Research
Journal articles, working papers, op-eds and blog posts, newest first. Filter by type, or jump to a year. For older work and chapters see Books, the CV and the Finance Research Group website.
-
2026
Abstract
Information technology (IT) systems are now ubiquitous in government. The Government of India has tripled its spending in real terms on IT procurement over the last fifteen years. IT procurement is tricky because these systems need continuous maintenance and refinement, and traditional systems of procurement are unsuitable for solving the incomplete contract problem that follows IT procurement. We build on three case studies from India – Aadhaar, Passport Seva, and MCA-21 – to make the case for why we need new thinking on IT procurement. Our solutions hinge around the fact that the government needs to build robust internal technical expertise even when system development has been outsourced. We make the case for flexible contracting, early concept verification, and strategic investments in state HR capabilities to successfully navigate the marketplace for digital services.
-
Abstract
Governments in developing countries spend heavily on subsidies, yet much of the money never reaches households: the macro–micro gap. We measure this gap in Karnataka, India, across a natural experiment. Karnataka has been delivering conditional Direct Benefit Transfers (DBT) for about a decade now. In July 2023 the state introduced new “guarantee” schemes, and directed much of its DBT allocation to a near-unconditional scheme. Because administrative records are endogenous to the state that produces them, we measure household receipts independently, from the CMIE Consumer Pyramids Household Survey. We compare robust population estimates from this survey against the state’s reported expenditure on DBTs in both periods. The gap averaged about 78% over the seven pre-2023 years and fell to about 17% after the shift. The paper contributes independently measured evidence on subsidy delivery in India and a replicable method for auditing welfare programs in the presence of longitudinal data.
-
Abstract
We examine the impact of borrowing on consumption smoothing for households entering credit markets for the first time, using a large panel dataset with income, expenditure and financial choices of 150,000 households in India. Compared to households that never borrow, first-time borrowers are poorer, younger, less educated, and face higher income volatility. Their decision to borrow is often preceded by a change in income or consumption, regardless of whether it is a decrease or an increase. The act of borrowing increases both the level and the volatility of consumption expenditure. Credit does not help to smooth consumption, even for first time borrowers with no shocks to income and consumption. Higher volatility of consumption persists after borrowing, particularly for expenditure on non-durable goods such as food or planned household expenditure. This suggests that new borrowers use credit as one component of a broader risk-management strategy that involves multiple adjustments within their consumption portfolio.
-
2025
Abstract
The proliferation of large-scale conditional and unconditional cash transfers in emerging economies necessitates novel measurement frameworks that move eyond administrative delivery tracking to capture complex socio-economic impacts. In July 2023, the Government of Karnataka implemented five major ‘guarantee schemes’ designed to alleviate poverty through cash transfers and targeted utility subsidies. This paper establishes an evaluation framework to assess these schemes, shifting the analytical lens toward household welfare, labour market dynamics, and macroeconomic opportunity costs. We use a high-frequency monthly panel dataset of approximately 6,000 households in Karnataka, contrasted against a control group of 12,000 households from bordering states (January 2022 to December 2024) to measure both direct consumption smoothing, and structural economic shifts. While the data indicates successful, large-scale consumption smoothing, we find negligible improvements in structural indicators, such as women’s intra-household agency and female Labour Force articipation Rates. We argue that comprehensive scheme evaluation must prioritise structural and macroeconomic indicators to prevent welfare programs from inadvertently substituting necessary, long-term economic growth strategies.
-
Abstract
Large Language Models (LLMs) have demonstrated remarkable capabilities in text comprehension, but their ability to process complex, hierarchical tabular data remains underexplored. We present a novel approach to extracting structured data from multi-page government fiscal documents using LLM-based techniques. Applied to large annual fiscal documents from the State of Karnataka in India, our method achieves high accuracy through a multi-stage pipeline that leverages domain knowledge, sequential context, and algorithmic validation. Traditional OCR methods work poorly with errors that are hard to detect. The inherent structure of fiscal tables, with totals at each level of the hierarchy, allows for robust internal validation of the extracted data. We use these hierarchical relationships to create multi-level validation checks. We demonstrate that LLMs can read tables and also process document-specific structural hierarchies, offering a scalable process for converting PDF-based fiscal disclosures into research-ready databases. Our implementation shows promise for broader applications across developing country contexts.
