Home Menu
icon-menu-close.svg

Faculty & Research

Faculty & Research

Research Productive

Show result

Clear
Search Query :
Area :
Search Query :
3721 items in total found

Working Papers | 2019

Buy, Sell or Hold: Entity-Aware Classification of Business News

Ankur Sinha, Satishwar Kedas, Rishu Kumar, and Pekka Malo

Financial sector is expected to be at the forefront of the adoption of machine learning methods, driven by the superior performance of the data-driven approaches over traditional modelling approaches. There has been a widespread interest in automatically extracting information from financial news flow as the signals might be useful for investment decisions. While quantitative finance focuses on analysis of structured financial data for investment decisions, the potential of utilizing unstructured news flow in decision making is not fully tapped. Research in financial news analytics tries to address this gap by detecting events and aspects that provide buy, sell or hold information in news, commonly interpreted as financial sentiments. In this paper, we develop a framework utilizing information theoretic concepts and machine learning methods that understands the context and is capable of extracting buy, sell or hold information contained within news headlines. The proposed framework is also capable of detecting conflicting sentiments on multiple companies within the same news headline, which to our best knowledge has not been studied earlier. Further, we develop an information system which analyzes the news flow in real-time, allowing users to track financial sentiments by company, sector and index via a dashboard. Through this study we make three dataset related contributions - firstly, a training dataset consisting of more than 12,000 news headlines annotated for entities and their relevant financial sentiments by multiple annotators, secondly, an entity database of over 1,000 financial and economic entities relevant to Indian economy and their forms of appearance in news media amounting to over 5,000 phrases and thirdly, make improvements in existing financial dictionaries. Using the proposed system, we study the effect of the information derived from daily news flow in the years 2012 to 2017, over the Indian broad market equity index NSE 500, and show that the information has predictive value.

Go to link
Read More

Working Papers | 2019

Does IT work? Information Technology (IT) in Welfare in India*

Reetika Khera and Vineeth Patibandla

The use of information technology (IT) in public administration is viewed as a significant tool for enhancing transparency and accountability. In popular rhetoric, these continue to be heralded as necessary and sufficient conditions for increasing transparency and accountability. This paper studies the use of various forms of IT such as computerization, public management information systems (MIS), digital ID and biometrics in two welfare programmes in India. This paper aims to (a) use government MIS portals to shed light on the performance of welfare programmes, (b) understand whether recent IT applications have been beneficial or detrimental to programme performance and (c) comment on what extent IT has fulfilled its potential to enhance transparency. We find support for two earlier findings: one, there is no automaticity between use of IT and enhanced transparency or accountability and two, the use of IT may reinforce, even exacerbate, existing power imbalances rather than mitigating them. Further, we find some evidence of 'too much' technology being detrimental to improving administration and accountability.

Go to link
Read More

Working Papers | 2019

Financing Infrastructure in India – Issues and the Way Forward

Sebastian Morris

Optimal approaches that recognize the specific kind of market failure/s, in the policy and design of infrastructure, greatly reduce the financing costs and improves the ability of to attract finance in the private provisioning of infrastructure. When state systems are weak organizationally it is first best to strengthen the state capacity so that it can minimally perform the roles of design, regulation, development of frameworks, and of monitoring, for the private provisioning of infrastructure. This is particularly so in the case where there are dual market failures arising out of both the natural monopoly and the appropriability failure aspect. Thus sewerage and water, city roads, multimodal facilities, solid waste, public health care, the challenges have proven beyond the current ability of the state, despite its large commitment to the use of private capital. The challenges in design and policy are large and with many false starts it is only now barely beginning to be considered in India. Thus infrastructure design rather than debilities in financial markets remain the key problem.
The need to develop capital markets and institutions to lend long is vital, but much of the challenge is really in having good projects that are financed keeping in mind the capacity limitations within banks and financial institutions. The potential to use of foreign capital to finance infrastructure is often overstated. Reform of financial institutions (FIs) and banks is vital today, as also the necessary incorporation of interest rate (change) risks into the project cost to overcome adverse selection. The forces leading to the current mess-up of the Indian banks and FIs in lending to infrastructure are brought in perspective. The key issues in developing state capacity, and the changes required for getting the design of infrastructure right, as also to bring functionality to the role of financial institutions in the private development of infrastructure are highlighted.

