Quantitative trading - page 28

 

Rajib Ranjan Borah at POC2015 - Trade Options and be ahead of the markets



Rajib Ranjan Borah at POC2015 - Trade Options and be ahead of the markets

Contents:

  • 04:27 Definitions
  • 07:12 Two types of options
  • 09:56 Option Types elaborated
  • 15:38 Terminology - Moneyness
  • 23:35 Option Premium (price) components
  • 33:28 Directional strategies
  • 36:54 Volatility strategies
  • 46:30 Hedging strategies
  • 50:38 Volatility Curve Skew Trades
  • 59:14 Risk Evaluations
Rajib Ranjan Borah at POC2015 - Trade Options and be ahead of the markets
Rajib Ranjan Borah at POC2015 - Trade Options and be ahead of the markets
  • 2015.09.22
  • www.youtube.com
QuantInsti is a pioneer training institute to learn algorithmic trading - http://www.quantinsti.com/Know more about QuantInsti's Algorithmic Trading Course -...
 

Momentum Based Strategies for Low and High Frequency Trading | Webinar



Momentum Based Strategies for Low and High Frequency Trading | Webinar

This webinar focused on the various aspects of Momentum Trading Strategies for both Conventional/Low Frequency as well as High Frequency (HFT). Some popular strategies in momentum based trading were also dug deeper into to select niche momentum trading strategies. The webinar aimed to evaluate how HFT momentum strategies differ from conventional momentum strategies both from logic and deployment perspective.

Points discussed in detail:

  • Popular Momentum Trading Strategies
  • Momentum Trading in HFT
  • Risks in Momentum Trading
  • Sample Model
Momentum Based Strategies for Low and High Frequency Trading | Webinar
Momentum Based Strategies for Low and High Frequency Trading | Webinar
  • 2015.12.04
  • www.youtube.com
This webinar focused on the various aspects of Momentum Trading Strategies for both Conventional/Low Frequency as well as High Frequency (HFT). Some popular ...
 

[WEBINAR] Changing Notions of Risk Management in Current Markets



[WEBINAR] Changing Notions of Risk Management in Current Markets

In this video Mr. Rajib Borah, Director and Faculty at QuantInsti, talks about a few major risk oversight issues in algorithmic trading, like:

  • How did Knight Capital lose $460 in 45 minutes?
  • Why was Deutsche Bank forced to close their Algorithmic Trading desk in Tokyo?
  • What went wrong at Infinium Capital while trading Crude ETFs and why were they fined $850,000?
  • What mistake caused HanMag Securities of Korea to lose 57 billion Korean Won in a few minutes?
[WEBINAR] Changing Notions of Risk Management in Current Markets
[WEBINAR] Changing Notions of Risk Management in Current Markets
  • 2015.08.11
  • www.youtube.com
The video is a recording of QuantInsti's Webinar on "Changing Notions of Risk Management in Current Markets" which was conducted on 10th August, 2015. In thi...
 

Nitesh Khandelwal at POC2015 - Trade Futures and be ahead of the markets



Nitesh Khandelwal at POC2015 - Trade Futures and be ahead of the markets

Nitesh Khandelwal delivers an overview of futures and options trading, emphasizing that futures are financial instruments whose value depends on the price of an underlying financial instrument. He differentiates between futures, which are standardized contracts traded on exchanges, and forwards, which are traded on the over-the-counter market. Khandelwal highlights the various participants in futures trading, including hedgers, speculators, and arbitrageurs, and explains how each group can benefit from engaging in futures trading. The pricing of futures contracts and the modeling of trading strategies using futures are also discussed.

Moving on, Khandelwal delves into the types of market participants in the futures market, namely hedgers, arbitrageurs, and speculators. Hedgers utilize futures contracts to safeguard themselves against potential price increases in the physical market, effectively minimizing their risk. Arbitrageurs seek profit opportunities by exploiting price discrepancies between different exchanges, while speculators participate in futures trading solely to capitalize on price fluctuations. Khandelwal proceeds to define two essential characteristics of futures markets: the spot price, which represents the underlying asset's current price, and the contract or lot size, which specifies the predetermined size of the futures contract.

The concept of futures trading is then explained, with Khandelwal highlighting that futures contracts come in various sizes and have expiration dates. Settlements can be made in cash or through cross settlements, with cash settlements being the most common. Margins are used to initiate and maintain positions, and each asset has specific margin requirements based on price expectations. Futures trading allows for substantial leverage, as only a small percentage of the underlying asset's value is required to take a position. However, this also increases the risk for traders and clearinghouses, especially during periods of extreme market volatility.

Delivery aspects in futures trading are discussed, as certain contracts can be deliverable while others cannot. Commodities and stock futures can be delivered, but index futures cannot, as indices are merely numerical representations without a physical counterpart. During delivery, the exchange provides a list of accepted parameters for the underlying asset to ensure quality standards. Khandelwal underscores the advantages of trading futures, such as the ability to leverage positions by paying a margin instead of the full asset price and the wider range of trading strategies available compared to the cash market.

Khandelwal then explores the benefits of trading futures over cash markets, including enhanced liquidity for larger quantities and a fair and transparent price discovery process at different points in time. He explains that futures prices are determined by various factors, including spot prices, the date of expiry, risk-free rates of return, storage and delivery costs, and the convenience yield. The convenience yield represents the price businesses are willing to pay to possess an asset physically, thereby avoiding supply and demand issues and potential delivery defaults upon expiry.

The speaker provides insights into the concept of convenience yield, particularly in relation to investable assets like gold, where physical ownership is often preferred due to its symbolic value. A formula for calculating the expected price of equity futures or index futures is presented, taking into account the current spot price and the potential return from investing the money elsewhere. Storage costs and convenience yield, reflecting the premium investors are willing to pay to hold the physical asset, are also factored into the equation. Khandelwal notes that rational investors consider convenience yield when formulating their market views.

The concept of cash-future strategy is introduced, which involves trading in both cash and futures in opposite directions to generate profits. This strategy requires sufficient liquidity in the cash market for the stocks being held and access to mechanisms for delivery if shorting is allowed. However, Khandelwal advises caution regarding the high returns observed in short cash and long futures positions, as the feasibility of such options depends on the available delivery mechanisms.

Factors affecting the volatility of spreads as the expiration date approaches are explained by Khandelwal. These include the lack of returns during the zero period, potential erratic spreads due to delivery mechanisms during futures delivery, the impact of prevailing interest rates, and market sentiment during periods of high volatility or news announcements that can cause swift price movements. Two spread trading strategies are discussed: calendar spreads, which are highly effective and offer risk-free opportunities on the spot market, and intermarket spreads, which involve arbitrage or statistical arbitrage across different but related asset classes.

Khandelwal delves into the analysis of correlations between different asset classes, such as commodities, equities, and currencies. He highlights that movements in one instrument can indicate potential movements in others, although the direct or inverse correlation depends on thorough analysis. Correlations can also exist within the same asset class, as exemplified by the inverse relationship between food prices and gold prices. Market sentiment and fundamental analysis play crucial roles for investors in taking positions based on these correlations. Khandelwal introduces interexchange spreads, which can be pure arbitrage or statistical arbitrage, depending on their connection to each other, even if they do not belong to precisely the same asset class.

Nitesh Khandelwal further discusses the importance of understanding correlations between different asset classes in trading. By recognizing the relationships between commodities, equities, or currencies, traders can gain valuable insights into potential market movements. When one instrument experiences a shift, there is a likelihood of similar movement in related assets. However, the nature of the correlation, whether direct or inverse, depends on in-depth analysis and market conditions. Khandelwal emphasizes that correlations can also exist within the same asset class, as demonstrated by the inverse relationship between food prices and gold prices. This type of correlation indicates the market sentiment and provides opportunities for fundamental analysis-based positions.

Additionally, Khandelwal introduces the concept of interexchange spreads, which involve trading strategies that exploit price discrepancies between different exchanges. These spreads can be categorized as pure arbitrage or statistical arbitrage, depending on the nature of the connection between the involved assets. Despite not belonging to the same asset class, interexchange spreads offer opportunities for profit if traders can identify and capitalize on the pricing disparities.

Nitesh Khandelwal's comprehensive overview of futures and options trading covers essential aspects such as the participants in the futures market, characteristics of futures contracts, trading strategies, pricing factors, convenience yield, spread volatility, correlations between asset classes, and interexchange spreads. By understanding these concepts and their interplay, traders can make informed decisions and potentially optimize their trading strategies in the dynamic and ever-changing financial markets.