-
Abstract
With the Russia–Ukraine war evolving into a prolonged war of attrition, the functioning of the two economies has become central. A small but growing literature uses alternative data to measure economic activity, responding to the limitations of official statistics. In this paper, we use nighttime lights (“NTL”) data to track economic changes in both countries. Ukraine’s NTL has fallen by 50% compared to 2022 levels. In contrast, Russia’s aggregate NTL showed virtually no change between 2022 and 2025. Within Russia, the Yamal–Nenets Autonomous Okrug, which produces 90% of the country’s gas, recorded substantial ntl growth despite reports of declining national gas production. We also identify growth reversals in regions hundreds of kilometres from the Ukrainian border, likely reflecting the effects of standoff weapons. Similar reversals appear along Russia’s Western European border but not along other international frontiers, suggesting uneven enforcement of sanctions. Over the war years, Ukraine’s economic activity has shifted westward, while Russia’s has moved eastward.
-
Abstract
A banking sector regulator must be non-partisan and separate from government. However, the banking sector in India comprises banks of many kinds: private banks, public sector banks, cooperative banks, and regional rural banks. The laws that govern the sector have evolved to grant special powers over some categories of banks to the Union government, thereby, diluting or sometimes even nullifying the power of the Indian banking regulator, the Reserve Bank of India (RBI), over these banks. In most cases, the power of the RBI has been diluted or nullified in respect of public sector banks and the State Bank of India. This is not desirable from the perspective of ownership neutral regulation and creates an unfair playing field for banks in the sector. It also raises serious conflict of interest concerns, where the government is both a borrower and regulator, can influence the functioning of some banks. This article examines the legal provisions that undermine the regulatory powers of the RBI including on shareholding, appointment of directors, winding up and making schemes. It argues for an ownership-neutral regulatory framework emphasising the need for greater autonomy for the RBI, and the creation of a clear separation between the governments borrowing and regulatory role. The aim is to chart the pathway towards a more robust regulation of the banking sector in India.
-
2024
Abstract
At independence, India committed itself fully to foster, promote and sustain the creation of science and scientific research in all aspects. In this paper, we present first principles reasoning on the case for state action, and the optimal mechanisms for using taxpayer resources, to fulfill this commitment. We argue in favour of a reorientation of public spending on innovation: away from building vertical government organisations, and in favour of a contracting-out strategy. Such a strategy would induce knowledge and capabilities in the society, and through this, induce greater gains for the people of India. We present a preliminary sketch of the path to implementation.
-
Abstract
This paper examines the prerequisites for successful drug procurement within the Indian public healthcare system. It argues that the primal objective is timely availability of medicines, and analyses two successful procurement cases, from which four common elements emerge. One, an information system on stocks and flows of medicine. Two, processes and manuals covering the full pipeline of the procurement contract. Three, procedures and work allocation during emergencies. Four, an information system on present and prospective vendors. These elements, coupled with a supply-chain focused mandate on delivery to the end user, and investments in regular training of personnel, enable a public healthcare system that can deliver medicines reliably.
-
Abstract
Much of Indian public finance research has focused on the level of debt and deficits. In this paper, we examine the structure of lenders to the Indian state. To what extent is this lending coerced? Is the present debt management strategy consistent with the objectives of low cost borrowing for the government in the long run, while preserving efficiency in the economy, and retaining the optionality of surging borrowing when faced with rare events? We find 5% of the lending to the Indian state comes from voluntary sources. While financial repression for banks eased de jure with a decline in the SLR from 33% in 1988 to 18% in 2021, lending to the state beyond regulatory requirements was Rs.30 trillion in 2021. Alongside this, the growth of the pension and insurance industries created new pools of assets where financial repression generated bountiful lending to the government. These facts help re-examine debt management strategy in India.
