Trading and Exchanges

Trading and exchanges, by Larry Harris

===================

## Chapter 1: introduction

- a brief overview of trading and exchanges

- this is an excellent overview, checkout text

- key recurrent themes

- information asymmetries

- options to trade

- externalities

- market structure

- competition with free entry and exit

- communications and computing technologies

- price correlations

- principal-agent problems

- trustworthiness and creditworthiness

- the zero-sum game

# Part 1: the structure of trading

## Chapter 4: orders and order properties

- what are orders, and why do people use them?

- orders are necessary for traders who do not personally arrange their trades (one exception is dealers), which specify instrument, side, and some conditions

- traders who arrange their own trades have an advantage over those who use orders, as they can directly respond to market conditions as they change. Therefore it is crucial to **properly set order conditions** for the latter

- market orders

- market orders pay the spread

- market impact

- execution price uncertainty

- limit orders

- limit order placement terms

- marketable, in/at/behind the market

- standing limit orders are trading options that offer liquidity

- standing buy limit orders can be seen as **put options** for other traders with strike equal to the limit price. The value of the option depends on price, expiration date, and price volatility

- for instance, in volatile market, the option value of the limit order can change substantially before traders can cancel them, so traders often place limit orders **far behind the market** to reduce their option values, which causes wide bid/ask spreads

- the compensation that limit order traders hope to receive for giving away free trading options is a better price

- stop orders

- stop orders, limit orders, stop limit orders

- stop orders and liquidity

- stop orders accelerate price changes by adding buying/selling pressure when prices are rising/falling. Therefore stop orders **add momentum to the market**

- market-if-touched orders

- tick-sensitive orders

- tick-sensitive orders are limit orders with dynamically adjusting limit prices

- market-not-held orders

- validity and expiration instructions

- Table 4-3

- quantity instructions

- other order instructions

- spread orders

- e.g. use spread orders to roll futures contract to next expiration date

- display instructions

- substitution orders

- special settlement instructions

## Chapter 5: market structures

- trading sessions

- continuous vs call trading sessions

- execution systems

- quote-driven dealer markets

- order-driven markets

- brokered markets

- hybrid markets

- market information systems

## Chapter 6: order-driven markets

- oral auctions

- order precedence rules

- price priority

- time precedence

- time precedence is meaningful only when the **minimum price increment** (tick) is not trivially small, which is the smallest amount by which a trader must pay to acquire precedence

- public order precedence

- trade pricing rule

- rule-based order-matching systems

- order precedence rules

- the matching procedure

- the uniform pricing rule and single price auctions

- example: most continuous order-driven stock markets open the trading sessions with a single price call market auction, where all trades take place at the same **market-clearing price**. It also happens when market halts.

- supply and demand

- the intersection of supply and demand schedule curves determines the market-clearing price

- trader surpluses

- trader surplus is the difference between the trade price and the trader's valuation of the item

- single price auction maximizes **total trader surpluses**

- in markets that arrange trades at multiple prices, a successful buyer in one trade might value the item less than does a successful seller in another trade, which reduces total trader surpluses, **since the buyer could have bought it at a lower price**

- discriminatory pricing rule and continuous two-sided auctions

- discriminatory vs uniform pricing rules

- standing limit order traders prefer uniform pricing, since they could often trade at a better price, therefore traders are more aggressive under uniform pricing

- given a set of standing orders, large impatient traders prefer discriminatory pricing since they can trade first part of their orders at better prices than the last parts

- continuous vs call markets

- continuous markets produce less trader surplus, but it allows traders to trade whenever they want

- derivative pricing rule and crossing networks

- prices are completely independent of the orders, crossing networks only discover whether traders are willing to trade at the crossing prices (e.g. NYSE's after-hour trading session I use 4 pm closing prices as the crossing prices)

- traders use crossing networks because they want to trade **without having impacts on prices**. Although not all orders can be filled

- markets are subject to price manipulation

## Chapter 7: brokers

- what brokers do

- clients use brokers to arrange their trades because brokers usually can arrange trades **at a much lower cost**

- the principal-agent problem

- performance measurement

- traders cannot easily measure the quality of service, rating agencies evaluate brokerages and sell results to interested parties

- best execution

- dual trading problem

- conflicts of interest arise when broker also acts as a dealer

- when they internalize orders, client wants to buy at lower prices, while dealer wants to sell at higher prices

- when dealer and client wants to trade on the same side, dealer would benefit by being a **front runner**

