Market Liquidity And Its Risk - An Overview

Market Liquidity and Its Risk An Overview

Sebastian Stange and Prof. Christoph Kaserer Chair of Financial Management and Capital Markets Technische Universität München Arcisstr. 21 D-80290 München Tel.: +49 89 / 289 - 25485 Mail: [email protected] URL: www.ifm.wi.tum.de

Presentation aims to convey state of research on market liquidity Goals of this presentation

•

Quick introduction to important aspects and developments

•

Summary of current state of research

•

Starting point for further research

•

Practitioners

•

Students

•

Researchers

•

Those interested

for

General financial background required

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

2

This presentation presents an overview on market liquidity Agenda •

Introduction

•

Market liquidity – Definition – Characteristics – Measurement – Empirical facts

•

Market liquidity risk – Introduction – Models

•

Summary and open research questions

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

3

Introduction

Liquidity has continuous attention in practice Press clippings "Credit crisis puts heat on liquidity Financial market liquidity can take months to build but just seconds to evaporate [...] Credit traders have received a painful reminder of [...] strangled trading liquidity" Financial Times, 30.07.2007

Liquidity always a hot topic – especially in crises • Liquidity always scarce when needed most

NY Times, 28.08.2008

Institutional investors engage in illiquid strategies • Sometimes large, concentrated position to exploit market inefficiencies • Returns due to liquidity risk compensation?

'Some less liquid strategies which also provided genuine portfolio diversification [...] have romped past the S&P 500'

Trading strategies rely on tradability of assets • Hedging often requires frequent trading

'Panicked Traders Take VW Shares on a Wild Ride stock soared to as high as 1,005 euros a share ... after last week at 210 euros.... short sellers were forced to act"

Michael Goldman, Global Investor, 04/2007

"Unless you include liquidity in your models, which some small funds don't, then the model may not always work." Euromoney, June 2007

'[Quantitative strategies] rely on their ability to trade with high frequency […] What if the model is built to sell a company at 20, but there is no buyer?'

Risk management still needs to account for illiquidity • Often neglected • Strong recent pressure from regulators

Euromoney, June 2007

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

4

Introduction

Liquidity is approached from various research directions

Volume / volatilities literature Supply-demand curve

Anomaly explanation Factor in asset pricing

Market microstructure Transaction cost literature

Liquidity Limits to arbitrage Asset management

Risk management

= Causes = Effect Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

5

Introduction

This presentation focuses on market liquidity only Delimitation of topic area

Market liquidity

Funding liquidity

• Also "asset liquidity"

• Liquidity of liabilities

• Liquidity of assets only

• Perspective of a firm (solvency)

– "Marketability" – "Ease of trading an asset"

Monetary liquidity • Liquidity of an economy

• Corporate finance view

• Perspective of an investor • Capital market view

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

6

This presentation presents an overview on market liquidity Agenda •

Introduction

•

Market liquidity – Definition – Characteristics – Measurement – Empirical facts

•

Market liquidity risk – Introduction – Models

•

Summary and open research questions

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

7

Market liquidity – Definition

Liquidity can be defined as the cost of trading an asset Liquidity cost has three components • Cost L of trading a quantity q of an asset • Relative to fair value • Fair value often set at mid-price, i.e. mid-point of bid-ask-spread

Liquidity Lt(q)

= Direct trading costs D(q)

• Includes exchange fees, commissions, taxes • Deterministic • Small for institutional investors

+ Price impact costs PIt(q)

• Difference between transaction price and mid-price • Depends on order size q and point of time

+ Delay costs Dt(q)

• Includes search cost, cost due to add. risk during delay

Note: Definition similar to Amihud (2006), XXX noch ergänzen Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

8

Market liquidity – Definition

Backup

The cost perspective integrates older, more elusive definitions Definition

Source

Comment

• "Ease of trading an asset" • "Marketability"

Longstaff (1995)

• Very general • Cost perspective more specific

• Tightness = "cost of turning a position around in a short time" • Depth = "size of an order flow innovation required to change prices a given amount" • Resiliency = "speed with which prices recover from a random, uninformative shock"

