of Pakistan for FY 2020 to FY 2021 by using a dataset having Energy, Labor and Capital as exogenous variables. The data set has been attached for reference. As its evident from the data that three values of GDP are missing (for FY 2020 to 2022) as I have to forecast these values. Rstudio is treating these missing values as NAs which poses a problem for neuralnet () function. Is there any way for Rstudio to recognize these missing values as dynamic values so that it doesn't prompt an error? Or is there any other way for me to forecast these values using ML/AI in R?
Any help in this regard would be highly appreciated.
Thanks.
  • asked a question related to Economic Forecasting
Question
26 answers
What kind of scientific research dominate in the field of Futurology, forecasting, future, technologies of the future?
Please reply. I invite you to the discussion
Relevant answer
Answer
Foresight of global trends and processes (globalistics).
The common rules of scientific methodology do apply to futuristics (as expertise).
Tech-know-togical evolution is technically more simple to predict than the implied socio-ethical and cultural dilemmas, e.g. medical technology
'Bucky' is still important: https://www.bfi.org/
  • asked a question related to Economic Forecasting
Question
12 answers
Are there banks that take into account the forecasted climate changes related to the ongoing global warming process in the creditworthiness assessment process?
Are there banks that in lending in the field of granting long-term investment and mortgage loans in the process of assessing the creditworthiness of potential borrowers include forecasted climate changes related to the ongoing global warming process?
For long-term investment loans and mortgages granted for 20, 30 years or more, predicted climate changes related to the ongoing global warming process may already be significant.
Please reply
Best wishes
Relevant answer
Answer
None that I know
  • asked a question related to Economic Forecasting
Question
49 answers
What do you think about the possibility of forecasting economic processes based on the analysis of large data sets in Big Data database systems?
Will Big Data help in precisely forecasting future economic processes, including in terms of forecasting stock exchange trends on stock exchanges?
Will Big Data help in forecasting future, next financial, economic and other crises such as climate disasters, weather anomalies, earthquakes, etc.?
Please reply
Best wishes
Relevant answer
Answer
The answer to this straight question is the straight resounding NO.
Data (big or small) represent the past, i.e. what is already happen. We are trying to predict the future using solely the past. Is it always possible?
It is well known that in general the future events are NOT defined solely by their past.
When sensors or social media are used to collect data without planned experimentation, then the data usually contain a lot of noise and correlations.
The future "swan effects" or rare events that were not accounted for or were not happen in the past cannot be predicted no matter how much past data are used.
Only the stable signal (pattern) that can be extended into the future can provide meaningful prediction using some validated mathematical models. Noise is a data pattern that is unstable (disappears) in the longer run. Prediction made using mostly noisy and correlated data will not be reliable, or prediction (say, classification) will be inundated by false positives or false negatives.
In summary, economics, financial markets, climate and natural events are highly nonlinear chaotic systems that cannot be reliably predicted in the long run using only past data.
  • asked a question related to Economic Forecasting
Question
5 answers
Dear Researchers/Scholars,
Suppose we have time series variable X1, X2 and Y1. where Y1 is dependent on these two. They are more or less linearly related. Data for all these variables are given from 1970 to 2018. We have to forecast values of Y1 for 2040 or 2060 based on these two variables.
What method would you like to suggest (other than a linear regression)?
We have a fact that these series es have a different pattern since 1990. I want to make this 1990-2018 data as prior information and then to find a posterior for Y1. Now, please let me know how to asses this prior distribution?
or any suggestions?
Best Regards,
Abhay
Relevant answer
Answer
Let me play the devil's advocate:
You have data for the past 50 years. However, you say that there is a mayor break or change in the pattern around 1990, so that you want to use only the more recent 30 years ... to predict what will be in 30 or 50 years in the future?
I doubt that this makes any sense. Toss some dice. It will be as reliable as your model predictions.
If "phase changes" like around 1990 can happen, they can happen in the future, too. Additionally, many other things can happen that we are not even aware of today. The uncertainty about such things must be considerd. Further, as you don't have any model that might be justifed by subject matter arguments, there is a universe of possibilities, again adding to the uncertainty of the prediction. If you consider all this, you will almost surely find that the predcition interval 30 or 50 years ahead will be so wide that it can't have any practical benefit.
You can surely grab one possible model, select some subset of your data, and neglect anything else, then you can make a possibly sufficiently precise forecast, which applies to this model fitted on this data, assuming that nothing else happens or can impact the dependent variable. Nice. But typically practically useless. It's a bit different when you has a model, based on a theory. Then you could at least say that this theory would predict this and that. But if you select a model just because the data looks like it's fitting, you actually have nothing.
It's important to think about all this before you invest a lot of work and time in such projects! It may turn out, in the end, that your approach is still good and helpful. But many such "data-driven forecast models" I have seen in my life have benn completely worthless, pure waste. Good enough to give a useful forecast for the next 2-3 years, but not for decades.
  • asked a question related to Economic Forecasting
Question
9 answers
I am running ARIMA model with the natural log values of 10 years daily frequency data of a stock market index i.e., Nifty
the variable is integrated at first differencing .
the lag lengths of AR and MA are decided based on partial autocorrelation and autocorrelation functions. the decided lag length is '1' for both AR and MA terms.
my doubt is whether the dependent variable should be level form(i.e., LN_Nifty) or in first differenced Nifty while running ARIMA . Because the variable is integrated at first order.
if i run the ARIMA with the level form of Ln_Nifty, the AR and MA terms are significant.
if the first differenced form of LN_Nifty is used, the AR and MA terms are not signficant .
please anyboday clarify my doubt
thanks in advance
Relevant answer
Answer
Try to use auto.arima function in R package "forecast" to help you in identifying the best model, and also be aware that stock markets data sometimes have volatility clustering so you might need to use ARIMA-GARCH models, not only ARIMA models.
  • asked a question related to Economic Forecasting
Question
11 answers
Is it possible to heal the banking banking supervision of corporate investment banking to significantly reduce the dramatic effects of the next global financial crisis for the national economy and society? Is it too late for that?
Of course, this question should be answered in the negative, that it is never too late to repair the operation of any system that is supposed to serve people. But whether the scale of mistakes made in the past has not generated the unavoidable pursuit of a global financial crisis that is even more dramatic in the negative consequences for entire economies and societies. A crisis that will start with the spectacular collapse of one of the largest financial institutions, a bank or an investment fund. A globally operating financial institution that will lose playing "poker" on international capital markets with other investment banks. Some of these others earn from this crisis by winning this "global poker" and real economies will again plunge into a multifaceted economic crisis, debt crisis, a period of deep recession, rising unemployment and falling income of citizens. Is the capital flow in this way through these games, games in "global poker" on the capital markets played between the largest investment banks is economically effective? Well, it is not an economically effective process, it is a process harmful to economic development. So why are these games in "global poker" conducted? Is it only because in the process of excessive, secondarily realized liberalization of supervisory standards over financial systems implemented in the 90's, allowed to create too large, increasingly globally and monopolistic investment banks? In my opinion, not only because. Not only the scale of operations, not only the large share of capital compared to the financial system and the entire economy is a serious threat and a crisis-generating factor. Also important are the elementary rules of risk management, which are forgotten, ignored at certain organizational levels of the financial institution or financial system management.