Go to link
Read More

Working Papers | 2019

Does GST in India Hurt Producing Regions?
A New Estimate of the Tax Base Under GST of Select States

Sebastian Morris, Ajay Pandey, Sobhesh Kumar Agarwalla, and Astha Agarwalla

GST as introduced in India being a destination based tax, does not encourage regions to vigorously promote manufacturing and tradable services industries. Being in the midst of its economic transformation, and given the subnational character of most states (regions), it is important that the states engage in locational tournaments to attract investments, not through tax concessions, but through the provision of infrastructure services, governance, and other intangible services. A new consumption based approach that adjusts the detailed consumer expenditure figures of the National Sample Surveys at the state level is developed. This is shown to be robust and is used to estimate the RNR (Revenue Neutral Rate of Taxes) at the State level. This reveals that there are stark differences between the rates for the producing states and the consumption oriented states amounting to as much as 10% of GDP. These differences cannot be bridged by the proposed compensation scheme. As the impact of GST goes on to the next stage of determining the locational choices of new investments, the lack of fiscal incentives for states to attract and nurture investments, unless corrected would have deleterious effects on the investment process.
As much as 50% of the Centre's collection of GST may have to be distributed based on economic activity centered around manufacturing and tradable services production, if the country is not to lose the steam of high and growing investments to take it through its economic transformation. The contrast with China is remarkable, China's GST is only partial covering only manufacturing and associated labour services, allowing states to tax and retain many services irrespective of the location of the consumer of the service. More importantly as much as 25% of the central collections on account of GST( in manufacturing) go to the provinces based on their public goods production.

Go to link
Read More

Working Papers | 2019

Whose Empowerment? National Digital Infrastructure and India's Healthcare sector

Rajesh Chandwani, Saneesh Edacherian, and Mukesh Sud

Patient-centric digital infrastructure can potentially enhance the efficiency of the healthcare systems. Even in developed nations evidence suggests low adoption rates for such infrastructure. The Indian government, piggybacking off biometric identity, is setting up digital infrastructure to enable the provision of universal healthcare. Invoking an information ecology perspective, we investigate the physician's perception to this initiative. We find that, equipped with a unique patient identifier and stakeholders' registry, this initiative is perceived to be a game changer and could significantly impact the power dynamics in the healthcare sector. Physicians, who are the key stakeholders in this initiative, are skeptical about the change in the locus of the power, with power residing in 'data' rather than 'professional expertise'. The changes are expected to manifest through monitoring, controlling and managing the data rather than the provision of knowledge-based services. We present recommendations for the design and implementation of this large-scale patient-centric digital infrastructure.

Go to link
Read More

Working Papers | 2019

Overestimation in the Growth Rates of National Income in Recent Years? – An Analyses Based on Extending GDP04-05 through Other Indicators of Output

Sebastian Morris and Tejshwi Kumari

Quarterly indices of output like those of Industrial Production, other measures of production like net sales, exports, of companies for which data is available, besides proxies like credit to the sector, and indices of price levels have been used to forward project the growth rates of GDP04-05, for the principal sectors of the National Income Accounts (NAS). These were then compared with the growth rates given by the new Gross Value Added (GVA11-12) at constant price measure. It is very highly probable that some sectors of the National Accounts Statistics (NAS) notably Manufacturing, and Trade, Transport, Storage, Hotels and Communication Sectors were overestimated, especially in periods when the output (economic activity) was slowing down. This questions the use of the new GVA series for macroeconomic (policy) actions, wherein more than extensiveness of coverage, the movements over time of the measure have to be reliable and accurate. This is especially so because manufacturing and its related sector-trade etc., are the core sectors which are responsive to changes in policy and to shocks (that could be countered), wherein there is deep overestimation. Some evening out of the overestimation is noticed over the upswing in the business cycle since 2017:2. However the demonetization which rudely reduced demand did not allow the "phase shifting" and "flattening" aspects, which the new GVA series possibly imposes to be examined in detail, although the same is suggested.