Kyle (1985)

• Several dimensions • No unifying perspective

• Immediacy = time between order submission and settlement

Black (1971)

• A new dimension • No unifying perspective

• Price • Time

Kempf (1999)

• Very general

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

9

Market liquidity – Characteristics

endogenous

endogenous

exogenous

Asset type, order size and horizon are major liquidity determinants

Type of asset

Order size

Liquidation horizon

• Liquidity costs decrease with – Trading volume – Value certainty

• Liquidity costs increase with order size • Reasons • Heterogeneous opinion • Delay probability • Capital restrictions • Liquidity costs decrease with liquidation horizon • Costs relative to return are small when held over long period • Costs are zero for assets held to maturity • Hence, liquidity is a characteristic of the trading process

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

10

Market liquidity – Characteristics

Liquidity is a continuous characteristic

Relative liquidity costs

Liquidity degrees and categories

illiquid

Main issue

Ex.

Category

liquid Costless trading

Degree of illiquidity

Costly, continuous trading

Costly, interrupted trading

No trading

• All order sizes of the asset can be traded at zero costs

• Order sizes can be traded at a cost • Price impact cost important

• Asset are traded from time to time • Zero trading days occur • Delay cost important

• No order size of the asset is traded • Prohibitive Liquidity costs

• Cash

• Limit order book markets of stocks

• OTC markets of more exotic bonds

• Rare art, CDOs?

• None, no liquidity adjustment necessary

• Precise liquidity cost adjustment

• Liquidity cost and delay adjustment

• Intrinsic value determination because price non-observable

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

11

Market liquidity – Characteristics

Cost can be measured as price impact curve Liquidity costs increase with quantity transacted Price

buy price function

ask price Bid-ask- mid price spread bid price

area = absolute liquidity costs

sell price function Market depth Quote depth / Size of best limit orders

Size of next-best bid order

Quantity transacted

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

12

Market liquidity – Characteristics

Optimal trading strategies •

Optimal trading strategies can be applied – Delaying parts of the transaction reduces order size and hence liquidity costs – But, uncertain future price and liquidity costs, i.e. delay costs, for delayed parts of the position – Optimal strategy balances liquidity cost saving against increased delay costs

•

Several optimization objectives possible – Maximize expected liquidation proceeds – Maximize expected liquidation proceeds with penalty for potential shortfall (proceed variance) • Corresponds to maximizing proceeds VaR

•

Immediate liquidation as benchmark strategy – No optimization strategy is applied – Proceeds are certain Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets

Market liquidity v090127b.ppt

13

Market liquidity – Measurement

Measurement of market liquidity is still difficult Multitude of direct measures and proxies available

Direct measures

Indirect measures

• • • • •

Fees Quoted bid-ask-spread (Amihud, Mendelson, JFinEco 86) Traded bid-ask-spread Effective bid-ask-spread (Roll, JoF 84) Relation between price change and order flow (Brennan, Subrahmanyam, JFinEco, 96) • Price response to turnover (Amihud, JFinMar 02) • Volume related reversal (Pastor, Stambaugh, JPolEco 03) • Weighted spread by order size (Irvine et. al., working paper 00) • • • • • •

Depth Volume Number of transactions Turnover rate (Datar et al, JFinMar 98)(1) Proportion of zero-trading days (Bekaert et al., WP, 03) Turnover-adjusted zero trading days (Liu, JFinEco 06)

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

14

Market liquidity – Measurement

Example: Weighted spread measures precise size-specific liquidity cost

Weigthed spread is precise measure...