Analysis of the origin of the next global economic crisis
Currently, forecasting systems are being developed regarding forecasting future trends of economic processes based on various analytical, not only economic, determinants. Personally, I also support the thesis about the impact of various cosmic and atmospheric phenomena on various events that take place on Earth in the field of economy, economics, politics, etc. On the other hand, because sources of the global financial crisis I mainly researched in terms of progress (or rather lack of it) ) in the field of improving the credit risk management process, implementation of modern IT solutions streamlining these processes, filling gaps in legal regulations developed in financial supervision institutions in relation to technological development of transactional, corporate and investment banking, creation of new derivatives etc., so I add to this type of analysis the issue of the analysis of the process of improvement of systemic management, banking credit risk. Unfortunately, the strong investment banking banking lobby influencing the politics of the world's largest economies is accepted by the government establishment, because monetary policy, periodically regulated lending policy, increasingly liberalized, transactional modernization, electronically and disseminated investment banking are areas treated as "universal magical tools" that can be used as a determinant for economic growth as part of state intervention. In this respect, there is a lack of full information flow in the area of ​​growing credit risk and the fast approaching new global financial crisis between the transactional level of sales of banking products and the level of monetary, credit and financial system security at national and supranational level. According to the demands of the classical economy, liberalism at the transactional level of the sale of banking products should not be limited by state intervention at the level of the entire financial system. But the exception in this regard is the issue of the security of the financial system. If, secondarily, the extremely liberalized principles of systemic security periodically lead to an increasing financial crisis in investment and credit banking, why should the costs of these errors be spread across entire economies? Why is it that investment banks in economic crises, which often cause them to earn money from them, and the costs are repaid by entire societies, people lose their jobs and many years of experience of their lives? Therefore, because these investment banks have genuinely monopolized the systemic credit risk management system. They no longer serve the economy, but try to shape economies according to their investment strategies. The question that now arises is whether this harmful and crisis-provoking process can be reversed, corrected before the emergence of the next global financial crisis? Is it already too late and only one of the next financial crises, which will lead to the collapse of not one but a few major banks and investment funds will make it possible to repair damages resulting from errors that politicians began to make in the 1990s liberalizing then secondary issues of banking supervision systems? If it is only in the situation of the next global economic crisis, then how dramatic are the consequences for entire national economies, for societies, for people? It is not easy to predict this issue, but it is almost certain that it will be very dramatic, above all economically and socially, but perhaps also politically, strategically and militarily for many countries.
Relevant answer
Answer
Analysis of the origin of the next global economic crisis
Currently, forecasting systems are being developed regarding forecasting future trends of economic processes based on various analytical, not only economic, determinants.
So, how do you or anyone for that matter even have a hang of what is NEXT Global Financial crisis going to be about?
Any Forecasting model is close to being useless and analyst have over confidence/bias in their findings because they are either not aware of do not appreciate the fact that we are living in a highly complex Systems which is getting even more complex non-linearly.
Any forecasting model that try to drive order of chaos in complex systems in a fallacy. It's very disappointing that academicians and economists disregard this very fundamental understanding and try to fit their observations in a way that their pre-conceived conclusions remain intact.
  • asked a question related to Economic Forecasting
Question
8 answers
Large amounts of information downloaded from comments, entries, posts from social media portals are processed in Big Data database systems to determine, for example, consumers' awareness of the offer of products and services of specific companies. This type of information is of great importance for the planning of advertising campaigns informing about the mission, idea, product offer, and the usability features of the company's offer. This type of data may be important for forecasting the changing preferences of consumers regarding the offer of specific companies.
In view of the above, I am asking you: To what extent the Big Data data sentiment analysis from social media portals can be used in forecasting the company's development?
Please reply. I invite you to the discussion
Relevant answer
  • asked a question related to Economic Forecasting
Question
8 answers
Technologically it becomes possible, only the level of precision of results and the usefulness of this type of analysis are still undetermined. It is connected with continuous improvement of tools, analytical systems, eg Business Intelligence, which in analytical processes would use large data sets, historical information and current real time. Improvement of these research techniques will in the future allow to create data based on the sentiment analysis of data contained in the Big Data resources, tools enriching the forecasting methods used to describe the trends of specific economic variables, including primarily macroeconomic ones. Does anyone from you conduct research on this topic? If so, I invite you to cooperation.
Relevant answer
Answer
Personally I don't do research on this topic, but I can suggest only an interesting paper from some Italian researchers:
  • asked a question related to Economic Forecasting
Question
2 answers
As per literature review, some of the studies included weather variables as input variables whereas there are also studies which treated weather variables as determining factors of technical inefficiency. Any econometric/ economic reasoning to decide this?
Relevant answer
Answer
The weather variable can be treated input or deterministic factor and it will depend on the crop yield due to weather conditions, if weather is affecting the yield positively then it should be treated as input factor but too much dependency (eg. Heavy rain) can ruin the crop yield then it will consider as determine factor of inefficiency.
  • asked a question related to Economic Forecasting
Question
62 answers
What is your opinion regarding to the above question?
It is important to note, therefore, that demographic projections should not be confused with economic forecasts. Changes in the number of people, families, or households do not necessarily relate to the social and economic well-being of an area.
Relevant answer
Answer
The Relationship Between Economic Growth and Population Growth. ... On the other hand, if population growth affects per capita output growth, higher population growth rates would contribute to either higher or lower overall economic growthdepending on the nature of its effects on per capita GDP.
  • asked a question related to Economic Forecasting
Question
7 answers
I want to forecast volatility with GARCH, EGARCH and GJR-GARCH. How do I obtain the RMSE, MAE and MAPE. The problem now is that I am using a mean equation and the values reported in the little table on side are for stock returns. I need the same thing for volatility. Any idea except from calculating them by hand?
Relevant answer
Answer
The command accuracy() under forecast package in R returns a range of forecast accuracy measures including:
  • ME: Mean Error
  • RMSE: Root Mean Squared Error
  • MAE: Mean Absolute Error
  • MPE: Mean Percentage Error
  • MAPE: Mean Absolute Percentage Error
  • MASE: Mean Absolute Scaled Error
  • ACF1: Autocorrelation of errors at lag 1.
You can generate the necessary forecasts in EViews and use forecast and actual values to calculate loss measures in R.
accuracy(f, x, test = NULL, d = NULL, D = NULL, ...)
where argument f - forecasts and x- is realized values.