  • 00:00:00 Nitesh Khandelwal provides an overview of futures and options trading, explaining that futures are financial instruments whose price is based on the price of some other underlying financial instrument. Futures are standardized contracts traded on exchanges, while forwards are traded on the over-the-counter market. Market participants, including hedgers, speculators, and arbitrageurs, can benefit from trading in futures. Khandelwal also discusses the pricing of futures and the modeling of different trading strategies using futures.

  • 00:05:00 Nitesh Khandelwal explains the three types of market participants in the futures market: hedgers, arbitrageurs, and speculators. Hedgers participate in the futures market to protect themselves from buying goods at higher prices in the physical market by buying futures, thereby minimizing the risk of loss. Arbitrageurs are market participants who seek to make a profit by buying an asset on an exchange where the price is cheaper and selling it on another exchange where the price is higher. Speculators participate in the futures market solely to profit from the price fluctuations of assets. Khandelwal then defines two key characteristics of futures markets: the spot price, which is the price of the underlying asset, and the contract or lot size, which is the predetermined size of the futures contract.

  • 00:10:00 Nitesh Khandelwal explains the concept of futures trading. Futures contracts come in different sizes depending on the asset being traded and all expire on a specific date. Settlements can be cash settlements or cross settlements with the former being the most common. Margins are used to initiate and maintain a position, and each asset has a margin requirement based on the expectation of price movement. Futures can be highly leveraged, as a small percentage of the underlying asset's value is required to take a position. However, the risk for traders and clearinghouses increases in the event of extreme market volatility.

  • 00:15:00 Nitesh Khandelwal explains that the delivery aspect is a key characteristic of trading futures, as some contracts may be deliverable while others may not be. Commodity and stock futures may be delivered, whereas index futures cannot, as an index is simply a number and not a physical asset. Khandelwal notes that during delivery, the exchange provides a list of accepted parameters for the underlying asset to ensure quality standards. Furthermore, trading futures allows traders to take advantage of leverage, as traders can pay a margin rather than paying the full price of an asset, making it more capital-efficient. Trading futures also allows for a greater variety of trading strategies, compared to trading in the cash market.

  • 00:20:00 Nitesh Khandelwal discusses the advantages of trading in futures over cash markets, such as better liquidity for larger quantities and a more fair and transparent price discovery at different periods of time. He also talks about how futures are priced based on spot prices, date of expiry, risk-free rate of return, storage and delivery costs, and the convenience yield. This yield is the price that businesses are willing to pay to hold an asset in physical form to avoid supply and demand issues and default on delivery during expiry.

  • 00:25:00 The speaker discusses the concept of convenience yield and how it relates to investable assets, such as gold, where physical ownership is often preferred for the symbol of prestige it represents. The speaker provides a formula for calculating the expected price of equity futures or index futures, taking into account the current spot price and the return that could have been earned by investing the money elsewhere. The formula also factors in storage costs and convenience yield, which is the price an investor is willing to pay to hold the physical asset. The speaker then explains that rational investors are assumed to consider convenience yield when taking a view on the market.

  • 00:30:00 The speaker explains the concept of cash-future strategy, which seeks to make a profit by trading in cash as well as future in opposite directions. This strategy requires enough liquidity in the cash market for the stocks that one plans to hold onto, as well as access to a mechanism for delivery if the market allows shorting. The speaker also cautions against getting excited by high returns on short cash and long future, as it may not be a feasible option due to delivery mechanisms.

  • 00:35:00 Nitesh Khandelwal explains the factors that may make spreads more volatile as an expiry date approaches. Firstly, there is no return that you are making during the zero period so whatever deviation it is stays residual. Secondly, in the events of a delivery through the future, the delivery mechanism will be in play whereby the spreads can become erratic due to the availability or lack of stock in the warehouse for delivery. Additionally, the interest rates in the prevailing markets will affect the spirit. Finally, the market sentiment plays a significant role during high volatility periods or news announcements that can cause prices to move swiftly within minutes. He also explains the two different strategies for trading spreads, namely, the calendar spreads and intermarket spreads. Calendar spreads are highly effective since it is risk-free on the spot due to elimination, meaning you do not need to post much margin. Consequently, the returns are high even though they are less volatile.

  • 00:40:00 Nitesh Khandelwal discusses how to determine if there is a correlation between different asset classes, be it commodities, equities, or currencies. If there is any movement happening in one instrument, you can expect some movement in the other as well; however, whether they are directly or inversely correlated depends on your analysis. You can also have correlations within the same asset class. For example, if food prices are going up, gold prices are expected to go down, which suggests that the market sentiment is not very optimistic. If you are a fundamental player, you can take a position by looking at the fundamentals and the news. Additionally, Khandelwal introduces interexchange spreads, which can be pure arbitrage or statistical arbitrage, based on their linkage to each other, even though they are not precisely the same asset class.
Nitesh Khandelwal at POC2015 - Trade Futures and be ahead of the markets
Nitesh Khandelwal at POC2015 - Trade Futures and be ahead of the markets
  • 2015.09.22
  • www.youtube.com
Most Useful linksJoin EPAT – Executive Programme in Algorithmic Trading : https://goo.gl/3Oyf2BVisit us at: https://www.quantinsti.com/Like us on Facebook: h...
 

Webinar Topic: A sneak peek into Artificial Intelligence based HFT Trading Strategies



Webinar Topic: A sneak peek into Artificial Intelligence based HFT Trading Strategies

This video is a recording of our Webinar on "A sneak peek into Artificial Intelligence based HFT Trading Strategies" conducted by QuantInsti on 27th February, 2015.

In this video Mr. Sameer Kumar, Director and Faculty at QuantInsti, How machine learning techniques can help us design better trading strategies. He will cover alpha in trading and how we can extract it by applying the knowledge about the market structure and order flow. He will also explain how to use machine learning for predicting asset paths. Watch the video to understand the high frequency trading and using Artificial Intelligence for trading.

Sameer graduated from BITS Pilani with Masters in Economics and Information Systems. He started his career with Yahoo! where he gained expertise in technical architecture, design and development of highly scalable systems. A C++ evangelist and Perl poet with broad understanding of economics and market dynamics, he now designs and builds financial strategies with built-in intelligence. He leads the infrastructure development team along with the low latency programming division at iRageCapital Advisory Private Ltd.

At QuantInsti, he shares his experience on low latency systems as well as strategies involving artificial intelligence.

Webinar Topic: A sneak peek into Artificial Intelligence based HFT Trading Strategies
Webinar Topic: A sneak peek into Artificial Intelligence based HFT Trading Strategies
  • 2015.03.02
  • www.youtube.com
This video is a recording of our Webinar on "A sneak peek into Artificial Intelligence based HFT Trading Strategies" conducted by QuantInsti on 27th February...
 

Algorithmic Trading in Different Geographies



Algorithmic Trading in Different Geographies

In this video, Mr. Rajib Ranjan Borah, co-Founder QuantInsti & iRageCapital Advisory, compares algorithmic trading in different geographic across the globe. He shares his insights and experience of algorithmic trading across the major exchanges in Asia Pacific (APAC), Europe & Middle East (EMEA) and the Americas. The presentation has data of volumes of equity and options traded in more than 30 exchanges monthly and annually.

Algorithmic Trading in Different Geographies
Algorithmic Trading in Different Geographies
  • 2015.06.24
  • www.youtube.com
In this video, Mr. Rajib Ranjan Borah, co-Founder QuantInsti & iRageCapital Advisory, compares algorithmic trading in different geographic across the globe. ...
 

Order book dynamics in High Frequency Trading



Order book dynamics in High Frequency Trading

This Webinar on "Order book dynamics in High Frequency Trading" conducted by QuantInsti. In this video Mr. Gaurav Raizada, Director and Faculty at QuantInsti explains - How execution algorithms provide a price which is between Limit Order Execution and Market Order Execution.

An important task of high-frequency trading is to successfully capture the dynamics in the Data. Empirical Data on Indian Exchanges show that 95% of all NEW orders are placed within 5 ticks of best-bid and best-ask.

The Quantinsti® Replacement Matrix shows that most of the orders that are being replaced are among the top 3 levels and these replacements allow us to visualize and generalize about market behaviour. This matrix gives a visual representation of the cost metrics and replacement behaviour.

Order book dynamics in High Frequency Trading
Order book dynamics in High Frequency Trading
  • 2015.06.04
  • www.youtube.com
This Webinar on "Order book dynamics in High Frequency Trading" conducted by QuantInsti. In this video Mr. Gaurav Raizada, Director and Faculty at QuantInsti...
 

