Market Design of Trading with Blockchain Technology

Transcription

1 Market Design of Trading with Blockchain Technology Katya Malinova and Andreas Park P2P Financial Systems 2016 London, UK

2 The Buzzword in the Financial Press since mid-2015 Background %#!!" %!!!" $#!!" $!!!" #!!"!"

3 The Buzzword in the Financial Press since mid-2015 Background

4 Why is Blockchain Important? What can it do? An Analogy World Wide Web: frictionless transfer of information Blockchain: frictionless transfer of value Economic Implications: peer-to-peer transfer of value possible no need for trusted third party

5 What is Blockchain and why is it interesting? Securities Trading Today Verifies that Trader has money to buy Exchange - keeps book - displays limit order - facilitates match of Buy and sell Verifies that Trader has stock to sell Broker 2 Broker 1 Trader sends MKT buy order DTCC/CDS (Register holder; could be aggregated at broker level): End of day clearing & settlement Trader posted Limit sell Broadridge Maintains list of Beneficiary owner

6 What is Blockchain and why is it interesting? Transfer of value with Blockchain Stock 3 Pk6543 Money coin Pk5432 Makes offer to buy Pk6543 Pk5432 Agrees to buy

7 Design choice: Permissionless vs. Permissioned Blackchain Public Blockchain Private Blockchain => distributed network Features: - open access - irreversible blocks - public info of ownership (by public key) => select/limited number of network nodes Features: - filtered access - blocks can be changed - ownership info may be public or private

8 Information Environment Today vs. Distributed Ledger Ownership Information Current World Blockchain Broker A Acc A123 Acc A3214 Broker B Acc B432 Acc B123 Acc B654 Pk123 Pk234 Broker Pk321 Acc 3421 Broker A, Acc A123 PK123 Investor Investor Privacy is a choice Broker B, Acc B123 PK321

9 Information Environment Today vs. Distributed Ledger Blockchain = distributed ledger of all transactions of an electronic item Easiest way to think about it: a single, distributed database of past transactions for an item. Past transactions current ownership transactions are identifiable by public keys derive holdings public keys/ids are collections of numbers and letter not directly identify a person/entity. people/entities can choose to reveal identity or be known in permissioned network can achieve anonymity by using multiple IDs hard to achieve in, e.g., OTC markets. privacy is a choice. Main observations that motivate the paper: technology enables transparency of holdings and peer-to-peer trading.

10 Our Paper: Two market types Possible Markets Intermediated market maker absorbs orders at a cost Peer to peer contact other traders probabilistic fill complexity cost

11 Model Ingredients Normally distributed asset. No private information about asset value. Two large institutional investors, of the same size, one of whom will be hit with a liquidity shock. Continuum of small investors, trade small amounts with some probability. CARA intermediaries, willing to take on position (but expensive). Complexity cost of trading peer-to-peer.

12 Setting I: Non-transparent Concentrated Ownership One owner - one key Peer to peer 1: Ownership: one owner, one key => concentrate all holdings under one PK Visibility: NO Investor Large investor, but invisible!

13 Results: Relative to Benchmark with all trading through the intermediaries 1. Large investors split their large trade between small investors and the intermediaries. 2. Payoffs for large investors are higher (because of lower intermediation cost). 3. Price volatility is lower. Note: Nothing in this setup requires Blockchain Tech ITG s POSIT Alert or Algomi already facilitate anonymous peer-to-peer.

14 Setting II: Transparent Concentrated Ownership One owner - one key Peer to peer 2: Ownership: one owner, one key => concentrate all holdings under one PK Visibility of ownership: YES Investor Direct trade Between two Large investors

15 Results: Relative to Benchmark with all trading through the intermediaries I Pros and Cons of Transparency: (+) ability to locate/contact the other large trader escape the complexity costs & rents by the risk-averse intermediaries. ( ) reveal info about the trading needs & may get front-run by the other trader. Modelling Mechanics of Front-Running: (i) LT to LP: Buy quantity Q at price p? (ii) LP buys Q from intermediary and moves the public price P ր to P+x. (iii) LP to LT: sell you Q at price p?

16 Results: Relative to Benchmark with all trading through the intermediaries II Observations: To avoid front-running, offer price concession ( p high enough ) But: in single-shot setting, LP extracts all surplus from LT Repeated setting: front-running is punished in subsequent periods via trigger strategy punishment: Deviation large traders avoid each other; trade as in Setting I forever. Results: 1. In a repeated game, front-running can always be avoided. large traders trade only with each other. large traders share the cost savings. 2. For large enough discount factors ( frequent interactions), trading between large traders at p = 0 is an equilibrium.

17 Setting III: Non-Transparent Dispersed Ownership One share - one key Peer to peer 3: Ownership: one share, one key => disperse holdings over many PKs Visibility of ownership: Yes, but no memory Investor

18 Results: Relative to Benchmark with all trading through the intermediaries I Idea: LT approaches continuum and trades with small and large investors. Observations for equilibrium where large trade with each other: Can be welfare enhancing relative to non-transparent, concentrated ownership (less trade with intermediary). But: sometimes high price concession (which goes to small and large investors!) is necessary which lowers payoff to large traders.

19 Results: Relative to Benchmark with all trading through the intermediaries II Results: 1. Equilibrium with no front-running exists When the discount factor is large enough (frequent interactions). When the intermediated market is sufficiently liquid. 2. (Numerical): For small discount factors, the equilibrium with no front-running where large traders accept each other s offers does not exist. In equilibrium, the liquidity trader offers p = 0 to the continuum, and public keys of the other large trader reject the offer. Welfare-reducing for large traders relative to Setting I because the low probability of acceptance in the continuum leads to over-trading with the intermediaries.

20 Comparison of the three 1. Payoffs in peer-to-peer settings dominate pure intermediation. 2. Payoffs with transparent, concentrated ownership are highest. 3. Payoff ranking with non-transparent ownership depend on whether in equilibrium large investors trade with each other. (i) No? Payoffs with concentrated holdings are better. (ii) Yes? For small price concession p, the payoffs in the setting with dispersed holdings are better.

21 Summary Intrinsic Feature of Distributed Ledger: Information on holdings. Q: How much transparency is desirable? Q: What are there tradeoffs between anonymity and liquidity? Our paper: A framework for pros and cons of holdings transparency. Let s start talking about market design with blockchain technology!

22 Why academics should care about trading with blockchain securities 1. It an economically big question. HFT (still the hot microstructure topic) annual revenue: $1.3B (TABB Group estimate for 2014). IT cost savings from Blockchain for FI: $6B (Goldman), $20B (Santander). current post-trade and securities servicing fees are $100B (Oliver Wyman). 2. The technological change can give rise to new market structures Once-in-a-lifetime opportunity for reset and optimization of the organization and regulation of trading. 3. Academic research can make important contributions to set up solid market organization.