  • asked a question related to Economic Forecasting
Question
5 answers
I do know that Eviews has an add on for this model, But I am using a old version of the Eviews and therefore the add on feature cannot be incorporated in the same.
Relevant answer
Answer
The DCC model has become a standard model in many Econometric software: RATS, G@RCH under OxMetric, Microfit, MATLAB, rmgarch package in R.
  • asked a question related to Economic Forecasting
Question
5 answers
ARIMA models, from what I know, are atheoretical models in the sense that they generally don't provide us with meaningful economic interpretation of why a process is behaving the way it does.
However in population forecasting, can ARIMA models be interpreted in a meaningful way?
i.e.
  • can AR(1) process can be interpreted as: "population growth in period t is a function of population in period t−1"
  • can MA(1)process can be interpreted as: "population growth in period t is a function of some policy or ongoing event (a country's policy on immigration or emigration) which has occurred in t−1"
Relevant answer
Answer
Hello Jacob
I am not a time series expert, but I have used them in my forecasting work. You ask two questions:
  • can AR(1) process can be interpreted as: "population growth in period t is a function of population in period t−1"
I would not expect this to be the case, as there is no reason why population size should influence population growth.
  • can MA(1)process can be interpreted as: "population growth in period t is a function of some policy or ongoing event (a country's policy on immigration or emigration) which has occurred in t−1"
As a direct effect, a change of policy in period t-1 can be expected to also affect period t. However, a change of policy is not specifically contained in the population data for period t-1. Knowledge of the policy is additional information, which would have to be explicitly incorporated in the model. A substantial change in population size in period t-1 could be temporary (a migration event or change of rules that affects the pace of migrant acceptances - i.e. a tempo effect); this would be noise or could also be modelled by a time-specific dummy variable. So, I would answer 'no': as in economics, time series models involving only one step are atheoretical in population forecasting. However, it may be the case that large-scale in-migration of young adults in period t-1 generates additional population growth in a subsequent range of years through childbearing and the in-migration of fiances, spouses, children and other dependents, and this may possibly be modelled by time series, though forecasting would involve high uncertainty. It may be possible to establish more meaningful interpretations of population size with lags of a generation or so, for obvious reasons, but the age range of childbearing (even in low fertility populations) would weaken the association.
  • asked a question related to Economic Forecasting
Question
6 answers
For my research paper, i have to study the relationship between stock returns and macroeconomic variabels by using a suitable regression model. (stock return is DV and macroeconomic variabels IVs)
the stock returns data is available daily-wise and that of macroeconomic variables monthwise.
how to resolve the problem of different frequencies of the stock returns and macroeconomic variables?
thanks in advance
Relevant answer
Answer
Hi Srikanth,
If you do not have to stick to the regression model (I am not sure if it is the best tool to analyse daily time series of financia data), you may find the spline-GARCH model of Engle and Rangel useful. See: Engle, R.F., Rangel J.G."The spline-GARCH Model for Low Frequncy Volatility and Its Global Macroeconomic Causes".
  • asked a question related to Economic Forecasting
Question
9 answers
Doing research in the field of finance and investment, I found it difficult to understand various models, such as ARCH; GARCH and many more used in the contemporary research articles.
I am searching such a book(s) which can explain these models and methods in a very simple manner accompanied by example and data set. As I am not an expert in these areas of analysis, I want to learn those model from the very basic and beginners level.
Relevant answer
Answer
wooldridge introductory econometrics is also good
  • asked a question related to Economic Forecasting
Question
3 answers
Bubbles have a very tightly circumscribed economic meaning, but they have also been studied by scholars (and methods) from humanities and social sciences other than economics. I am seeking your expert help in identifying the very best multidisciplinary writings on bubbles.
Relevant answer
Answer
Hi,
You may be interested by this paper:
Birth or burst of financial bubbles: which one is easier to diagnose?
 By: Demos, G.; Sornette, D. Quantitative Finance. May2017, Vol. 17 Issue 5, p657-675. 19p. 5 Charts, 8 Graphs. DOI: 10.1080/14697688.2016.1231417.
Best regards
  • asked a question related to Economic Forecasting
Question
7 answers
I want to remove the trend from Time series for monthly data of water consumption.
I tried both methods
1- detrend 
2-moving average
which method is appropriate for this issue?
please see attached files
Thanks in advance for you cooperation  
Relevant answer
Answer
MA is one way of detrending. In macro data, one of the most popular detrending technique is HP-filter, which is actually an optimal MA.
  • asked a question related to Economic Forecasting
Question
7 answers
The Middle Eastern countries have experienced budget deficit due to lower oil revenues. Since their budget deficit is expected to persist for some periods due to lower forecasted oil prices, what options they to finance budget deficits in the short as well as long run?
Relevant answer
Answer
The oil prices remain sluggish for the last 13 months when this issue was raised in December 2015. Saudi Arabia responded with its long-term Vision 2030 economic transformation. Not sure how this economic blueprint will be implemented. Other oil-dependent economies are also struggling especially those that have slowly utilising their cash reserves.
The government revenue displacement as a result from the decrease in resource rents has prompted many government to introduce new taxes. Other measures may include government austerity measures but these measures may have some adverse impact on the growth.
The more diversified the oil economy, the better it is in terms of absorbing the oil price shocks. So the key operative word is diversification.
  • asked a question related to Economic Forecasting
Question
3 answers
I am currently working with 60 observations time series data in eviews.
After applying log to my model, the residuals were correlated and heteroscedastic. In order to treat this issue i created logs. After creating 2 lags i got my ADL(2) model which showed up the problem of non normality of residuals.The same was treated by taking another lag and thus ADL(3) was formed.
The problem popped up when Lag selection was to be done and information criterion was used. Thie information criterion approved that ADL(1) should be taken. Whereas that model has NON NORMALITY AND SERIALLY CORRELATED residuals. Please guide what should be done.
which ADL model should be considered?
Relevant answer
Answer
The information criterion is telling you (roughly) that adding lags 2 & 3 does not give a worthwhile improvement in fit. Behaviour of the residuals is (largely) a separate matter. Serial correlation could be a clue that additional explanatory variables are required. If you don't have them then retain the lags but note that the model may be under-specified.
Do take a look at the residuals time-graph and histogram to assess the nature of non-normality. It might be that you have some residual outliers and adding 0/1 dummies at those dates might then cure the non-normality and maybe even the serial correlation. You would then need to consider whether the dummies indicate unusual events (if so then try to identify them)  or some missing variable(s).
  • asked a question related to Economic Forecasting
Question
4 answers
In my research, I want to calculate realized volatility, which necessitates the availability of intraday data series. unfortunately the data at hand are only daily frequency. 
