Moore’s Law versus Murphy’s Law: Algorithmic Trading and Its Discontents Flashcards
Moore’s Law
growth of semiconductor industry which had profound impact on the financial system.
Murphy’s law
“whatever can go wrong will go wrong” and it corollary “whatever can go wrong will go wrong faster and bigger when computers are involved”
Algorithmic trading
the use of mathematical models, computers and telecommunication networks to automate the buying and selling of financial securities.
Authors’ idea
A more systematic and adaptive approach to regulating this system is needed, one that fosters the technological advances of the industry while protecting these who are not as technologically advanced.”
Good aspects of algorithmic tradingo The regulatory framework doesn’t adjust so fast to technological innovation.
o Lower costs (bid ask-spread)
o Reducing human error (ex: broker doesn’t buy the needed stock)
o Increasing productivity
Bad aspects of algorithmic trading
o The regulatory framework doesn’t adjust so fast to technological innovation.
Facilitators of development of algorithmic trading:
1) Financial markets are becoming more complex
2) Developments in the financial technology/financial modeling.
3) Breakthroughs in computer technology.
Five major developments that have fueled algorithmic trading’s popularity:
1) Quantitative models
2) Emergence and proliferation of index funds
3) Arbitrage trading activities
4) Push for lower costs of intermediation and execution
a. One doesn’t have to go to the market to sell all his stocks/bonds/CDOs etc.
5) Proliferation of high-frequency trading (40-60% of all trading activities)
2 critical roles of arbitrage
liquidity provisions and price discovery
High-Frequency trading
A form of automated trading that consumes extremely many trades per day/ second. It is an innovation in financial inter mediation that does not fit neatly into a standard liquidity-provision framework.
“Spoofing”
placing an order to buy or sell a security and then cancelling it shortly thereafter.
“Layering”
placing a sequence of limited orders at successively increasing or decreasing prices to give the appearance of change in demand or artificially increase or decrease the price.
5 policies can be implemented for algorithmic trading
1) Do nothing
2) Banning high-frequency trading
3) Change the definition and requirements of a market maker
4) Force all trades to occur at distinct time intervals
5) Tobin tax (small transaction tax on all financial transactions)
+ and - of doing nothing
+ Allow for intermediaries to find their own ways to optimize costs, leading to an even greater supply of immediacy and efficient trading
- Unlikely to address investor’s concerns about fair and orderly markets.
+ and - of banning high-frequency trading
+ More fair and orderly markets in the short-run
- Reduce market liquidity, efficiency, capital formation
+ and - to of Change the definition and requirements of a market maker
+ Fair and orderly market (No withdrawal when their services are needed)
- Increase the cost of intermediaries of being present, greater legal costs
+ and - of forcing all trades to occur at distinct time intervals
+ Higher supply of immediacy
- Investors seemed to prove having preference for continuous systems
+ and - of Tobin tax
+ Increase government budget
- Decrease liquidity, transactions will migrate to other countries
four types of financial regulations
1) System engineered
2) Safeguards-Heavy
3) Transparency-Rich
4) Platform-Neutral
System engineered
Financial regulations should approach automated markets as complex systems composed of multiple software, not only from human side perspective.
Safeguards-Heavy
Regulation should encourage safeguards at multiple levels of the system to check the fairness and the order in all transactions.
Transparency-Rich
Financial regulation should aim to make the financial operations more transparent and accessible.
Platform-Neutral
Financial regulation should be designed to encourage innovation in technology and finance and should be neutral how computer systems in this financial system work.
Examples how algorithm trading destabilized the situation, market
August 2007: Arbitrage Gone Wild - Record loss of hdge fund
May 6, 2010: The Perfect Financial Storm - Order imbalance; At some point of time investors saw the highest volatility (e.g. Apple stock) orders were really enhanced
March and May 2012: Pricing Initial Public Offerings in the Digital Age - i. Facebook IPO delayed by half an hour. Took more time to calculate initial price. And when it was recalculating new offers came in place and many were cancelled.; BATS Global Market, cancelled IPO, due to “computer” glitch
August 2012: Trading Errors at the Speed of Light - i. Knight Capital lost 400m USD and went bankrupt. Algorithm dint manage to pick errors and had to sell hundreds of assets in open markets. In half an hour, the whole capital was wiped out. (150 stocks decreased in their price)
September 2012: High-Frequency Manipulation - Put the orders, so that price would decrease
Quantitative models
- Portfolio theory
- CAPM
- BS option pricing
- Optimization models
Which three developments in the financial industry have greatly facilitated the rise of algorithmic trading.
- The financial system becoming more complex over time
- Breakthroughs in the quantitative modeling of financial market
- Breakthroughs in computer technologies
