Research
Copies of many of my papers are available on my SSRN page
Working Papers
Public and private information in quotes and trades with Terry Hendershott
In financial markets, traders race to react first to new public information. We identify races on public information via different traders placing identical orders at the same time. Races comprise roughly 20\% of both trades and quotes. While the price impact of races is greater than non-races, most price discovery occurs in non-races. Races occur in the direction of index price movements, particularly in stocks with higher market betas. While high-frequency traders cause almost all races, only half of their price discovery occurs in races. Overall, not accounting for races can lead to private information being overestimated in trades and underestimated in quotes.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4674014
Benchmarking benchmarks with Marta Khomhyn and Talis Putniņš - Revise and resubmit (2nd round), Journal of Financial Economics
Financial benchmarks such as LIBOR underpin the pricing of trillions of dollars of contracts around the world. We evaluate the quality of benchmark prices using a state-space model to separate information from noise. Applying the method to LIBOR benchmarks and their replacements, we find that alternative reference rates (ARRs) are less noisy in four of the five currencies. However, the USD ARR is considerably more noisy, resulting in billions of dollars of noise-related wealth transfers between contract counterparties. We show that benchmark reforms such as expanding the reference market and using a trimmed mean can reduce noise in ARRs.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4136093
Differential access to dark markets and execution outcomes with Carole Comerton-Forde - Under review
We compare execution outcomes in dark pools that are open to all investors against outcomes in pools that restrict high frequency trader access. Conditional on execution, trades in restricted pools have less order flow information leakage and adverse selection risk than trades in pools with unrestricted access. Dark pools that completely prohibit high frequency traders have less order flow information leakage than those that allow customers to opt-out of interacting with this type of order flow. Differences in execution outcomes are concentrated in smaller trades. We conclude that the ability to segment order flow can improve execution outcomes for investors.
This paper was previously circulated under the title "Differential access to dark markets and execution quality".
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3910952
Competition and exchange data fees with Jonathan Brogaard and Dominik Roesch - Under review
We investigate how the introduction of market data fees impacts trading and market quality. We find that data fees decrease the fee-introducing exchange’s market volume, time with competitive quotes, visible liquidity, and role in price discovery. Brokers route their market and limit orders away from the impacted exchange. Market-wide, the fee introduction decreases visible depth and price discovery. The results indicate that some traders have an elastic demand for data and that their response to the introduction of data fees on one exchange can reverberate throughout the stock market.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3703431
Estimating background risk hedging demands from cross-sectional data with Joachim Inkmann and Adrian Rizzo - Conditionally accepted - Journal of Financial Research
Based on a theory of portfolio choice with non-tradable assets, we estimate hedging demands due to background risks before and after the Great Recession for U.S households. Background risks related to human capital, residential property and business assets reduce financial risk-taking, but these effects decline over the Great Recession, as does expected risk-adjusted stock market performance. We also estimate the appropriate discount rate to compute the risk-adjusted value of human capital, which declines by around 8% over the period. Unlike previous literature requiring panel data with large time dimensions, our approach only requires cross-sectional data to identify hedging demands.
This paper was previously circulated under the title "Did background risk hedging demands change over the Great Recession?".
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3428088
Publications
Secondary market transparency and corporate bond issuing costs with Carole Comerton-Forde and Spencer Martin
Review of Finance (2022)
Mandated post-trade transparency in secondary markets lowers the cost of issuing corporate bonds. We show that costs are lower due to the mitigation of information asymmetry in the issuing process. Three pieces of evidence support this finding. First, new issues with higher information asymmetry experience relatively larger reductions in issuing costs. These bonds also experience lower reductions in trading activity than lower information asymmetry bonds, so liquidity cannot explain these results. Second, when a larger fraction of trades in comparable bonds are made post-trade transparent, new issue pricing improves. This holds when conditioning on expected bond liquidity. Third, transparency raises prices in the secondary market, but not by as much as it does for newly issued bonds.
This paper was previously circulated under the title "Do you see what I see? Transparency and bond issuing costs".
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2875165
Does financial market structure impact the cost of capital? with Carole Comerton-Forde and Terry Hendershott
Journal of Financial and Quantitative Analysis (2021)
We provide evidence on market structure and the cost of raising capital by examining market structure changes in US equity markets. Only the Nasdaq's Order Handling Rules (OHR), the one reform that reduced institutional trading costs, lowered the cost of raising capital. Using a difference-in-differences framework relative to the NYSE and that exploits the OHR's staggered implementation, we find that the OHR reduced the underpricing of seasoned equity offerings by one to two percentage points compared to a pre-OHR average of 3.6 percent. The effect is largest in stocks with the largest reduction in institutional trading costs after the OHR.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3074097
Link to the OHR Roll-out Schedule (csv file): here
Comment on: Price Discovery in High Resolution with Carole Comerton-Forde
Journal of Financial Econometrics (2019)
In this comment, we briefly describe the evolution of markets and detail the challenges that these changes create for microstructure researchers and highlight the solution that Hasbrouck (2018) offers for these problems. We survey the literature that uses linear multivariate time-series models to understand high-frequency markets. We focus on three examples from the literature to discuss how estimation constraints have affected their modelling choices, describe the potential drawbacks of these choices and how Hasbrouck’s (2018) method can alleviate these constraints. We deliberately select papers that cover different asset classes: cash equities, fixed income and equity options. We hope that our discussion will help provide guidance about the costs and benefits of different modelling choices for future researchers confronted with a variety of methods to answer related research questions. We conclude by considering the implications of Hasbrouck’s 2018 paper for the current policy debate on market data costs.