Relevant answer
Answer
The dayly changes can be seen as a sum of intraday (e.g. hourly) changes. Under the assumption that these are independent random variables, one could get the hourly variance by dividing the dayly variance by the number of dayly hours. This procedure can, of course, be supplied for any time units per day. This variance can be seen as a maximum, because if the changes are not totally independent, the variance would be lower. The calculation of the volatility would be more complicated, but I am sure that random walk theory has solutions for that. A propos, you could use Bloomberg data to find the relation between dayly and interday data or to see if the calculation according to my proposal leads to similar results. It would also make it possible to detect intraday patterns (if these exist).
  • asked a question related to Economic Forecasting
Question
6 answers
Do you know any articles where the number of trades on stock exchange is predicted? I am particulary interested in the prediction of number of trades in 1-5 minute intervals.
  • asked a question related to Economic Forecasting
Question
3 answers
The company i42 GmbH, Mannheim, developed MoneyBee: a system to predict stock market values, basing on artificial intelligence (neural networks), distributed computing and different applications to optimize the input- and output-data (e.g. genetic algorithms, statistical methods).
Is there any similar project like MoneyBee?
Relevant answer
Answer
 Thanks for the answer. Is it an open-source project?
  • asked a question related to Economic Forecasting
Question
5 answers
In International trade, to know about economic progress for the future. we have to measure by forecasting analysis. Could you possibly explain forecast of the economic in International trade?
Relevant answer
Answer
Two approaches for forecasting future in economics are possible. The first approach is used more frequently and it uses statistical methods. Future is extrapolated using past trend. Technically, there are many methods of implementation, but neural network just makes this forecast a bit more complicated. The basic principle here is that future dynamics is the function of the past dynamics.
The second approach is fundamental, but is used not so often. It may be based on calculation of general equilibrium and convergence to it. However, we often have the factors that emerge from nowhere (the so called “black swans”) that work as shocks for economic systems. Analysts usually write a lot about that, but it is technically difficult to calibrate the scale of such effects even if one knows that such event is highly likely to happen.
  • asked a question related to Economic Forecasting
Question
3 answers
How to measure the so-called "upgrading of consumption" ? 
Relevant answer
Answer
1) The income elasticity of demand for luxury goods in China. Not just on the sensitivities of the Chinese on existing product lines but also the enhanced savviness in relation to more innovative lines or emergent trends that seem to undermine existing preconceptions about consumption of luxury items (e.g. fashion).
2) The income elasticity of demand for health and wellness products.
3) The income elasticity of demand for long haul overseas travel (i.e. outbound tourism in China)
The premise and assumption is that these three categories of goods seem to be in higher demand as economies advance with wealthier average citizens. So if we can demonstrate a corresponding increased demand in these categories due to shifts in wealth, it may help us gauge the "consumption upgrade" in China.Though even in the advanced economies there is a structural change towards experiential products rather than ownership.  One should be cautious to take income elasticities at face value. The so-called "accessibility versus ownership" phenomena which characterizes the sharing/experience economy.
  • asked a question related to Economic Forecasting
Question
3 answers
I'm doing some research using som relatively high-frequency(quarterly or monthly) panel data set. In this data setting, seasonality and data availability are both concern. I want to refer to some classics in the area but find simple google is not enough. If you are familiar with this area, please throw some literatures (application is enough) for me . Thanks
Relevant answer
Answer
The classic article in dynamic panel analysis is Arellano and Bond (1991)
Here they outline the most efficient method of dealing with endogeneity generated by including lagged dependent variable in your panel model.  See Cameron and Trivedi "Microeconometrics Using Stata" for how to implement the model.
  • asked a question related to Economic Forecasting
Question
2 answers
For example, how to define a monetary structure of a car industry in the USA or any other country, which may consist of production of rubber, steel, glass etc. In other words, I'm interested to know how the Leontiev input-output intersectional tables are formed. Do you know any websites where such statistics is available?
Relevant answer
Answer
Perhaps you have checked the SNA 2008 and ESS 2010 guidelines for the input-output methodological details?
Also check the Eurostat input-output manual and tables of member-states:
  • asked a question related to Economic Forecasting
Question
2 answers
This is a very general question. I have seen a lot of repeated public goods game which have different round numbers. Some have more than 20 rounds, some are less 10 rounds. To test for public goods contribution in different mechanisms, do experimental economists have a rule of thumb to determine how many rounds is enough in one treatment? Thanks in advance.
Relevant answer
Answer
In my opinion, it is never a good idea to have a fixed number of rounds. If you do, you will have two different kinds of effects: learning effects in the starting rounds of the game and end effects in the final rounds of the game. In the final rounds, participants tend to behave more selfishly. Refraining from being too selfish makes sense if you do not want to destroy the prospects of long-term future cooperation. If the future is very brief, this incentive disappears. You can prevent end effects (to a certain extent) by not specifying the number of rounds N beforehand, but by terminating the game with a certain probability alpha per round. If you choose alpha=1/N, the expected number of rounds is N, but at every stage of the game the expected time horizon of the game is still the same.
I would make the (expected) number of rounds dependent on the complexity of the setting. In complex settings, it may take a long time before the players understand the situation and before they can make sense of the behaviour of the others. In other words, there will be a long learning phase, that is, it takes a long time before a group of players arrives at a stable pattern of behaviour. You can reduce this phase by changing the description of the game or by giving the players the opportunity of gaining experience (e.g. in separate training sessions). But be careful that this does not interfere with the intentions you have with your experiment. In very simple settings, it does not take more than 3 to 5 rounds before most groups settle at a stable behavioural pattern. If you are mainly interested in such equilibrium behaviour (studying the learning phase might actually be more interesting...), then it suffices to run, say, 10 rounds on average, giving you about 5 rounds for studying this equilibrium behaviour. Increasing the number of rounds does often more harm than good: participants get bored by the situtation and start doing "strange" things just in order to create novel situations.
Franjo Weissing
  • asked a question related to Economic Forecasting
Question
3 answers
I am new in the field of developing forecasting models. 
I want to develop a forecasting model to estimate the electricity prices of IEX, India. I read some research papers and found that work on MSARIMA-EGARCH has already been done.
I want to know about some suggestion about any other model from GARCH family which can be used. 
Please also point out the important points which should be kept in mind while developing a model as I had not developed any till now.
  • asked a question related to Economic Forecasting
Question
1 answer
One has to compare the delicate economic reasoning which led Keynes to assume that interest did not affect MPC, or the complex considerations on which Hicks constructs Capital and Time, with the modern axiomatic method that reasons downwards. The former seems to retain a grasp of economic reality even after mathematical cloaking, while the latter seems arid.
Relevant answer
Answer
It IS arid.  To "assume an axiom" and then apply reason to develop a theory from which you then proscribe policy, as so many mathematical economists of the post '70's generation appear to do, is no different to the approach taken by the scientists of the Spanish Inquisition and those impositions of belief upon others which are no more than a belief in the rightness of one's own beliefs.
