Quantitative trading - page 30

 

Lecture 8, part 1: Market Fragmentation (Financial Markets Microstructure)



Lecture 8, part 1: Market Fragmentation (Financial Markets Microstructure)

The lecturer starts by providing a brief review of the previous classes, emphasizing the models and measures related to order-driven markets and market design that were discussed. They highlight the potential trade-offs and unintended consequences of implementing measures to improve liquidity.

The focus of the current class is on market fragmentation, which refers to the existence of multiple markets trading the same asset. The lecturer delves into the costs and benefits associated with market fragmentation and provides historical and regulatory context to better understand its impact.

The lecture explores how the financial markets' microstructure has evolved, leading to market fragmentation. In the past, assets were only traded on the exchange where they were listed. However, with cross-listing and being admitted for trading, assets can now be traded on multiple exchanges. The lecturer explains the concepts of cross-listing, where a company fulfills requirements to be listed on another exchange, and being admitted for trading, where European exchanges allow companies to trade without an explicit procedure. This change has led to most stocks being traded on multiple exchanges.

Policymakers have responded to the challenge of market fragmentation in different ways. Some have opted for artificial consolidation, aiming to reduce fragmentation through virtual means or establishing connections between multiple markets. In the United States, regulations such as order protection require market orders to be automatically routed to the national best bid or offer, ensuring a unified order book. On the other hand, European Union regulations prohibit concentration rules, allowing national companies to trade on exchanges of their choice and fostering fragmentation. The lecturer examines the potential effects of fragmentation, including violations of priority rules in markets with limit order books, where different priority rules can exist for orders within the same price.

The lecture delves into order priority rules and the concept of visibility priority in financial market microstructure. Visibility priority refers to hidden limit orders being executed before visible ones, which can lead to priority rule violations. Additionally, market fragmentation can make it challenging to search for the best price, potentially resulting in worse price discovery as information about the asset's fundamental value becomes dispersed across different markets. This dispersion leads to higher trading costs and hinders price discovery.

The concept of market fragmentation is further explored in terms of its impact on trading costs and liquidity. While fragmented markets may reduce overall liquidity, they can also lead to lower trading costs due to increased competition among exchanges and platforms. Traders may also benefit from improved price discovery as information is distributed across multiple markets. Additionally, fragmented markets may result in greater total liquidity as more liquidity providers participate, potentially attracting more traders. The lecture provides an example of the Dutch stock market before 2003, where the entry of new competitors led to lower trading costs for traders.

The video emphasizes how market fragmentation, characterized by the presence of multiple trading platforms for the same instrument, can influence competition and prices in financial markets. The lecturer cites the example of Euronext, a dominant market player in trading Dutch stocks, facing competition from Deutsche Bursa and the London Stock Exchange. In response, Euronext reduced order entry and execution fees, leading to price reductions that benefited traders. However, fragmentation also increases search costs for traders who need to navigate various exchanges to find the best price before placing their orders.

The lecturer discusses the challenges posed by market fragmentation, particularly the difficulty in searching for the best price in financial markets. Factors such as the depth of different markets, hidden orders, and dark pools of liquidity contribute to the complexity and costliness of the search process. Additionally, there is a misalignment of incentives between brokers and traders, and implementing performance-based contracts becomes challenging. Exchanges may also influence broker incentives by offering payment to direct order flow to a particular exchange, potentially giving rise to conflicts of interest.

The speaker highlights how order protection rules can break down, leading to agency problems, and emphasizes the role of regulations in addressing such issues. In the US, order protection rules require orders to be executed at the best price, but this mechanism works effectively for small orders. For larger orders, the protection rules necessitate climbing up the order book or allowing brokers to route orders as they see fit. Challenges also arise from incorporating exchange fees and different tick sizes across exchanges. The US regulation mandates a minimum tick size of one cent for all exchanges participating in the order protection system, while Europe imposes best execution rules on brokers.

The formulation of broker execution requirements is discussed, highlighting how brokers can consider factors beyond price, such as fees and execution times. The lecture then revisits the Kyle model, which involves a risky asset with a normal distribution of fundamental value, three types of agents, and a market maker who observes aggregate order flow and prices the asset based on the expected fundamental value.

The lecturer explains that the model consists of two equations, one for the pricing schedule and one for the dealer's optimal order size. At this point, the only unknown variables remaining are beta and lambda, which can be solved for. This leads to the derivation of a linear trading strategy and expresses beta and lambda in terms of model parameters such as sigma's variance of view and variance of V. Furthermore, the speculator's profit and average trading cost can be computed. The lecture mentions that the model encompasses not just one market but two, which will be further elaborated on after the break.

  • 00:00:00 The lecturer provides a quick recap of the previous classes, where they examined various models related to order-driven markets and market design, along with the measures that could improve liquidity, but backfire due to their effect on traders' incentives. The focus of this class is on market fragmentation, which refers to the co-existence of multiple markets trading in the same asset. The lecture explores the potential costs and benefits of fragmentation and also provides historical and regulatory context. The students are invited to revisit some of the previously discussed models to better understand the impact of fragmentation.|

  • 00:05:00 The lecturer discusses the changing dynamics of the financial markets microstructure and how market fragmentation is occurring due to the availability of trading various assets on multiple exchanges. In the past, assets were only traded on the exchange where they were listed. However, cross-listing and being admitted for trading have changed this setup. Cross-listing is when a company fulfills all requirements in order to be listed on another exchange, which can be an expensive procedure. Admitted for trading is when European exchanges allow companies to trade on their platform without requiring them to go through an explicit procedure. This change means that being listed on one exchange is no longer a requirement for being able to trade on another. The lecturer states that nowadays, most stocks are traded on multiple exchanges.

  • 00:10:00 The lecturer explains how policymakers have responded to the challenge of market fragmentation by artificially reducing it through virtual consolidation or establishing connections between multiple markets. For example, the US regulation imposes order protection, where market orders are automatically routed to the national best bid or offer, ensuring that investors are acting against one unified order book. On the other hand, EU regulation prohibits concentration rules and requires that national companies need not trade on national exchanges, fostering fragmentation. The lecturer explores the possible effects of fragmentation, such as violating priority rules established in markets with limit order books, where there can be different priority rules for orders within the same price.

  • 00:15:00 The lecture discusses order priority rules and the concept of visibility priority in financial market microstructure. Visibility priority refers to when hidden limit orders are executed before visible ones, and this can lead to violations of priority rules. Additionally, market fragmentation can make it difficult to search for the best price and may lead to worse price discovery because information about the asset's fundamental value is dispersed among different markets. This results in higher trading costs for traders and worse price discovery.

  • 00:20:00 The lecturer discusses the concept of market fragmentation and its potential impact on trading costs and liquidity. Although fragmented markets can lead to lower liquidity overall, it may also result in lower trading costs due to increased competition among different exchanges and platforms. Additionally, traders may benefit from better price discovery as information and signals are dispersed across multiple markets. Lastly, fragmented markets may lead to greater total liquidity as liquidity providers are less concentrated and can potentially earn higher profits, attracting more traders to become liquidity providers. The lecture provides an example of the Dutch stock market before 2003, where a single exchange dominated the market until new competitors emerged, leading to lower trading costs for traders.

  • 00:25:00 The video discusses how market fragmentation, or the presence of multiple trading platforms for the same instrument, can affect competition and prices in financial markets. An example is given of Euronext, a conglomerate that had a dominant market share in trading Dutch stocks until competitors Deutsche Bursa and the London Stock Exchange launched their own platforms. To compete, Euronext reduced its order entry and execution fees substantially, which drove down prices and benefited traders. However, the downside of fragmentation is that it also increases search costs for traders who need to find the best price on various exchanges before placing their orders.

  • 00:30:00 The lecturer discusses the challenges of market fragmentation and the difficulty in searching for the best price in financial markets. The depth of different markets, hidden orders, and dark pools of liquidity make search difficult and costly. There is also a misalignment of incentives between brokers and traders, and performance-based contracts are difficult to implement. Additionally, exchanges may distort incentives for brokers by paying them to direct order flow towards a particular exchange, which can lead to conflicts of interest.

  • 00:35:00 The speaker discusses how order protection rules can break and lead to agency problems, and how these problems can be solved through regulations. The US has order protection rules in place that require orders to be executed at the best price, but this only works for small orders as larger orders require the protection rules to climb up the book or be routed in any way the broker wants. There are also issues with not incorporating exchange fees and different tick sizes across exchanges. The US regulation mandates that all exchanges participating in the order protection system must have a tick size of at least one cent while Europe has best execution rules imposed on brokers.

  • 00:40:00 The speaker discusses the formulation of broker execution requirements and how they allow for brokers to take other factors into account beyond just price, such as fees and execution times. They then move on to discussing trading costs and how they are difficult to compare in abstract due to varying market circumstances. The lecture then transitions to a refresher on the Kyle model, which includes one risky asset with a normal distribution of fundamental value, three types of agents, and a market maker who observes aggregate order flow and prices the asset according to the expected fundamental value.

  • 00:45:00 The speaker explains that the model mentioned previously has two equations- one for the pricing schedule and one for the dealer's optimal order size. The only unknown variables left at this point are beta and lambda, both of which can be solved for. This leads to the derivation of a linear trading strategy and expresses beta and lambda in terms of model parameters such as sigma's variance of view and variance of V. Moreover, the profit of the speculator and average trading cost can be computed as well. The speaker also mentions that in the model, there is not just one market but two markets in which trading happens, which will be further explained after the break.
Lecture 8, part 1: Market Fragmentation (Financial Markets Microstructure)
Lecture 8, part 1: Market Fragmentation (Financial Markets Microstructure)
  • 2020.03.25
  • www.youtube.com
Lecture 8, part 1: Market FragmentationFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www....
 

Lecture 8, part 2: Market Fragmentation (Financial Markets Microstructure)


Lecture 8, part 2: Market Fragmentation (Financial Markets Microstructure)

Let's return to our chiral model, but this time with fragmented markets. Instead of a single market, we now have two markets. Each market has a competitive dealer, as well as an insider who knows the asset value precisely. The noise traders are split into two groups, U1 and U2, and they are assumed to be independent. The idea is to compare the case of two fragmented markets with the case when everyone participates in the same market or only noise traders participate in the same market.

To solve this new model, we follow the same approach as before, but with some modifications. The main difference is that we now have volatilities (Sigma UI) in each market. We can compute the prices that would result in each market by using an expression similar to the consolidated market case, except for an additional term.

If we take the expected value of the price, the last term vanishes because the expectation of U is zero. So, the average prices will be the same in both markets, just like in the consolidated market. However, in the short run, the prices may differ due to this additional term.

When considering the variance of prices (P), we find that the variance of each price in the fragmented markets will be the same as the variance in the consolidated market. This is because the Sigma UI term cancels out with the variance of U.

Moving on to the more interesting aspects, we explore the impact of fragmentation on trading volumes and profits. The trading volumes by informed traders in each market follow linear strategies, with beta values given by the ratio of volatilities (Sigma UI / Sigma V). If we sum the total trading volumes in the two markets, we obtain an expression that can be compared to the order size in the consolidated market. The comparison reveals that the total trading volume in fragmented markets is higher than in the consolidated market.

However, when we compute the profits of informed traders, which are equal to the expected loss of uninformed traders, we find that the expected loss in the fragmented market is greater than in the consolidated market. This implies that noise traders suffer a larger loss in fragmented markets, and this can lead to fewer noise traders participating in the long run.

On the other hand, informed traders thrive in fragmented markets, as their profits increase. This may not be desirable, as it distorts market prices, but it does contribute to price discovery. So, while there are pros and cons to fragmented markets regarding informed trading, it is crucial to consider the impact on noise traders' losses and overall market liquidity.

Another aspect to examine is market depth. In fragmented markets, the depth in each market is lower than in the consolidated market. However, when considering the aggregate depth across both markets, the fragmented market can be deeper in terms of total depth.

Regarding price discovery, informed trading is often considered a proxy for price discovery. The more informed trading occurs, the more price discovery is expected. The linear form of the price equation holds even when considering the distribution of the fundamental value (V) conditional on the information revealed in both markets. This information can be observed through trade quantities or resulting prices.

o, we have a trader who can trade in both markets, acting as an insurance device for the dealers. Fragmentation doesn't affect this insurance mechanism. However, it's important to note that dealers can still trade and provide insurance to each other even in fragmented markets, so the risk-sharing motive is not a strong reason for consolidation.

Now, let's briefly discuss Clausten's model of limit or order-driven markets. In this model, we find that the aggregate depth of the fragmented market is larger than the depth of the consolidated market. This conclusion aligns with Kyle's model, although the underlying reasons are different.

Moving on to the specific features of Clausten's model, we assume two asymmetric markets: an incumbent market (I) and a new entry market (II). Market participants behave similarly to before, with market orders being split between the two markets based on certain probabilities.

The model reveals that the total depth in the fragmented market (Y bar) is greater than the depth in a consolidated market. This is because fragmentation allows traders to bypass price priority, leading to larger depths. The intuition here is similar to the concept of pro rata allocation versus time priority, where pro rata allocation can result in deeper markets. Additionally, there is a critical value of trader sophistication (gamma) below which the entry market cannot survive, highlighting the importance of attracting a critical mass of traders for market viability.