-
2023
Abstract
Credit rating agencies have long been criticized for being slow to downgrade ratings, when the financial health of the firm deteriorates. This stands out in sharp contrast to how security market prices of the firm respond more rapidly, during the same time. In this paper, we ask whether this disparity between the two indicators can be bridged. We estimate a logit model for the probability of a rating downgrade, using a large dataset of firms in India which combines both high frequency market price data and lower frequency accounting data. We find that changes in measures such as the Distance to Default, in combination with firm characteristics such as ownership structure, can predict a higher probability of ratings downgrades, before they are announced. This will be particularly useful for regulated financial firms, which have to re-weight their portfolios to satisfy micro-prudential requirements, based on downgrades in credit ratings.
-
Abstract
Information systems about dispute resolution processes and forums that can aid litigants, a key stakeholder in the justice delivery system, are absent. We propose the development of a system that allows a litigant to comparatively evaluate the different forums where he can take his dispute for resolution. To develop this system, we first identify the types of information that can support the decision of a potential litigant on whether to take her matter to court and if yes, which court. We find that information on efficiency, effectiveness, fairness, predictability, and access to courts are crucial for a litigant's decision-making process. Subsequently, we locate data sources required to quantify such measures from the viewpoint of periodically publishing such measures in the public domain. For this purpose, the case type was narrowed down to debt disputes. We systematically collect quantifiable information about cases from the select courts' websites, namely, the Bombay High Court (Bombay HC), the Mumbai bench of the National Company Law Tribunal (NCLT), and the Debt Recovery Tribunal (DRT). We use data from these courts up to September 2022. We find that a litigant has the lowest chance of getting a first hearing within the first year of filing a case at the Bombay HC (37%) as compared to the NCLT or the DRT. The chance for disposal within the first year is even slimmer. There is a 1/3rd chance of disposal at the NCLT within one year of filing, while it is less than 1/5th at both the Bombay HC and the DRT. However, a case is likely to have the highest number of hearings at the NCLT. A survey questionnaire is designed and deployed to capture features of a case that cannot be quantified. We deploy the survey to capture the relative performance of each of the three courts to a sample of respondents who have litigated frequently in the courts of our interest. The results of the survey indicate that the NCLT is ranked the highest by survey participants. Bombay HC comes second, and the DRT is ranked third.
-
Abstract
A central choice in public administration is whether to ‘make’, which is production within the public sector, or ‘buy’, which is public procurement. While ‘buy’ has the advantage of harnessing the energy of private firms, it requires state capacity in contracting and leads to questions on how to obtain state capacity in contracting. One measure of state capacity in procurement is the extent of failure to spend resources budgeted for procurement. We construct a novel dataset about spending gaps (between what is budgeted and what is spent) in procurement by the Indian union government. We find that the spending gaps are the smallest when buying goods, and the highest with works. We find that ministries that have a sustained engagement with procurement fare better on obtaining minimum gaps in procurement spending, which suggests a process of learning by doing.
-
Measuring Financial Inclusion: A project report
Abstract
Despite significant policy support for financial inclusion, there is little evidence about what has been achieved by way of knowing which households and individuals are financially included, or what is the impact of such inclusion on these persons. We propose a measurement framework that aims to capture these aspects of financial inclusion using an input-output-outcome perspective. The input to financial inclusion is measured as participation in formal financial sector instruments. The output is the extent to which these are used. The outcome is how those who hold and use these instruments perceive their own well-being. A survey questionnaire is designed to collect this information, with two distinct features: It is conducted for a household, since financial holdings benefit more than one individual in a household. It is designed to facilitate regular collection, in order to aid financial service providers and policy makers in their thinking about improving financial inclusion. We deploy the questionnaire to understand the financial inclusion of a sample of 300 low-income households in two geographic areas in India. We find that financial participation has a positive correlation with financial well-being, a relationship which is significant even after controlling for income level and ownership of physical assets.