Financial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020) - Lecture 1: Concepts and Institutions (Financial Markets Microstructure)



Lecture 1: Concepts and Institutions (Financial Markets Microstructure)

The instructor begins by setting the stage for the financial markets microstructure course, explaining that the lectures were primarily conducted as live streams and uploaded to YouTube due to the COVID-19 pandemic. The recordings, along with the slides, problem sets, and reading list, can be accessed on the instructor's personal website. The course heavily relies on a textbook authored by Terry Foucault, Marco Pagano, and Ilse Hoyle. Viewers are advised to start at lecture 11 if they prefer to skip material that can be easily read in the textbook. The introductory video establishes the course as a study of financial markets and aims to provide a comprehensive understanding of their functioning.

The concept of markets is introduced as institutions where property rights are exchanged, and people engage in trading activities. The main objective of studying markets is to ensure the efficient allocation of property rights and that market transactions contribute to the overall increase in social welfare. Financial markets, specifically, are highlighted as a distinct type of market that facilitates the trading of financial assets such as stocks, bonds, and derivatives. The purpose of investing in these assets is either to reallocate wealth over time or to navigate various contingencies or outcomes.

The instructor explains the concept of financial assets and how they serve as a means to transfer wealth across different time periods and contingencies. An example is given, illustrating how investing in renewable energy companies can help mitigate potential job losses in the coal industry if renewable energy becomes more prevalent. The video emphasizes that financial markets involve asymmetric information, where different market participants possess varying degrees of knowledge about different prospects of the world. The instructor also discusses the institutional details specific to financial markets and highlights their purpose of facilitating profitable trades between agents with opposing desires.

The multiple values of financial markets are then explored, focusing on their role as platforms for traders to compare their private evaluations and aggregate dispersed information. Financial markets also provide a degree of security. The distinction between primary and secondary markets is explained. Primary markets enable the allocation of savings to investments, with the final user of the money ensuring that it works to fulfill financial promises. In contrast, secondary markets serve the purpose of reallocating investments among savers, allowing trades between different owners and potential holders of assets on fixed platforms like exchanges.

The video specifically emphasizes secondary markets, such as stock markets, bond markets, derivatives markets, currency or foreign exchange markets, and commodity markets that function as derivative markets. It states that understanding market efficiency and the process of price formation are crucial aspects addressed in the course. The role of traders' behavior and the environment they operate in, as well as how they act on their information in comparison to the broader market, will be examined to understand the microstructure of financial markets.

The course is presented as utilizing different methods and approaches to answer questions related to market organization, design, and policy issues. Real-world markets will be discussed, and the knowledge gained will be used to build theories for analyzing policies within the framework of these institutions. Additionally, the course will touch upon empirical issues related to applying these theories and concepts to real-life data.

Prerequisites for the course are outlined, including a basic understanding of finance, microeconomics, game theory, and mathematics. While the course primarily focuses on rational models in finance, students are encouraged to explore the complementary field of behavioral finance. The section then delves into the fundamental concept of prices and how it differs from the idealized model of market prices without arbitrage.

The video explains the concept of bid-ask spreads and how they violate the law of one price in financial markets. It clarifies that in nearly all financial markets, there exist two prices: the bid price and the ask price. The difference between these prices is known as the bid-ask spread, which typically ensures the absence of arbitrage. However, bid-ask spreads can create inefficiencies in the market, resulting in less efficient market outcomes. The investigation of market efficiency is closely linked to the study of bid-ask spreads. The lecturer draws a parallel between bid-ask spreads and the difference in prices when buying or selling foreign currency, providing real-world examples to enhance understanding.

The video further explains that actual trading prices differ from the idealized market prices, as the former are forward-looking while the latter are backward-looking, often representing the price of the last trade. The fundamental value of a stock is described as arising from the future income streams it can generate, such as dividends or price appreciation. This value is influenced by managerial decisions within the company. The course aims to analyze how this fundamental value translates into market prices and whether prices accurately reflect it. The concept of price discovery is introduced, focusing on how quickly new information about the fundamental value is incorporated into market prices.

The lecturer proceeds to discuss how prices and asset allocations are established within the microstructure of financial markets. It is emphasized that not all agents seeking to trade a particular asset are present in the market simultaneously, leading to limited supply and demand at any given time. These limitations can result in temporary imbalances, which can impact the market price in the short term. However, the price eventually returns to its long-run level once the imbalance is resolved. Analyzing these market imbalances is crucial for determining the degree to which prices reflect the fundamental value and the speed at which they incorporate relevant information.

The concept of liquidity and its relationship to market depth is then explored. Liquidity refers to the market's ability to facilitate the sale of an asset quickly without significantly impacting its price. A more liquid market is characterized by a greater number of buyers and sellers, reducing the impact of individual orders on the price. Market depth, on the other hand, measures the amount of order volume required to cause a fixed price change. The lecturer highlights the importance of understanding liquidity and market depth for traders, as they affect the prices they receive for their trades. Liquidity can also influence the fundamental value of an asset. In the subsequent lecture, the measurement of liquidity will be discussed.

The concept of market depth is further examined, referring to the potential volume of buy and sell orders beyond the best quote visible in the market. Understanding market depth enables traders to gauge the extent to which their trades can impact market movement without causing significant price fluctuations. The video provides a broad overview of two types of financial markets: order-driven markets, where orders are submitted to a common limit order book, and dealer markets, where trades are facilitated through a centralized intermediary. The lecture delves into the sub-categories of each market type, including continuous markets and call auctions.

The lecturer elaborates on the dimensions that differentiate order-driven markets from one another. One such dimension is order precedence, wherein the highest bidder receives priority in buying. In case of two buyers offering the same amount, time priority is followed, giving priority to the order that was submitted first. Another dimension is price interval, where discriminatory pricing allows different trades to occur at different prices, rather than enforcing a single market price. Furthermore, markets differ in their trading day's beginning and end, with pre-trade call auctions potentially taking place before continuous trading commences. These concepts and institutions are essential for understanding the microstructure of financial markets.

The lecturer goes on to discuss how markets have opening and closing hours, along with specific trading rules that can vary across different exchanges. Limit orders, which are submitted to a limit order book, remain there until suitable trading opportunities arise. In contrast, market orders are executed immediately at the best available price. Patient traders tend to use limit orders, while impatient traders opt for market orders, depleting the limit order book. The pricing mechanism in markets is often discriminatory and depends on the timing of trades.

Two types of financial markets are then introduced: continuous limit or book exchanges and call auctions. Continuous limit or book exchanges, such as the New York Stock Exchange and the London Stock Exchange, are popular market structures where trading occurs through a limit order book. On the other hand, call auctions involve trades happening at specific intervals, and the price for the trade is determined to maximize the number of executed orders. However, call auctions have their drawbacks, including slower trading times and the absence of impatient traders, which can have long-lasting effects. Examples of call auction exchanges include Nasdaq, LSE, and Euronext, which may operate in parallel with continuous trading for certain assets.

Moving on, the lecturer explains the distinction between order-driven markets and dealer markets. In dealer markets, a central intermediary known as a market maker or dealer buys assets from sellers and sells them to buyers, setting prices that balance supply and demand. Dealers profit by quoting positive bid-ask spreads, but they must also compete with other dealers by narrowing their bid-ask spreads enough to attract business while generating sufficient trading profits. Exchanges are the most regulated and formalized markets, while alternative trading systems and multilateral trading facilities are less regulated and formal.

The video then delves into the comparison between exchanges and over-the-counter (OTC) trading, which represent two distinct types of financial markets. Exchanges provide a range of services, including security, clearing and settlement services, liquidity, stability, and transparency. On the other hand, OTC trading refers to transactions that are not conducted through exchanges but are still highly formalized platforms. However, OTC platforms may require less financial disclosure compared to major exchanges. While reduced transparency is a trade-off, it comes with corresponding benefits. Additionally, the existence of dark pools of liquidity is mentioned. These are internal platforms that enable large investment banks to match orders from their clients internally. The video notes that markets can differ in various dimensions and examines the factors that should be considered when comparing these markets.

The speaker addresses different perspectives on financial market microstructure. From a regulator's standpoint, ensuring competition on all sides of the market is crucial for achieving efficient allocation. Traders, on the other hand, prioritize liquidity, transparency, and information availability in the market to determine the optimal value for their assets. The speaker identifies three groups of agents in the market: retail investors, institutional investors, and for-profit traders. Retail investors are often amateurs, while institutional investors are professionals who are compensated for devising optimal trading strategies.