It's worth mentioning that the model assumes positive tick size, unlike the real world where we often have negative display costs for limit orders. Finally, we acknowledge that market fragmentation has both advantages and costs, impacting trading costs and market depth.

To summarize, fragmented markets have implications for trading volumes, profits, market depth, and price discovery. Informed traders tend to benefit from fragmented markets, while noise traders suffer greater losses. Market depth may decrease in individual markets but can increase in aggregate. Price discovery is influenced by the level of informed trading in both markets.

At this point, I conclude the discussion on market fragmentation. I recommend solving exercise three in chapter seven on brokers receiving order flow payments to explore how these payments affect market outcomes. I will also upload a couple of related articles on epsilon for further reading. Thank you for today, and I apologize for going over time. Remember, there will be no class this Friday, but we will meet next week on Twitch. Feel free to ask any questions before we wrap up. Goodbye and take care!

Lecture 8, part 2: Market Fragmentation (Financial Markets Microstructure)
Lecture 8, part 2: Market Fragmentation (Financial Markets Microstructure)
  • 2020.03.25
  • www.youtube.com
Lecture 8, part 2: Market FragmentationFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www....
 

Lecture 9, part 1: Market Transparency (Financial Markets Microstructure)



Lecture 9, part 1: Market Transparency (Financial Markets Microstructure)

The lecture begins by reviewing the previous session's discussion on market fragmentation and its costs and benefits. The focus of the current lecture is market transparency and its impact on market outcomes. While financial markets are generally considered transparent due to the availability of historical price and trade data, there are still important information asymmetries that exist. Different markets have different levels of transparency, and the type of transparency can have varying effects on the market.

The lecturer explains that market transparency means all participants observe the same type of information, eliminating the issue of fragmentation. Transparency can be categorized into three types: pre-trade information, information available during the trade, and post-trade information. Exchanges profit from selling this data, but they aim to strike a balance between releasing enough information to set a reasonable price and not giving away data for free or aiding their competitors. It is important to note that different traders possess different pieces of information, leading to asymmetric information in the market, which can result in market frictions.

Regulations play a crucial role in ensuring market transparency in financial markets. The lecturer discusses how laws and rules in both Europe and the US govern market transparency. The objective of these regulations is to ensure that sufficient information is released pre-trade, and firms are required to disclose relevant information to decrease the degree of information asymmetry between informed and less informed traders. In the US, a centralized system called the National Marketplace System (NMS) collects information about all trades in financial assets, promoting transparency.

To illustrate the impact of market transparency, the lecture provides a real-world example involving rapper Jay-Z's attempt to buy back the streaming service Tidal. The stock price of Tidal surged to an unprecedented level before trading was halted, leaving some traders buying stocks at 11 kroner that they would later have to sell at one kroner. This example highlights that transparency is not only about information being available but also about it being easily accessible, affordable, and understandable.

The lecturer introduces the concept of market transparency in relation to the Diamond Chain Store Paradox. The paradox states that in a market where consumers sequentially search for the best price, all firms need to set the same price to remain competitive. However, by doing so, each firm obtains market power and can charge a higher-than-equilibrium price. In financial markets, this translates to dealers charging profit-maximizing bid and ask quotes, eliminating the usual competition-driven undercutting. Traders are then required to approach multiple dealers to obtain the best price, resulting in wider price spreads. The solution to this issue is market transparency, where dealers can publicly post their prices for everyone to see.

The lecturer delves into the market transparency's impact on search costs in financial markets. Search costs affect the market power of dealers and traders. Dealers, who have more market power due to less visibility, would prefer a lack of transparency. In contrast, traders face higher search costs and suffer from wider spreads when transparency is absent. Lack of transparency decreases market efficiency due to increased transaction costs. Regulators enforce transparency to compel dealers and market makers to provide efficiency by requiring them to publish quotes. While the best bid and ask prices are available in the market, assessing the market's depth and its response to changes in order size becomes challenging.

The lecture introduces a reduced-form Kyle model to discuss the effects of depth uncertainty on financial markets. The model assumes that depth, represented by lambda, determines the pricing rule for market makers. However, traders are uncertain about the value of lambda, which affects their trading behavior. In transparent markets, traders can observe lambda, while in opaque markets, they cannot. The optimal trade size is inversely proportional to 1/lambda in a transparent market and 1/expected lambda in an opaque market. The lecture also introduces Jensen's inequality, which states that the expected value of a convex function is greater than or equal to the convex function of the expected value.

The lecturer explains how market transparency impacts trading volume. In transparent markets, the expected trading volume is higher than in opaque markets due to risk aversion among informed traders. The lecture utilizes a graph to demonstrate the relationship between trader profits and order size for different lambda values, showcasing how uncertainty about lambda influences trading behavior. When informal traders are uncertain about market depth, they trade based on the expected value of X, resulting in lower trading volume compared to transparent markets.

The speaker emphasizes the significance of lambda, the price impact coefficient, in decreasing the average level of X and its effect on market transparency. In situations where lambda is high, even a slight decrease in X leads to a stronger price effect. On the other hand, if lambda is low, a small price decrease has a limited impact. Traders are more concerned about lambda being high rather than low. The lecture concludes by hinting at the next section, which will focus on order flow transparency and the debate over whether this information should be made available to all dealers.

In the upcoming section of the lecture, the speaker delves into the concept of order flow transparency and the ongoing debate surrounding its availability to all dealers in the market. Order flow transparency refers to the visibility of information regarding the flow of orders, including the identities of buyers and sellers, the quantity of orders, and the timing of trades.

The lecturer acknowledges that there are differing opinions on whether order flow transparency should be universally accessible. Proponents argue that increased transparency allows for a more efficient market by reducing information asymmetry and facilitating fairer price discovery. They believe that making order flow information available to all market participants promotes healthy competition and improves overall market outcomes.

However, opponents argue that unrestricted access to order flow information may lead to negative consequences. They assert that large institutional investors, such as market makers or high-frequency traders, may exploit this information advantage to their benefit, potentially harming smaller investors or retail traders. Additionally, concerns about front-running, where traders with access to order flow information can exploit it for personal gain, further fuel the debate.

The lecturer proceeds to explore the different approaches taken by regulators regarding order flow transparency. In some jurisdictions, regulations mandate the disclosure of order flow information to ensure a level playing field for all market participants. This type of transparency aims to prevent unfair advantages and promote market integrity.

However, there are alternative approaches as well. For instance, some regulators opt for a more controlled dissemination of order flow information. They may limit access to this data or introduce delayed reporting to mitigate potential negative impacts.

The lecturer emphasizes that achieving the right balance between order flow transparency and market efficiency is a complex task. Regulators need to consider various factors, including the size and structure of the market, the nature of the participants, and the potential risks associated with unrestricted access to order flow information.

To illustrate the practical implications of order flow transparency, the lecturer provides a real-world example. They discuss a hypothetical scenario where a market with full order flow transparency experiences increased trading activity, reduced bid-ask spreads, and improved liquidity. In this case, market participants have access to comprehensive information about the flow of orders, allowing them to make more informed trading decisions.

On the other hand, the lecturer also highlights potential drawbacks. They explain how certain market participants, such as institutional investors, may strategically withhold their order flow information to maintain a competitive advantage. This behavior can hinder transparency and lead to distorted market outcomes.

The lecture concludes with the lecturer posing thought-provoking questions to encourage further reflection and discussion. They encourage the audience to contemplate the trade-offs associated with order flow transparency, the potential impact on different market participants, and the role of regulations in striking the right balance.

By examining the nuances and implications of order flow transparency, the lecture provides valuable insights into the ongoing debate surrounding this topic and prompts the audience to critically evaluate the significance of transparency in financial markets.

Following the discussion on order flow transparency, the lecturer shifts the focus to the broader concept of market transparency and its impact on market outcomes. Market transparency refers to the availability and accessibility of information within financial markets, which plays a crucial role in shaping market dynamics and participant behavior.

The lecturer explains that while financial markets are generally regarded as transparent due to the abundance of historical price and trade data, it's important to recognize that not all relevant information is equally accessible. Different markets may vary in terms of the type and extent of information they make observable or readily available to market participants.

To further explore the effects of market transparency, the lecturer distinguishes between three categories of information: pre-trade information, information available during the trade, and post-trade information. Pre-trade information encompasses data on the bid-ask spread, order book depth, and pending orders, which can influence trading decisions and price formation. Information available during the trade refers to real-time updates on trades being executed, while post-trade information includes details about completed transactions, such as prices and volumes.

The lecturer highlights that market transparency is not a one-size-fits-all concept. Different types of transparency can have varying effects on market outcomes. For example, increased pre-trade transparency may enhance price efficiency and reduce information asymmetry among market participants, leading to more accurate pricing. On the other hand, excessive transparency during the trade can potentially expose traders' intentions and strategies, negatively impacting their ability to execute trades at favorable prices.

The lecturer also acknowledges that exchanges derive profits by selling market data. While exchanges aim to strike a balance between releasing sufficient information to establish fair prices and avoiding giving away data for free or aiding their competitors, conflicts of interest can arise. The lecturer explains that this dynamic contributes to the presence of asymmetric information in the market, which can create frictions and impact trading behavior.

To address the challenges associated with market transparency, the lecturer highlights the regulatory framework implemented in both Europe and the United States. These regulations aim to ensure that relevant information is disclosed before trades occur, reducing the degree of information asymmetry between informed and less informed traders. In the United States, the National Marketplace System (NMS) serves as a centralized system that collects information on trades across various financial assets, fostering transparency and enhancing market integrity.

To illustrate the practical implications of market transparency, the lecturer presents a real-world example involving a music performer's acquisition of a music service. The consequences of transparency, particularly regarding the buyback of stocks by the performer, demonstrate how market participants' access to information can shape their decision-making and subsequent market outcomes.

By examining the nuances of market transparency and its regulation, the lecture provides a comprehensive understanding of its impact on financial markets. It emphasizes the importance of striking a balance between transparency and market efficiency, as well as the role of regulations in ensuring fair and transparent market practices.

As the lecture concludes, the audience is encouraged to critically evaluate the benefits and challenges associated with market transparency. The lecturer emphasizes the dynamic nature of market transparency and the ongoing need for regulators, market participants, and scholars to adapt and address emerging issues in order to foster transparent and efficient financial markets.

  • 00:00:00 In this section of the lecture on financial markets microstructure, the focus is on market transparency. The lecturer starts with a quick review of the previous lecture which talked about the costs and benefits of market fragmentation. Today's lecture will explore a related topic, market transparency, and how it affects market outcomes. Although financial markets are generally seen as transparent because of the availability of historical price and trade data, some important information is not available. Markets differ in terms of what they make observable or accessible. Different types of transparency can also have different effects on the market.

  • 00:05:00 The lecturer discusses market transparency in financial markets microstructure. It is explained that market transparency means all market participants observe the same type of information, negating the issue of fragmentation. The three categories of transparency are pre-trade information, during the trade, and post-trade information. The lecturer highlights that exchanges profit from selling this data, but they aim to balance releasing enough information to set a reasonable price while not giving away data for free or helping their competitors. The lecturer explains that different traders will have different pieces of information, leading to asymmetric information in the market, which can lead to frictions.

  • 00:10:00 The lecturer discusses the concept of market transparency and how it is regulated in financial markets through laws and rules in both Europe and the US. The goal of these regulations is to ensure that enough information is released pre-trade and that firms must disclose relevant information to decrease the degree of asymmetry of information in the market between informed and less informed traders. The lecture mentions a centralized system in the US called the National Marketplace System (NMS) which collects information about all trades that have happened in financial assets. The section concludes with a real-world example of how a music performer acquired a music service and the consequences of transparency with respect to the buyback of stocks by the performer.

  • 00:15:00 The speaker discusses an example of market transparency, or lack thereof, relating to the rapper Jay-Z's attempt to buy back the streaming service Tidal. The stock price climbed to an unprecedented level before trading was halted, leaving some traders buying stocks at 11 kroner that they would have to sell at one kroner. The speaker notes that transparency is not just about information being available but also that it must be accessible, cheap, and digestible. The lecture then delves into three kinds of transparency and how they vary across markets. In some markets, quotes are readily available, while in others, like OTC markets, you must actively contact dealers to inquire about prices.

  • 00:20:00 The lecturer discusses the Diamond Chain Store Paradox and how it applies to financial markets microstructure. The paradox states that in a market where consumers sequentially search for the best price, all firms need to set the same price to remain competitive, but in doing so, each firm has market power and can charge a higher than equilibrium price. In the case of financial markets, this means that dealers charge profit-maximizing bid and ask quotes, and the undercutting that usually leads to competition no longer applies. Traders have to approach multiple dealers to get the best price, leading to a wide spread in prices. The solution is market transparency, where dealers can post their prices for everyone to see.

  • 00:25:00 The instructor discusses the market transparency in microstructure. The concept of search cost in financial markets is introduced and how it affects the market power of dealers and traders is explained. Dealers, who have more market power due to less visibility, would be happy if the market was not transparent, whereas traders face a larger search cost and suffer due to wider spreads in the absence of transparency. The efficiency of the market will be lower due to transaction costs in this type of situation. To force dealers and market makers to provide efficiency, regulators enforce transparency and require them to publish the quotes. While the best bid and ask price are available in the market, it is difficult to gauge the market's depth and its reaction to a change in order size.