-
Abstract
Regulators use measures such as a fee on high order-to-trade ratio (OTR) to slow down high-frequency trading. Their impact on market quality is, however, mixed. We study a natural experiment in the Indian stock market where such a fee was introduced twice, with differences in motivation and implementation. Using a difference-in-difference approach, we find that the fee decreased OTR and improved market quality when it was imposed on all orders, while it had little effect when it was imposed selectively on some orders. Improvement in liquidity was driven by a reduction in adverse selection costs following lower OTR.
-
2022
Abstract
Nighttime lights captured from satellites has emerged as an important way to measure prosperity. Every researcher who uses the files released by NASA and NOAA faces the challenge of pre-processing it in addressing data quality issues. We present NighttimeLights.jl, a package written in Julia, which implements conventional and novel methods for cleaning the data. The package also serves as a platform for methodological research in remote sensing.
-
Abstract
The literature on the performance evaluation of the judiciary captures the perspectives of judges, researchers and court administrators. However, it is not obvious if a litigant who proposes to access the judiciary for the resolution of a dispute would use the same or similar metrics when evaluating the performance of courts. In this paper, we review the global literature that evaluates the performance of the judiciary and identify which of the metrics in the literature would directly matter to a litigant who proposes to access the courts for redress. Using the litigant’s expectations as performance metrics, we develop an evaluative framework for comparing similar courts. Our work creates the foundation for developing an information system that could potentially help litigants make informed choices when approaching courts.
-
2021
Abstract
An estimation of the total value of central government procurement in India using data from the Government e-Marketplace (GeM) and departmental accounts, providing a baseline measurement of the state's footprint as a buyer in the economy.
-
Abstract
The VIIRS nighttime lights dataset constitutes progress in the measurement of night lights radiance, with monthly data at a pixel of roughly 0.5km ¡Á 0.5km. We identify a downward bias in the reported radiance when the number of cloud-free images in a month is low. This bias often takes on large values from -10% to -30%. We develop a cautious bias-correction scheme which partially addresses this problem. This scheme is applied upon the pixel-level dataset to create an improved dataset. The bias-corrected data hews closer to the ground truth as seen in household survey data.
-
Abstract
Using household survey data matched with administrative land records quality indicators, this paper estimates whether digitisation and completeness of land record systems affect households' access to formal credit — with implications for land reform policy.
-
2020
Abstract
The article asks what drives the time to repurchase for life and accident insurance contracts in low income households. We use data on customers of a financial services provider from three states in India and find that the probability of repurchase is highest in the first 2 months after the contract expires, and steadily declines after. This suggests a window of opportunity for financial firms and governments to target customers to ensure continuous insurance purchase. Nonmembership of microfinance groups and poor rainfall in the month of expiry affect the time to repurchase adversely. Customers who take longer to repurchase tend to increase the amount of insurance cover.
-
2019
Abstract
This paper revisits the role of leverage in price discovery, using one of the most liquid single-stock futures (SSFs) markets in the world. Price discovery is analysed as a dynamic intraday process. We find that the information share of the SSFs is 55% during news arrivals. It increases to 61%, when the news is negative and the futures is preferred because of short-sales restrictions on the spot. A partial equilibrium analysis predicts that the trade-off between leverage and market liquidity determines price discovery across securities. These predictions are validated by empirical evidence.
- 2018
-
2017
Abstract
This paper studies how firms respond after getting listed on small and medium enterprise (SME) exchanges in India, examining changes in capital structure, growth, and profitability — and whether listing confers benefits beyond access to equity capital.
-
Abstract
An empirical evaluation of the commodity transactions tax introduced in India in 2013, examining its impact on trading volumes, liquidity, and price discovery in commodity derivatives markets.