The video proceeds by discussing the various types of investors in financial markets, distinguishing between informed and uninformed traders. Informed traders possess private information that is not accessible to the rest of the market, while uninformed traders have the same information about asset value as the overall market. Brokers are introduced as intermediaries who facilitate orders between traders and investors. The video briefly touches upon conflicts of interest between traders and brokers and explores the role of regulation in achieving efficient market outcomes.

The lecturer moves on to explore the different goals of financial markets, which include protecting uninformed traders against insider trading, ensuring price discovery and efficiency, and stabilizing the market during sudden shocks. Selecting the optimal trading structure for various types of assets is also essential. Methods to achieve these goals encompass requiring interaction between fragmented markets, imposing transaction taxes or subsidies, mandating collateral, regulating algorithmic and high-frequency trading, and overseeing competition between exchanges. The trade-off between liquidity and natural monopoly must be considered, as excessive fragmentation can hinder the goals of the markets.

The video delves into how platforms can enhance the terms of trade for traders and the potential benefits of increased competition among exchanges. Statistics on different stock exchanges worldwide are presented, highlighting the concentration of exchanges in the US compared to Asia and Europe. The lecturer poses open questions for regulators regarding market structure and emphasizes the significance of comprehensive analysis in addressing the trade-offs associated with market design.

In conclusion, this lecture provides an overview of financial market microstructure, discussing concepts such as bid-ask spreads, price discovery, liquidity, market depth, and different market types. The speaker also touches upon the roles of various market participants, the distinction between exchanges and OTC trading, and the goals of financial markets. Understanding these concepts and structures is crucial for comprehending the dynamics of financial markets and designing efficient market systems that balance the interests of traders, investors, and regulators.

  • 00:00:00 In this section of the video, the instructor sets the stage for the financial markets microstructure course. The lectures were mostly held as live streams and were uploaded to YouTube due to COVID-19. The instructor warns that the recordings are one-pass with no editing and can be found on his personal website, along with the slides, problem sets, and reading list. The course is largely based on a textbook by Terry Foucault, Marco Pagano, and Ilse Hoyle, and the instructor invites viewers to start at lecture 11 if they prefer to skip anything that is easily read in the textbook. Finally, the video introduces the course as a study of financial markets, beginning with a step back to understand what financial markets are all about.

  • 00:05:00 In this section, the concept of markets is introduced as an institution where property rights are exchanged and people can trade. The main goal of studying markets is to ensure that property rights are allocated efficiently, and that market transactions lead to an increase in social welfare. Financial markets are a particular type of market for trading financial assets, such as stocks, bonds, and derivatives. The purpose of buying these assets is to either reallocate wealth over time or across different contingencies of the world.

  • 00:10:00 In this section, the lecturer explains the concept of financial assets and how they are used to move wealth across time and different outcomes or contingencies. He gives the example of investing in renewable energy companies to balance out the potential job loss of working in the coal industry if renewable energy becomes more prevalent. Financial markets involve asymmetric information where different agents have different knowledge about various prospects of the world. The lecturer also discusses the institutional details specific to financial markets and the purpose of financial markets which is to match agents who have opposing desires for profitable trade.

  • 00:15:00 In this section, the lecturer explains the multiple values of financial markets as platforms for traders to compare their private evaluations and aggregate dispersed information, and for ensuring a degree of security. There are two types of financial markets; primary and secondary. Primary markets allow savings to be allocated to investment, and the final user of the money attracted through trade ensures that the money will work to repay financial promises. On the other hand, secondary markets serve to reallocate investments over savers, and the trade occurs between different owners and potential holders of assets, happening on fixed platforms like exchanges.

  • 00:20:00 In this section, the focus is on secondary markets, such as stock markets, bond markets, derivatives markets, currency or foreign exchange markets, and commodity markets that function as derivative markets. The market's efficiency and how market prices are created are key questions addressed in this course. The role of traders' behavior and the environment they operate in, in addition to how they act on their information compared to the market, will be examined to understand the microstructure of the markets.

  • 00:25:00 can use market liquidity as a core concept to measure how well financial markets work, along with related concepts such as market depth, trading volume, efficiency, and stability. The course will use different methods and approaches to answer questions about market organization, design, and policy issues. Additionally, the course will discuss real-world markets and use that knowledge to build theories to analyze policies within the frame of these institutions. There will also be some discussion of empirical issues related to applying these theories and concepts to real-life data.

  • 00:30:00 one market, its price can differ across markets, which can create opportunities for arbitrage. This section of the lecture establishes some prerequisites for the course, including basic knowledge of finance, microeconomics, game theory, and math. The course focuses on rational models in finance, but students may also explore the complementary field of behavioral finance. The section then goes on to discuss the fundamental concept of prices and how it differs from the idealized model of market prices without arbitrage.

  • 00:35:00 In this section, the concept of bid-ask spreads and the violation of the law of one price in financial markets are discussed. It is explained that in almost every financial market, there are two prices, a bid price, and an ask price. The difference between the two prices is known as the bid-ask spread, and it typically holds the no arbitrage condition. The bid-ask spread can create inefficiencies in the market, making the market outcomes less efficient. The investigation of market efficiency is intimately connected to the investigation of the bid-ask spread. It is further explained that the bid-ask spread is similar to the difference in prices when buying or selling foreign currency. Real-world examples are given to understand the concept better.

  • 00:40:00 In this section, the video explains that the actual prices at which you can trade differ and are forward-looking, while the outline price is a backward-looking price, which is often the price of the last trade. The fundamental value of a stock is said to come from the stream of future incomes that it can give, such as dividends or price appreciations. This value is determined by various managerial decisions within the company, and in this course, they will not discuss how to determine this value but will analyze how this fundamental value translates to market prices and if prices accurately reflect this fundamental value. One of the concepts that they will look at is price discovery, which deals with how quickly new information about the fundamental value gets incorporated into the market prices.

  • 00:45:00 In this section, the lecturer discusses how the prices and allocations of assets are established in financial markets microstructure. He emphasizes the importance of understanding that not all agents who want to trade in a given asset are present in the market at the same time, resulting in a limited capacity of both supply and demand in the market at any given time. These limitations can lead to temporary imbalances, which can affect the market price in the short term, but the price does return to its long-run level once the imbalance is resolved. The lecturer highlights that analyzing these market imbalances is crucial in determining how well the price reflects the fundamental value and how quickly it incorporates any information about the fundamental value.

  • 00:50:00 In this section of the lecture, the professor discusses the concept of liquidity and its relationship to market depth. Liquidity is defined as the market's ability to quickly facilitate the sale of an asset without significantly reducing its price. The more buyers and sellers there are in the market, the more liquid it is and the less impact any one order will have on the price. Market depth, on the other hand, measures how big of an order is required to change the price of an asset by a fixed amount. The professor explains that understanding liquidity and market depth is crucial for traders because it affects the price they receive for their trades. Additionally, he notes that liquidity can impact the fundamental value of an asset, and in the next lecture, he will discuss how to measure liquidity.

  • 00:55:00 In this section, the concept of market depth and its relation to market liquidity is discussed. Market depth refers to the potential volume of buy and sell orders beyond the best quote visible in the market. Understanding market depth is important because it allows traders to know how much market movement they can create without causing major price fluctuations. Additionally, this section provides a broad overview of two types of financial markets: order driven markets, where orders are submitted into a common limit order book, and dealer markets, where trades happen through a centralized intermediary. The discussion goes into further detail of the sub-categories of each type of market, including continuous markets and call auctions.

  • 01:00:00 In this section, the lecturer discusses several dimensions in which order-driven markets can differ from one another. One dimension is order precedence, where the highest bidder gets to buy first. In case of two buyers offering the same amount, time priority is followed, meaning the one who submitted the order first gets the first execution. Another dimension is price interval, where discriminatory pricing allows different trades to occur at different prices instead of forcing everyone to trade at a single market price. Finally, markets differ in what happens at the beginning and end of the trading day, where a pre-trade call auction might take place before continuous trading begins. Overall, the lecture introduces several concepts and institutions that are crucial to understanding financial market microstructure.

  • 01:05:00 In this section, the lecturer discusses how markets have opening and closing hours with special trading rules that can differ across different exchanges. Limit orders are submitted to a limit order book and stay there until a suitable trading opportunity arises, while market orders are executed immediately at the best available price. Patient traders use limit orders, while impatient traders use market orders and deplete the limit order book. The pricing in markets is usually discriminatory and depends on when one decides to trade.