  • 00:30:00 The speaker discusses the effects of depth uncertainty in financial markets using a reduced-form Kyle model. The model assumes that depth lambda represents the pricing rule for market makers, but traders are uncertain about the value of lambda, which affects their trading behavior. In transparent markets, traders can observe lambda, but in opaque ones, they cannot. The optimal trade size is inversely proportional to 1/lambda in a transparent market and 1/expected lambda in an opaque one. The use of Jensen's inequality, which states that the expected value of a convex function is greater than or equal to the convex function of the expected value, is also introduced.

  • 00:35:00 The lecturer explains how market transparency affects trading volume. The expected volume of trade in a transparent market is greater than in an opaque one because of risk aversion among informal traders. The lecturer uses a graph to illustrate the relationship between trader profits and order size for different values of lambda, and how uncertainty about lambda affects trading behavior. If an informal trader is uncertain about the depth of the market, they will trade the expected value of X, which results in lower trading volume than in a transparent market.

  • 00:40:00 The speaker discusses the impact of lambda, or the price impact coefficient, on decreasing the average level of X and how it affects the market transparency. He explains that in scenarios where lambda is high, decreasing X even slightly will result in a stronger price effect. On the other hand, if lambda is low, decreasing the price by a small amount won't have a significant impact. The speaker emphasizes that traders are more concerned about lambda being high than being low. Furthermore, the speaker teases the next section, which will focus on order flow transparency and whether or not this information should be made available to all dealers.
Lecture 9, part 1: Market Transparency (Financial Markets Microstructure)
Lecture 9, part 1: Market Transparency (Financial Markets Microstructure)
  • 2020.04.01
  • www.youtube.com
Lecture 9, part 1: Market TransparencyFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.y...
 

Lecture 9, part 2: Market Transparency (Financial Markets Microstructure)



Lecture 9, part 2: Market Transparency (Financial Markets Microstructure)

In order to understand the consequences of flow transparency, the lecturer introduces a simple model. The model assumes the existence of one asset with a fundamental value that can be either high or low with equal probability. Additionally, there are at least two dealers in the market, and two traders submitting orders. The traders can either both be informed or both be uninformed. Based on this model, the lecturer draws conclusions about the correlation of order flow and the behavior of liquidity traders.

The lecturer explains that when traders are informed, there will be a higher correlation of order flow. This means that the orders submitted by informed traders will be more similar to each other. On the other hand, the orders coming from liquidity traders, who are typically uninformed, will be less correlated with each other.

Moving on, the lecturer discusses two scenarios: market opacity and market transparency. In an opaque market, dealers quote prices without having full visibility of the entire market order flow. They rely on a probability to obtain the bid price. In contrast, in a transparent market, dealers can see both orders and base their quotes on the total order flow. This leads to better price discovery and a more dispersed market.

The lecturer emphasizes that accidental correlations in the market may lead to less drastic versions of these conclusions. However, the overall theme remains the same—transparency promotes better price discovery and market efficiency.

Furthermore, the lecturer explains how market transparency can affect different types of traders. In transparent markets, uninformed traders are better off as they are easily identified as such. Consequently, they do not have to pay adverse selection premiums and face a zero spread. On the other hand, informed traders are worse off in transparent markets because they can be more easily identified and, as a result, pay wider spreads.

The lecturer notes that order flow transparency can act as a substitute for trader identification transparency. However, there may be a reduction in informed order flow in transparent markets. It's a delicate balance between promoting transparency and maintaining a sufficient level of informed trading activity.

Shifting focus to dealer behavior, the lecturer analyzes the profitability of dealers in markets where information is not transparent. In such scenarios, dealers can make profits despite quoting the highest price. The uninformed trader's average profit would be zero. This model demonstrates how attracting order flow gives dealers an informational advantage, leading to their profitability. The spread—the difference between bid and ask prices—would be smaller than that of the static limit-order model.

The lecturer points out that this situation is observable in reality, where dealers often provide negative spreads to large traders in exchange for future trades. This practice underscores the strategic importance of attracting order flow.

Additionally, the lecturer discusses the effects of market transparency on dealer behavior. Attempts to attract order flow may force dealers to quote narrower spreads in order to gain an information advantage over other dealers. However, dealers are not always inclined to commit to transparency. Revealing their past trade information would divulge their competitive advantage, potentially leading to collusion and wider spreads among dealers.

The lecture then delves into the impact of market transparency on market organization and trader reputation. Transparency enables firms in the market to identify who is trading and on what terms, making it easier to detect deviations from pre-established agreements. Greater transparency can result in price improvements for uninformed traders and allow limit traders to react quickly to market orders from institutional investors. However, it can also lead to informed traders receiving bad prices and unfavorable spreads due to their reputation. This creates a separation in the market.

The lecturer concludes by discussing the reallocation of wealth and welfare from insiders to uninformed traders due to market transparency. While insiders may suffer from transparency, uninformed traders benefit by obtaining better terms of trade. This explains why regulators often advocate for transparency in markets, aiming to protect uninformed traders. However, the market may resist transparency as insiders typically have more influence on market organization than uninformed traders. The negative impact on informed traders outweighs the benefits for uninformed traders. Ultimately, transparency has significant consequences and favors the liquidity-providing uninformed traders.

Lastly, the lecturer acknowledges the potential benefits of opaqueness in limiting the adverse effects of symmetrically distributed knowledge. Hidden limit orders serve as an example where traders can submit orders that are executed without being visible to others in the market. This provides insurance for uninformed traders, allowing them to submit limit sell orders on stocks they hold long positions in without affecting the market price. Opaqueness can be beneficial for society as it reduces asymmetric knowledge and promotes a more balanced market environment.

The lecturer further expands on the concept of opaqueness and its positive effects on limiting the adverse consequences of symmetrically distributed knowledge. By allowing traders to submit hidden limit orders, the market can avoid sudden price movements caused by the immediate execution of visible orders. This insurance-like mechanism benefits uninformed traders, as they can execute their orders without impacting market prices.

Opaque market conditions also help reduce the asymmetry of knowledge among market participants. When certain orders are hidden from others, it prevents the immediate dissemination of information, allowing traders to make decisions based on their own analysis rather than reacting to every trade in the market. This can contribute to a more stable and balanced market environment.

However, the lecturer emphasizes that opaqueness should be carefully balanced with the need for transparency in certain areas. While opaqueness can provide benefits in terms of reducing adverse selection and limiting information asymmetry, excessive opaqueness can also create opportunities for market manipulation and unfair practices.

Regulators and market participants must strike a balance between transparency and opaqueness to ensure a fair and efficient market. Transparency promotes price discovery and protects uninformed traders, while opaqueness helps mitigate the adverse effects of symmetrically distributed knowledge. Finding the right combination of transparency and opaqueness is crucial for maintaining market integrity and promoting overall market welfare.

The lecturer's discussion on market transparency and opaqueness highlights their significant impacts on market outcomes, trader behavior, and overall market welfare. Transparency improves price discovery and benefits uninformed traders while potentially disadvantaging informed traders. Opaqueness, on the other hand, can limit adverse consequences and promote stability but should be carefully balanced to avoid market manipulation. Finding the right level of transparency and opaqueness is essential for creating a fair, efficient, and robust market environment.

  • 00:00:00 The lecturer discusses a simple model to understand the consequences of flow transparency. The model assumes that there is one asset with a fundamental value that can be high or low with equal probability, and there are at least two dealers in the market. The model also assumes two traders submitting orders, with both orders coming from informed traders or both orders coming from uninformed traders. Based on this model, the lecturer concludes that there will be higher correlation of order flow when traders are informed, and the orders coming from liquidity traders will be less correlated with each other.

  • 00:05:00 The lecturer discusses the two scenarios of market opacity and market transparency. In the opaque market, dealers quote without seeing the entire market order flow and form a probability to obtain the bid price. On the other hand, the transparent market allows dealers to see both orders and condition their quotes on the total order flow, leading to better price discovering and a more dispersed market. The lecturer also points out that accidental correlations in the market may lead to less drastic versions of these conclusions, but the overall theme remains the same.

  • 00:10:00 The professor discusses how market transparency can affect market outcomes. Transparent markets are better for uninformed traders, as they are identified as such, and therefore do not have to pay adverse selection premiums and face zero spread. On the other hand, informed traders are worse off in transparent markets because they can more easily be identified and pay wider spreads. The professor notes that order flow transparency can act as a substitute for trader identification transparency, though there may be a reduction in informed order flow in transparent markets. The discussion then shifts to the effects of revealing post-trade information and how transparent markets compare to opaque markets in terms of dealer behavior.

  • 00:15:00 The analysis begins with the assumption that dealers will have different levels of information: one dealer (I) will know what the first-period order was, while the other will not. The dealers move in sequence, with the uninformed dealer setting the code first, followed by the informed trader. The dealer who is not informed will not know what the first-period trade was. The analysis focuses on what happens in this market in the second period, specifically on the prices that the dealer would be willing to set based on their expectations of V, the asset's value. The dealer would be willing to charge any price between Miu and VH, with the maximal number being VH.

  • 00:20:00 The lecturer explains the scenarios and possible outcomes of an uninformed trader interacting with an informed dealer. The uninformed trader could quote a price below the highest ask price, but the informed dealer would choose not to trade. Therefore, the only possible case in equilibrium is for the uninformed dealer to quote the highest ask price, and the informed dealer will barely undercut the price. This latter case will result in the informed dealer earning a significant profit knowing that the asset is worth the price of the highest ask.

  • 00:25:00 The instructor discusses the profitability of dealers in a market where information is not transparent. He explains that in period two, if the spread is large enough, dealers can make profits despite quoting the H price, and an uninformed trader's average profit would be zero. In period one, dealers would bid lower to attract order flow, thus leading to a narrower spread in competitive markets. This model shows how attracting order flow gives dealers an informational advantage, leading to their profitability, and the spread would be smaller than that of the static limit-order model. The instructor notes that this situation is observable in reality, where dealers often put negative spreads to large traders in exchange for future trades.

  • 00:30:00 The lecturer discusses the market transparency in financial markets microstructure and how attempts to attract order flow may force dealers to quote narrower spreads to gain an information advantage over other dealers. The lecturer also explains that opacity increases the aggregate trading cost for the uninformed traders, while transparency benefits them. However, he points out that dealers would not like to commit to transparency because it would involve divulging their past trade information, which gives them a competitive advantage. Additionally, he explains how the availability of historical data resulting from transparency may lead to collusion and wider spreads among dealers.

  • 00:35:00 The professor discusses the concept of market transparency and how it fosters collusion in financial markets. Successful collusion requires a physical threat of punishment which relies on the ability to detect deviations from the pre-established agreements. Greater transparency allows firms in the market to identify who is trading and on what terms, making it easier to detect deviations and punish them if necessary. Furthermore, market transparency affects the prices that traders are offered, with uninformed traders receiving good prices and price improvements from dealers, while limit traders try to react quickly to the market orders they see coming from institutional investors. The entry information, which identifies the traders, also affects the prices they are offered.

  • 00:40:00 This is because informed traders are likely to get bad prices and unfavorable spreads due to their reputation, creating a separation in the market. However, this transparency can also be advantageous for building trader reputation or using signaling mechanisms such as sunshine trading to disclose the information. Similar outcomes can occur due to cream-skimming practices by large banks, leading to an equilibrium that is self-enforcing. Thus, transparency regarding trader identity can have significant impacts on market microstructure.

  • 00:45:00 The video discusses how market transparency leads to a reallocation of wealth and welfare from insiders to uninformed traders. While insiders suffer from transparency, the uninformed actually benefit as they get better terms of trade. This helps explain why regulators often push for transparency in markets, aiming to protect the uninformed traders. However, the market may resist transparency as insiders often have more influence on market organization than the uninformed traders. Additionally, the loss that informed traders suffer from being identified as insiders hits them more significantly than the benefits of the uninformed traders. Ultimately, transparency has reallocated consequences and benefits the uninformed traders who generally provide or foster liquidity in the market.

  • 00:50:00 The professor discusses how opaqueness can be beneficial in limiting the adverse effects of symmetrically distributed knowledge. Hidden limit orders are an example of this where traders can submit orders that are executed without being visible to others in the market. These hidden orders can act as a form of insurance for uninformed traders who can use them to submit limit sell orders on stocks they have a long position in without affecting the market price. Opaqueness can be good for society as some orders can be submitted without moving the market price and therefore lead to less asymmetric knowledge.
Lecture 9, part 2: Market Transparency (Financial Markets Microstructure)
Lecture 9, part 2: Market Transparency (Financial Markets Microstructure)
  • 2020.04.01
  • www.youtube.com
Lecture 9, part 2: Market TransparencyFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.y...
 

Lecture 10, part 1: Value of Liquidity (Financial Markets Microstructure)



Lecture 10, part 1: Value of Liquidity (Financial Markets Microstructure)

During the lecture, the instructor introduces several announcements and engages the audience through interactive activities. Firstly, the instructor informs the students about the inclusion of small blitz quizzes throughout the lecture to enhance interactivity and active learning in the course. These quizzes are designed to test students' understanding of the material and encourage participation.