Interactions among High-Frequency Traders with Evangelos Benos, Erik Hjalmarsson and Filip Zikes
Journal of Financial and Quantitative Analysis (2017)
Using unique transactions data for individual high-frequency trading (HFT) firms in the UK equity market, we examine the extent to which the trading activity of individual HFT firms is correlated with each other, and what impact this has on price efficiency. We find that HFT order flow and net positions exhibit significantly higher commonality than those of a comparison group of investment banks. However, intraday HFT order flow commonality is associated with a permanent price impact, suggesting that commonality in HFT activity is information-based and so does not generally contribute to undue price pressure and price dislocations.
Press Coverage:
http://www.ft.com/cms/s/0/0e54ea02-c31b-11e4-ac3d-00144feab7de.html#axzz3TViT846D
Testing Preference Stability After Marriage
Economics Letters (2016)
Estimation of non-unitary models of household demands is often complicated by the fact that the majority of consumer expenditure surveys only record consumption at the household level. Identification strategies of some recent non-unitary models assume that the consumption preferences of singles who have never married are representative of the consumption preferences of couples. In this note I use data from the US Consumer Expenditure Survey to test whether this assumption might fail due to dependence between selection into marriage and consumption preferences. The testing strategy relies on comparing the preferences of the non-random sample of couples who divorce with those of singles who have never married. The data strongly reject the assumption that selection into marriage is independent of consumption preferences for both men and women.
Policy research and collaborations with policy institutions
Liquidity shocks and pension fund performance: Evidence from early access with Minsoo Kim and Zhuo Zhong
Forthcoming at the Australian Journal of Management
We study how expectations of fund flows causally affect fund performance by exploiting a quasi-natural experiment in the Australian pension system where an unexpected policy change temporarily allowed fund withdrawals from a prespecified date in the future. Using fractions of young members, middle-aged members, and government co-contributions for low-income earners as instrumental variables, we find an insignificant effect of expected fund outflows on performance. A potential explanation is that Australian superannuation funds preemptively engage in liquidity management in response to changes in expectations of future fund flows and this helps to limit direct and indirect costs in the rebalancing process.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3745990
Press coverage:
https://www.financialstandard.com.au/news/ers-had-little-effect-on-funds-research-177163697
Forecasting the Australian Yield Curve with Bonnie Li (DTF, Victoria), Maryam Nasiri (DTF, Victoria) and Ravi Sastry
The Australasian Journal of Applied Finance (2019)
We apply a number of forecasting models to Australian Government Bond yields. All methods rely solely on the history of yields. Consistent with findings from US Treasury data, we show that the simplest forecasting models across all maturities and forecasting horizons are also generally the best: the forward yield (when available) and the random walk model. Models with more structure - e.g. principal components and Bayesian vector autoregression - can help forecast overnight yields at very short horizons, but provide little or no improvement in other cases.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3543214
The Spillover Effect of Single Stock Circuit Breakers on the London Stock Exchange with Oliver Linton, Joseph Noss (Bank of England) and Lucas Pedace (FCA, UK)
Market Microstructure and Liquidity (2018)
This paper uses transaction data to estimate how single stock circuit breakers on the London Stock Exchange affect other stocks that remain in continuous trading. This ‘spillover’ effect is estimated by calculating the effect of a trading halt on the market quality of stocks that remain in continuous trading and comparing this with the effect of a stock whose absolute returns are of a magnitude nearly sufficient to trigger a trading halt but do not do so. Market quality is measured using a combination of trading costs, volatility and volume. In the two-month period we study, characterised by a relatively volatile trading environment, we find that circuit breakers lead to a significant improvement in the liquidity, and reduction in the volatility, of stocks that remain in continuous trading. This suggests that – at least over the period covered by our data – single stock circuit breakers can play an important role in reducing the spillover of poor market quality across stocks.
Permanent Working Papers
Into the Light: Dark Pool Trading and Intraday Market Quality on the Primary Exchange
This paper uses proprietary transaction data to investigate how trading in dark pools affects intraday market quality on the limit order book of the primary exchange for members of the FTSE-100 index. Using trading patterns from execution algorithms as instrumental variables, I show that dark trading leads to improved liquidity on the primary exchange, both in absolute terms and relative to trading on the limit order book. Although these relationships differ across stocks of different size, dark trading does not lead to worse market quality at the intraday level for either small or large stocks during the sample period.
SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2660119