...under realistic assumptions

• Discount vs. achievable price in limit order book

• Direct trading costs are zero – Ok for institutional investors

• Equals area between limit order curves

• Asset position continuously tradable • Ok for developed markets and positions smaller than market depth

• Generalization of bid-ask-spread to rest of the limit order book • Precise price impact measure L(q)= WS(q) / 2

• Immediate liquidation • Ok, if optimal trading strategies neglectable • Weighted spread data available • Ok in many electronic limit order book markets

Source: Stange, Kaserer (2008a, b) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

15

Market liquidity – Measurement

Liquidity cost quickly rise with order size Example: Weighted spread for German stocks (I/III) Average liquidity cost in bp 500 454 SDAX

400 331 TECDAX

300 214 MDAX

200 146 110

DAX

100 52 15 0 vol 10

vol 25

vol 50

vol 75

vol 100

vol 150

vol 250

27

vol 500

vol 750

vol 1000 vol 2000 vol 3000 vol 4000 vol 5000

Note: Sample = daily data for 160 stocks in four major German stock indices over 5.5 years (II/02-1/08) Source: Stange, Kaserer (2008a) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

Volume in k € (not to scale) 16

Market liquidity – Measurement

Liquidity cost have strongly declined in 2002-2008 Example: Average weighted spread for German stock indices (II/III)

Source: Stange, Kaserer (2008a) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

17

Market liquidity – Measurement

Liquidity cost decline evident for all order sizes Example: Weighted spread for German stocks by selected order size, indexed (II/III)

Source: Stange, Kaserer (2008a) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

18

Market liquidity – Empirical facts

Backup

Further empirical facts •

•

•

•

Liquidity costs can be substantial – Price impact is concave (Hasbrouck 1991, Stange/Kaserer 2008) – Over 50% of total trading cost from price impact and delay (Kritzman, Myrgren and Page 2006) There is strong variation over time – Flight-to-liquidity asymmetry: Liquid assets become more liquid and less liquid less liquid in crises (Acharya and Pedersen (2005), Longstaff (2004)) Liquidity measures determine asset prices – Market liquidity esp. important for pricing – no full diversification possible (Acharya and Pedersen (2005) for stock market; Goyenko (2005) for integration between stock and bond market) – Further sources: Amihud and Mendelson (1986); Datar, Y. Naik, and Radcliffe (1998); Amihud (2002); Pastor and Stambaugh (2003); Acharya and Pedersen (2005); Ang, Chen, and Xing (2006), Keene and Peterson (2007) and others There is strong liquidity commonality Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets

Market liquidity v090127b.ppt

19

This presentation presents an overview on market liquidity Agenda •

Introduction

•

Market liquidity – Definition – Characteristics – Measurement – Empirical facts

•

Market liquidity risk – Introduction – Models

•

Summary and open research questions

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

20

Market liquidity risk – Introduction

Traditional VaR neglects liquidity effects Definition classical VaR

Critique

• Value-at-Risk (VaR) measures worst loss in α% of the cases over a specific forecast horizon h – (1-α) is confidence level – Loss is generally the forecasted loss in market prices

• Assumes that position can be liquidated without significant cost – Bid-ask-spread neglected – Price impact of position size neglected – Market conditions and liquidation horizon neglected

Future market price distribution at horizon h

• Assumes that position can be liquidated within any given horizon – No delay costs • Sometimes ad-hoc adjustments from expert estimates used, but only rough proxy for true liquidity risk – Flat lengthening of horizon – Artificial increase of volatilities – Flat liquidity cost deduction ("hair cut") by asset class

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

21

Market liquidity risk – Introduction

Traditional VaR assumes 'mark-to-market'-price by convention Several price definitions can be used Before transaction prices Mid-price = mark-to-market price • Middle of bid-ask-spread Bid- / ask-price • Usually quoted by market maker – But sometimes best limit order prices if no market maker coverage • Ex-ante achievable transaction price for buy/sell-transaction – But for small volume (=bid-askdepth) only Achievable price in limit order book • Ex-ante price when transacting larger orders as market order against limit order book

After transaction prices Transaction price • Ex-post achieved transaction price • Can be inside bid-ask-spread if market maker transacts within spread • Usually outside bid-ask-spreads for larger order sizes

'Mark-to-market' simple, but not most realistic assumption Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

22

Market liquidity risk – Introduction

Market liquidity risk is worst loss due to liquidity cost •

Liquidity component usually neglected in standard risk models – Usual assumption that liquidity cost can be neglected if horizon is long enough – Partially accounted for via asset-class-specific "hair cuts" • Expert estimates of liquidity cost deduction