It is not only arid, it is also totally illegitimate in an academic context.  Such an approach to be presented, and even worse when it is accepted, is a demonstration of the extent to which the mainstream of academia has been infiltrated and subverted by commercial, political and other selfish interests seeking fraudulent 'endorsements' and 'validations' from supposedly authoritative sources for their own exploitative schemes.
It should be a criminal offence.....
Rob
  • asked a question related to Economic Forecasting
Question
9 answers
Can anyone assist me to get research papers on financial planning and forecasting?
Relevant answer
Answer
Abrahama: You can find a number of papers on the psychological basis of financial planning for retirement on my ResearchGate page.  Or, visit my website at: http://psychology.okstate.edu/faculty-page?id=35   Best, Doug Hershey
  • asked a question related to Economic Forecasting
Question
3 answers
I want to examine the implication of ethanol in India and its impact
Relevant answer
Answer
Farah,
You can use an univariate model, like ARIMA, or a multivariate model, like VAR.
Good job.
Cleiton
  • asked a question related to Economic Forecasting
Question
5 answers
I am preparing an introductory guide and course on revenue forecasting for a developing country. While I have a depth of experience in this field in more developed countries, I want to make sure that i cover the basics very well.
I am looking for advice on texts and articles with broad coverage, rather than narrow coverage of specific technical topics and applications.
Thanks, Michael
Relevant answer
Answer
Thanks, John.  There was a less explicit review of forecasting performance by the NZ Treasury in the late 1990's. 
  • asked a question related to Economic Forecasting
Question
3 answers
1. i want to forecast health care expenditure(HCE) for 2020 and 2030 for Asian developing countries. I am using annual data ranging form 1995-2014 n the following variables per capita HCE, GDP per capita, education, life expectancy at birth, population under 15 and above 65 years and some other variables.
2. How can i get the future values for GDP and HCE like2015,2016, 2017.....?
Appreciated if some one tell me the codes in E-Views or Grtle or SPSS
3. How can i forecast HCE and GDP ?
Relevant answer
Answer
Projecting the future using data from the past is a dangerous game with a question like this, because the effect of "era" has an enormous impact on something like HCE, and it is unlikely that data from 1995, 1996, 1997 have almost any relevance to the amount of HCE that will occur in 2020 and beyond (things have just changed so much).  It is certainly something that we often resort to but I have some serious concerns about the simplistic approach you seem to be taking.
One very important thing to keep in mind: many of the variables you have collected DO have an effect on HCE but it will not be immediately apparent.  The classic example of this is smoking and cancer incidence: a decrease in smoking TODAY does not mean lower cancer incidence TODAY, but most likely lower cancer incidence in 10, 15, 20 years from now.
An aging population  may have immediate effects on HCE.  But changes in education and life expectancy at birth are not likely to manifest themselves immediately.  The statistical model you create to project future HCE in 2020-2030 should account for this.
  • asked a question related to Economic Forecasting
Question
4 answers
In the Global Knowledge Economy, the role of human talent reigns supreme.  Future national performance in Human Development Indices and GDP growth is determined by the talent pool available to nation states. Economic Development policy-making and growth projections are best girded with relevant Labour Market Information and Intelligence.  Which states are the leading avatars of LMIS globally ?  What are the features and characteristics of their LMIS ?  How does one objectively measure, compare and contrast national LMIS's ?
Relevant answer
Answer
The Canadian government is currently reforming its LMI service, trying to improve something that was already efficient, focussing on international best practices. The new system should be available shortly, but no evaluation will be possible at the medium term ...  
In a federal system, like Canada, the development of LMI systems must be done with regional governments, which slow down the process, but the last iteration will give more power to the central government in the production of LMI.  
You are right, LMI now is clearly associated with learning, and the Canadian government is looking for the best ways to connect both.
  • asked a question related to Economic Forecasting
Question
9 answers
Hello,
I want to do a forecast project on inflation for one of my classes, however I have some questions for it. Firstly, I have data which exhibits seasonal and trending pattern. 
I seasonally adjusted my data by using x-13 Census in Eviews, though just very minor changes occurred and it stills seem seasonal according to ACF and PACF. After this, I took differences of observations at lag 1 and now I am confused. My data points were walking around 6 and 10. Now it is between -1 and 1. When I apply Winter's Method to differenced data, I get results between -1 and 1. How can I report my forecasts in normal level, that is between 6 and 10. If I use ARIMA, it automatically take differences and it is okey. The only thing I have to specify is base and seasonality parameters. However, for Winter's Method, how should I proceed?
Relevant answer
Answer
In Eviews, if you have a series (FD) of forecast first differences then "SERIES FL=@CUMSUM(FD)" will produce the forecast levels.
  • asked a question related to Economic Forecasting
Question
5 answers
Hello,
lots of papers in this property cycle research use statistical forecasting methods to predict the property cycle.[1][2]
However, some researchers argue that the property cycle is deeply dependent on the fundamentals.
Still, I have not found any specific models, which use fundamentals as input(I could imagine these inputs would be GDP, interest-rate, price-to-earnings ratio etc.) to measure the property cycle.
Any papers or research that you could suggest that use fundamentals to forecast property cycles?
I appreciate your replies!
References:
Relevant answer
Answer
I guess you are looking at the property sector in a general sense, as property markets are usually fragmented with unique local fundamentals (residential property, commercial property which itself can be broken down into retail, office, warehouse etc). I would have also thought any consideration of fundamentals would need to take account of landlord and tenant changing sentiments, so inclusion of behavioural factors may help.  But using that to forecast future property cycles would require quite a lot of subjectivity I would have thought. 
  • asked a question related to Economic Forecasting
Question
9 answers
I have collected last 10 year's monthly average prices of an item.I am working on a forecasting model.I used 2005 to 2014 data to forecast prices for 2014-15.Forecasts I got from Winter,Arima,TBATS and naive approach are not at all close to actual values.I have attached monthly price plot.
1) Which model should I use? 2) How can I improve my accuracy.
Output:
Actual Winter ARIMA TBATS
679 668 667 671
654 663 639 670
657 659 646 668
653 654 651 667
685 649 656 665
687 644 661 664
689 640 666 662
691 635 670 661
695 630 673 660
731 625 677 659
751 620 680 657
705 616 683 656
Note: I am using R.
Relevant answer
Answer
I suspect it will help many of us if you add more detail.  The graph is interesting; however, without knowing the variables, it can be challenging to understand.  For example, I created an Excel spreadsheet with your output and I created a graph of it adding time for reference here, which seems appropriate (and if it is not, please say so).
If I am correct when looking at the graph of the "Actual" it looks like seasonality may be occurring.  For example, time periods 10 and 11 are considerably above the others, while 2 through 4 are somewhat less. The contiguous nature of these suggests some "seasonal" effects.
If this is correct, and it may not be, consider adding dummy variables to account for weather, holiday season, or other factors that may be in play.  Including time as a variable, like I have have set it up might be useful as a variable if there is a trend.  