Next, the instructor addresses some administrative matters. They mention the cancellation of the exercise class on Easter Friday, and they propose the possibility of rescheduling the class to a later date, approximately two or three weeks after Easter. This ensures that students will have the opportunity to cover the missed material and maintain the continuity of the course.

The instructor also announces the upcoming release of problem set number two, indicating that students should expect to receive it soon. This allows students to prepare and allocate sufficient time to work on the problem set, promoting effective learning and timely completion of assignments.

Furthermore, the instructor acknowledges the importance of audio quality during the lecture and assures the audience that they have resolved any issues with their sound setup. However, the instructor encourages the students to provide feedback if they notice any problems to ensure a seamless learning experience for everyone.

Shifting focus to the lecture content, the instructor delves into the topic of liquidity and its impact on asset value. They initiate the discussion by providing a brief review of transparency, building upon the previous week's session. To illustrate the concept of limited liquidity and its effect on prices, the instructor presents a motivating example involving US Treasury notes and bills. This real-world scenario demonstrates how constrained liquidity can introduce inefficiencies in pricing.

The lecture progresses, emphasizing the relationship between liquidity and asset value. The instructor explains that less liquid assets tend to be traded at a discount due to the additional costs associated with selling them in the future, owing to limited liquidity. Investors factor in these costs when valuing the asset and require a higher return to compensate for the risks associated with liquidity constraints. Moreover, the instructor highlights that liquidity can vary over time, causing fluctuations in asset prices.

Delving deeper into the topic, the lecturer explores liquidity risk and its implications for asset pricing. They emphasize that liquidity risk can indeed impact asset prices, and this phenomenon can be observed in empirical data. Introducing a toy model of liquidity premium by Mendelson, the lecture focuses on determining the rate of return on an asset, specifically how the mid-price grows in the market. Various factors influencing the rate of return are discussed, contributing to a comprehensive understanding of liquidity risk's influence on asset pricing.

The lecture proceeds with an explanation of how to compute the nominal rate of return on an asset using the required rate of return formula. The nominal rate of return is derived based on the mid-quote of the asset and the expected future payout, adjusted for half-spread. Through the derivation of this formula, students gain insights into the mathematical relationship between these variables.

The instructor introduces the concept of approximation to analyze the difference between the real return (small R) and the average return based on price growth (big R) in financial markets. Utilizing logarithmic expressions and making suitable assumptions, such as the approximation log of 1 + X = X for small values of X, the instructor derives an expression for the difference between big R and small R. This clarifies how the real return is generally smaller than the average return due to factors such as the spread between buying and selling prices and the incorporation of trading costs.

Building upon this understanding, the lecture delves into the influence of limited liquidity on asset pricing. The instructor highlights that the nominal return, representing the rate at which the asset price grows, tends to exceed the real return due to trading costs, spreads, and the fact that investors buy at prices above the mid-quote and sell at prices below it. The cost of trading is considered a fixed cost, and as investors hold the asset for longer periods, the impact of this cost diminishes. The discrepancy between the nominal return and the real return is termed the liquidity premium, representing the rate at which the asset price must grow for traders to be willing to trade given the fixed liquidity of the asset.

Moving forward, the lecture addresses the process of deducing the small R required rate of return from the nominal rate of return and asset growth rates. The instructor tackles the question of how to determine small R from big R, considering the presence of two expressions and whether to use the first one, the second one, or disregard both and rely on the nominal rate average. The instructor clarifies that the choice depends on the presence of positive or negative supply. When positive supply exists, buyers benefit from a high nominal rate of return but suffer from a low small R, whereas sellers benefit from a low small R but suffer when it is high.

The discussion continues by exploring the value of liquidity in financial market microstructure and its influence on the required rate of return. The instructor explains that the preferred rate of return is determined by buyers with greater bargaining power in the market, leading to positive or negative aggregate supply. Empirical evidence reveals a positive liquidity premium for stocks and bonds, indicating its significance. Additionally, the impact of heterogeneity in holding periods is briefly mentioned, suggesting potential avenues for further investigation.

Focusing on the value of liquidity in financial market microstructure, the lecture employs a simple example involving two assets with different spreads and two types of investors with varying holding periods. The instructor highlights how investors self-select into trading different assets based on their characteristics. Those with shorter holding periods opt for assets with smaller spreads, despite lower nominal returns and higher trading costs. In contrast, investors with longer holding periods choose less liquid assets with higher spreads and larger nominal returns, ultimately yielding a higher real rate of return. The concept of equilibrium is introduced, indicating that it can only exist when investors with lower holding periods trade in low-spread assets and those with longer holding periods trade in less liquid assets.

Concluding the lecture, the instructor reflects on the topic of specialization based on investor characteristics. While acknowledging that the conclusion suggesting pension funds trade in riskier assets while hedge funds trade in less risky assets due to adverse selection may not fully explain the situation, the instructor acknowledges the intriguing concept of specialization based on investor traits. They suggest exploring this aspect further within the context of financial market microstructure.

  • 00:00:00 The instructor announces a few small blitz quizzes during the lecture to make the course more interactive. He also mentions the cancellation of the exercise class on Easter Friday and the possible rescheduling of the class in two or three weeks, and the upcoming release of problem set number two. The instructor also notes that he has figured out his sound setup and asks the audience to let him know if something is wrong.

  • 00:05:00 The lecturer discusses the value of liquidity and how limited liquidity can affect the asset value itself. The lecture begins with a refresher on transparency from the previous week's session. The lecturer then gives a motivating example regarding US Treasury notes and bills to explain how limited liquidity can introduce inefficiency in prices. The lecture concludes with a blitz quiz asking which between two assets, given the same face values, was cheaper in the secondary market.

  • 00:10:00 The lecturer discusses the relationship between liquidity and asset value. He explains that less liquid assets are generally traded at a discount because of the additional trading costs incurred when trying to sell them in the future due to limited liquidity. Investors incorporate these costs into their current valuation of the asset and require a higher return to compensate. Additionally, liquidity can change over time, so fluctuations in liquidity can also affect the asset price.

  • 00:15:00 The lecturer discusses the concept of liquidity risk and its impact on asset pricing, highlighting that liquidity risk may affect asset prices, and it's possible to see it in the data. The lecture moves to a toy model of liquidity premium due to Mendelson, where agents explicitly care about the resale value of the asset they are buying. The model focuses on finding the rate of return on the asset, which is the rate at which the mid-price grows in the market. The investor requires a fixed return level, and the lecture explores the different factors that affect the rate of return on the asset. The lecture concludes by outlining the general plan for the day, which includes further discussion on liquidity risk and its impact on asset pricing.

  • 00:20:00 The lecturer explains how to determine the nominal rate of return on an asset by using the formula for required rate of return, which is the minimum rate of return an investor requires to invest in a particular asset. The nominal rate of return is computed in terms of the mid-quote of the asset and the expected future payout adjusted for half-spread. The formula for nominal rate of return is derived using the definition of required rate of return, and it is expressed as an equation with the mid-quote of the asset and expected future payout adjusted for half-spread.

  • 00:25:00 The speaker explains the use of an approximation to show the difference between the real return (small R) and the average return based on price growth (big R) in financial markets. To do this, they use logarithmic expressions and make the assumption that all values are small enough to use the approximation log of 1 + X = X. They then show how to derive an expression for the difference between big R and small R, based on the spread between the buying and selling prices of an asset, and demonstrate that the real return is smaller than the average return. The speaker also explains the difference between small R and big R and how small R represents the real return that investors get after trading costs are factored in, whereas big R is how the price of the asset grows on average.

  • 00:30:00 The speaker explains that the nominal return (the rate at which the asset price grows) will be above the real return due to limited liquidity, spread, and the fact that investors buy an asset at a price above the mid-quote and sell it at a price below the mid-quote. The cost of trading is basically a fixed cost, and the longer an investor holds the asset, the less relevant this cost becomes. The liquidity premium is the difference between the nominal return and the real return, and it is by how much faster the asset price must grow for traders to be willing to trade in it, given the fixed liquidity of the asset.

  • 00:35:00 The speaker discusses the process of backtracking the small R required rate of return from the nominal rate of return and the growth rate of assets. The question arises on how to learn the small R from big R given that there are two expressions for it, and whether to use the first one, the second one, or ignore both and use the nominal rate average. According to the speaker, the answer is to use the first one when there is a positive supply and the second one when there is negative supply. He explains that when there is positive supply, the buyers benefit from high nominal rate of return and suffer from low R, while the sellers benefit when R is small but suffer when it is large.

  • 00:40:00 The speaker discusses the value of liquidity in financial markets microstructure and the impact of the number of buyers and sellers on the required rate of return. They explain that the preferred rate of return is chosen by the buyers with more bargaining power in the market, which can result in a positive or negative aggregate supply. The speaker notes that there are many different layers to this model and that the empirical evidence shows a positive liquidity premium for stocks and bonds. Finally, the speaker considers the impact of heterogeneity in holding periods on the model and suggests further exploration of this topic.

  • 00:45:00 The speaker begins by using a simple example of two assets with different spreads and two different types of investors with different holding periods. The investor's utility function is interpreted as their choice within a realm of assets with different nominal returns and spreads. The investor chooses a bundle of s and R, representing their budget set, and their indifference curves are generated from a utility function given by the required rate of return given nominal rate of return and spread. The trader who trades frequently has a low holding period and their indifference curves are linear with a negative slope while the trader who trades less frequently has a large holding period, resulting in flatter indifference curves.

  • 00:50:00 The lecturer discusses how different types of investors will self-select into trading different assets based on their holding periods and the size of trading costs. Investors with shorter holding periods will trade in assets with smaller spreads and suffer more from trading costs, even though they will have a lower nominal return. In contrast, investors with longer holding periods will trade in less liquid assets with higher spreads and larger nominal returns, but will ultimately get a higher real rate of return. The example given demonstrates that an equilibrium can only exist if investors with lower holding periods trade in assets with low spreads and those with higher holding periods trade in less liquid assets.

  • 00:55:00 The lecturer discusses how investors can specialize based on their characteristics. However, the conclusion that pension funds trade in riskier assets while hedge funds trade in less risky assets due to adverse selection is not entirely accurate. This model does not account for the explicit risk aversion of the agents and therefore does not fully explain why pension funds invest in less risky assets. Despite this, the idea of specialization based on characteristics is an interesting concept to explore in the context of financial markets microstructure.
Lecture 10, part 1: Value of Liquidity (Financial Markets Microstructure)
Lecture 10, part 1: Value of Liquidity (Financial Markets Microstructure)
  • 2020.04.08
  • www.youtube.com
Lecture 10, part 1: Value of LiquidityFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.y...
 

Lecture 10, part 2: Value of Liquidity (Financial Markets Microstructure)



Lecture 10, part 2: Value of Liquidity (Financial Markets Microstructure)

The lecture transitions to the topic of liquidity risk and its impact on asset returns, highlighting the fluctuation of liquidity over time and the resulting unpredictability in market correlations. The liquidity CAPM model is introduced as a tool to understand how liquidity affects the expected return of assets. The speaker emphasizes that only systematic risk influences the surprise mean, and this knowledge is applied to account for liquidity in the market.

The lecture then explores the effect of liquidity on the CAPM equation and how it alters beta in a new context. Specifically, the return on an asset (SJ) is calculated by subtracting the spread of SJ from the nominal return on the market (SG), while the risk-free rate remains unchanged. The beta coefficient is determined by the covariance between the nominal returns on a given asset (SHA) and the nominal market returns. The total beta consists of four individual betas, with beta 2 influenced by the covariance between the liquidity spreads of an asset and the overall market liquidity. Betas 3 and 4 are negatively correlated, implying that higher betas are advantageous for a secure asset.

The lecture emphasizes the value of liquidity in financial market microstructure, particularly in the Liquidity CAPM model, which quantifies the impact of liquidity on asset returns and highlights the role of beta coefficients in depicting an asset's sensitivity to market liquidity. Empirically, all beta coefficients have significance, but beta four makes the most significant contribution to returns as it primarily explains how investors consider hedging market returns with individual asset liquidity. The lecture concludes with a question regarding the existence of arbitrage opportunities in the market, citing the example of US Treasury bills versus bonds. The options presented include the notion that arbitrage opportunities may be costly to exploit due to market frictions, the possibility that the fundamental principles of economics and finance are incorrect, or the absence of arbitrage opportunities in the market.

The lecturer examines the concept of arbitrage, which involves capitalizing on price differences by buying and selling assets in different markets. While acknowledging the costs associated with arbitrage, such as collateral and financing expenses, the lecturer argues that these costs alone do not account for the absence of arbitrage opportunities. Arbitrageurs face the same trading costs as regular traders, including limited liquidity, spreads, and deviations from mid-quotes. Consequently, the argument that arbitrage opportunities do not exist solely due to the costs of arbitrage is insufficient. The lecturer asserts that the given examples demonstrate the nonexistence of arbitrage opportunities.

Introducing a new model by Duffy Colonel Patterson, the presenters attempt to simultaneously calculate the mid-price and spread using the cash flow approach in over-the-counter (OTC) markets. This model accounts for heterogeneous traders or investors with varying dividend valuations, assuming that they can hold zero or one unit of the asset and have an outside option of earning interest at the required rate of return. Additionally, the model assumes that the asset is supplied to a fraction of the population less than one-half, a crucial factor in the model's formulation.