•

However, liquidity costs are substantial and strongly vary between securities – 25-30 % underestimation of total risk in emerging market currencies (Bangia et al. 1999) – Bid-ask-spread component over 50 % of total risk for illiquid stocks (Le Saout 2002) – 30 % liquidity contribution to intraday risk in small price stocks (Lai 2007) – Up to 25-30 % liquidity impact on price risk at 10-day horizons (Stange/Kaserer 2008a) • More than doubles at daily horizons • Large variance between stocks and position sizes Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets

Market liquidity v090127b.ppt

23

Market liquidity risk – Models

We will survey eight different liquidity risk models Model overview Model categories

Models 1.

Bangia et al. (1999): Add-on model with bid-ask-spread

2.

Ernst et al. (2008): Modified add-on model with bid-ask-spread

3.

Berkowitz (2000): Transaction regression model

4.

Cosandey (2001): Volume based price impact

5.

Angelidis, Benos (2006): Structurally implied spread

6.

Francois-Heude, Wynendaele (2001): Limit order model

7.

Giot, Gramming (2008): T-distributed net return model with weighted spread

8.

Stange, Kaserer (2008): Empirical netreturn model with weighted spread

• Available data • Type of liquidity measurement • Assumptions on the structure of liquidity

Note: Untraceable models excluded from overview. Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

24

Market liquidity risk – Models

A simple model deducts worst bid-ask-spread 1

Bangia et al. (1999): Add-on model with bid-ask-spread Liquidity Measure • Bid-ask-spread

Advantages and Disadvantages • Data available in many markets • Simple add-on, hence quickly implementable

Liquidity Risk Measurement • Worst spread added as cost to worst price

• Empirical distribution ( ) used for worst spread estimation • Normal distribution used for price risk

– Assumes that position can be traded at bidask-spread • However, liquidity costs can quickly rise beyond bid-ask-spread when trading positions larger than bid-ask-depth • Tends to underestimate risk – Assumes perfect liquidity – price correlation • Worst spread and worst price occur simultaneously • Tends to overestimate risk – Logically inconsistent because worst spread deducted from current, not worst prices • Easily correctable • Tends to overestimate risk – Historical spread distribution might poorly proxy for future distribution

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

25

Market liquidity risk – Models

Non-normality of liquidity can be explicitly taken into account 2

Ernst et al. (2008): Modified add-on model with bid-ask-spread (I/II) Liquidity Measure • Bid-ask-spread • Other liquidity measures can be used analogously

Liquidity Risk Measurement • Worst cost added to worst price

Advantages and Disadvantages • Similar to Bangia et al. (1999), but specifically accounting for non-normality via CornishFisher approximation • Provides empirically more precise results than Bangia et al. (1999) • See Ernst et al. (2008) • Can be applied to other liquidity cost measures as well • See Ernst et al. (2009) – Skewness and kurtosis need to be additionally estimated

• Cornish-Fisher approximated percentiles, which take skewness ( ) and excess-kurtosis ( ) into account

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

26

Market liquidity risk – Models

Modified risk model performs better than Bangia et al. (1999) 2

Ernst et al. (2008): Modified add-on model with bid-ask-spread (II/II)

Note: "Risk correctly estimated" is determined via standard Kupiec (1995)-statistic; Sample = 160 stocks in four major German stock indices over 5.5 years Source: Ernst, Stange, Kaserer (2008) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

27

Market liquidity risk – Models

Price impact of order size distilled from transaction data 3

Berkowitz (2000): Transaction regression model Liquidity Measure • Abs. liquidity cost per share ( ) derived from linear regression of transaction prices when controlling for other risk factor changes (x)

Advantages and Disadvantages • Integrates price impact of order size – High data requirements – Liquidity measure highly approximate • Measurement very noisy – Assumption of zero price-liquidity correlation • Independence between price risk and liquidity impact

Liquidity Risk Measurement • Not explicitly defined

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

28

Market liquidity risk – Models

Simple price impact approximation via relative traded shares 4

Cosandey (2001): Price impact derived from volume Liquidity Measure • Relative increase of traded shares by position