  • 01:10:00 In this section, the lecturer discusses two different types of markets: continuous limit or book exchanges and call auctions. Continuous limit or book exchanges, such as the New York Stock Exchange and the London Stock Exchange, are popular ways of organizing markets where trading is done through a limit order book. Meanwhile, call auctions are auctions in which trade happens at a given frequency and the price of the trade is chosen to maximize the number of executed orders. However, these auctions have some drawbacks, such as slower trading times and the absence of impatient traders, which can have long-running repercussions. Some examples of call auction exchanges are Nasdaq, LSE, and Euronext, which may operate in parallel with continuous trading for some assets.

  • 01:15:00 In this section, the lecturer explains the difference between order driven markets and dealer markets. In dealer markets, a central intermediary called a market maker or dealer buys assets from those who want to sell them and sells assets to those who want to buy them, setting prices that equalize demand and supply. Dealers profit by quoting positive bid-ask spreads, but they must still compete with other dealers by making their bid-ask spreads narrow enough to attract business while still generating enough trading profits for subsistence. Exchanges are the most regulated and formal markets, while alternative trading systems and multilateral trading facilities are less regulated and less formal.

  • 01:20:00 In this section, the video discusses exchanges and over-the-counter (OTC) trading as two different kinds of financial markets. Exchanges offer a range of services including security, clearing and settlement services, liquidity, stability, and transparency. On the other hand, OTC trading generally refers to transactions that are not traded through exchanges, but are still very formalized platforms. However, OTC platforms may require less financial disclosure than big exchanges, but the perk of less transparency comes with corresponding benefits. Additionally, there are dark pools of liquidity, which are internal platforms that allow large investment banks to match orders of their clients against each other. The video notes that markets can differ in various dimensions and examines what factors should be considered when comparing these markets.

  • 01:25:00 In this section, the speaker discusses different perspectives on financial market microstructure. From a regulator's standpoint, there should be competition on all sides of the market to result in efficient allocation. From a trader's perspective, liquidity, transparency, and information available in the market are essential in determining the best value for their assets. The speaker identifies three groups of agents in the market: retail investors, institutional investors, and for-profit traders. Retail investors are amateurs, whereas institutional investors are professionals who get paid for devising optimal trading strategies.

  • 01:30:00 In this section, the video discusses the different types of investors in financial markets, including informed and uninformed traders. Informed traders possess private information that is not available to the rest of the market, while uninformed traders have the same information about asset value as the market. Additionally, brokers are introduced as intermediaries between traders and investors, who facilitates orders to the market. The video also briefly touches upon conflicts of interest between traders and brokers and discusses the role of regulation in achieving efficient market outcomes.

  • 01:35:00 In this section, the lecturer discusses the various goals of a financial market, including protecting uninformed traders against insider traders while ensuring price discovery and efficiency, stabilizing the market during sudden shocks, and selecting the optimal trading structure for different types of assets. Methods to achieve these goals include requiring interaction between fragmented markets, imposing transaction taxes or subsidies, requiring collateral, regulating algorithmic and high-frequency trading, and regulating competition between exchanges. The trade-off between liquidity and natural monopoly must also be considered, as fragmentation can go against the goals of the markets.

  • 01:40:00 In this section, the video discusses how platforms can improve the terms of trade offered to traders and the potential benefits of having more competition among exchanges. The video also presents statistics on different stock exchanges around the world, highlighting the concentration of exchanges in the US compared to Asia and Europe. The lecturer poses open questions for regulators on market structure and recommends exercises for students to get more immersed in the world of finance. These include finding share prices, identifying the stock exchanges they come from, reading an article on the London Metal Exchange, and solving exercises from chapter 1 of the textbook.
Lecture 1: Concepts and Institutions (Financial Markets Microstructure)
Lecture 1: Concepts and Institutions (Financial Markets Microstructure)
  • 2020.07.27
  • www.youtube.com
Lecture 1: Concepts and InstitutionsFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.you...
 

Stochastic Market Microstructure Models of Limit Order Books


Stochastic Market Microstructure Models of Limit Order Books

During the lecture, the speaker explains the process of executing a large trade order through an algorithm designed to achieve optimal execution quality. When a trader submits an order, it is sent to a trading engine that breaks it down into smaller blocks. These smaller order chunks are then sent into the market, which includes various venues such as exchanges and dark pools. To execute the order successfully, traders must decide where to route it. The market participants involved in the trading process include institutional investors, market makers, retail flow, and opportunistic or active liquidity providers.

Executing large orders can be challenging due to the limited liquidity available in the market. To mitigate the impact on prices and minimize information leakage, large orders are chopped into smaller chunks and executed over time. Market makers, on the other hand, have a different role as intermediaries, providing liquidity and avoiding adverse selection.

To trade a large position effectively, traders need to make market variable forecasts, such as bid-ask spread, volatility, market depth, and available liquidity. They also solve an optimization problem that guides them in sequencing their trades. The execution of small order chunks is carried out by a micro trader, who aims to minimize the impact on the market during each five-minute slice.

The lecturer further discusses the behavior of volumes, volatility, spreads, and liquidity in the S&P 500 security universe throughout the trading day. They observe that volumes exhibit a small spike at the beginning of the day due to news and then flatten out until increased activity occurs towards the end of the day. Volatility, on the other hand, tends to be high at the start of the day due to overnight news but gradually decreases as the day progresses. Spreads, which represent the difference between bid and ask prices, are wider in the morning due to uncertainty but narrow as the day unfolds. Liquidity follows a similar pattern, increasing towards the end of the day and decreasing at the beginning due to concerns about large position exposure.

The lecture also delves into the concept of a limit order book, which represents the queue of orders at different price levels. Each price level in the order book operates on a first-come, first-served basis, allowing orders to trade against arriving market orders. The lecturer explains that the structure of a limit order book creates a queuing control problem, and they highlight some of the challenges that arise in this context.

The speaker emphasizes the significance of stochastic modeling and multi-class queues for understanding high-dimensional systems with strategic interactions between market participants. Visual representations of the limit order books in the S&P 500 are showcased to illustrate the difference between trading rates and the number of limit orders placed and cancelled at the top prices.

The lecture continues by discussing inter-arrival times between events in a limit order book, focusing on trade frequency and cancellations. The speaker notes that these events do not occur randomly but exhibit predictable behavior, such as spikes every half a second for certain algorithms. Confidence intervals are used to check the stationary parameters of the system, indicating that the parameters typically change within five to ten minutes.

Waiting times in the limit order book are typically in the range of 1 to 100 seconds, suggesting that modeling should consider short horizons due to the difficulty of predicting parameter changes in the book. The tick period is also mentioned as being comparable to the queuing delay, highlighting the importance of modeling cancellations, which occur at a higher rate than trading. The lecturer suggests incorporating trading strategies and mathematical devices to capture jumps or bursts of events in the limit order book.

The lecture further explores the behavior of trades in limit order books, particularly when large orders are executed, resulting in instantaneous and simultaneous trades. The speaker emphasizes the importance of decomposing orders and rolling up trades to understand the different types of trades and their dependence on the state of the book. The modeling of cancellations, including exponential or state-dependent approaches, is also discussed, highlighting the tradeoffs between tractability and realism.

The lecture dives into the heterogeneous behavior of market participants in the queuing context. Some market participants constantly monitor the market and exit quickly when something seems unsettling, while others rely on alarms to send orders. The speaker suggests modeling this heterogeneity to estimate the length of time it takes for an order to be executed. This control problem and its queuing implications are deemed essential in algorithmic execution systems.

Estimating the waiting time for an order to get a trade is a crucial aspect of order placement. The speaker presents two methods: a simple calculation that disregards cancellation rates and a more sophisticated method that models cancellation rates. The latter approach involves solving a logarithmic formula that estimates how long it takes for the queue length to be depleted. The two methods are tested on an actual dataset of orders placed by an algorithmic trading system.

The lecture also addresses the biases in stochastic market microstructure models, pointing out that certain assumptions can lead to incorrect estimates. The use of exponential alarm clock models, for example, can be overly optimistic as they assume everyone cancels ahead of the trader. Disregarding cancellations altogether is also problematic, as different cancellation methods exist in the market. The speaker suggests modeling cancellations as a stopping time to account for the impact of market makers and other traders on cancel rates.