- there are some regulations to resolve this issue

- dishonest brokers

- front running

- inappropriate order exposure

- fraudulent trade assignment

- prearranged trading and kickback schemes

- unauthorized trading and churning

# part 2: the benefits of trade

## Chapter 8: why people trade

- utilitarian traders

- investors and borrowers

- they move money from one point in time to another, when incomes and expenses do not coincide

- asset exchangers

- hedgers

- **price risk** can be hedged with forward or (liquid standardized) futures contracts

- **market risk** can be hedged with index futures or options

- **interest rate risk** can be hedged with swaps

- gamblers

- fledglings

- trade to learn whether they can trade profitably

- cross-subsidizers

- tax avoiders

- profit-motivated traders

- speculators

- informed traders

- they are the only traders who cause prices to move towards fundamental values, others are **noise traders**

- parasitic traders

- order anticipators, bluffers

- technical traders

- dealers

- market makers, block facilitators

- futile traders

## Chapter 9: good markets

- welfare economics

- aims to find policies that maximizes specific measure of social welfare

- private benefits of trading

- since trading is a zero-sum game, speculators and dealers can only profit if utilitarian traders are willing to trade. This suggests that welfare of utilitarian traders is ultimately more important than that of profit-motivated traders

- public benefits from informative prices

- production and allocation decisions

- informative prices (those close to fundamental values) help better allocate resources such that **marginal benefits** of a resource is the same for all projects that use it, this improves overall efficiency

- capital allocation in primary markets

- manager allocation problem in secondary markets

- public benefits of liquid markets

- public benefits of exchange

- public benefits of hedging

- with low cost of hedging using financial instruments, wheat producers can specialize in their single crop without need to produce other products (which they may not have comparative advantage) to diversify risks, which improves overall efficiency

- public benefits of risk sharing

- easier to raise capital at low cost

- other public benefits of liquidity

- provide alternative options so people can arrange their affairs differently

- facilitate profitable informed trading

- attract gamblers who would lose to informed traders on average, therefore encouraging more informed traders to make the prices **more informative**, which benefits the public. Also, they add liquidity

# part 3: speculators

## Chapter 10: informed traders and market efficiency

- fundamental values

- price = fundamental value + noise

- informed traders make prices informative

- markets are essentially statistical calculators that aggregate value estimates from various informed traders to obtain more accurate estimates of value

- informed trading strategies

- traders want to minimize price impacts (such that trading costs are low, and that other people do not know they are informed) while getting orders executed

- styles of informed trading

- value traders

- news traders

- information-oriented technical traders

- arbitrageurs

- will lose money if they mistakenly conclude that instruments are mis-priced relative to each other

- competition, trading profits and informative prices

- informed trading profits

- depends on prediction ability and the impacts their trading has on prices

- informed traders make profits with uninformed traders, the latter are willing to tolerate some loss since they obtain other valuable services from the market

- competition among informed traders

- market efficiency

- trade-off between liquidity and informative prices

- informative prices benefit overall economic efficiency, but at the price that **uninformed traders lose to informed traders**, which could discourage uninformed traders who provide liquidity

- benefits of public information

- release of some public information could increase both market liquidity (since uninformed traders lose less) and price efficiency (as cost to get information is lower), therefore some markets require regular releases of substantial fundamental information

## Chapter 11: order anticipators

- introduction

- order anticipators are **parasitic traders**, they do not make prices more informative and they do not make markets more liquid

- front runners

- front running aggressive traders

- front running passive traders

- also known as "quote matchers", they try to extract **option value** of passive traders' orders

- example: place a limit buy order **one tick higher than best bid**, suppose a market sell order fills the order, if price goes up, he takes the profit of price change, if price goes down, he can immediately sell it to (existing) best bid and lost one tick. The return is one-side, **like an option** with price being one tick (so front runners are less profitable in markets with large tick). Note that quote matchers should be **faster than the passive trader** (with best bid), otherwise they may have trouble getting out of their positions

- sentiment-oriented technical traders

- they try to predict behavior of **uninformed traders**, and trade before them

- note that uninformed traders could push price away from fundamental value, so sentiment-oriented technical traders tend to lose to value treaders, therefore they need to know when to close positions

- squeezers

- manipulation of stop orders

- manipulators can push prices up or down to activate stop orders, which further accelerates price changes and manipulator can close positions with profits

## Chapter 12: bluffers and market manipulations

- summary

- bluffers can generally profit when price impact is different for their purchases and sales. If purchases have greater impacts, they buy first then sell, and vice versa. To avoid losing to bluffers, liquid suppliers must be disciplined when they adjust prices in response to buy and sells

