lukestein’s Twitter Archive—№ 2,619

Really looking forward to spending two days at the 2019 Northeastern University Finance Conference. Thanks so much to @nikir1 and her @NU_Business colleagues for having me. Thread warning; mute if boring. damore-mckim.northeastern.edu/events/finance-conference-2019/ #NUfin2019
Permalink On twitter.com ♻️ 8 Retweets ❤️ 18 Favorites 2019 May 10 Mood -4 🙁

…in reply to @lukestein
PAPER 1: “Overcoming Borrowing Stigma: The Design of Lending-of-last-resort Policies” by Yunzhi Hu (@kenanflagler) with Hanzhe Zhang (@MSUBroadCollege). Model of discount window stigma and the term-auction facility, with confirmatory empirical evidence. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood 0

…in reply to @lukestein
Endogenous sorting of banks → lower quality use discount window, higher quality use TAF. Robin Greenwood (HBS) discussion asks whether “stigma” = “signal”? And looks like TAF premium right after Lehman was largely large foreign banks. Why? #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood +6 🙂

…in reply to @lukestein
PAPER 2: “Disagreement and Liquidity” by Samuel Kruger @UTexasMcCombs Model (≈Grossman ’76+overconfidence)→ -Trade entirely overconfident disagreement? Private info inc. trade and liquidity. -Some trade from liquidity shocks? Liquidity non-monotone in private info. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 10 Mood -4 🙁

…in reply to @lukestein
Ronnie Sadka @BCCarrollSchool discussion. Runs a cool new analysis: try to precict So and Johnson (Mgmt Sci ’18) market asymetry information measure using analyst dispersion (private) vs. Ravenpack media coverage disagreement (private). doi.org/10.1287/mnsc.2016.2608 #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood -1 🙁

…in reply to @lukestein
PAPER 3: “Persuasion in Relationship Finance” by Ehsan Azarmsa with Lin William Cong (both @ChicagoBooth) and Uday Rajan (@MichiganRoss). How should entrepreneurs finance experimentation? Model of staged financing with imperfect competition between potential funders. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 10 Mood -2 🙁

…in reply to @lukestein
CORRECTION🤦🏻‍♂️: Uday Rajan (@MichiganRoss) is discussant not coauthor Azarmsa/Cong model suggests 1. Endogeneous info production matters 2. Outsiders are important 3. Interim competition improves info production 4. Optimality of warrants on convertible securities #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood +5 🙂

…in reply to @lukestein
PAPER 4: “More Cash Flows, More Options? The Effect of Cash Windfalls on Small Firms” by Jacelly Cespedes (@CarlsonNews) with Xing Huang (@WUSTLbusiness) and Carlos Parra (@ucatolica) @arpitrage did a nice thread on this paper a few weeks ago. @arpitrage/1121798183640813568 #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 3 Favorites 2019 May 10 Mood +3 🙂

…in reply to @lukestein
I’ll get to the paper, but note I met Jacelly when she presented as a PhD student at the Colorado Finance Summit. (At same great conference I met @paulgp, @sgil1122, and @EmilVerner for the first times). She’s a great scholar and presenter. sites.google.com/site/coloradofinancesummit/2017-program #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 6 Favorites 2019 May 10 Mood +6 🙂

…in reply to @lukestein
Hand-collected data on retailers/owners that win $$$ bonus for selling a jackpot lottery ticket Bigger bonus (random size x̄≈$150k CONDITIONAL on winning a bonus) → owners *close* this store and open a non-retail business (but if store stays open, it grows) #NUfin2019
Permalink On twitter.com ♻️ 5 Retweets ❤️ 5 Favorites 2019 May 10 Mood +8 🙂

…in reply to @lukestein
Discussion by Edie Hotchkiss @BCCarrollSchool. Thinks about distinguishing financial illiteracy from credit constraints. Founded business not necessarily capital intensive… #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood -2 🙁

…in reply to @lukestein
Not enough time for all the great audience questions. One thing I wonder (but couldn’t ask): Given that these bonuses are public—and in fact there’s associated marketing of stores that sold winners—how much is getting captured by stores’ landlords? #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 10 Mood +3 🙂

…in reply to @lukestein
PAPER 5: “Bank Market Power and Monetary Policy Transmission: Evidence from a Structural Estimation” by Yufeng Wu (@giesbusiness) w Yifei Wang (@MichiganRoss), @toniwhited, @kairong_xiao Banks’ market power creates imperfect pass through of rate changes to depositors. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 10 Mood -2 🙁