Go to link
Read More

Journal Articles | 2018

Reforming agricultural markets in India: A tale of two model Acts

Sukhpal Singh

Economic & Political Weekly

The union Ministry of Agriculture and Farmers’ Welfare had prescribed a model Agricultural Produce Marketing Committee Act in 2003. The state-level adoption of the act has been tardy and varied in terms of both the magnitude and content of agricultural market reforms. Yet, the ministry under the current central government has come up with another model act, the Agricultural Produce and Livestock Marketing (Promotion and Facilitation) Act, 2017, supposedly an improvement over the 2003 act. Among other things, the provision that has grabbed much attention is the removal of contract farming from the APMC domain to a separate model act of Agricultural Produce and Livestock Contract Farming and Services (Promotion and Facilitation). Analysing these draft acts, the paper finds that both the model acts suffer from serious conceptual lacunae that have implications for their application and governance, and, consequently, for inclusive and sustainable agricultural development.

Go to link
Read More

Journal Articles | 2018

Using Karush-Kuhn-Tucker proximity measure for solving bilevel optimization problems

Ankur Sarin, Tharo Soun, and Kalyanmoy Deb

Swarm and Evolutionary Computation

A common technique to solve bilevel optimization problems is by reducing the problem to a single level and then solving it as a standard optimization problem. A number of single level reduction formulations exist, but one of the most common ways is to replace the lower level optimization problem with its Karush-Kuhn-Tucker (KKT) conditions. Such a reduction strategy has been widely used in the classical optimization as well as the evolutionary computation literature. However, KKT conditions contain a set of non-linear equality constraints that are often found hard to satisfy. In this paper, we discuss a single level reduction of a bilevel problem using recently proposed relaxed KKT conditions. The conditions are relaxed; therefore, approximate, but the error in terms of distance from the true lower level KKT point is bounded. There is a proximity measure associated to the new KKT conditions, which gives an idea of the KKT error and distance from the optimum. We utilize this reduction method within an evolutionary algorithm to solve bilevel optimization problems. The proposed algorithm is compared against a number of recently proposed approaches. The idea is found to lead to significant computational savings, especially, in the lower level function evaluations. The idea is promising and might be useful for further developments on bilevel optimization both in the domain of classical as well as evolutionary optimization research.

Go to link
Read More

Journal Articles | 2018

Modelling subjective utility through entropy

Manish Aggarwal

Journal of the Operational Research Society

We introduce a novel entropy framework for the computation of utility on the basis of an agent’s subjective evaluation of the granularised information source values. A concept of evaluating agent as an information gain function of this entropy framework is presented, which takes as its arguments both an information source value and the agent’s evaluation of the same. A method to model the agent’s perceived utility values is proposed. Based on these values, several new measures are designed for the evaluation of the information source values, perceived utilities, and the evaluating agent. A real application is included.

Go to link
Read More

Journal Articles | 2018

Preferences-based learning of multinomial logit model

Manish Aggarwal

Knowledge and Information Systyems

We learn the parameters of the popular multinomial logit model to gain insights about a DM’s decision process. We accomplish this objective through the recent algorithmic advances in the emerging field of preference learning. The empirical evaluation of the proposed approach is performed on a set of 12 publicly available benchmark datasets. First experimental results suggest that our approach is not only intuitively appealing, but also competitive to state-of-the-art preference learning methods in terms of the prediction accuracy.

Go to link
Read More
IIMA