- 2016
-
Abstract
The paper studies the effect of a law that banned micro-credit lending in the state of Andhra Pradesh in India. Regions in Andhra Pradesh are matched to regions that did not face the ban. A difference-in-difference estimation of changes in matched regions is used to establish a causal impact on average household consumption in the region. The results show that the average household consumption in the ban-affected regions dropped by 15% immediately after the ban compared to the matched regions, and persisted for four quarters. The result is robust to cross-sectional variations in regional exposure to micro-finance prior to the ban, variation in rural and urban locations and variations in matching strategy. The analysis points to a ban as a sub-optimal intervention to improve customer welfare.
- 2015
-
Abstract
This paper examines who contributes and who persists in contributing in a national, voluntary, defined contributory pension programme, where the government provides the incentive of matching contributions of a minimum amount (USD 16). The paper uses proprietary data from a financial services firm where 12 per cent of customers (37,000 individuals) chose to participate in this programme. The evidence shows that only about 50 per cent of contributors reach the minimum amount for the co-contribution, but that participants persist in contributing even if they failed to contribute the minimum amount in a given year. While this paper does not provide causal estimates, it does present evidence of considerable interest among the informal sector in a state-run voluntary pension programme in an emerging market where access to formal finance is otherwise poor.
-
2014
Abstract
Regulation of retail finance has been the subject of policy interventions in several countries, including India. Much of the regulatory change in India has been carried out with little support of empirical evidence. This paper is motivated by questions of the evidence of losses due to mis-sales of financial products. It constructs two measures of the loss to customers due to mis-selling of life insurance policies. The first is calculated using the value of lapsed policies, and the second uses the persistence of premium payments. Both arrive at estimates of around USD 28 billion lost between 2004 and 2011.
-
The way forward for India's National Pension System
Abstract
This paper examines the existing implementation of the National Pension Scheme against the goals with which it was created. It finds that there are certain critical areas in which the NPS has deviated. These include multiplicity of schemes, lack of investment choice, low transparency of the system, and a lack of focus on keeping asset management fees low. These gaps are well understood and can be corrected with regulatory interventions. There remain other policy issues that need to be addressed. These include well designed payout policies, and occupational pension systems that will leverage the institutional development of the NPS to include the informalworkforce.
-
Abstract
The causal impact of algorithmic trading on market quality has been difficult to establish due to endogeneity bias. We address this problem by using the introduction of co-location, an exogenous event after which algorithmic trading is known to increase. Matching procedures are used to identify a matched set of firms and set of dates that are used in a difference-in-difference regression to estimate causal impact. We find that securities with higher algorithmic trading have lower liquidity costs, order imbalance, and order volatility. There is new evidence that higher algorithmic trading leads to lower intraday liquidity risk and a lower incidence of extreme intraday price movements.
-
Abstract
In 2003, trading of commodity futures shifted from single commodity, regional exchanges to national exchanges that trade multiple commodities. This paper examines price discovery and hedging effectiveness of commodity futures after this change and concludes that,on average, futures prices do discover information relatively efficiently, but helps to manage risk less efficiently. The paper uses the viewpoint of the hedger to conjecture what factors may improve hedging effectiveness. These include high settlement costs caused by few and widely dispersed delivery centers and an unreliability of warehouse receipts,a mismatch between the grade specified in the futures contract and what is available for delivery in the market, and disruptions caused by various policy interventions in both commodities spot and futures markets.
-
2013
Abstract
This paper proposes and applies a network-based methodology to identify systemically important non-financial firms in India, using balance sheet linkages and equity return correlations to rank firms by their potential for systemic distress propagation.
-
What should regulation do in the field of micro-finance?
Abstract
Recent events in India have brought a fresh focus on the appropriate regulatory stance towards micro-finance. In this paper, we review facts and recent experience about Indian microfinance. We analyse the puzzles of financial regulation in this field from first principles, and argue that the mainstream mechanisms of consumer protection and micro-prudential regulation need to be modified owing to joint-liability groups. From this perspective, we suggest regulatory strategies that need to be adopted for dealing with micro-credit and financial distribution that focuses on the poor. This analysis and conceptual framework also helps analyse the two policy responses till date, the Malegam report and the draft Microfinance Bill, 2011.