- traders most vulnerable to bluffers are momentum traders and liquidity suppliers

- value traders can call a bluffer's bluff

# part 4: liquidity suppliers

## Chapter 13: dealers

- dealer quotations

- quoted vs realized spreads

- attracting order flow

- dealer inventories

- why need to control inventories

- if too high: have significant **inventory risk** (lose if prices move against them)

- if too low: may cost too much selling in some markets

- inventory risk

- diversifiable inventory risk

- they are due to changes that no one can predict, since price changes are **uncorrelated** with inventory imbalances, dealers can gain or lose with equal probabilities

- it can be diversified by **holding many instruments**

- adverse selection risk

- this happens when dealer trades with **informed traders**: they buy when they expect price to rise, which causes one-sided order flows, narrower or negative spread, and that inventory imbalances become **inversely correlated** with future price changes

## Chapter 14: bid/ask spreads

- spread components

- transaction cost component

- used to compensate for normal costs of doing business

- adverse selection spread component

- used to compensate for losses when trading with well-informed traders. This component allows dealer to **earn from uninformed traders** what they lose to informed traders

- adverse selection and uninformed traders

- adverse selection explains why uninformed traders lose to informed traders, regardless of whether they place limit or market orders, see Table 14-1

- equilibrium spreads in continuous order-driven auction markets

- simple analysis

- if spreads are too wide, traders tend to place limit orders and thereby narrow the spread, and vice versa

- more realistic analysis

- differential commissions

- other costs for limit orders

- order management

- limit order traders give valuable **timing option** to market order traders

- spreads when public traders compete with dealers

- public limit order traders can quote more aggressive prices since they do not have the business costs, however, dealers can see more of the order flow and can **change their quotes faster**

- cross-sectional spread predictions

- 3 primary spread determinants

- asymmetric information: which causes dealer to lose to informed traders

- volatility

- utilitarian trading interest: when it is high the market would be more active and spread would be narrower

## Chapter 15: block traders

- block trading problems

- latent demand problem

- block trades must discover the **latent demands** of responsive traders (who are willing to trade but do not initiate their trades) when they cannot find adequate liquidity in the market

- order exposure problem

- price discrimination problem

- block liquidity suppliers do not want to be the first to offer liquidity to a large trader, only to see **prices move against them** when the large trader continues to trade

- asymmetric information problem

- liquidity suppliers suspect block initiators are well informed and are therefore reluctant to trade with them

## Chapter 16: value traders

- value traders supply liquidity

- the outside spread and its determinants

- risks of value trading

- adverse selection risk: when they supply liquidity to better-informed traders

- winner's curse and value trading

- winner's curse: when buys compete in an auction, the winner (who bid the highest) would probably be cursed by the price they pay since winners tend to be those who have **overestimated values**

- value traders and winners' curse: they lose when their value estimates are wrong, when they are uncertain about the value, they widen the **outside spread** (the prices at which they are willing to buy and sell)

- outside vs dealer spreads

## Chapter 17: arbitrageurs

- types of arbitrage

- pure arbitrages

- def: arbitrages involving instruments for which the value of hedge portfolio is strictly **mean-reverting**

- types: shipping/delivery/conversion arbitrage

- speculative arbitrages

- def: involving value that is non-stationary

- pair trading

- statistical arbitrage

- risk arbitrage

- arbitrage risks

- implementation risk

- def: transaction costs are greater than they expect, due to competition between arbitrageurs (market order could trade at worse prices, and limit order may fail to be executed)

- to minimize this risk, an arbitrageur needs to **synchronize both legs** of the portfolio, e.g. if one leg is in illiquid market and another in liquid market, it would be better to trade the illiquid one first

- basis risk

- def: risk that the basis will move in the wrong direction

- model risk

- carrying cost risk

- due to unexpected cost of carry, unexpected price increases, slow convergence, and unexpected buy-ins

- causes of arbitrage opportunities

- different speeds of adjustment for similar instruments to the changes of factor

- uninformed traders buy in some markets and sell in others

- arbitrageurs, dealers and brokers

- dealers connect buyers and sellers arriving at different times, and arbitrageurs connect them at different places (markets)

## Chapter 18: buy-side traders

- order exposure decision

- benefit of exposure

- trades are easier to arrange

- can attract **latent traders**

- costs of exposure

- may reveal trader motives so that others can compete with them, withhold liquidity from them, or act against their interests

- may reveal future price impacts, and front runners can profit from this

- may reveal valuable trading options, quote matcher can profit

- defensive strategies

- evasive strategies: avoid revealing information to those who would act against them