To achieve more accurate results, the speaker presents a model that estimates the number of market participants with alarm clocks and those who cancel orders when the queue length becomes small. By incorporating heterogeneity in the behavior of orders within the queue, more precise estimations can be obtained. The lecture highlights the importance of modeling heterogeneity in trading systems, distinguishing it as a novel aspect compared to queuing models studied in other settings. Characterizing queuing behavior is deemed important and plays a crucial role in algorithmic trading systems. The next section of the lecture will focus on routing and diffusion approximations.

The speaker explores the application of heavy traffic approximations in modeling the dynamics of limit order books at high frequencies. This approach allows for analytical approximations that are more manageable compared to discrete models. By treating the limit order book as a queuing system, it becomes possible to estimate waiting time distributions and rates while maintaining analytical tractability. The speaker emphasizes the wide range of time scales involved in the problem, from ultra-high frequencies to daily time scales, and highlights the importance of developing models that can be applied to different applications, such as optimal trade execution.

Building upon the previous discussion, the speaker describes how familiar techniques from heavy traffic limits of queues can be used to derive effective quantities on larger time scales. The focus is placed on the best queues, which have the highest bids and the lowest asks, to understand price dynamics. The rest of the order book is treated as a stationary reservoir of liquidity. Whenever the liquidity in the best queue is depleted, a new value is drawn from the distribution of the size of the next best queue. The bid-ask spread is assumed to be tight and equal to one tick for liquid stocks. The dynamics of the price are entirely determined by the interaction between the two best queues and the hitting times.

In the lecture, the speaker discusses a queuing model that incorporates the arrival and cancellation of orders while also considering price changes. The model assumes a diffusion scaling limit and features a covariance matrix that incorporates the variance of order sizes per unit time and the correlation between the order flow at the bid and ask. The queues exhibit diffusive behavior as long as they are not depleted. However, when a queue gets depleted, the price either increases or decreases. The price dynamics are modeled as a discrete process that jumps by one unit at the hitting time of the ask or bid queue. This model is particularly useful for analyzing high-frequency trading and exhibits interesting properties, such as diffusive dynamics interrupted by discontinuous reflections.

The lecture highlights that the diffusion limit allows for the computation of practically anything, even when starting from a complex discrete model. The duration between price changes can be characterized by a closed-form distribution, enabling precise price forecasting based on the orders in the queue. Additionally, a second diffusion limit is discussed, which explains that while the price undergoes discrete jumps at hitting times, it exhibits diffusive dynamics at longer time scales, such as daily or hourly. The lecture concludes by presenting a formula that expresses volatility in terms of features extracted from order flow. This formula can be tested against the empirical standard deviation of stocks in the S&P 500, showing good agreement.

The lecture acknowledges that there are numerous extensions and more sophisticated models beyond the basic two-queue model. These extensions include state-dependent arrival rates, explicit modeling of the next best queues, and more complex approaches such as modeling the entire order book or utilizing stochastic partial differential equations to model the order book as a density. While these models may be intricate, they can yield explicit formulas for various quantities of interest and provide analytical insights into the relationship between liquidity and price behavior in financial markets.

  • 00:00:00 The order is sent to a trading engine that chops the order down into smaller blocks and sends them into the market. The market includes various venues, such as exchanges and dark pools, and, to execute an order, traders need to choose where to route it. Market participants include institutional investors, market makers, retail flow, and opportunistic or active liquidity providers. In the next section, the speaker will highlight some queuing control problems that arise in limit order books.

  • 00:05:00 The speaker explains the process of executing a large trade order through an algorithm that focuses on providing the best execution quality. The trader needs to access the market with minimal information leakage and avoid drastic price fluctuations. Large orders need to be chopped into small chunks to be executed over time, as liquidity is scarce in the market. Market makers, on the other hand, have different considerations such as providing liquidity as intermediaries and avoiding adverse election. To trade a large position, traders need to build market variable forecasts, including bid-ask spread, volatility, market depth, and available liquidity, and solve an optimization problem that guides them on how to sequence their trades. The micro trader takes the five-minute slices and executes the small order chunks without significantly impacting the market.

  • 00:10:00 The lecturer discusses the behavior of volumes, volatility, spreads, and liquidity in the S&P 500 security universe for any given minute of the trading day. They show how volumes have a small spike at the beginning of the day due to news and flatten out until there is a lot of activity towards the end of the day. Additionally, the lecturer talks about how volatility is high at the beginning of the day due to news that has been released overnight but then declines throughout the day. Furthermore, they show that spreads are wider at the beginning of the day due to uncertainty but narrow as the day progresses, and liquidity increases towards the end of the day and decreases at the beginning of the day due to the fear of posting big positions. Finally, the lecturer discusses the limit order book and how each price level of the order book is a queue where orders wait in a first-come, first-serve manner with cancellation and orders can trade against arriving market orders.

  • 00:15:00 The speaker discusses the structure and mechanics of a limit order book market, which includes queues of bids and asks for a particular security at various price levels. The speaker introduces arrival rates of limit orders, cancellation rates, and market order rates into these queues, which create a multi-class queuing system that is FIFO within each price level and prioritized by price. The speaker also mentions that placing a limit order in such a market requires estimating how long it will take for the order to execute and understanding which orders will be executed first based on time and price priority. Additionally, the speaker notes the coupling between limit order book markets across different exchanges.

  • 00:20:00 The speaker discusses the control problem for institutional investors when executing trades in the stock market. The horizon of the control problem is similar to the queuing time, which means they only need to think of small multiples of the queuing time when making decisions about placing orders in the market. The cueing delays are of great significance to both institutional investors and market makers because if market makers place an order and have to wait for a minute or longer for it to fill, they could be exposed to news or events that may adversely affect pricing. The speaker emphasizes the importance of stochastic modeling and multi-class queues in modeling high-dimensional systems with strategic interactions between market participants. They also showcase visual representations of limit order books in the S&P 500, which show that trade rates are much lower than the number of limit orders placed and cancelled at the top prices.

  • 00:25:00 The speaker discusses inter-arrival times between events in a limit order book using the example of trade frequency and cancellations. The top plot shows trade frequency and the time it takes for the next trade to occur, with a hump around 40 microseconds, which was the technological performance limit in 2017. The bottom plot shows cancellations and the time it takes for the next cancellation, with two statistical signatures at 40 microseconds and 20 microseconds. The speaker also notes that orders do not happen randomly, with predictable behavior, such as spikes every half a second for certain algorithms. The system's stationary parameters are checked using a confidence interval, with the result that most of the time the parameter has changed within five to ten minutes.

  • 00:30:00 The speaker discusses the typical waiting times for trade in the limit order book, which are usually of the order of 1 to 100 seconds, meaning that the parameters stay constant for a time scale that is of the order of the queuing delay. Therefore, it is recommended to build a model with a horizon that is short, as it's hard to predict how the parameters in the book change. It is also mentioned that the tick period is almost the same as the queuing delay, and we need to consider modeling cancellations, as they are occurring at a higher rate than trading rates. Finally, the speaker suggests modeling people's trading strategies (such as HFT and algorithmic orders) and using a mathematical device to capture jumps or bursts of events in the limit order book.

  • 00:35:00 The speaker discusses the behavior of trades in limit order books by breaking them down into smaller quantities such as round lots or odd lots, which are orders less than a hundred shares. When large orders are sent to sell, they can consume multiple orders, resulting in instantaneous, simultaneous trades. The speaker mentions the importance of rolling up these trades and decomposing orders to understand the different types and how they depend on the state of the book. They also explore different models for cancellations, including those that are exponential or state-dependent, and the tradeoffs between tractability and realism. Finally, the speaker suggests modeling events that drive cancellations in addition to other factors.

  • 00:40:00 The speaker discusses the importance of differentiating between market participants who constantly monitor the state of the market and quickly exit when something appears unsettling versus those who use alarms to send orders to the market. He proposes the idea of modeling the heterogeneous behavior of market participants in terms of queuing and provides an example of a simple problem: estimating the length of time it takes for an order to be executed in the market. The speaker suggests that this control problem and its queuing implications are crucial building blocks in algorithmic execution systems.

  • 00:45:00 The speaker discusses the process of placing an order and estimating how long it will take for the order to get a trade, taking into account adverse selection and opportunity cost. To estimate the waiting time, the speaker presents two methods: a simple calculation that disregards cancellation rates and a more sophisticated method that models cancellation rates. The latter approach involves solving the logarithmic formula for how long it takes until the queue length gets wiped out. The speaker then tests the two methods on an actual dataset of orders placed by an algorithmic trading system.