-
2012
Abstract
Option markets have significant variation in liquidity across different option series. Illiquidity reduces the informativeness of the price. Price information for illiquid options is more noisy, and thus the implied volatilities (IVs) based on these prices are more noisy. In this study, we propose weighting schemes to estimate IV, which reduce the importance attached to illiquid options. The two indexes using liquidity weights are SVIX, which is a spread-adjusted volatility index, and TVVIX, which is a traded volume weighted VIX. We find SVIX outperforms TVVIX, the conventional schemes such as the traditional VXO, or vega weights, and volatility elasticity weights.
-
Abstract
Distance-to-default (DtD) from the Merton model has been used in the credit risk literature, most successfully as an input into reduced form models for forecasting default. In this paper, we suggest that the change in the DtD is informative for predicting change in the credit rating. This is directly useful for situations where forecasts of credit rating changes are required. More generally, it contributes to our knowledge about reduced form models of credit risk.
-
Abstract
Develops measures of asymmetry in market liquidity — situations where the cost of buying differs from the cost of selling — and applies these to Indian equity markets to understand market stress and one-sided order flow dynamics.
-
How is financial regulation different for micro-finance?
Abstract
What is the role of financial regulation in the field of micro-finance? This paper identiles two features of micro-finance which call for unique treatment in policy considerations as compared to policy thinking in the mainstream body of financial law. These features are credit recovery and the credit risk of the MFI, when credit access is enabled through the structure of the joint liability group. The paper goes on to offer draft law which embeds a regulatory treatment of micro-finance that flows from this analysis.
- 2011
-
2010
Abstract
The Indian financial system has been revolutionised by the application of a new market design : continuous trading with an anonymous limit order book at NSE and BSE. However, in certain situations, this market design has limitations. Call auctions represent an alternative strategy, where the order flow over a certain time period is pooled, and the market-clearing price obtained through an aggregated supply and demand curve. Call auctions trade off instantaneity of order execution in favour of elimination of impact cost, and can achieve a more trusted price. They can improve the functioning of the market on issues such as market opening, market close, extreme news events, and potentially for illiquid securities including bonds. Call auctions could usefully replace some existing market rules such as `circuit breakers\'. At the same time, there are many subtle elements in making a call auction market work, which require care in market design.
-
2006
Politics holds up full implementation
-
Resilience of liquidity in Indian securities markets
Abstract
Liquid securities markets are a core goal of financial sector reforms, and this has two dimensions: low impact cost and high resilience. We employ elements of evidence to suggest that equity market liquidity is robust to negative price shocks. However, bond market liquidity appears fragile when faced with negative price shocks.
-
2004
Financial sector readiness and the growth of pension funds: the case of South Asia
-
Financial sector regulation in India
-
2003
Institutional mechanisms in pension fund management: Lessons from three Indian case studies
Abstract
Retirement income plans in India face one or more of three common problems: poor funding, inadequate diversification, and weak governance. This paper discusses policy issues raised by these problems in the context of three case studies - the Indian Railways Pension Fund, the Employees Provident Fund, and the Seamen's Provident Fund. It examines means to reform these pension funds, drawing upon both Indian and international evidence and experience. Using these case studies as a basis, it offers general recommendations for pension reform in India.
-
Abstract
Value-at-Risk (VaR) is widely used as a tool for measuring the market risk of asset portfolios. However, alternative VaR implementations are known to yield fairly different VaR forecasts. Hence, every use of VaR requires choosing among alternative forecasting models. This paper undertakes two case studies in model selection, for the S&P 500 index and India's NSE-50 index, at the 95\% and 99\% levels. We employ a two-stage model selection procedure. In the first stage we test a class of models for statistical accuracy. If multiple models survive rejection with the tests, we perform a second stage filtering of the surviving models using subjective loss functions. This two-stage model selection procedure does prove to be useful in choosing a VaR model, while only incompletely addressing the problem. These case studies give us some evidence about the strengths and limitations of present knowledge on estimation and testing for VaR.
-
2002
The stock market response to the Union Budget
- 2000
- 1996