06 Inst Trad Liq II - lecture notes PDF

Preview

Summary

lecture notes...

Description

Lecture 6 Institutional Trading and Liquidity II • • • • •

How do institutions trade? What are the institutional trading needs? Non-displayed orders in lit venue Dark pool trading Internalization

Dr. Yuanji Wen [email protected]

Tuesdays: 4PM – 6PM UWA Business School Wesfarmers LT

Chapters 3 & 4

FINA3307 Trading in Securities Markets

Recap  What is liquidity • Black, Kyle’s definition

 Who are institutional traders? • US • Australian

 Price impact of institutional trades  How do they trade? This week continues here…

The University of Western Australia

2

Related Chapters  Teall, 2018, Chps 3 & 4  Hasbrouck, 2017, Chps 8, 13

The University of Western Australia

3

What are the institutional trading needs?  Best execution

Best price

• “traders receiving the most favourable terms available for their trades” • Execution cost: Order processing costs + Market impact costs • More in Week 9

 Transparency • “Everyone wants to see the liquidity, but no one is actually going to put his or her order there. Everyone wants markets to be transparent, but nobody wants anyone else to see what they themselves want to do.” – Day trader, hedge funds and others pilfered research done by institutional investors. The University of Western Australia

4

What are the institutional trading needs?  Consolidation of order flow • Consolidation refers to the pooling of order flow in one market center. – Increases order interaction, concentrates liquidity – Improves accuracy of price discovery • Fragmentation both “spatial” and “temporal” due to the “slice and dice” – Affects quantity discovery and price discovery – Increases intraday price volatility

The University of Western Australia

5

How do institutions trade? Algorithmic trading  Algorithmic trading refers to automated trading with the use of live market data and rule-driven computer programs • For automatically submitting and allocating trade orders among markets and brokers. • As well as over time so as to minimise price impact of large trades.

 Many institutional traders use algo trading to reduce execution risk, preserve anonymity, and to minimise trade slippage. • Trade slippage = market impact of trade (i.e., Pt+k – Pt) • Relation between trade slippage and transaction size is likely to be nonlinear.

The University of Western Australia

6

Hendershott Jones Menkveld 2011 http://albertjmenkveld.org/ Q1- Largest cap quitile

The University of Western Australia

7

How do institutions trade?

The University of Western Australia

8

How do institutions trade? Trading strategies  Minimise trading costs by revealing minimal information about their trading intentions to the market. • What do institutions require in terms of transparency? – Anonymity! – However, they want to see orders of everyone else.

 Pre-trade transparency vs post-trade transparency • Ex-ante transparency vs ex-post transparency

The University of Western Australia

9

Dark Executions  “Darkness” generally refers to pre-trade opacity.  Dark executions typically arise from one of the following mechanisms: 1.

2.

3.

Trades that involves nondisplayed orders in lit venue Dark pool trading Internalization

U.S. EQUITIES TRADE VOLUME (%), DEC. 2018 Exchange (Lit) 27%

1. Exchange (Hidden) 5%

OffExchange 68%

2. Dark pool: 12 -14% 3. Other off-exchange: 56 - 58% (Rosenblatt Securities estimates) The University of Western Australia

1. Non-displayed orders in lit venue  Non-displayed orders are limit orders that allow traders to hide partially or fully the intended volume at the time of order submission. • Hidden orders: Min disp. shares (peak size) = 0 • Iceberg orders: Min disp. shares > 0 – only a small part of an order is shown in the limit order book, while the larger part is hidden.

 They interact with other orders in a limit order market that also handles visible orders. – Executions in visible markets The University of Western Australia

1. Nondisplayed orders in lit venue (cont’d)  How does this type of order work? • The “price-visibility-time” priority rule is imposed on lit venue. • Non-displayed shares have lower execution priority over displayed shares. • Once the peak part is traded, it would be refilled by the remaining volume.

 Example (ASX, LSE) • An iceberg order to buy 25,000 @ 100p with peak = 10,000 (15,000) 10,000 @ 100p

The University of Western Australia

12

1. Non-displayed orders in lit venue (cont’d)  Example (cont’d): what if an aggressive sell order of 11,500 shares arrived at 8:26:00? (5,000) 8,500 @ 100p, new time label 8:26:00

• Transactions (“Time & Sales”) are recorded as: – 10,000 @100p – 1,500 @100p

 Anti-gaming feature: Randomize the peak size • NASDAQ reserve order – https://business.nasdaq.com/media/iceberg-order-fs_a4_tcm504459476.pdf The University of Western Australia

13

2. Dark pool trading  Dark pools are venues/forums where buyers and sellers trade securities without displaying orders to market participants. • Function in parallel with traditional markets.  Classified by operator: • Exchange-owned ones such as BATS Europe Dark Pool, Turquoise Dark and et al. • Broker-operated ones such as UBS ATS, Credit Suisse’s CrossFinder, Goldman Sachs’ Sigma X and et al.

 Classified by execution algorithm: • Mid-quote execution (one-sided market): US; prevails in European markets • Dark limit order book trading (two-sided market): US The University of Western Australia

2. Dark pool trading  Classified by execution algorithm (cont’d) Types

Examples

Typical Features

One-sided market ITG Posit, Liquidnet, Instinet

Mostly owned by agency brokers and exchanges; typically execute orders at midpoint or VWAP of exchanges, and customerto-customer

Two-sided market Credit Suisse Crossfinder, (dominate in Goldman Sachs Sigma X, terms of ADV) City Match, Baclays LX, Morgan Stanley MS Pool, UBS PIN

Most broker-dealer dark pools; may offer some price discovery and contain proprietary order flow.