The speaker discusses a liquidity model in financial markets where traders can assign different values to the asset. The model assumes a steady state with both high-value and low-value investors. Traders are subject to a Markov process, where the probability of a trader's value changing in a given period is denoted as SCI. In the steady state, the shares of high-value and low-value investors are equal, summing up to one. The speaker derives equations for the shares of high-value and low-value investors in the steady state, demonstrating their equality.

The video examines how changes in asset value generate trade in an economy. The assumption is made that high-value traders desire to hold the asset, while low-value traders do not. However, due to aggregate supply being less than one and less than one-half, it is

impossible for all agents to hold the asset in equilibrium. This generates a willingness to trade, as traders actively seek out dealers to exchange their assets. Dealers possess market power due to the difficulty of finding them, allowing them to quote different prices. The spread between bid and ask prices arises from the dealers' bargaining power parameter. Moreover, there are gains from trade within a given relationship, and the distribution of the surplus is determined by a parameter denoted as Z.

The lecturer introduces the values of "bar" and "B bar" as the highest possible ask price and the lowest possible bid price, respectively. The mid-price is defined as the center point between these values. The focus then shifts to the assumption that not all high-value investors can hold the asset in equilibrium due to the aggregate supply being less than half. In such a scenario, when a trader is willing to buy the asset and is quoted the ask price, they must be indifferent between buying and not buying. The probability of trading must equalize the trade flows in the market, ensuring that all dealers can clear their positions by the end of the period.

It is acknowledged that there is market power on the sell side due to fewer sellers, enabling them to make some profit. However, dealers also possess power in this interaction, resulting in the sell price falling between B bar and mu, with the bargaining parameter Z determining the split of the surplus between the seller and the dealer. A larger value of Z implies a smaller profit for the seller and a greater profit for the dealer. This relationship between profits is interpreted as market power.

The presenter discusses the use of value functions to determine the equilibrium of a model. The value functions, denoted as VG, represent the discounted lifetime utility of a trader who either owns or does not own the asset. The VG values are related to the maximum price that a trader would be willing to pay for the asset. However, these values are distinct from the bid and ask prices as they are linked to the trader's valuation and ownership of the asset. The presenter explains how the bid price can be computed using the value functions based on the trader's initial ownership of the asset and their valuation.

Next, the speaker delves into computing the value functions for both the buy and sell sides of the market. The buy side consists of high-value traders who aim to purchase the asset for its dividends, while the sell side comprises low-value traders who intend to sell their assets to receive dividend payouts. The lecture derives the lifetime utility functions for high-value traders who own the asset, encompassing scenarios where they receive a dividend and subsequently either remain a high-value trader or transition into a low-value trader seeking to sell the asset. This recursive formulation of values includes a normalization constant of 1 plus R.

The speaker highlights the significance of liquidity in financial market microstructure. They explain the process of determining the value of a high valuation investor who currently does not own the asset. This involves calculating the probabilities of the investor becoming a low-value trader and receiving dividends or remaining a high-value trader and participating in market trades. These probabilities are then employed to compute the values of the trader and the asset, subsequently influencing the ask and bid prices in the market. The ask price incorporates a discount due to the liquidity premium, which represents the cost of market frictions, including search costs for dealers. Overall, this section emphasizes how the value of liquidity impacts asset pricing in financial markets.

The speaker further discusses the spread and its correlation with the ask price in financial market microstructure. Reduced liquidity leads to a decreased evaluation of assets, necessitating a liquidity premium and increasing liquidity risk for investors. The lecturer recommends analyzing exercise one in chapter nine, specifically comparing zero coupon bonds and dividends, to further understand these concepts.

In exercise one of chapter nine, the lecturer prompts an analysis of zero coupon bonds and dividends to gain insights into liquidity and its impact on asset evaluation. Zero coupon bonds are financial instruments that do not pay regular interest or dividends but are sold at a discount to their face value. Dividends, on the other hand, refer to the periodic payments made by companies to their shareholders as a distribution of profits.

The exercise aims to examine the differences in liquidity and valuation between zero coupon bonds and dividends. Liquidity plays a crucial role in determining the ease with which an asset can be bought or sold without significantly impacting its price. Assets with higher liquidity tend to have lower bid-ask spreads, implying that they can be traded more easily and at a narrower price range.

When comparing zero coupon bonds and dividends, it is essential to consider their liquidity characteristics. Zero coupon bonds are typically traded in organized markets, such as bond markets, where their prices are determined based on market supply and demand. These bonds have a known future cash flow, making their valuation relatively straightforward. In contrast, dividends are distributed by individual companies, and their payment is contingent on the company's profitability and management decisions.

The liquidity premium associated with zero coupon bonds is typically lower compared to dividends. This is because zero coupon bonds have a predetermined maturity date and a known cash flow, which enhances their tradability. On the other hand, dividends are subject to various uncertainties, such as changes in company performance, dividend policies, and market conditions, which can impact their liquidity and valuation.

Investors, when evaluating zero coupon bonds and dividends, consider their respective liquidity risks. Liquidity risk refers to the potential for an asset's market liquidity to fluctuate, affecting its ease of trading and price volatility. Higher liquidity risk is associated with assets that are more difficult to buy or sell, leading to wider bid-ask spreads and potentially impacting their valuation.

Understanding the relationship between liquidity and asset valuation is crucial for investors and market participants. Liquidity considerations play a significant role in asset pricing models, such as the Liquidity CAPM model, which takes into account the effect of liquidity on expected returns and beta coefficients.

Analyzing exercise one in chapter nine involves examining the liquidity and valuation differences between zero coupon bonds and dividends. Liquidity, as a key factor in asset pricing, influences the ease of trading, bid-ask spreads, and the overall valuation of assets. By understanding these dynamics, investors can make informed decisions based on their risk tolerance, investment goals, and market conditions.

  • 00:00:00 The lecture shifts to discuss the concept of liquidity risk and how liquidity fluctuates over time leading to arbitrariness in correlation across the market. Liquidity risk factors can be priced in real markets and the lecture uses a liquidity CAPM model to explain how liquidity affects the expected return on any given asset. The lecture notes that only the systematic risk affects the surprise mean and how this applies to accounting for liquidity in the market.

  • 00:05:00 The speaker discusses the impact of liquidity on the CAPM equation and how it affects beta in the new world. The return on SJ is equal to the nominal return on SG minus the spread of SJ, with the risk-free rate remaining unchanged. The beta is equal to the covariance of real returns, which is based on the covariance between the nominal returns on SHA and nominal market returns. The total beta is made up of four individual betas, with beta 2 being affected by the covariance between the spreads of the liquidity of a given asset and liquidity of the market asset. Betas 3 and 4 enter with a negative sign, meaning that it is good to have high betas for a good safe asset.

  • 00:10:00 The focus is on the value of liquidity in financial markets microstructure, particularly on the Liquidity CAPM model, which estimates the effect of liquidity on asset returns and how beta coefficients play a role in illustrating an asset's sensitivity to market liquidity. Empirically, all beta coefficients matter, but the most significant contribution to returns is given by beta four, which mainly explains how investors pay attention to the issue of hedging market returns with individual asset liquidity. The lecture ends with a question about how arbitrage opportunities exist in the market, citing the example of US Treasury bills versus bonds, and the options include arbitrage opportunities being costly to exercise due to market frictions, that the fundamental law of economics and finance is wrong, or that there are no arbitrage opportunities in the market.

  • 00:15:00 The lecturer examines the concept of arbitrage, which is the practice of buying and selling assets in different markets to profit from price differences. While it is true that there are costs associated with arbitrage, such as collateral and financing costs, the lecturer argues that this is not the sole reason why arbitrage opportunities may not exist in certain situations. Arbitrageurs are subject to the same trading costs as regular traders, which include limited liquidity, spreads, and deviations from mid-quotes. Therefore, the argument that arbitrage opportunities do not exist solely due to the costs of arbitrage is not a good explanation, and the lecturer claims that no arbitrage opportunities exist in the examples given.

  • 00:20:00 The presenters introduce a new model by Duffy Colonel Patterson that attempts to calculate both the mid-price and spread simultaneously by focusing on the cash flow approach in OTC markets. In this market, traders or investors are heterogeneous, with different values placed on dividends, and there is a continuum of them. Traders can either hold zero or one unit of the asset, and they always have an outside option of going to the bank and earning interest, which will be the required rate of return on the asset. Furthermore, it is assumed that the asset is supplied to a fraction of the population less than one-half, which is a crucial factor for this model.

  • 00:25:00 The speaker discusses a model of liquidity in financial markets where traders can change the value assigned to the asset and where there is a steady state of high and low-value investors. The model assumes that traders are subject to a Markov process where, in any given period, the probability of a trader's value changing is SCI. The shares of high and low-value investors in the steady state are equal and they must sum up to one. The speaker derives equations for the shares of high and low-value investors in the steady state and shows that they are equal.

  • 00:30:00 The video discusses the concept of how changes in asset value can generate trade in an economy. The assumption is made that high-value traders will want to hold the asset while low-value traders will not, but with aggregate supply less than 1 and less than 1/2, it is impossible that all agents will hold the asset in equilibrium. This generates willingness to trade, where traders will search for a dealer with some probability to exchange their assets. The dealers have market power due to their difficulty in finding, and they can quote different prices, where the spread arises from their bargaining power parameter. Finally, there are gains from trade in a given relationship, and the split of the surplus occurs according to a parameter Z.

  • 00:35:00 The lecturer introduces the values of bar and B bar as the highest possible ask and the lowest possible bid, respectively. They define the mid-price as the center point between these values. The focus then shifts to the assumption that not all high-value investors can hold the asset in equilibrium, as the aggregate supply is less than half. In such a scenario, when a trader is willing to buy the asset and they're quoted the ask price, they must be indifferent between buying or not buying. The probability of trading must equalize the trade flows in the market, and all dealers at the end of the period should clear their positions.

  • 00:40:00 We know that there is market power on the sell side due to fewer sellers, which means they can make some profit. However, dealers also have power in this interaction, and the resulting sell price will be between B-bar and mu, with a bargaining parameter Z determining the split of the surplus between seller and dealer. The larger Z is, the smaller the seller's profit and the greater the dealer's profit. This relationship between profits is our interpretation of market power.

  • 00:45:00 The presenter discusses the use of value functions to determine the equilibrium of a model. The value functions, denoted as VG, represent the discounted lifetime utility of a trader who owns or does not own the asset in question. The VG values are related to the maximum price that a trader would be willing to pay for the asset, but these values are not the same as the bid and ask prices, as they are tied to the trader's valuation and ownership of the asset. The presenter goes on to explain how the bid price can be computed using the value functions based on the trader's initial ownership of the asset and their valuation.

  • 00:50:00 The speaker discusses how to compute the value functions for both the buy and sell sides of the market. The buy side is composed of high-value traders who want to purchase the asset for its dividends, while the sell side is made up of low-value traders who want to sell their assets for their dividend payouts. The speaker derives the lifetime utility functions for the high-value traders who own the asset, where they receive a dividend, followed by the possibility of either remaining a high-valued trader or becoming a low-valued trader who wants to sell the asset. This leads to a recursive formulation of values, with a normalization constant of 1 plus R.

  • 00:55:00 The speaker discusses the value of liquidity in financial markets microstructure. They explain the process of finding the value of a high valuation investor who currently does not own an asset, which involves determining probabilities of the investor becoming a low value trader and receiving dividends or remaining a high value trader and going to the market to trade. The probabilities are used to calculate the values of the trader and the asset, which are then used to find the ask and bid prices in the market. The ask price includes a discount due to the liquidity premium, which represents the cost of friction in the market due to search costs for dealers. Overall, the section highlights how the value of liquidity affects asset pricing in financial markets.

  • 01:00:00 The speaker discusses the spread and its correlation to the ask price in financial markets microstructure. Decreased liquidity can lead to a lowered evaluation of assets, requiring a liquidity premium and increasing liquidity risk for investors. The speaker advises an exercise to analyze exercise one in chapter nine, comparing zero coupon bonds and dividends.
Lecture 10, part 2: Value of Liquidity (Financial Markets Microstructure)
Lecture 10, part 2: Value of Liquidity (Financial Markets Microstructure)
  • 2020.04.08
  • www.youtube.com
Lecture 10, part 2: Value of LiquidityFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.y...
 

Lecture 11, part 1: Corporate Governance (Financial Markets Microstructure)



Lecture 11, part 1: Corporate Governance (Financial Markets Microstructure)

In this section of the lecture, the professor begins by reviewing the previous week's topic, which focused on the influence of liquidity on market valuation and the different approaches used to determine prices under conditions of limited liquidity. The importance of liquidity in financial markets is highlighted, particularly in terms of its impact on the cost of capital and the efficiency of transactions.

The lecture then transitions to the intersection of liquidity and corporate policy, examining how market liquidity and organizational factors affect corporate policies and the implications for corporate governance. The professor emphasizes the significance of liquidity for firms' access to capital in primary markets. Liquidity plays a crucial role in funding initiatives, attracting investors, and facilitating transitions in ownership throughout a firm's lifecycle. The lecture includes a graph illustrating the various funding sources available to firms at different stages of their growth, with early-stage projects being funded by business angels and venture capitalists paving the way for initial public offerings (IPOs).