The ARIMA results are close to what I expected since it often picks up turning points.  Of course, that does not necessarily mean ARIMA models are the best for forecasting; however, if turning points are critical, ARIMA models can be helpful.  Several of our articles use these and if you are interested, I can guide you to some of them.  
I hope this helps and if I did not understand, adding more details might help me and others. 
Best,
Rob
  • asked a question related to Economic Forecasting
Question
3 answers
I have supplied 2 stationary series (as per adf,test) to build a VAR model with appropriate lag order. I got the results. But when I tested for stationary of this VAR model using 'roots' command in R that tests for values of eigen values, it shows that VAR model is not stationary.
How can it happen and why?
How to take on this issue?
Relevant answer
Answer
Probably the series is just NOT stationary,m even if it 'looks' like one... . 
  • asked a question related to Economic Forecasting
Question
1 answer
Hello,
I read several papers regarding forecasting methods [1]. One of the most recent works is "AN EVALUATION OF THE USE OF COMBINATION TECHNIQUES IN IMPROVING FORECASTING ACCURACY FOR COMMERCIAL PROPERTY CYCLES IN THE UK" from Jadevicius[2].
He argues that combination forecasting is currently the state of the art to forecast property cycles. However, his proposed model uses a property index as an independent variables and he tries to find a model to forecast this index. To make his forecast more stable, he uses two methods and "simply" said combines them for a more accurate forecast.
However, I was wondering if there are more sophisticated methods than the use of combination forecasting?
I would apprecaite your replies!
References:
Relevant answer
Answer
The research in combination forecasting indicates that forecasting is still an art. It takes forecast from standard theories and use an average. It also means that forecasting model is validated or falsified only by the accuracy of the forecast, and not by the techniques or assumptions used. We judge a technique by the results it produces, and this cannot be overcome maybe because the future is unknown. But some people have a taste for technique such as parametric vs. nonparametric. With the failure of forecasts recently people are blaming the quants and moving more towards judgmental forecasts. However, statistical works using the "Generalized Extreme Value Distribution" (GEV) that corrects for skewness and kurtosis look promising.
  • asked a question related to Economic Forecasting
Question
4 answers
Hi all,
I have monthly data for several years and for all regions of my country. The task is to make forecast basing on this data and compare the result with usual time series models. Does anybody has an idea about references? Are there any papers were such forecasts for fiscal parameters were made? Will be grateful for any help, links, thoughts.
Relevant answer
Answer
Hi Artem Vdovychenko
Look at the followink ling. You find many papers about "time series prediction"
  • asked a question related to Economic Forecasting
Question
20 answers
China's global economic activities account for 12% of global GDP, yet the Chinese currency is used only in about 2 - 3% of global trade volumes. Currently, the US dollars, Euro and Yen still dominates the international trade scenes.
-What are the implications of the "internationalization" of RMB?
-How would the US dollar and Euro be affected?
-What challenges are/will be faced by China if RMB is internationalized?
Relevant answer
Answer
Dear Bruno & David,
interesting is the question of the Central Bank. The Fed can make use of the monetary policy with the operation of the monetary base (increase in the money supply by buying bonds) to finance the deficit. So the US has bailed out the banks and have come out of the crisis.
Until recently, the European Central Bank could not buy the bonds to cover the deficit of the countries in crisis. The EU countries had only the lever of fiscal policy. To cover the deficit, the income and assets of households and businesses have been squeezed (very high levels of taxation). The latter, by reaction, have opted for the liquidity preference and have fallen into a liquidity trap. The result was the reduction of the velocity of circulation of money, and the deepening of the crisis with firm failures and higher unemployment.
Moral: it is not enough that there is a central bank, but it must have the power to expand the monetary base and then to do a real monetary policy. If a country do not have this power, they do not trigger neither investments nor consumptions and the Keynesian multiplier can not operate.
Have a good night. 
  • asked a question related to Economic Forecasting
Question
4 answers
Recent research suggests that Bayesian Model Averaging (BMA) is a useful method for combining forecasts. I am looking for prior evidence on the relative out-of-sample forecast accuracy of BMA compared to the simple average. Can you point me to relevant studies that provide empirical evidence? I am looking for evidence from all fields, although my particular interest is in social science problems.
Relevant answer
Answer
Hi Andreas, I know a few empirical case studies coming from supply chain forecasting literature combining statistical with judgmental forecasting, all of them using empirical data from different companies.. Initialiallly, Blattberg and Hoch proposed the 50% manager, 50% statistical. (Blattberg RC, Hoch SJ. 1990. Database models and managerial intuition: 50% model + 50% manager. Management Science 36: 887–899. )
Then, Fildes et al, they proposed an optimal model, (Fildes R, Goodwin P, Lawrence M, Nikolopoulos K. 2009. Effective forecasting and jugdmental adjustments: an empirical evaluation and strategies for improvement in supply-chain planning. International Journal of Forecasting 25: 3–23)
And finally, we identified even non-linearities in the combination that could improve the forecasting accuracy (http://onlinelibrary.wiley.com/doi/10.1002/for.1184/pdf)
JR
  • asked a question related to Economic Forecasting
Question
1 answer
I already have written one article relating to this issue. I want to know do you have the same experience?
Relevant answer
Answer
EVALUATING FORECAST: Forecasting techniques are evaluated on the basis of its ability to match the forecast value to the observed value: (θ^ - θ). A better forecast technique is one which could better predict the data. Statistical tests are tests used to evaluate the forecasting techniques or function. Most statistical tests assume that the data is normally distributed. However, in real life not all data are normally distributed. These data could be transformed to approximate a normal distribution so that generally applicable tests statistic (requiring approximate normal distribution) could be used. Fuzzy data is a case in point. Box-Cox transformation is a method on point.
BOX-COX TRANSFORMATION: The transformation is given by:
T(Y) = (Yʎ - 1) / ʎ
If ʎ = 0 use the natural log method:
X'ʎ = [eʎ log(x) - 1] / ʎ
DOES BOX-COX TRANSFORMATION IMPROVE RESULT OF FORECASTING? No, it does not improve the "result" of forecasting. The Box-Cox transformation is a tool to turn (transform) non-normally distributed data into an 'approximately normal" by means of correlation coefficients plot analysis. The normality here is not the data distribution, but the linearity of the series of normality plot of the correlation coefficients of the series of plots. This provides a standard against which the proposed forecast function may be assessed. It is an assessment tool, not a tool to "improve a forecasting technique." What does it assess? Given a fuzzy data set that is difficult (or seemingly impossible to run a regression for the forecast), the fuzzy data is transformed by equation (1) or (2) from which series of correlation coefficients are obtained. Then, these correlation coefficients are plotted---hopefully, a straightly would emerge and it is from there that a normal distribution is approximated for purposes of forecasting. Thus, Box-Cos improves a forecasting technique, not the result.