  • 00:50:00 The speaker discusses the biases in stochastic market microstructure models and how certain assumptions can lead to incorrect estimates. The use of exponential alarm clock models can lead to optimistic estimations because it assumes everyone cancels ahead of the trader. Disregarding cancellations altogether is also not ideal, as there are different cancellation methods in the market. The speaker suggests that cancellations should be modeled as a stopping time. The cancelling behavior of some market makers and other traders can significantly impact cancel rates, and this phenomenon should be accounted for in models.

  • 00:55:00 The speaker presents a simple model to estimate the number of market participants that have alarm clocks and those that cancel orders when the queue length becomes small. They put their order at the back of the queue, and some people never cancel, while others cancel exponentially. By modeling the heterogeneous behavior of orders inside the queue, they get more accurate results. They emphasize the importance of modeling heterogeneity in trading systems, which is a novel aspect compared to queuing models studied in other settings. The speaker highlights that characterizing queuing behavior is important, and it is crucial in algorithmic trading systems. In the next section, they plan to discuss routing and diffusion approximations.

  • 01:00:00 The speaker discusses the use of heavy traffic approximations in modeling the high frequency dynamics of limit order books. This approach leads to analytical approximations that are easier to manage than discrete models. By modeling the limit order book as a queuing system, the model can accurately estimate distributions of waiting times and rates while maintaining some analytical tractability. The speaker notes that there is a wide range of time scales at play in this problem, from ultra high frequencies of microseconds to daily time scales. By separating the different time scales, it is possible to develop models that can be applied to different applications, such as optimal trade execution.

  • 01:05:00 The speaker explains how to use familiar techniques from heavy traffic limits of queues to derive effective quantities on larger time scales. He suggests focusing on the best queues, which have the highest bids and the lowest asks, in order to understand price dynamics, and only modeling these two queues in a reduced form model. The rest of the order book is then treated as a stationary reservoir of liquidity, and whenever the liquidity and the best queue gets depleted, a new value is drawn from the distribution of the size of the next best queue. The speaker also assumes that the bid ask spread is very tight and equal to one tick for liquid stocks, and notes that the dynamics of the price is determined entirely by the interaction between the two best queues and the hitting times.

  • 01:10:00 The speakers discuss modelling the dynamics of two queues in a market point process, where orders arrive, get cancelled, and get executed. The event times and their sizes, which can be positive or negative, are modelled as a point process. The literature has studied these models, where each level is modelled using the MM1Q model. Still, the data shows that the exponential interarrival times and the unit sizes for orders do not seem to hold, making them unreliable assumptions. Thus, research has turned to look at the heavy traffic limit of a queuing system, which tells us that the models' details do not matter much in certain scaling regimes. Thus, a diffusion limit is explored to model the intraday dynamics of order flows, showing that a two-dimensional motion in the queue size is an appropriate model.

  • 01:15:00 The speaker describes a queuing model that takes into account the arrival and cancellation of orders while also incorporating price changes. The model assumes a diffusion scaling limit and has a covariance matrix that includes the variance of the order sizes per unit time and the correlation between the order flow at the bid and the ask. The queues have a diffusive limit as long as they are not touching zero, but when a queue gets depleted, the price either increases or decreases. The price is a discrete process that increases or decreases by one unit at the hitting time of the ask or bit queue, respectively. This model is useful for analyzing high-frequency trading and has interesting properties such as diffusive dynamics interjected by discontinuous reflections.

  • 01:20:00 The speaker discusses the diffusion limit for computing practically anything, even if starting from a messy discrete model. The diffusion limit of this model shows that the duration between price changes has a closed form distribution and that forecasting prices based on the orders in the queue is possible and precise. Additionally, the speaker explains the second diffusion limit, which explains that the discrete price jumps at hitting times but has a diffusive dynamic at the daily level or hours. Lastly, the formula provided for volatility expressed in features extracted from order flow can be tested against the empirical standard deviation of stocks in the S&P 500, providing good agreement.

  • 01:25:00 The speaker discusses different extensions to the basic model with two queues, including making the arrival rates state-dependent and explicitly modeling the next best queues. The speaker also mentions more sophisticated models, such as modeling the entire order book or using stochastic partial differential equations to model the order book as a density. These models, although complex, can lead to explicit formulas for various quantities of interest and provide analytical insights into the link between liquidity and price behavior in financial markets.
Stochastic Market Microstructure Models of Limit Order Books
Stochastic Market Microstructure Models of Limit Order Books
  • 2020.12.08
  • www.youtube.com
Authors: Costis Maglaras, Columbia University; Rama Cont, University of Oxford Many financial markets are operated as electronic limit order books (LOB). Ove...
 

Lecture 2: Measuring Liquidity (Financial Markets Microstructure)


Lecture 2: Measuring Liquidity (Financial Markets Microstructure)

In the lecture, the concept of liquidity is introduced and defined as the market's ability to facilitate the quick sale of an asset without significantly reducing its price. Liquidity is seen as a characteristic of the market that determines how easy it is to trade in a particular market, and it can vary depending on the type of asset or the specific market being examined. The lecturer also mentions two other types of liquidity: monetary liquidity and funding liquidity, which are interconnected with the broader concept of liquidity.

The lecturer explains the importance of liquidity in relation to market efficiency. Liquidity affects the efficient allocation of assets in the economy. When a market is illiquid, it results in an inefficient allocation where extra costs are incurred to buy or sell items without significantly impacting their prices. This inefficiency limits access to assets for willing buyers and hampers market efficiency. Regulators are concerned about market efficiency and stability, and liquidity serves as a measure to assess market efficiency and identify inefficiencies. Therefore, reducing market illiquidity becomes a crucial goal for regulators.

The concept of liquidity is further explored, distinguishing between the fair price and the efficient price in a perfectly liquid market. Illiquidity can indicate structural issues in the market that may require regulatory intervention to address the inefficiencies. Market depth, which measures the amount that must be traded to move a price by a certain amount, is discussed as an important indicator of liquidity. The lecturer notes that liquidity is not constant over time and often decreases during times of adversity. The ideal scenario would be for markets to be more efficient during crisis times when assets need to be quickly exchanged.

Different measures to quantify liquidity in financial markets are introduced. These measures include spread measures, price measures, and non-trading measures. The lecturer demonstrates their application using a dataset from Krispy Kreme stock. The importance of accurately estimating liquidity is emphasized, and the lecturer explains that prices can sometimes fall within the spread. This occurrence can be attributed to hidden limit orders and individual price improvements offered by dealers.

The lecture delves into specific measures of liquidity, such as the quoted spread, normalized quoted spread, effective spread, normalized effective half spread, and realized spread. The quoted spread, which is the difference between the ask and bid prices, can be misleading, leading to the use of the normalized quoted spread, which considers the average price of the asset in the market. The effective spread, which takes into account the actual execution prices of transactions, is regarded as a better measure of liquidity. It captures the price improvements that occur in the market during transactions, providing a more reliable indicator. The normalized effective half spread and the effective spread offer a more consistent representation of spread behavior in the market. The realized spread measures the cost of taking a given position in an asset, incorporating the mid-quote with a delay to allow prices to adjust to new information.

The relationship between liquidity, transaction prices, and the mid-quote is discussed. The lecture explains how transaction prices and the mid-quote interact when an investor places a buy order at the ask price. The subsequent trade can cause the transaction price to remain the same or increase, depending on the next order being a sell order or another buy order. The lecture highlights the negative covariance between changes in the direction of trade, indicating that trade directions are mean-reverting.

Other measures of liquidity, such as price impact coefficient, bid-ask bounce, and volume-weighted average price (VWAP), are introduced. These measures provide insights into market liquidity and microstructures. The lecture emphasizes the need for careful application of these measures based on the level of data aggregation.

The lecture concludes by summarizing the different measures of liquidity and their variations based on data requirements and specific goals. It emphasizes that liquidity is not a static concept and can change throughout the trading day, with significant impacts during major events. The lecturer provides exercises for viewers to practice, including recreating figures and examining the implementation shortfall in the textbook. A link to an article comparing corporate bond markets to equity markets is shared to illustrate the differences in liquidity due to market structures. The next lecture is announced to focus on analyzing the determinants of the spread and illiquidity in the market.

  • 00:00:00 The lecturer introduces the concept of liquidity, which he defines as a market's ability to facilitate an asset being sold quickly without having to reduce its price much, if at all. He notes that liquidity is a feature of the market and tells us how easy it is to trade in a given market, and may differ depending on the type of asset traded or the specific market being examined. He also touches upon two other types of liquidity, namely monetary liquidity and funding liquidity, and explains how they relate to the broader concept of liquidity.