To illustrate the impact of liquidity on ownership transition, the lecturer shares the story of the social network Tumblr. Verizon's decision to ban all forms of pornography on the platform resulted in a significant loss of users, prompting Verizon to seek another buyer. Potential bids, such as one from Pornhub, did not materialize, and eventually, Tumblr was acquired by Automattic, the company responsible for WordPress. This real-world example highlights the influence of market liquidity on ownership changes and the subsequent impact on corporate policies.

The lecture then delves into the process of initial public offerings (IPOs). When a company decides to go public, it engages an investment bank to act as an underwriter. The investment bank approaches potential investors, asking them to submit limit orders to buy stock at a certain price. The investment bank aggregates these orders into a book and continues the process of bookbuilding until the IPO price is set, and shares are sold to investors. The concept of underpricing in IPOs is also explained, with the lecturer noting that illiquid assets tend to exhibit a more pronounced underpricing effect compared to liquid assets, as supported by empirical evidence.

Next, the lecture explores the links between financial markets and corporate governance. One issue raised is the potential misalignment of incentives between company owners and managers, particularly when ownership and control are separated. This divergence can create a wedge between the goals of owners and the actions of managers. Compensation schemes are discussed as a means to alleviate this misalignment, but ultimately, owners must be willing to intervene and replace managers if necessary. However, questions arise regarding whether shareholders prioritize long-term profitability over short-term gains and if they are genuinely committed to improving company governance. The lecture emphasizes the importance of shareholders' role in influencing corporate governance and its impact on overall company value.

The problem of corporate governance is traced back to the 1930s, where it was recognized that shareholders may not always act in the best interests of the company, leading to a decrease in its value. In widely-held corporations with numerous small shareholders, there may be a lack of responsibility for company performance and management, resulting in imperfect decision-making and governance. The lecture suggests that concentrated ownership with a majority investor who is committed to improving governance could be a potential solution. Additionally, it is noted that in illiquid markets, it is less attractive for activists to buy shares but more beneficial for corporate activism due to the difficulty of selling shares. The goal is to create asymmetric liquidity, making it easy to enter the company but challenging to sell shares, thereby promoting corporate activism.

The lecture also explores the regulation of buying and selling stocks in relation to corporate governance in financial markets. Laws require investors with significant holdings in a company to disclose their buying and selling activities to ensure transparency. However, this creates a situation where informed investors are less likely to sell their stock, although they may face adverse market reactions. The relationship between company managers and the market in terms of information is discussed, highlighting the mechanism by which companies can extract market information to inform managerial decisions by observing market reactions. However, both the lecturer and the chat participants agree that the market rarely possesses better information than the firm due to the greater access to internal indicators such as sales, revenue, and margins.

In conclusion, this section of the lecture covered the interplay between liquidity and market valuation, the impact of liquidity on corporate policies and governance, the process of IPOs and underpricing, and the relationship between financial markets and corporate governance. The lecture emphasized the importance of liquidity in primary markets, the role of market liquidity in funding and ownership transitions, and the challenges and implications of corporate governance in different market conditions. The overall goal was to provide insights into how firms' actions can affect secondary markets and why firms care about what happens in those markets.

  • 00:00:00 The professor reviews the previous week's topic of how liquidity affects market valuation and the various approaches used to determine prices given limited liquidity conditions. The focus then shifts to the intersection of liquidity and corporate policy, specifically how market liquidity and organization affect corporate policies and the implications for corporate governance. The lecture will also cover the topic of digital markets.

  • 00:05:00 The first half discussing the connection between liquidity and firms' access to capital in primary markets, while the latter part focuses on how the digital revolution has transformed financial markets and a brief overview of the course to date, touching on topics including blockchain and cryptocurrency. The section also explains why the course has primarily focused on secondary markets, and the challenges in generalizing primary markets due to the wide range of different market formats. Overall, the section aims to break down independence and explain why firms care about what happens in secondary markets, and how their actions can affect the market.

  • 00:10:00 The lecturer emphasizes the importance of market liquidity in terms of the cost of capital and efficiency of transactions. He explains how the more liquid the markets, the easier it is to fund initiatives and attract investors, and how liquidity helps firms progress through their lifecycle and transition ownership. A graph shows the different funding sources available to firms at different stages of their growth, with business angels funding early-stage projects and venture capitalists making way for IPOs. The lecturer also shares a story about the social network Tumblr and how market liquidity played a role in its ownership transition.

  • 00:15:00 The lecturer discusses the story of Tumblr and how Verizon's decision to ban all kinds of pornography on the platform resulted in a loss of almost a third of its users, prompting Verizon to look for another buyer. One bid came from Pornhub, but it never materialized. Eventually, Tumblr was acquired by Automattic, the company responsible for WordPress. The lecture then delves into how IPOs work, with a company going to an investment bank who acts as an underwriter, and approaches potential investors to buy stocks of the company. The investment bank forms a limit or a book, and the process of bookbuilding continues until the IPO price is set, and the shares are sold to investors.

  • 00:20:00 The concept of book building in IPOs is explained. The investment bank approaches various investors and asks them to submit a limit order to buy stock at a certain price, with the buy side being filled by the investment bank. This establishes a uniform price for shares at the start of trading, but typically this price is lower than what is established later in the day. This underpricing effect is more pronounced for illiquid assets compared to liquid ones, as shown by empirical evidence.

  • 00:25:00 The lecturer discusses the possible links between financial markets and corporate governance. One issue is the potential misalignment of incentives between company owners and managers. This can occur when ownership and control are separated, leading to a wedge between what owners want and what managers do. Compensation schemes can help alleviate this, but ultimately, owners must be willing to intervene and replace managers if necessary. However, there is a question of whether shareholders are willing to prioritize long-term profit over short-term gains and if they are truly committed to improving company governance. If investors are purely speculative, then managers are essentially out of control, which can be detrimental to the company's overall value.

  • 00:30:00 The speaker discusses how the problem of corporate governance has been identified since 1930, where shareholders do not work in the best interests of the company, leading to a decrease in company value. For widely-held corporations with many small shareholders, there is no sense of responsibility for company performance or management, making decision-making and governance imperfect. Concentrated ownership with a majority investor who is willing to improve governance may be needed. In illiquid markets, it is less attractive for activists to buy shares, but it is also more difficult to sell shares, making it more beneficial for corporate activism. The goal is to make liquidity asymmetric by making it easy to enter the company, but difficult to sell shares in order to promote corporate activism.

  • 00:35:00 The speaker discusses the regulation of buying and selling stocks in relation to corporate governance in financial markets. Laws dictate that investors who hold a large interest in a company are required to disclose their buying and selling activities, which allows them to accumulate a large portion of the company's shares without transparency. This creates a situation where informed investors are less likely to sell their stock, but they are also likely to face adverse market reactions. Additionally, the speaker explores the relationship between company managers and the market in terms of information. The book outlines a mechanism where the company can extract market information to inform managerial decisions by observing the market's reaction to a decision. However, the speaker and the chat participants agree that the market rarely knows better than the firm due to the greater information at the firm's disposal.

  • 00:40:00 The lecturer discusses the information advantage of firms over the market, except for the firm's competitors. The firm has access to internal indicators such as their sales, revenue, and margins. However, analyzing all the data can be costly, so the firm relies on the market to analyze and make decisions. The stock price is an indicator of a company's performance, and companies care about their stock prices since it can impact the CEO's job. The lecture also mentions a case study of the American space shuttle Challenger crash, where the market was quicker to determine which supplier was guilty than NASA, which took four months to investigate.
Lecture 11, part 1: Corporate Governance (Financial Markets Microstructure)
Lecture 11, part 1: Corporate Governance (Financial Markets Microstructure)
  • 2020.04.15
  • www.youtube.com
Lecture 11, part 1: Corporate GovernanceFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www...
 

Lecture 11, part 2: Digital Markets (Financial Markets Microstructure)



Lecture 11, part 2: Digital Markets (Financial Markets Microstructure)

In this section of the lecture, the professor discusses the concept of managerial compensation schemes as a means to alleviate incentive problems between company owners and managers. The ideal scheme is one that rewards managers for doing the right thing and punishes them for doing the wrong thing, while being cost-effective for shareholders. However, evaluating managerial performance and incentivizing managers can be challenging.

To illustrate the concept, the lecturer presents a quick and simple model where a manager's effort affects the probability of a good outcome, and exerting effort comes at a cost. In an ideal world, the best contract would be to pay the manager a salary based on their effort, with zero pay if they didn't exert effort. However, in reality, effort is not always contractible, meaning it cannot be perfectly observed or measured. As a result, the manager's compensation can be made contingent on company value or realized profits.

The lecturer explains that the optimal contract for a manager in a company where effort is not contractible is one that is contingent on the stock price. This is because the stock price is more sensitive to the manager's effort and, therefore, cheaper for shareholders. Such a contract pays the manager nothing if the company fails, but offers high payment if it performs well, aligning with the concept of the first-best contract.

However, the lecturer acknowledges that there can be unintended consequences of tying manager compensation to stock price. One such consequence is the problem of career concerns, where managers may prioritize maximizing their reputation rather than making decisions that are in the best long-term interest of the company. This behavior can lead to various inefficiencies.

To address this issue, the lecturer suggests that if a company cares about its stock price, it may be willing to improve the liquidity of its stocks. Higher liquidity makes the stocks more valuable, and this increased value can indirectly incentivize the manager. The lecturer presents three instruments that companies have in affecting liquidity: conducting an initial public offering (IPO), listing on another exchange, and improving transparency and financial reporting.

Listing on an exchange, though it comes with transparency requirements, can increase the accessibility of a company's stocks. Additionally, hiring a dedicated market maker who posts relatively aggressive limit orders can improve liquidity. Furthermore, companies can choose a capital structure that is optimal for liquidity, depending on the liquidity levels of their assets.

The lecture concludes by mentioning that corporate finance is a field that explores primary capital markets in more detail and can provide further insights for those interested in studying this topic.

This section of the lecture focused on the concept of managerial compensation schemes to address incentive problems between company owners and managers. The lecturer explained the challenges of evaluating managerial performance and presented the idea of contingent compensation based on company value or stock price. The potential drawbacks of this approach were discussed, along with the role of liquidity in indirectly incentivizing managers. The lecture also highlighted the importance of understanding primary capital markets and corporate finance for a comprehensive understanding of these concepts.

  • 00:00:00 The lecturer discusses the concept of managerial compensation schemes to alleviate incentive problems between company owners and managers. The ideal scheme would reward managers for doing the right thing and punish them for doing the bad thing, while being cheap for shareholders. The challenge lies in evaluating managerial performance to incentivize the manager. The lecturer creates a quick and simple model where a manager's effort affects the probability of a good outcome, and exerting effort is costly. He suggests that the salary scheme can be contingent on company value or stock price, which is visible to everyone and can, therefore, indirectly incentivize the manager.

  • 00:05:00 The lecturer discusses the optimal contract for a manager in a company where effort is not contractible. They start by explaining that the best contract would be to pay the manager a salary based on their effort, with zero pay if they didn't exert effort. However, in the real world, effort is not always contractible, so the manager's compensation can be made contingent on company value or realized profits. The optimal contract is one that is contingent on the stock price, as it is more sensitive to the manager's effort and therefore cheaper for shareholders. This contract pays the manager nothing if the company fails, and high payment if it performs well, making it consistent with the first best contract.

  • 00:10:00 The lecturer discusses the issue of time manager compensation to stock price and how it may not be a perfect solution to the problem as it can lead to unintended consequences. One of these consequences is the problem of career concerns where managers will try to maximize their reputation and this might lead to various inefficiencies. However, if the company cares about its stock price, it may be willing to improve the liquidity of its stocks as higher liquidity means that the stocks are more valuable. The lecturer offers three instruments that the firm has in affecting this liquidity, including doing an IPO, listing on another exchange, and improving transparency and financial reporting.

  • 00:15:00 The speaker discusses ways in which companies can increase their liquidity through financial markets microstructure. One option is for a company to list on an exchange, although this comes with transparency requirements. Another option is for the company to hire a dedicated market maker who will post relatively aggressive limit orders to improve liquidity. The company may also increase its overall liquidity by choosing the capital structure which is optimal for liquidity, depending on the different liquidity levels of its assets. Finally, the speaker notes that corporate finance is a field that explores the primary capital markets in more detail and may be useful for those interested in learning more about this topic.

  • 00:20:00 The video discusses how the digitization and computerization of everything have transformed financial markets. The lecture refers to the doubling of the US stock market capitalization every decade and doubling of the trading volume in the Dow Jones Industrial Average every seven and a half years from 1929 to 2009, and how in the most recent decade, this pace has accelerated. The lecture compares this progression to Moore's law and explains how the computerization of financial markets has allowed for more trading. However, this has also led to more integration and interdependence in the markets, which can result in failures being more significant. The lecture concludes by discussing how this information will be used to evaluate the different factors that determine how the trading proceeds in the markets.