The University of Western Australia

15

Source: https://tabbforum.com/liquidity-matrix/equities/ The University of Western Australia

16

2.1 Dark pool – Mid-quote execution  How does it work? • Derivative pricing rule

 Theoretically, because informed traders trade on one side, the dark pool has high execution risk for informed traders.  Benefits for venue operators: • Example of price matcher – First, reduce the cost of determining and updating advertised prices – Second, in the presence of a price matcher, the advertiser has a reduced incentive to post an aggressive price. When the advertiser lowers the posted price, the additional customers will be split with the price matcher.

 Impact on the pegged lit venue • Low or even no volume? The University of Western Australia

17

2.2 Dark pool – dark limit order book market  How does it work? • Still kind of derivative pricing rule via complex orders such as mid-pegged order, far-pegged, near-pegged order et al.

 Theoretically, because it is two-sided market, the execution risk concerned by informed traders is lower.  Issues: • Price manipulation

The University of Western Australia

18

2.3 Other issues of dark pools  Principal- agent problem • Venue operator maximizes their profit instead of acting in the best interest of their clients • E.g., Barclay’s misleading advertisement of its dark pool

 Rule-violation • Sub-penny trading • E.g., UBS paid $12 million fine to settle SEC complaint that in its dark pool allowed high-frequency traders and market makers to jump ahead of customer orders

The University of Western Australia

19

ASIDE: Trading game – dark market  No price tracking function! No derivative pricing rule!

3. Internalization  A crude way of distinguishing dark pools from brokerdealer internalization is that the former are often marketplaces that allow direct customer-to-customer trades, whereas the latter typically involves brokerdealers as intermediaries.  An example: • A NASDAQ market maker (MM) trades against a customer order at the NBBO at a time when the MM’s own quotes are behind the NBBO.

The University of Western Australia

21

In-class exercises  In each of the following situations, determine if the trade is regarded as “dark”, and why or why not? Assuming, the NBBO is 10.00 (bid) and 10.10 (ask)  A: MM in SPQR, whose own offer is 10.15, receives a customer order to buy 100 shares at the market. A1) He routes the order to an exchange offering 10.10, where the order is executed at 10.10. A2) He sells 100 shares to the customer at the NBO, 10.10.  B: The INET book receives an order to buy 100 shares limit 10.20. The order executes against a hidden limit order priced at 10.09  C: Pension fund LSM and mutual fund EFT are paired off in the Credit Suisse Crossfinder system. LSM buys, EFT sells, 10,000 shares, at the midpoint of NBBO, 10.05

The University of Western Australia

22

Regulations in US  Volcker Rule • Part of the Dodd-Frank Wall Street Reform and Consumer Protection Act. • Former chairman of the Fed, Paul Volcker • It prohibits banks and institutions that own a bank from engaging in proprietary trading or even investing in or owning a hedge fund or private equity fund. • Risk concern

 SEC requires ATS to regularly disclose information on how they operate. • Principal- agent problem, rule violation et al. The University of Western Australia

23

Proprietary Trading  What is Proprietary Trading? • Proprietary Trading (Prop Trading) occurs when a bank or firm trades stocks, derivatives, bonds, commodities, or other financial instruments in its own account, using its own money instead of using clients’ money. • This enables the firm to earn full profits from a trade rather than just the commission it receives from processing trades for clients.

 Prop traders use various strategies such as merger arbitrage, index arbitrage, global macro-trading, and volatility arbitrage to maximize returns. Proprietary traders have access to sophisticated software and pools of information to help them make critical decisions. The University of Western Australia

24

Regulations in Europe  double caps mechanism for dark pools (DVC) • 4% - single venue • 8% - all venue

 Venues that use periodic auction • Is it a dark pool? • Does it result dark execution? • Recent debate and potential issues Hans Gulyas, “Periodic auctions under MiFID II a loophole to circumvent transparency obligations”, Posted on LMS The University of Western Australia

25

Limit Order Book (LOB) trading game

The University of Western Australia

26

How to use logs?  Note that logs in our trading game contain more information than what is offered by stock exchanges/data providers in practice. • E.g., broker ID, True Value.

 This allows you to make comprehensive analysis that could not even be done with archival data from practice.  Key column: “Type” The University of Western Australia

27

How to use logs?  Related information that could be retrieved: • True Value: Type = TRUE_VALUE_UPDATE • VWAP: next slide • Your own transaction: Type = TRANSACTION and then search your trading account in columns F and G (“bidder” and “asker”) • Market information such as price series (column H), bid-ask spread (columns K and L); venue level comparison? • Individual information: not only yours but others’ transaction/order submission The University of Western Australia

28

How to compute VWAP? EXCEL  Start from log file (downloadable from LMS)  Filter data: • Type = TRANSASCTION • Copy and paste into a new spreadsheet, named “Clean” sheet;

 Choice of variables: • Venue = Mkt%, Mkt# • Price • Volume

 Command: (suppose there are only 19 trades) “= SUM(H2:H20*I2:I20)/SUM(I2:I20)” • Press ctrl+shift+enter • Shown as “{=SUM(H2:H20*I2:I20)/SUM(I2:I20)}”

 You may want to take only 2-digits rounded number for further computation. The University of Western Australia

29

Last slide  Lecture quiz

 Next week: Mid-semester Exam • See information on LMS

 Week 8: trading gone awry • Illegal insider trading • HFT undesirable strategies

The University of Western Australia

30...