- deceptive strategies: attempt to confuse these traders

- offensive strategies: attempt to attack these traders

- how markets help traders control exposure costs

# part 5: origins of liquidity and volatility

## Chapter 19: liquidity

- the search for liquidity

- unilateral searches

- main decision: when to stop searching, we want to search when **expected benefits** > expected cost of search

- bilateral searches

- main differences

- may search actively or passively

- may not always be able to return to the best match identified before. Therefore, need to take into account this potential loss

- liquidity dimensions

- immediacy/width/depth/resiliency

- who, how and why of liquidity

- market makers

- they avoid large inventory positions to avoid inventory risks, they supply liquidity **in the form of immediacy**

- block dealers

- they supply liquidity in form of **depth**

- value traders

- they are best able to supply **depth** because they are best able to solve the adverse selection problem, by knowing value better than anyone else

- they make market **resilient**

- pre-committed traders

- arbitrageurs

- they are **porters of liquidity** (from one market to another)

## Chapter 20: volatility

- fundamental volatility

- def: volatility due to unexpected change in fundamental factors (e.g. interest rate, inflation, storage costs)

- transitory volatility

- def: due to the demands of impatient uninformed traders that cause prices to diverge from fundamental values

- measuring volatility and its components

- fundamental volatility consists of seemingly random price changes **that do not revert**, while transitory volatility consists of price changes that **ultimately revert** (so has negative serial correlation)

# Part 6: evaluation and prediction

## Chapter 21: liquidity and transaction cost measurement

- transaction cost components

- explicit costs

- implicit costs

- missed trade opportunity costs

- implicit cost estimation

- methods

- specified price benchmark method: compute the difference between trade price and the benchmark price (e.g. VWAP, opening/closing price)

- econometric transaction cost estimation

- measuring transaction costs with specified price benchmarks

- benchmark prices

- quoted/effective/realized spreads: realized spread tend to be smaller than effective spread, since aggregate buy (sell) push up (down) best offer (bid)

- implementation shortfalls: two components

- transaction cost of **completed trades**

- missed trade opportunity costs

## Chapter 22: performance evaluation and prediction

- performance evaluation methods

- absolute performance measurement

- internal rate of return, holding period return

- current yield component, capital gain component

- estimation of value could be different for those that do not trade often (e.g. real estate)

- relative performance measurement

- risk/market adjusted return

- the performance prediction problem

- statistical performance evaluation

- example: is Buffet a skilled manager? consider 100000 unskilled managers (with normal distribution with std = 7%), the probability that the best performer of them over 3 decades beat the market by 11.8% is 5%, this suggests that Buffet is very likely a skilled manager

- more important problems with statistical performance evaluation

- distribution shape: normality, fat-tailed, asymmetric distributions

- fraudulent returns

- return smoothing: artificial smoothing may reduce variation and overestimate risk-adjusted performance

- Ponzi scheme

- sample selection bias

- survivorship bias

- e.g. a mutual fund company may kill poorly-performing funds and only report well-performing ones

# Part 7: market structures

## Chapter 23: index and portfolio markets

- price indexes

- index funds

- index funds generally slightly underperform their target indexes, due to various **frictions** (e.g. trading cost, management fees)

- liquidity and price formation in index markets

- package trading: traders sometimes do package trading since they can often get better prices than trading individual stocks, since **informed traders are less likely to trade portfolios** therefore dealers are willing to quote better prices

- index products

## Chapter 24: specialists

- specialists as dealers

- affirmative obligations

- specialists are required to maintain **price continuity**, which could be expensive since they have to trade when no one else is willing to do so

- negative obligations

- order precedence rules require that they **yield to public orders** at the same price or better

- public liquidity preservation principle

- specialists as brokers

- specialist privileges

## Chapter 26: competition within and among markets

- market consolidation

- market tend to consolidate without regulatory intervention (if all traders are identical and have same needs), since more liquidity attracts traders and more traders add liquidity. A new market needs to be substantially better than the old market to attract order flow

- market fragmentation

- market fragment because traders are not identical and their problems are very different. e.g.

- large traders do not want to expose their orders, which is easier in some markets than others

- uninformed traders prefer to trade in markets that expose identities since they do not want to trade with informed traders

- some markets serve impatient traders better

- market segmentation: how fragmented markets consolidate

- when the information in different segments is publicly available at low costs, there exists mechanisms to consolidate segments (e.g. transferring of liquidity by arbitrageurs)

见到＂Bernard L. Madoff Investment Securities＂作为当时美国著名dealer（本书2002年出版），难免不让人＂呵呵＂。