• Assumes that position increases number of shares but total traded value Liquidity Risk Measurement

Advantages and Disadvantages • Market volume data available • Accounts for price impact of order size via simple theoretical assumption – No time variation of liquidity (here measured with traded share volume) – Price impact linear to relative traded shares • Concavity of price impact function neglected

• Worst price risk adjusted for relative increase of traded shares

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

29

Market liquidity risk – Models

Liquidity cost derived from theoretical drivers 5

Angelidis, Benos (2006): Structurally-implied spread Liquidity Measure • Spread model estimated from intraday data • Model derived from theoretical ideas on liquidity drivers • Traded shares N, degree of info assym. theta, price elasticity to volume kappa, liquidity fixed cost Phi

Advantages and Disadvantages • Partially accounts for price impact of order size via increased volume percentile – Unclear if structural model is correct – Complicated estimation of parameters in intraday data required – Possible overestimation because liquidityreturn correlation assumed perfect

Liquidity Risk Measurement • Structurally-implied spread added to price risk • Calculated with top-percentile of volume • Assumes that individual position size dissipates in market

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

30

Market liquidity risk – Models

Using more limit order book data increases measure preciseness 6

Francois-Heude and Wynendaele (2001): Price impact from best limit orders Liquidity Measure • Price impact curve estimated from best five limit orders • Available from Paris Bourse • Extrapolated for larger sizes

Liquidity Risk Measurement • Normal price risk adjusted for average price impact • Correction term for difference between average and stock-specific price impact

Advantages and Disadvantages • Accounts for price impact of order size • Price impact more precise because directly measured in limit order book – Only applicable in electronic limit order book markets – Time variation of liquidity neglected – Assumes perfect correlation between liquidity and price – Five best limit orders need to be available – Intraday only – Liquidity cost only extrapolated for medium to large order sizes – Somewhat arbitrary spread adjustment – Provides empirically imprecise results than other models using limit order data • See Ernst et al. (2008)

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

31

Market liquidity risk – Models

Precise price impact measurement with limit order book data 7

Giot, Gramming (2005): T-distributed net-return model with weighted spread Liquidity Measure • Size-specific liquidity cost extracted as weighted spread from limit order book

• at(n) and bt(n) are weighted ask and bidprices for trading n = q/Pmid shares Liquidity Risk Measurement • T-distribution percentile of historical net return distribution • Net return is mid-price return net of weighted spread liquidity costs

Advantages and Disadvantages • Accounts for price impact of order size in very precise way • Correctly accounts for liquidity-return correlation – Only applicable in electronic limit order book markets – Intraday data required if lower frequencies not provided by exchange – Possible overestimation if instant liquidation assumption is suboptimal – Possible distortion through assumption of tdistributed net-returns

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

32

Market liquidity risk – Models

Precise empirical price impact measurement with limit order book data 8

Stange and Kaserer (2008): Price impact measured with weighted spread Liquidity Measure • Size-specific liquidity cost extracted as weighted spread from limit order book

• at(n) and bt(n) are weighted ask and bidprices for trading n = q/Pmid shares Liquidity Risk Measurement • Empirical percentile of historical net return distribution • Net return is mid-price return net of weighted spread liquidity costs

Advantages and Disadvantages • Accounts for price impact of order size in very precise way • Correctly accounts for liquidity-return correlation – Only applicable in electronic limit order book markets – Intraday data required if lower frequencies not provided by exchange – Possible overestimation if instant liquidation assumption is suboptimal

Source: Stange, Kaserer (2008b) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

33

Market liquidity risk – Models

Model overview

Source: Stange, Kaserer (2009) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

34

Market liquidity risk – Models

Limit order data models with superior performance in stocks Model performance

However, models including delay costs not developed yet Note: "Acceptance rate" is fraction of stocks where Kupiec (1995)-statistic could not reject deviation between realized and predicted loss frequency; ; Sample = 160 stocks in four major German stock indices over 5.5 years Source: Ernst, Stange, Kaserer (2009) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