EXTENT OF IMPROVEMENT (EFFECT OF TRANSFORMATION): The transformation allows approximately equal to a situation as if the  "fuzzy data set" is approximately normally distributed. If after the transformation the plotting of series of correlation coefficients is still not a straight line (see link below), then the improvement of the forecast affected by the transformation is not optimal. One needs to look at other alternative techniques to deal with the fuzzy data set. See attached articles.
  • asked a question related to Economic Forecasting
Question
32 answers
Hey everyone,
I want to create an AR process to forecast electricity generation by photovoltaic plants. I have the data of three sample photovoltaic plants to develop identify the best number of AR-lags. Later I want to use that model to predict the electricity generation of over 300 photovoltaic plants.
I started with determining the order of my AR process by calculating AIC and BIC for each sample plant. Then I eliminated insignificant lags for each sample plant.
Now I am left with different significant lags for each of my three sample photovoltaic plants:
Plant 1: Significant Lags are 1, 2, 3, 4, 23
Plant 2: Significant Lags are 1, 2, 4, 23
Plant 3: Significant Lags are 1, 2, 3, 4, 20, 23
I've got two options now: Either I can include ALL of these lags in my final model, or I can only include lags that are significant for ALL of my three sample plants.
I would be very grateful to get your opinion on this matter.
Thanks!!
Relevant answer
Answer
In such a case, using an aggregate model will not probably produce sensible results. In similar cases of electricity consumption using hourly data, we have estimated one model for each hour of the day, which means that you have to estimate 24 models. It may be time consuming, but the dissaggregated results were much better. In my experience, aggregation tends to distort the dynamics in the data.
  • asked a question related to Economic Forecasting
Question
13 answers
I want to fit a SARIMA model to some photovoltaic electricity generation time series to develop a forecasting model. Since the data are highly seasonal I differentiated two times to get stationary data. The kpss test now also tells me I have stationary data.
At the moment I am using 2 AR lags, 2 SAR lags, 1 MA lag and 1 SMA lag with a seasonality of 24. This configuration has so far created the best forecasts in comparison to many other configurations i have tried. When looking at the ACF and PACF of the residuals of the model there are no significant lags (except for lag 24, which I cannot seem to get rid of).
When I estimate my coefficients the value of the MA and SMA coefficients is exactly 1. This means I have a unit root in my moving average part of the process, right? And this is bad, right?
How can I get rid of this unit root while maintaining the accuracy of my forecasts? I have already tried differentiating my data only once and eliminating the MA and SMA lags so that I only have 2 AR and 2 SAR lags. The forecast accuracy is not as good as before but acceptable but the ACF and PACF both have many significant lags now...
Is it better to have a model without unit root and therefore a less accurate forecasat? Or is it better to have the best forecast quality I can get while maintaining the unit root?
Thanks a lot!!
Relevant answer
Answer
Thank you everyone for your answers! Do I have to use the ADF test on the residuals or on the acutal values of my time series?
  • asked a question related to Economic Forecasting
Question
3 answers
diaspora direct investments (DDIs) are international flows similar to FDIs 
Relevant answer
Answer
ISSUE: Diaspora Direct Investment (DDI)
"Diaspora" refers to people who emigrate (go out) of their home countries to seek better opportunities or education abroad. This out-going is called diaspora. After they are successful overseas, they return to the home countries to invest---bringing with them money and knowledge. This part is called DDI.
MODEL: In general, DDI model is been proposed by prior studies (see attached articles: USAID) as:
Yddi = B0 + B1X1 + ... + B5X5 + e
... where B0 = intercept; X1 = brain gainer; X2 = technology transfer; X3 = capital investor; X4 = catalyst; and X5 = commercial diplomacy.
PREQREQUISITES: In order for X1, X2, ..., X5 to occur, the following must be in place: (i) institutionl reform in the home (old) country to provide equal treatment to DDI and FDI; (ii) financial incentives; and (iii) political risk insurance.
REFERENCES: See attached 3 articles with case studies.
  • asked a question related to Economic Forecasting
Question
29 answers
Hello,
I am currently working on my thesis - my model is as follows:
LGDP=LFDI +LGFC +LIKG + LIIG+ LICG +LLF +D1
The variables are integrated I(0) and I(1), so l must use ARDL. However, I have a big problem because my data are only available from 1990 to 2012 (23 years).
My questions are:
Is it possible to use the ARDL approach?
How can I get lag length of multiple variables by EViews 8?
Relevant answer
Answer
ARDL: Yes, given that the variables are cointegrated. You should check this property.
23 years: In principle, you should collect as much data as possible. However, in your appication 23 data points should be sufficient. Note that you have annual data, i.e. 23 observations cover a rather long time span (in contrast to 23 months, for example).
Lags: Because of the annual frequency, only short lags (1, perhaps 2 years) might be plausible. You should check the appropriate specification by estimating different model variants. The optimal model is identified by the information criteria, given that the residuals fulfill the required properties (no autocorrelation, homoscedasticity, and so on) 
  • asked a question related to Economic Forecasting
Question
1 answer
Please help me to identify the current problem in Financial Market Forecasting and the Efficient Market Hypothesis of UK Market. This looks like a very interesting research topic for my PhD. However, I have to be more precise with my research topic. Please advise.
Relevant answer
Answer
The Efficient Market Hypothesis (EMH) developed in finance developed in parallel with the Rational Expectation Hypothesis in economics. I suppose you want to specify some financial equation for your research. First you need to choose a topic such as the UK stock market, or a segment of the UK market such as housing financing. Then you need to put in EMH specifications in those equation. How to do that? You need to check the literature on how one builds these expectation for the variables you choose. For example, if you are studying efficiency in the stock market, then you may need to use the standard deviation of stock prices as a measure of efficiency. You may even scale it by the standard deviation of stock return. You can include other variables that are potentially associated with price efficiency in your equation.
Good luck in your studies
Lall Ramrattan
  • asked a question related to Economic Forecasting
Question
8 answers
Most frequently, (economic) forecasters do not only provide the expectations of their forecasts but also some measures for uncertainty of the forecasts (like the 66% probability interval). However, the forecasters' customers (like politicians, managers, or journalists) frequently complain about the poor ex-post performance of the forecasts, about the gaps between real development and the mean of the forecast distribution – even if real development lies within the 66% confidence band. Likewise, they complain about large ex-ante forecast uncertainty (and uselessness of the forecast) when a 80% probability interval is reported, and are satisfied with accuracy of the narrower 66% probability interval even though both are calculated from the same distribution.
Do you have any suggestion on how to communicate forecast uncertainty in a way that is more easily understandable by persons without statistical training?
Should we make it a standard to report something like a `recession probability' in forecasts on economic growth or unemployment, of a `deflation probability', and so on? Similar to the precipitation probability frequently found in weather forecasts...