  • 00:25:00 The lecturer discusses how digitization has impacted different factors in financial markets through a series of quizzes. The first quiz focuses on trading costs, which have decreased significantly due to market digitization. The second quiz relates to investors' risk aversion, which has decreased to some extent because they can diversify risk more easily. However, investors' risk aversion may also have become more stringent due to access to a wider range of assets. Finally, the lecture emphasizes how algorithms have transformed the way markets are organized, allowing for automatic order matching and reducing the role of dealers.

  • 00:30:00 The lecturer discusses the impact of market digitization on market fragmentation and transparency. Digitization has reduced the significance of some original markets as distance is now less of a factor and consolidation is possible. Furthermore, the role of fragmentation has greatly diminished as traders can now easily access information from multiple exchanges. However, the impact on market transparency is more complex as accessing information is now easier than ever, but physical trading floors still offer advantages in terms of visibility. Overall, while digitization has changed the dynamics of financial markets, the effects are not easily categorized as entirely positive or negative.

  • 00:35:00 The video discusses how digital markets have increased the spectrum of possible ways to decrease transparency and make trading more anonymous. Additionally, the decreased latency in trading has created new trading strategies and approaches to trading, such as arbitrage, which has fostered price discovery and balanced prices across markets. However, the heterogeneity of latency across investors has become an issue in high-frequency trading. The video also addresses the upcoming exam and mentions that it will likely consist of problems that deal with models and essay questions but will not include multiple-choice questions.

  • 00:40:00 The professor discusses the concept of algorithmic trading, which allows traders to use algorithms to automatically execute trades on their behalf based on certain criteria, such as prices or portfolio value. This has greatly improved market liquidity and made traders more willing to take risks, but also made the market more fragile as the similar algorithmic orders can lead to abrupt market crashes. The professor also mentions that the next lecture will be devoted to high frequency trading, and whether it brings any benefits or harms to the society.

  • 00:45:00 The speaker talks about the potential for significant market impact due to algorithmic trading. He mentions a survey that chronicles the history of algorithmic trading and the various screw-ups that have occurred as a result of it. Although the speaker opts not to assign mandatory readings on digital markets in this course, he encourages students to read academic journals and consider taking the digital economics seminar to gain a better understanding of the impacts of digitization in markets. Finally, the speaker touches on the topic of blockchain and cryptocurrencies, citing the Bitcoin price chart and the problems associated with these concepts.

  • 00:50:00 The lecturer explains the concept of blockchain technology and its relation to cryptocurrencies, which function as distributed payment systems. The unique feature of blockchain technology is that it allows for transactions to be recorded and verified in a decentralized manner without the need for intermediaries like exchanges or trading platforms, which can improve the efficiency of the market by reducing transaction costs. While this technology has potential applications in financial markets, there are challenges related to accurately determining who possesses stocks and shares due to transactions occurring outside of the platform.

  • 00:55:00 The concept of transparency in blockchain technology is examined. While the history of transactions is available to everyone, it is not truly anonymous. Additionally, the limited processing capacity of blockchain, particularly Bitcoin, is considered a major drawback. However, the use of smart contracts, as an advanced version of algorithmic trading, is seen as a significant benefit. The potential to use smart contracts in forward or future contracts could potentially alleviate counterparty risks.

  • 01:00:00 The lecturer discusses the drawbacks and challenges of using blockchain technology for financial transactions, particularly in terms of the competition and execution risk that arises from bidding for transactions. Along with the order costs, this bidding process creates an extra layer of uncertainty and delay, as transactions can be outbid and delayed in being included in the next block. This, along with the volatility and unpredictability of transaction costs, makes incorporating blockchain technology into decision-making for financial transactions somewhat risky and uncertain. The lecturer concludes that although blockchain technology has great potential, it also has significant drawbacks that need to be addressed before it can be used more widely.

  • 01:05:00 The lecturer talks about the drawbacks of Bitcoin blockchain due to the lack of intermediaries in trading, which can lead to increased counterparty risk. The establishment of exchanges and trusted intermediaries in financial markets can help to absorb this risk and provide financial transparency by enforcing disclosure of financial reports. While OTC markets can be accessed without intermediaries, they may have weaker transparency requirements and can be less liquid as a result. Overall, the discussion highlights the importance of intermediaries in financial markets.
Lecture 11, part 2: Digital Markets (Financial Markets Microstructure)
Lecture 11, part 2: Digital Markets (Financial Markets Microstructure)
  • 2020.04.15
  • www.youtube.com
Lecture 11, part 2: Digital MarketsFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://www.yout...
 

Lecture 12, part 1: High-Frequency and Algorithmic Trading (Financial Markets Microstructure)



Lecture 12, part 1: High-Frequency and Algorithmic Trading (Financial Markets Microstructure)

The lecturer begins the session by summarizing the previous week's topics, highlighting the relationship between liquidity and corporate governance, as well as the transformative impact of digital markets on trading. They briefly mention cryptocurrency and blockchain, cautioning that these technologies may have been excessively advertised. The lecturer then moves on to the main focus of the day: high-frequency trading. However, before delving into the subject, they discuss a recent event involving crude oil futures contracts trading at negative prices. The audience is presented with a quiz, asking them to consider whether this anomaly was caused by algorithmic trading or strategic human traders. Ultimately, the lecturer reveals that the contracts were indeed traded at negative prices, ruling out algorithmic failure or a mere joke as the cause.

Next, the speaker dives into two interconnected topics. First, they discuss a predictable trading pattern in the commodity market involving the US oil fund and the subsequent negative prices caused by traders anticipating and capitalizing on the rollover of these contracts. The second topic explored is algorithmic trading, which extends beyond high-frequency and professional traders to include institutional and retail traders who employ algorithms for more efficient order execution. The lecturer refers to a paper by Beeson and Warhol that investigates the various applications of algorithmic trading.

Building on this, the speaker introduces another research paper that examines how algorithmic trading impacts the modeling of uninformed traders in modern markets. The paper analyzes data from a brokerage company that employs widely-used algorithms to execute trades. The algorithms split parent orders, submitted by institutional investors, into child orders to minimize price impact. The data reveals that, on average, each parent order generates 63 runs, with 3-9 children per run, resulting in over 500 child orders per parent order. This data highlights the sophistication of uninformed traders and suggests that models may need to be adjusted accordingly.

The lecturer further emphasizes the increasing sophistication of traders and the practice of splitting market orders into child orders to minimize market impact. They present a thought-provoking question to the audience, asking them to guess the composition of market orders and limit orders for retail investors versus institutional investors. The reveal shows that institutional investors heavily rely on limit orders, with 80% of their orders being limit orders, while less than 0.4% are market orders. The concept of bag orders, tied to market prices, is introduced to further illustrate this aspect of trading.

The concept of marketable limit orders is then explained as a safer alternative to market orders. Marketable limit orders are submitted at prices within the bid-ask spread, allowing for immediate execution. In contrast, traditional limit orders are passive and placed at prices outside the bid-ask spread, anticipating execution at a later time. The advantage of marketable limit orders lies in their reduced susceptibility to sudden price changes and delays, as they are executed promptly at the best available price. However, there are instances where marketable limit orders may go unfilled due to specific volume or price restrictions set by the trader.

The speaker elaborates on the idea that even unfilled limit orders can have an impact on the market. They discuss a research paper that demonstrates how cancelled orders, both filled and unfilled, can influence market prices. Unfilled orders, in particular, have a more substantial impact than filled orders, and this impact occurs within seconds, emphasizing the speed of today's market. The lecture then transitions to the main topic of high-frequency trading, underscoring the importance of reading research papers and providing guidance on how to approach them effectively. The speaker emphasizes the significance of understanding the drawbacks associated with the assumptions made in these models.

The lecturer proceeds to discuss high-frequency and algorithmic trading (HFT) in financial market microstructure. HFT refers to the computerized execution of trading strategies at a rapid pace and has become prevalent in modern markets. They mention that HFT accounts for over 50% of trading volume in the US and more than 25% in Europe, but there is still uncertainty within the scientific community regarding its effects on the market and whether it requires regulation. To shed light on these questions, the lecture explores theoretical papers that investigate the advantages and investments associated with gaining speed in HFT. While earlier models focused on informed traders, recent research has examined the use of HFT by uninformed traders.

To illustrate the advantages of speed in trading, the speaker introduces a simple two-period model where profit-maximizing institutions, categorized as having either high or low private values, engage in trading. These traders observe their private values before trading and combine approaches based on their previous encounters with heterogeneous valuations. A fundamental value, which can be high or low with equal probability, is also introduced. Fast institutions invest, while slow institutions remain slow, with the former gaining an advantage by submitting orders earlier and acquiring more knowledge and information from the market during the interim.

The lecturer explains how high-frequency trading provides advantages in identifying profitable trading opportunities. Fast traders are able to observe the fundamental value (V) at the time of order submission, whereas slow traders may not observe V until after submitting their orders. Furthermore, fast traders have a higher probability of discovering lucrative trading opportunities because they have more visibility into the limit order book if they delay order submission. The speaker delves into the various types of private information that both fast and slow traders may possess and how their behavior is influenced by this information within an equilibrium framework.

The professor discusses a model for trading, highlighting the distinction between traders who have knowledge of the asset's value and those who do not. The traders also possess a private valuation element that affects the trading behavior of uninformed traders. The model draws a parallel to the Gloucester Milgram model and can be solved using similar methods. In scenarios where only slow traders are present, all orders execute at the mid-quote. However, when both fast and slow traders participate in the market, the lecturer focuses on the most extreme trader types. In a symmetric equilibrium, fast traders with a high private valuation buy the asset, while those with a low valuation and knowledge of bad news sell it, forming six distinct strategies.

The speaker proceeds to discuss the computation of equilibrium prices for the buy-side. By calculating the probabilities of receiving buy orders from fast and uninformed traders, equivalent to the informed traders in their model, the equilibrium price for buy orders can be derived. The ask price, quoted by the dealer, is determined by the conditional valuation of the asset upon receiving a buy order. The section concludes with the lecturer posing questions regarding trader behavior and announcing a break in the lecture.

After the break, the lecture resumes with a discussion on the impact of high-frequency trading (HFT) on market outcomes. The speaker presents another research paper that explores the effects of HFT on market liquidity and price efficiency. The paper examines how the presence of HFT traders, who have access to faster information and execution capabilities, influences market dynamics.

The lecturer introduces a model that incorporates HFT traders alongside other market participants. They explain that HFT traders are characterized by their ability to observe the fundamental value of the asset before submitting their orders. In contrast, non-HFT traders, referred to as "regular traders," are unable to observe the fundamental value and make decisions based on their private valuations and available market information.

The lecture delves into the equilibrium analysis of the model, considering both the behavior of HFT traders and regular traders. The speaker highlights the importance of understanding the strategic interactions between these different types of traders and how they impact market outcomes. They emphasize that HFT traders' ability to access information faster and make quicker trading decisions can significantly affect market liquidity and price efficiency.

The lecturer presents key findings from the research paper, highlighting that the presence of HFT traders can lead to improved price efficiency and narrower bid-ask spreads in the market. The increased trading activity and faster information processing by HFT traders contribute to enhanced liquidity and the more rapid incorporation of new information into prices.

However, the speaker also notes potential concerns related to HFT, such as the possibility of increased market volatility and the potential for HFT strategies to amplify market movements. They stress the importance of further research to better understand these dynamics and assess whether regulatory measures are necessary to mitigate any negative consequences associated with HFT.

The lecture concludes by summarizing the main points discussed, including the advantages and potential drawbacks of high-frequency trading. The speaker encourages the audience to continue exploring research papers and academic literature on the topic to gain a deeper understanding of the complex dynamics at play in modern financial markets. They emphasize the importance of staying informed and critically analyzing the implications of different trading strategies and technologies for market functioning and stability.

  • 00:00:00 The lecturer begins by summarizing last week's topics on how liquidity interacts with corporate governance and how digital markets have transformed trading. He briefly touches on cryptocurrency and blockchain, noting that although they have their uses, they may have been over-advertised. He then moves on to the main topic of the day, high-frequency trading, but first discusses a recent event where crude oil futures one month forward contracts were traded at negative prices. He quizzes the audience on how this could have happened and suggests thinking about whether it was due to algorithmic trading or strategic human traders. Ultimately, he reveals that contracts were actually traded at that price and it was not an algorithmic failure or a joke.

  • 00:05:00 The speaker discusses two topics related to trading. The first is a recent event in the commodity market where the US oil fund had a predictable trade happening every month in oil futures, and many traders ran ahead of these contracts, profiting off the rollover but causing negative prices for some. The second topic is algorithmic trading, where algorithms are not only used by high-frequency traders and professional for-profit traders but also by larger institutional and retail traders for better execution of orders. The speaker presents a paper by Beeson and Warhol that explores the uses of algorithmic trading.

  • 00:10:00 The speaker discusses a paper on algorithmic trading and how it affects the modeling of uninformed traders in modern markets. The paper analyzes data from a brokerage company that executed trades through widely-used algorithms. Institutional investors submit parent orders, which the algorithms split into child orders to minimize price impact. On average, each parent order produces 63 runs, with about 3-9 children per run, totaling over 500 child orders per parent order. The paper's value lays in showing that uninformed traders are still sophisticated and trade in advanced ways, indicating that models may need to be adjusted.