35

This presentation presents an overview on market liquidity Agenda •

Introduction

•

Market liquidity – Definition – Characteristics – Measurement – Empirical facts

•

Market liquidity risk – Introduction – Models

•

Summary and open research questions

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

36

Summary Market Liquidity Market liquidity is cost of trading an asset • Components: Direct trading costs, price impact costs, delay costs • Determinants: type of asset, order size, liquidation horizon • Continuous liquidity degrees – different treatment required • Optimal trading strategies developed to minimize sum of cost components – Balance increased delay cost against reduced price impact – Less relevant for risk management • Several liquidity measures available – Direct liquidity measure promise increased precise liquidity measurement Several models can be chosen to measure market liquidity risk • Market liquidity risk is substantial, but often neglected • Models based on limit order data perform better in backtests • Modified risk model (Ernst et al. 2008) consistently outperforms

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

37

Backup

Interesting open research topics 1. 2.

3.

4. 5.

6. 7. 8.

9.

Can the superiority of the model of Ernst et al. (2008) be improved via better moment estimation techniques?** Estimation and validity of optimal trading strategies (OTS)* – How can parameters of theoretical models with optimal trading strategies be estimated? – In which situations are optimal trading strategies beneficial compared with instant liquidation - in normal times or also in crises? In which situations are liquidity costs efficient? • How should arbitrage strategies be constructed? • What drives the efficiency of markets? Are total liquidation costs per stock an asset pricing factor? – Total liquidation costs can be estimated via weighted spread times traded size distribution Can the price impact curve be described via theoretical processes? – Similar to interest rate curve – Arbitrage conditions also apply if liquidity price are efficient – Can be helpful when describing unobservable / hidden liquidity Is it possible to construct liquidity options – in analogy to volatility options? How high are liquidity cost / risk levels between different asset classes? Can the stability of liquidity cost for relative order size be used in liquidity risk measurement? – Liquidity costs are relatively constant for order size in relation to volume and market value in the cross section (Stange, Kaserer 2008) – Is this helpful in implementing an asset class, not asset specific liquidity risk approach? How should liquidity risk be integrated into portfolios? – What is the role of liquidity commonality in the covariance matrix?

Source: Stange, Kaserer (2009) Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

38

Backup

Further references • • • • • • • • • •

Bervas (2006) "Market Liquidity and its Incorporation Into Risk Management", Banque de France technical report, http://www.gloriamundi.org Ernst, C., S. Stange. and C. Kaserer (2008a) "Accounting for Non-normality in Liquidity Risk", CEFS working paper 2008 No. 14, www.cefs.de Ernst, C., S. Stange. and C. Kaserer (2008b) "Empirical evaluation of market liquidity risk models", CEFS working paper 2009 No. 1, www.cefs.de Erzegovesi (2002) "VaR and Liquidity Risk. Impact on Market Behaviour and Measurement Issues", Alea Technical Reports Nr. 14, http://eprints.biblio.unitn.it Jorion (2007) "Value at risk: the new benchmark for managing Financial risk", 2. Ed., McGraw-Hill, New York Loebnitz (2006) "Market Liquidity Risk: Elusive no more - Defining and quantifying market liquidity risk", diploma thesis, University of Twente, http://purl.org/utwente/e582 Mahadevan, A. (2001) "Incorporating Liquidity Risk in VAR estimation", ICICI Working paper, http://www.gloriamundi.org Stange, S. and C. Kaserer (2008a): "The Impact of Order Size on Stock Liquidity - A Representative Study", CEFS working paper, www.cefs.de Stange, S. and C. Kaserer (2008b) "Why and How to Measure Liquidity Risk", CEFS working paper, www.cefs.de Stange, S. and C. Kaserer (2009) "Market Liquidity Risk – An Overview", CEFS working paper, www.cefs.de

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

39

Thank you for your attention! Comments are warmly welcome.

Sebastian Stange

Prof. Christoph Kaserer

Sebastian Stange and Christoph Kaserer, Chair of Financial Management and Capital Markets Market liquidity v090127b.ppt

40