Relevant answer
Answer
Hello Norbert,
As I am not an expert in statistics and probability, I can hardly give you a definite answer. Nevertheless, my feeling (based on some experience) is that the stochastic methods can deliver strongest results but they also need strongest assumptions (parameters of particular probability distributions, covariance functions, etc.), these, however, are not always fulfilled or are spoiled be severe uncertainty.
There are other approaches available to quantify the uncertainty in inputs and outputs of mathematical models. Let me mention Dempster-Shafer evidence theory that resembles probability theory but its assumptions are less demanding, or fuzzy set theory, or Ben-Haim's info-gap theory (that has some features similar to the fuzzy set approach), or even the worst-case and the best-case approach.
The theories I have just mentioned use weaker assumptions on input (uncertain) information than probabilistic methods, and, as a consequence, their outputs show, in some sense, more uncertainty than probabilistic methods, that is, their "confidence intervals" (though, in these theories, this term is not fully appropriate) are wider but perhaps more realistic.
With best wishes,
Jan
  • asked a question related to Economic Forecasting
Question
1 answer
What is the outlook for nickel demand in non-stainless first uses?
Relevant answer
Answer
This is an interesting question and since no one else has attempted an answer, let me try to provoke some discussion. The question specifically addresses less common uses of nickel since the most common use of nickel is to combine it with iron to form stainless steel (i.e., approximately 2/3 or more is used for stainless steel).
It is also used in other alloys or plating that account for another 20 percent or more. It is in this area that demand might change as a wide variety of metals can be used to form alloys or less commonly in plating.
Another source of a shift in the demand for nickel is in batteries (e.g., Nickel-metal-hydride or NiMH). Newer battery technology might replace these batteries because of the so-called "memory effect" of NiMH batteries. Although NiMH batteries still may be more cost effective per charge in some applications, many of us may prefer the Lithium-ion (Li-ion) batteries that seem to hold their charge longer. For example, I have multiple high-quality 18 volt drills. (Can a guy have too many drills?) I find myself using the Li-ion drill the most because its power does not fade; however, it just stops when discharged with little warning. Each has advantages, of course.
I strongly suspect that battery technology will continue to improve, especially in electric cars; however, it seems unlikely to me that the technology is moving toward nickel-intense batteries. Batteries probably account for around 2 percent of nickel, so only a large increase is likely to change the outlook for nickel overall.
Another issue with nickel is that a surprising number of people have allergy related issues with it. This is potentially an issue.
Perhaps the greatest change may have to do with alloys created in China. My experience with Chinese manufactured goods made of metal is mixed. For example, I used to think of metal brake rotors as a standardized good and although they typically contain little if any nickel, I suspect that the principle remains the same. I strongly suspect that those buying alloys will be more careful in their purchases as they recognize (or maybe just perceive) differences in quality.
Although the question by Erkan Yesel specifically addresses the non-stainless steel outlook for nickel, similar quality issues with stainless steel might change the overall outlook.
Since I have already veered outside the scope of the question to some extent, a couple of supply issues are worth noting. First is the relative abundance of nickel on earth and secondly as an export it seems to be exported from places like Ontario, Canada, Australia, Cuba , and Indonesia; that latter two may present some stable supply concerns, especially as Indonesia recently moved to ban or reduce exports in this area.
Considering all of these, I suspect some changes in demand; however, these changes do not seem to be dramatic changes. Short run changes in supply are another issue and they seem to be driving the price of nickel right now.
I trust others will improve my answer with corrections, better examples, or more information. Like most economist, I enjoy attempting to analyze things like this and look forward to the refinements others offer.
  • asked a question related to Economic Forecasting
Question
1 answer
What are the trends in the intensity of use of iron ore and steel?
Relevant answer
Answer
I analysed this questions recently together with a colleague in the paper "Long Term Trends in Steel Consumption" (Downloadable at ResearchGate, forthcoming in Mineral Economics). Quite often, intensity of use (IoU) of steel is defined as Steel consumption per unit of GDP. We think this definition to be problematic, since the IoU highly depends of on the concept of internationally comparable GDP employed. We, therefore, perfer to look at steel consumption per capita. We find a bell-shaped curve with steel sonsumption per capita reaching its maximum at the per capita income of about 30.000 $ (in PPP).However, in our paper we also hint at a second problem: IoU is typically measured only for apparent steel consumption, whereas indirect steel imports, i.e. steel incorporated in imported cars or in machines aree neglected. Therefore, IoU is most probably under estimated for small countries which tend to import steel intensive products instead of producing them.
  • asked a question related to Economic Forecasting
Question
3 answers
I am doing my PhD research on stock market forecasting and I need to know which is the best computer program to apply the Hybrid models. I am using Concordance and genetic, GARCH-HMM and other machine algorithm models. I haven't used any advanced computer programs before and I am trying to attend more courses and to read more books in MATLAB, R, and JAVA.
Relevant answer
Answer
Thank you Mohamamd, So how can I do that? I found an article about the GARCH-HMM.
  • asked a question related to Economic Forecasting
Question
6 answers
In our current paper, the "Golden Rule of Forecasting" we are searching for evidence on whether, over the past half-century, the practice of forecasting in any area of the social or management sciences has been shown to produce more accurate forecasts or less accurate forecasts or if there has been no measurable change. Please provide citations to relevant sources.
Relevant answer
Answer
Not included in Booth's review is my paper "European demographic forecasts have not become more accurate during the past 25 years", Population and Development Review 34(1)2008, 137-153. It supports her conclusion with the results from detailed analyses for 14 countries.
Note, however, that many forecast users do not distinguish between expected value and outcome. Before I cast a fair die, my best guess (expected value, median) is 3.5. But when I observe a result of 5 later on, does that mean that my forecast was wrong? No!
  • asked a question related to Economic Forecasting
Question
11 answers
I want to calculate long time series of forest product prices in current EURO value. Based on international trade prices in USD.
An answer on stack exchange :
suggests using "national currency weights to the ECU value" to obtain a time series back to 1979. It also suggest using a 1979-1999 time series called " Euro Community" from the St Louis FED. Are there other European sources of EUR/USD exchange rate before 1999?
Relevant answer
Answer
As I noted, the synthetic euro series uses 1997 GDP weights. If one is trying to put together a series for consistency (using the euro as a unit of account) then I see no issue with the data that I have. If, however, there is implied behavior (say lumber prices are determined in a euro framework) then the approach would make no sense.
  • asked a question related to Economic Forecasting
Question
1 answer
Which one of them is the best to validate different stock market volatility models (i.e. GA- HMM, HMM-GARCH, EGARCH, APARCH, GARCH normal, EWMA model) where all will be validated under full sample period and sub periods, I believe that it depends on the estimation conditions and on the tradeoff between bias and variance. But one of them must fit those model more than others. Also I am using daily, weekly, monthly and annual data. So I must select the best method from the beginning to prepare the data with different time horizons.
Relevant answer
Answer
I don't know what is the wrong with my question? Really strange. Good bye research gate.