  • 00:15:00 The speaker discusses the increasing sophistication of traders and the splitting of market orders into child orders to minimize market impact. The speaker also questions the data on run times for orders, but encourages viewers to read the paper for further insight. They then ask the audience to guess the composition of market orders and limit orders for retail investors versus institutional investors, and reveal that institutional investors' orders are 80% limit orders and less than 0.4% are market orders. They also explain the concept of bag orders and how they are tied to market prices.

  • 00:20:00 The concept of marketable limit orders is introduced, which are orders that are submitted at prices inside the bid-ask spread, creating an overlap, and are executed immediately. These orders are distinct from traditional limit orders, which are passive and are submitted at prices outside the bid-ask spread, with the expectation of being executed at some point in the future. The advantage of marketable limit orders over market orders is that they are less likely to be subject to sudden price changes and delays due to market makers holding them, as marketable limit orders are executed right away at the best possible price. However, there are still instances where marketable limit orders may not get executed, such as if the trader has set specific volume or price restrictions.

  • 00:25:00 The speaker explains the concept of limit orders as a safer alternative to market orders, providing insurance against sudden price shocks by giving traders an upper bound on the prices they are willing to execute. However, even marketable limit orders bear the risk of going unfilled, incurring some execution risk. The speaker provides a glimpse of how traders' private information may impact the price of the assets and introduces the idea that limit orders could be informative and reveal traders' private information about the fundamental value of the asset. Even unfilled orders could have some price impact, indicating the trader's information about the asset's fundamental value.

  • 00:30:00 The speaker discusses a research paper on the impact of cancelled orders on the market price. The findings show that even cancelled orders can have a price impact, with unfilled orders having a larger impact than filled ones. This impact was on the scale of seconds, which highlights the speed of today's market. The lecture then transitions to high-frequency trading and emphasizes the importance of reading research papers, starting with the abstract and introduction, browsing the contents, and focusing on the model setup and results. The speaker notes that authors may provide insight on the drawbacks of their assumptions.

  • 00:35:00 The lecturer discusses high-frequency and algorithmic trading (HFT) in financial markets microstructure. HFT refers to computerized algorithmic trading at a very high pace, which has become a prevalent practice in today's markets. The lecturer notes that HFT is estimated to account for over 50% of all trading volume in the US and more than 25% in Europe. However, the scientific community is still unsure about the effect of HFT on the market, and whether it necessitates regulation. The lecture explores a couple of theoretical papers that investigate this question, noting that while most models have focused on informed traders, recent work has considered the use of HFT by uninformed traders, with the aim of understanding the core concepts of its advantages and investments associated with gaining speed.

  • 00:40:00 The speaker introduces a simple two-period model with a binary version of profit maximizing institutions that have either a high or low private value and observe their private value before they trade. The traders also combine both approaches to heterogeneous valuations seen before, with a fundamental value that can be high or low, occurring with equal probability. Fast institutions are those who invest, while slow institutions remain slow. Speed gives fast institutions an advantage because they can submit orders earlier and learn the true value of the asset before other traders, resulting in more knowledge and information from the market in the meantime.

  • 00:45:00 The speaker discusses how high-frequency trading can provide advantages in finding profitable trading opportunities. Fast traders have the advantage of seeing the fundamental value (V) when they submit their orders, while slow traders may not observe V until after they submit their orders. Fast traders also have a higher probability of finding lucrative trading opportunities because they get to see more of the limit order book if they submit their order at a later time. The speaker then goes on to describe the various types of private information that both fast and slow traders might have, and how they might behave given this information in an equilibrium.

  • 00:50:00 The professor discusses a model for trading where some traders know the value of the asset they are trading while others do not. The traders also have a private valuation element which makes trade of the uninformed traders easier to understand. The model reminds the professor of the Gloucester Milgram model, and it can be solved similarly. If there are no fast traders, all orders will execute at the mid-quote, but if there are both fast and slow traders in the market, the professor looks at the most extreme types of traders. In a symmetric equilibrium, fast traders with a high private valuation will be buying the asset, while those with a low valuation and knowledge of bad news will be selling it, out of the six strategies.

  • 00:55:00 The speaker discusses how, after characterizing the equilibrium and determining optimal strategies, given a set of betas, the equilibrium price can be computed for the buy-side. This is done by calculating the probability of receiving a buy order from a fast trader, equivalent to the inform trader in their model, and the probability of receiving a buy order from an uninformed trader. The ask price, the price that the dealer will quote, is derived from the conditional valuation of the asset on receiving a buy order. The section ends with the speaker posing the question of how the traders will behave and taking a break.
Lecture 12, part 1: High-Frequency and Algorithmic Trading (Financial Markets Microstructure)
Lecture 12, part 1: High-Frequency and Algorithmic Trading (Financial Markets Microstructure)
  • 2020.04.22
  • www.youtube.com
Lecture 12, part 1: High-Frequency and Algorithmic TradingFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course pla...
 

Lecture 12, part 2: High-Frequency Trading (Financial Markets Microstructure)



Lecture 12, part 2: High-Frequency Trading (Financial Markets Microstructure)

Continuing after the break, the lecture focuses on the equilibrium analysis of a high-frequency trading model and explores the existence of multiple equilibria, which can arise due to self-fulfilling expectations in the market. The speaker explains that the pricing strategy for the dealer is formulated based on the remaining strategies employed by traders in the market.

To address the issue of multiple equilibria, the lecturer introduces the assumption that fundamentals play a more significant role than private valuations, although they do not completely overshadow them. Traders in the market rank the values of the asset based on their private valuations and news, which provides a narrower set of possible cases and helps guide their decision-making.

The lecture proceeds to discuss three distinct equilibria, labeled P1, P2, and P3, under specific conditions. In equilibrium P1, all three types of traders participate by buying the asset at a narrow spread. In P2, fast traders only buy if they have good news and high private valuations, while slow traders still engage in buying. P3 represents an equilibrium where only fast traders with extreme valuations participate, leading to a wider spread and effectively excluding slow traders from the market.

The speaker emphasizes that the existence of these equilibria depends on various parameter values, including the possibility of a spread becoming so wide that no trades occur in the market. The lecture highlights that while P3 always exists, P1's existence is contingent on a specific threshold of informed traders being present. It is found that P1 is Pareto dominant, providing better prices for all traders compared to P3. Consequently, uninformed traders no longer trade at a loss in this model, making the trading process more strategic and beneficial for all participants.

The professor further explores the implications of the P1 equilibrium on the profits of fast and slow traders. The profits of fast traders decrease as more fast competitors enter the market, indicating a negative impact of increased competition. Similarly, slow traders experience a similar outcome, but their profits depend on their private valuations. The lecture highlights that when the equity point crosses zero, the P1 equilibrium ceases to exist, resulting in a worse outcome for all market participants as it imposes an externality on others. Overall, profits for all traders decline as the alpha value increases.

The lecture introduces a more nuanced solution to the tragedy of the commons by considering the heterogeneity among institutions. The model assumes that institutions have different types, which determine their size and potential profits from being fast. This implies that not all traders necessarily become fast or slow, but rather it depends on the size of their institution and the number of markets in which they can participate.

The speaker delves into the decision-making process of institutions in choosing to become fast or slow, driven by the expected profit from being fast. They explain that the profit from being fast is the same across all markets and depends solely on the total share of fast institutions. Only institutions surpassing a certain cutoff in terms of type will opt to become fast. The lecture further discusses how, based on the assumed distribution, the distribution of trader types faced in any given market follows a uniform distribution from 0 to M. Additionally, the alpha value, representing the probability of informed trading in each market, is established.

The lecture refers to the findings of a research paper on high-frequency trading, which identifies an equilibrium where the probability of encountering a trader large enough to make it worthwhile to become fast is determined by the uniform distribution. The paper also reveals that the cost of becoming fast leads to fewer fast traders in the market, thereby decreasing alpha. Furthermore, the authors present a welfare result suggesting that markets without adverse selection generate more welfare compared to markets with adverse selection. The speaker interprets this as an indication that well-functioning markets may have an excessive amount of high-frequency trading in equilibrium and proposes that setting alpha to zero would be welfare-maximizing.

Towards the end of the lecture, the presenter mentions a proposal to conduct batch auctions every 0.1 seconds, which would not significantly delay traders but could potentially have adverse effects on high-frequency traders. They announce that the upcoming lecture will delve into this proposal in greater detail and provide empirical data to support it. The presenter acknowledges any confusion caused by the presentation and expresses gratitude to the audience for their attentiveness, concluding by announcing that the exercise class will take place on Friday.

Continuing with the lecture, the presenter moves on to discuss the proposed batch auction system in more detail. They explain that batch auctions involve grouping a set of orders together and executing them at a specific time interval, such as every 0.1 seconds. While this system may not cause significant delays for most traders, it could potentially disrupt the strategies and profitability of high-frequency traders.

The presenter acknowledges that high-frequency trading has become a controversial topic, with concerns about its impact on market stability and fairness. Batch auctions are seen as a potential solution to address some of these concerns by introducing a more structured and controlled trading environment.

The lecture then introduces the concept of empirical data, which will be presented in subsequent sessions to support the feasibility and effectiveness of the proposed batch auction system. The presenter emphasizes the importance of empirical evidence in understanding the real-world implications of market structures and trading strategies.

Apologizing again for any confusion caused during the lecture, the presenter expresses gratitude to the audience for their patience and engagement. They conclude the session by announcing that the exercise class, where students can further practice and apply the concepts discussed, will be held on Friday.

  • 00:00:00 The lecturer discusses the equilibrium of a high-frequency trading model and how markets can have multiple equilibria due to self-fulfilling expectations. The pricing strategy for the dealer is devised based on the remaining strategies, and the problem of multiple equilibria is addressed by assuming that the fundamentals are more important than private valuations but not to the extent that it overshadows them completely. Under this assumption, traders have a ranking of values for the asset based on their private valuations and news. The ranking provides a narrower set of possible cases and helps traders make decisions.

  • 00:05:00 The speaker discusses three possible equilibria, labeled P1, P2, and P3, for cases when a fundamental asset is not attractive to fast traders with good news and low valuations. For P1, all three types of traders buy the asset at a narrow spread, while for P2, fast traders only buy if they have good news and high private valuation. Slow traders still buy at P2, but fast traders with conflicting signals will now not trade due to a high ask price and low bid price. P3 is an equilibrium where only fast traders with extreme valuations trade, exiling slow traders from the market and creating a wider spread, making it more difficult for them to trade.

  • 00:10:00 The speaker discusses the different equilibria that can arise in the market under different parameter values, which may include a spread so wide that there are no trades in the market. The P1, P2, and P3 equilibria are discussed, with P3 always existing, while P1 depends on a specific threshold of informed traders in the market. P1 is found to be Pareto dominant, providing better prices for all traders compared to P3. As such, uninformed traders are no longer trading at a loss in this model, making the trade more strategic for everyone.

  • 00:15:00 The professor talks about how the profits of Fast and Slow traders are affected by the existence of a P1 equilibrium. The profits of Fast traders are decreasing in alpha, meaning that they suffer from having more fast competitors in the market. On the other hand, Slow traders have a similar outcome as Fast traders but depend on their private valuation. When the equity point crosses zero, P1 equilibrium stops existing, and this is worse for everybody as it imposes an externality on everyone else. Overall, profits for all traders decrease as alpha increases.

  • 00:20:00 This creates a more nuanced solution to the tragedy of the commons, as not all traders necessarily become fast or slow, but rather it depends on the size of their institution and the number of markets they can participate in. The model assumes heterogeneity among institutions, with each having a type that determines their size and potential profits from being fast.

  • 00:25:00 The speaker discusses the decision-making process of institutions in choosing to become fast or slow, which is driven by the expected profit from being fast. The speaker explains that the profit from being fast is the same in all markets and depends only on the total share of fast institutions, and only institutions above a certain cutoff in terms of type will choose to become fast. The speaker then discusses how due to the shape of the assumed distribution, the distribution of trader types faced in any given market is uniform from 0 to M, and establishes the alpha, which is the probability of informed trading in every market.

  • 00:30:00 The speaker discusses the findings of a paper on high-frequency trading. The authors find an equilibrium in which the probability of facing a trader large enough for it to be worth becoming fast is given by the uniform distribution. They also find that the cost of becoming fast leads to fewer fast traders in the market, which decreases alpha. In addition, the authors have a welfare result that says markets with no adverse selection generate more welfare than markets with adverse selection. The speaker interprets this as well-functioning markets having too much high-frequency trading in equilibrium, and that the welfare-maximizing way is to set alpha to zero.

  • 00:35:00 The presenter discusses a proposal to run batch auctions every 0.1 seconds which would not cause significant delays for traders but could potentially harm high-frequency traders. The upcoming lecture will delve into this proposal in greater detail and provide empirical data to support it. The presenter apologizes for any confusion caused by the presentation but thanks the audience for staying and announces that the exercise class will be held on Friday.
Lecture 12, part 2: High-Frequency Trading (Financial Markets Microstructure)
Lecture 12, part 2: High-Frequency Trading (Financial Markets Microstructure)
  • 2020.04.22
  • www.youtube.com
Lecture 12, part 2: High-Frequency TradingFinancial Markets Microstructure course (Masters in Economics, UCPH, Spring 2020)***Full course playlist: https://w...