Liquidity: Information Risk

Question 1

Kyle 1985

Answer 1

Setup: (1) 3 agents - informed insider, noise traders, risk-neutral market maker; (2) Continous auction; (3) Insider power/adverse selection modeled as insider knowing ex post liquidation value v of asset and setting his quantity x traded based on that knowledge, noise traders trade a random quantity u, and markets makers only see total quantity traded x + u.

Results: (1) modeling price innovations as functions of quantities traded equivalent to modelling as consequences of new info. (i.e. market makers facilitate efficient prices based on volume alone) (2) optimal trading by insider *not* inconsistent with market efficiency.

Question 2

French and Roll 1986

Answer 2

• Motivation: empirical puzzle of volatility being drastically higher during trading compared to nontrading hours
• Three competing hypotheses: (1) volatility caused by public info arriving during normal biz hours (2) caused by private info that affects prices when informed investors trade (3) caused by pricing errors that occur during trading
• Natural experiment: look at exchange holidays that fall on normal biz days –> since volatility drops significantly during holidays, rule out (1); further tests reveal that mispricing only accounts for 4-12% of volatility –> private info principle factor behind high trading-time variance.

Question 3

Easley and O’Hara 1987

Answer 3

• Model intuition: Conditional on wanting to trade, informed traders prefer to trade larger amts at any given price → adverse selection.
• Noise traders don’t share quantity bias → larger the trade size high prob market maker trading with informed trader.
• ⇒ market maker’s optimal pricing strategy also depends on quantity.
• Testable hypotheses: entire sequence of trades, not merely aggregate volume determines price-quantity relationship

Question 4

Lin, Sanger and Booth (1995)

Answer 4

• Motivation: empirical analysis of adverse selection and order processin costs for market makers.
• Hypotheses: (1) because theory predicts they are fixed costs, avg order processing cost should decrease with trade size
• (2) Easley and O’Hara (1987) predict that informed traders prefer to trade a larger amount –> adverse info component of spread should increase with trade size. Main
• Findings: (1) adverse info component of spread increases significantly and monotonically with trade size
• (2) order processing cost decreases monotonically as trade size increases

Question 5

Easley and O’Hara 2004

Answer 5

• Intuition: private info incr risk to noise traders b/c informed better able to shift pf weights b/c of new info –> noise traders hold too much bad news stock & too little good
• Diversifying can’t remove risk b/c noise traders always on wrong side.
• Informed & noise perceive diff risk/returns → hold diff pf (2FS fails). → private info induces sys risk ⇒private info incr cost of capital
• Results: (1) both public & private info affect firm’s req’d return (2) more info, even if private, better than none for reducing cost of capital
• Prediction: stocks with more private info & less public info, ceteris paribus, will have larger E[Rm - Rf]

Question 6

Chan, Menkveld and Yang 2008

Answer 6

• examine the effect of information asymmetry on equity prices in the local A- and foreign B-share market in China.
• construct measures of info asymm based on market microstructure models
• find that they explain a significant portion of cross-sectional variation in B-share discounts
• price impact measure and the adverse selection component of the bid-ask spread in the A- and B-share markets explains 44% and 46% of the variation in B discounts.
• once domestics allowed to trade Bs → B discount dramatically decreased

Question 7

Tookes 2008

Answer 7

• Develops a theoretical model, with empirical support, centered around idea that insiders have an incentive to trade their rivals stocks when they have private information about shocks to their product market.
• This is particularly true for insiders from large market share firms taking advantage of smaller rivals’ high sensitivities to industry shocks.
• Empirical tests show that order flows (net buyer-initiated volume) and returns in stocks of nonannouncing rivals have info content for announcing-firm returns.

Question 8

Duarte Young 2009

Answer 8

• examines whether PIN is priced because of info asymmetry or because of other liquidity effects that are unrelated to information asymmetry.
• starting point is a model that decomposes PIN into two components, one related to asymmetric information and one related to illiquidity.
• show that the PIN component related to asymmetric information is not priced, while the PIN component related to illiquidity is priced.
• ⇒ liquidity effects unrelated to information asymmetry explain the relation between PIN and the cross-section of expected returns.

Question 9

Glosten 2009

Answer 9

• Uninformed trade is elastic → some uninformed choose not to trade b/c their idiosyncratic valuation lies within the spread. → welfare loss
• Calculations show after some point welfare loss is decr in the amount of informed trade b/c with more informed trade info gets into prices faster & spreads decline.
• For info likely to be revealed in a short amount of time max loss occurs at a very high prob of informed trade → welfare loss is mostly incr in informed trade.
• For longer term info, max occurs at a relatively small prob of informed trade → over some range welfare loss is actually declining in informed trade.