…in reply to @lukestein
Super cool: HARD stuff explained clearly. They estimate structural model with monetary policy and bank regulation… First stage: Estimate deposit supply/demand system with local market power cf BLP Second stage: Estimate other parameters using SMM #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood +3 🙂

…in reply to @lukestein
Many findings, eg - Quantitatively, market power as important as regulation-based channels - Deposit market power (and need to compete with cash) + capital regulation creates reversals at low rates (≤~2%); consistent with reduced-form evidence on bank stock returns. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 10 Mood +2 🙂

…in reply to @lukestein
Gregor Matvos (@UTexasMcCombs) discussion helps place paper in broader Monetary Policy × Banking Regulation context. Value of structural approach connecting lots of important pieces (including reserves, which are largely missing from lit). #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 2 Favorites 2019 May 10 Mood +2 🙂

…in reply to @lukestein
Gregor gives incredibly clear summary of model. And a wishlist (stipulated: every model is—appropriately—a simplication) -Interest on reserves -Shadow banks/off-balance-sheet lending -Floating rate lending [Last paper of the afternoon🍻; see you here again tomorrow!] #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 3 Favorites 2019 May 10 Mood +1 🙂

…in reply to @lukestein
Back at #NUfin2019 for day 2 and PAPER 6: “Endogenous Price War Risks” by Winston Dou (@Wharton) with Yan Ji (@hkust) and Wei Wu (@maysbusiness). I saw this great paper in Kentucky two weeks ago but it’s deep and I’m looking forward to hearing it again. @lukestein/1121768115229143040
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +1 🙂

…in reply to @lukestein
See my ☝️linked summary of what Winston, Yan, and Wei do, but I didn't emphasize how much cool data work is here (even though it's a serious theory paper), including textual analysis of media coverage and patents. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +1 🙂

…in reply to @lukestein
Discussion from Andrey Malenko @MITSloan Incentives to compete @indsgemodels with asset pricing. Simplest intuition: Lower prices is an investment in persistent market share → collusion easier to sustain if - Market shares of firms are similar - Growth is higher #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +6 🙂

…in reply to @lukestein
Great connection to IO literatures: - Theory ambiguous: Price wars in recessions (Green and Porter 84)? In booms (Rotemberg and Saloner 86)? - Empirical evidence seems to be that collusion is procyclical but markups are countercyclical. #NUfin2019
Permalink On twitter.com ♻️ 5 Retweets ❤️ 1 Favorite 2019 May 11 Mood +3 🙂

…in reply to @lukestein
…and nice discussion of distinguishing from potential alternative channels - Operating leverage (higher margins in better times because fixed costs less relevant) - Financial leverage (debt can cause equity-maximizing firms to compete aggressively) #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +3 🙂

…in reply to @lukestein
PAPER 7: “Machines and Masterpieces: Predicting Prices in the Art Auction Market” by @CSpaenjers with Mathieu Aubry (@EcoledesPonts computer vision scientist), @RomanKraussl, and Gustavo Manso (@BerkeleyHaas) Looks at machine learning and automated valuation methods #NUfin2019
On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood +5 🙂

…in reply to @lukestein
Lots of idiosyncratic “valuation” problems where automated models can help. Lots are asset valuation questions, but more generally this is (part of) what all recommendation algorithms are trying to do. @lukestein/1099524337361551360 #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +2 🙂

…in reply to @lukestein
Use ~1M art auction lots 08-14 to train neural networks; test out of sample using 2015 auctions Assess - Auctioneers’ pre-sale estimates¹ - ML based on artwork-related text data - ML with text and images - Hedonic predictions ¹May be strategic, not just predictions! #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood 0

…in reply to @lukestein
Hedonic regressions don’t do great for prediction. (Still interesting: Large paintings sell for more. Oils sell for more. Bronzes sell for more…) #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood +5 🙂

…in reply to @lukestein
Standalone, ML doesn’t predict quite as accurately as auctioneer estimates, but ML info can help improve on accuracy of estimates alone. (Cf. chess?) Important cross-sectional variation in difficulty of prediction, too. Also in auctioneer estimates’ systematic biases. #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +4 🙂

…in reply to @lukestein
Always excited for a discussion by Ramana Nanda @HarvardHBS (who has cool work on ML evaluation of startup pitches with @ccatalini and Chris Foster). tuck.dartmouth.edu/uploads/centers/files/Paper_11_Machine_Intelligence_vs._Human_Judgement_in_New_Venture_Finance.pdf Goal - Nonlinear models vs. linear? - Help auctioneers? - Understand limitations of humans? #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 2 Favorites 2019 May 11 Mood +6 🙂

…in reply to @lukestein
PAPER 8: “Organized Crime and Firms: Evidence from Italy” by Pablo Slutzky @SmithSchool with Stefan Zeume @MichiganRoss Another paper I've seen before, have tweeted before, and love. Authors have a broader set of cool crime+finance papers, too. #NUfin2019 @lukestein/1099365700701835264
Permalink On twitter.com ♻️ 5 Retweets ❤️ 6 Favorites 2019 May 11 Mood +1 🙂

…in reply to @lukestein
Mafia affects firms a variety of ways, incl. - Cartel/collusion enforcement - Money laundering - Allocation of license agreements and contracts - Lending to firms Use panel variation in antimafia police efforts (30k asset confiscations) → fin statements of 80k firms #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 5 Favorites 2019 May 11 Mood +2 🙂

…in reply to @lukestein
In addition to baseline results👇, work/working on new stuff on - “Recycling” firms - Shifts in/out of organized economy (night lights, electricity/water usage 🤯) #NUfin2019 @lukestein/1099368639642230784
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood 0

…in reply to @lukestein
Great discussion by Carola Schenone @UVAMcIntire If cost of not paying protection fee ≈∞, why would mafia “protect” firm’s monopoly? Intermafia competition? Maybe mafia serves as alt govt→“protection fee” is just tax. We have lots of relevant models! #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +3 🙂

…in reply to @lukestein
Dear Conference Organizers Everywhere, Two days in a row, we’ve had *really* good food for conference lunch here at @NU_Business #NUfin2019. [Thank you!] The sandwiches were really good sandwiches. Participants notice. Love, Luke
Permalink On twitter.com ♻️ 4 Retweets ❤️ 2 Favorites 2019 May 11 Mood +11 🙂

…in reply to @lukestein
PAPER 9: My presentation (can’t tweet): “Financial Inclusion, Human Capital, and Wealth Accumulation: Evidence from the Freedman’s Savings Bank” (w Constantine Yannelis @ChicagoBooth) Excited for discussion from Charles Hadlock @MSUBroadCollege #NUfin2019 ssrn.com/abstract=3302996
Permalink On twitter.com ♻️ 5 Retweets ❤️ 7 Favorites 2019 May 11 Mood +7 🙂

…in reply to @lukestein
Thanks @nikir1 for tweeting my presentation! And Charlie’s discussion was *amazingly* thoughtful and very useful. #NUfin2019 @nikir1/1127250267945230336
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood +6 🙂

…in reply to @lukestein
PAPER 10/10: “Mobility Constraints and Labor Market Outcomes: Evidence from Credit Lotteries” @JanisSkrastins with Bernardus Van Doornik (Banco Central do Brasil), Armando Gomes (@WUSTLbusiness), and David Schoenherr (Princeton) 💰🇧🇷 🏍🛵🏍🛵 🇧🇷💰 #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood 0

…in reply to @lukestein
.@JanisSkrastins uses randomized timing of access to a ~$3,500 motorcycle loan (𝘤𝘰𝘯𝘴𝘰𝘳𝘤𝘪𝘰𝘴) in Brazil. Mobility matters! - Formal employment ↑ - Wages ↑ - Business ownership ↑ - Appears to reduce spatial mismatch #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets 2019 May 11 Mood +1 🙂

…in reply to @lukestein
This nicely complements @alexandraroulet’s work on randomized (financial) support for commuters in France. #NUfin2019 @lukestein/1048635807358181376
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood +2 🙂

…in reply to @lukestein
Final discussion from Joan Farre-Mensa (@NU_Business → @thisisUIC) Paper is finance + labor + urban economics Consorcio (cf. ROSCA): everyone gets a motorcycle, but early winners get high-NPV transfer from late winners. Shock isn't just mobility, it's financial! #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood -2 🙁

…in reply to @lukestein
Cool to think about potential channels, e.g., - 🛵 especially useful for job search? - 🛵 allows investment in education? - 🛵 allows buffering business cycle shocks? - Mobility works very differently for business owners? #NUfin2019
Permalink On twitter.com ♻️ 4 Retweets ❤️ 1 Favorite 2019 May 11 Mood +1 🙂