  • 00:05:00 The concept of liquidity is explained through examples of banks and individuals. Liquidity refers to the ease of converting an asset into cash. There are three different types of liquidity, including funding liquidity, market liquidity, and asset liquidity, which are all interconnected. Researchers are interested in liquidity because it affects market efficiency, as demonstrated by a simple graph of supply and demand curves. The equilibrium price, where supply meets demand, reflects market efficiency.

  • 00:10:00 The video discusses the relationship between market efficiency and liquidity. Liquidity is necessary for markets to achieve efficient allocation of assets in the economy. An inefficient allocation occurs when the market is illiquid, meaning that extra costs need to be paid for the items to be sold or purchased without dropping their price significantly. Such inefficiency restricts access to assets for willing buyers, which hinders market efficiency. Additionally, the video explains how regulators care about market efficiency and stability, and liquidity helps measure market efficiency and indicate how much inefficient allocation exists. Therefore, reducing illiquidity in the market is a crucial goal for regulators.

  • 00:15:00 The concept of liquidity is discussed, including the difference between the fair price and the efficient price established in a perfectly liquid market. Illiquidity may indicate structural issues in the market, and regulators may need to take action to reduce inefficiencies. Market depth measures how much must be traded to move a price by a certain amount, and liquidity is not constant over time, often decreasing in times of adversity. The ideal scenario would be for markets to be more efficient during crisis times when assets need to change hands quickly.

  • 00:20:00 The instructor introduces several measures for how to measure liquidity in financial markets. These measures include spread measures, price measures, and non-trading measures, all of which are applied to a data set from Krispy Kreme stock. The instructor also explains that prices can sometimes fall inside the spread and discusses two potential reasons for this occurrence: hidden limit orders and individual price improvements offered by dealers. Overall, the lecture provides an overview of the different measures used to quantify liquidity and how they can be applied in real-world scenarios.

  • 00:25:00 The lecturer discusses different measures used to analyze market liquidity. The first measure is the quoted spread, which is the difference between the ask and bid prices. However, because the quoted spread can be misleading, it is better to use the normalized quoted spread, which takes into account the average price of the asset in the market. Additionally, the effective spread is a better measure of liquidity because it takes into account the price at which actual transactions are executed. This is a more reliable indicator of market liquidity as it accounts for price improvements.

  • 00:30:00 The speaker explains different measures of liquidity, including the effective spread, the normalized effective half spread, and the realized spread. The effective spread compares the actual price with the mid-quote just before the transaction and captures all the price improvements that occurred in the market. It also takes into account the trade sizes, which can be good or bad depending on the goals and purposes of the measurement. The normalized effective half spread and the effective spread both tell a more uniform story of spread behavior in the market. Finally, the realized spread measures the cost of taking a given position in an asset and uses the mid-quote with a delay to allow prices to adjust to new information.

  • 00:35:00 The speaker explains the differences between using the realized spread, the quoted spread, and the effective spread to measure liquidity in financial markets. While the quoted spread and effective spread are more forward-looking measures of the costs of trading, the realized spread is a more relevant measurement for dealers and market makers because it takes into account the price effects on future trades. These effects can cause the dealer to make less profit or even experience a loss, as the market will adapt to the transaction and make inferences about the underlying asset's true value. Overall, the realized spread will typically be smaller than the effective spread.

  • 00:40:00 The lecturer discusses the role of intermediaries, specifically market makers, in holding inventory to balance trade flows, and how the realized spread is a measure of the cost or profit associated with holding this inventory. However, obtaining data to calculate the realized spread requires the price and direction of trades, as well as quote data for determining mid-quote prices. The lecturer explains Lee and Reedy's simple but effective algorithm for determining the direction of trades based on their proximity to the ask or bid price, or the direction of price changes if the trade occurs at the mid-quote.

  • 00:45:00 The speaker discusses the classification of trades and introduces a Lyrid algorithm that correctly classifies trades 85% of the time. The algorithm has difficulty classifying trades at the midpoint, small transactions, and stocks with large capitalization. The speaker notes that the Lyrid algorithm is useful in filling in missing data on trade direction. The section concludes with a discussion of estimating the spread when quote data is unavailable and introduces a popular way to estimate spread - the Roll measure from 1984.

  • 00:50:00 The speaker explains a simple model for estimating spread that only uses data on transaction prices. The model assumes that all traits have the same size and directions are random, and mid-quote follows a random walk. Market orders are not informative in this model. However, the model assumes that the bid-ask spread is constant. Using this model, transaction price can be written as mid-quote plus a half spread for a buy or minus half spread for a sell. The speaker elaborates on how the spread can be estimated using the mean reverting nature of trade directions and transaction prices.

  • 00:55:00 The speaker explains the relationship between liquidity, transaction prices, and the mid-quote. If an investor places a buy order at the ask price and the next order is either a sell order or a buy order at the same ask price, the transaction price will either remain the same or increase. If the next order is a sell order, the transaction price will decrease. As a result, prices are pressured to return to the mid-quote, which leads to a negative covariance between changes in the direction of trade. Directions of trade are mean reverting, meaning that if an investor buys an asset in one period, they will generally sell it in the next period.

  • 01:00:00 The video discusses how to estimate the spread, which is not the only measure of liquidity that one can use. Price depth, which is a similar concept, can be measured by the price impact coefficient, and it tells you how the mid-quote changes depending on the order size. Another measure of liquidity is the bid-ask bounce, which is the difference between the midpoint price and the average of the best bid and ask prices. Finally, the video discusses the effective spread, which is the difference between the execution price and the midpoint price. These measures of liquidity can be used to analyze financial markets and the microstructures within them.

  • 01:05:00 The lecturer discusses different measures for liquidity in financial markets microstructure. The first measure is the price impact coefficient, which estimates the impact of large buy or sell orders on the mid quote in the next period. The lecturer then introduces the Hasbrouck measure, which is almost the same as the price impact coefficient but measures the sensitivity of the price to trading volume instead of trading balance. The Hasbrouck measure is useful when the direction of trades is not known. Another related measure is the Amihud measure of liquidity, which takes the ratio between changes in the mid quote and trading volume and has a different functional form but a similar interpretation. The lecturer emphasizes that these measures must be employed carefully depending on the level at which data is aggregated.

  • 01:10:00 Volume-weighted average price (VWAP) is introduced as a measure of evaluating the performance of a broker in executing an order on behalf of their client. This benchmark is calculated as the average price of transactions in a given day, weighted by their volume. The VWAP is commonly used by large institutional investors aiming to execute trades with minimal price impact on the market. However, this measure is not perfect as it can be subject to manipulation and may depend excessively on a few orders. Another measure presented is implementation shortfall.

  • 01:15:00 The speaker discusses the cost of illiquidity in buying stocks. The realized gain from a transaction is the number of stocks purchased multiplied by the price gain on the current mid-quote. However, there is an opportunity cost to not having bought the stocks at the mid-quote at time zero. The implementation shortfall is the difference between the realized gain and the maximal benefit that could have been accounted for if the order was executed fully at the mid-quote at time zero. The implementation shortfall can be calculated by accepting parameters, and the speaker provides an example involving the purchase of 3,500 shares.

  • 01:20:00 The lecturer discusses additional measures of liquidity beyond execution measures, including trading volume, turnover rate, trade frequency, and price volatility. However, they caution that no single measure of liquidity is perfect due to the lack of a well-defined concept of liquidity, and different measures may contradict each other. For example, a stock's spread might increase after earnings reports, suggesting lower liquidity, while trading volumes often increase, indicating higher liquidity. The lecturer suggests using frequency of trading as a more relevant measure of liquidity for thin markets with limited data.

  • 01:25:00 The lecturer concludes by summarizing the different measures of liquidity and how they can vary depending on data requirements and specific goals. They also highlight that liquidity varies continuously throughout the trading day and can be affected more abruptly by big events. The lecture finishes by providing some exercises for viewers to try, including recreating the figures and looking at the implementation shortfall in the textbook and a link to an article comparing corporate bond markets to equity or stock markets to show how their liquidity differs due to specific market structures. The next lecture will analyze what drives the spread and the main determinants of illiquidity in the market.
Lecture 2: Measuring Liquidity (Financial Markets Microstructure)
Lecture 2: Measuring Liquidity (Financial Markets Microstructure)
  • 2020.07.28
  • www.youtube.com
Lecture 2: Measuring LiquidityFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.youtube.c...
Reason: