Session 15 - Trading, Performance Evaluation, and Manager Selection Flashcards
Motivations to Trade
1/ Profit Seeking
2/ Risk Management
3/ Cash Flow Needs
4/ Corporate Actions/Index Reconstitution/Margin Calls
Profit Seeking (under Motivations to Trade)
- active managers are trading based on information that is not reflected yet, believe securities are mispriced
- will try to hide their trades to prevent information leakage (executive in multiple or less transparent ventures)
- trade urgency is high when ST profit opportunity
- tries to execute close to the market
- occurs when alpha decay is high
Lit vs dark venues
lit - better execution likelihood
dark - less transparency, but higher likelihood of going unfilled
Risk management/hedging needs (under Motivations to Trade)
- may involve derivatives or just trades in the underlying securities
- investment mandate may not allow derivatives
- use ETFs or the underlying security
Cash flow needs (under Motivations to Trade)
➝ may involve high or low trade urgency
- collateral/margin calls ➝ high
- redemption ➝ low
- inflows from dividends and cash can be equitized using ETFs/futures to min. cash drag and until next rebalance or UL positions can be established
– client redemptions are based on NAV
- NAV based on closing prices
∴ trading at the closing price reduces redemption price risk
Corporate Actions/Index Reconstitution/Margin Calls (under Motivations to Trade)
- cash dividends/coupons need to be reinvested
- margin or collateral calls have high levels of trade urgency
- index tracking portfolios ➝ trades usually done ‘on close’ to minimize tracking error
- active managers with a benchmark may choose to rebalance with index reconstitution
Factors to consider for Trade Strategy Inputs
1/ Order characteristics
2/ Security characteristics
3/ Market conditions
4/ Individual risk aversion
Order characteristics (under Trade Strategy Inputs)
a) side of the order
- if buying in a rising market or selling in a falling market ➝ market risk exposure order may take longer to execute, or cost more
b) size of the order - large order sizes create market impact (i.e. adverse price movement in a security as a result of trading an order)
- larger orders take longer to trade
- will usually be traded with lower trade urgency to reduce market impact
Security characteristics (under Trade Strategy Inputs)
a) security type - liquidity and trading costs will vary by exchange
b) short-term alpha - the expected movement in the security price over the trading horizon (i.e. appreciation, depreciation, reversion)
c) price volatility - affects execution risk
d) security liquidity - greater liquidity ➝ lower execution risk & trading costs
- bid-ask spreads will indicate both costs and market depth
- larger trades ➝ broken down, traded more slowly
Alpha decay
➝ results from price movement in the direction of the trade forecast
➝ if information is reflected in prices quickly, decay is fast - requires faster trading
Execution risk
➝ risk of an adverse price movement over the trading horizon
Market conditions (under Trade Strategy Inputs)
- during market events or crisis ➝ volatility increases and liquidity decreases
- security liquidity may change bc of market liquidity
- lower market liquidity ➝ longer trading horizon but/ - higher volatility may heighten trade urgency
- longer trading horizons increase execution/market risk
Individual risk aversion (under Trade Strategy Inputs)
- high level of risk aversion, more concern about market risk, trade with greater urgency
Market Impact and execution risk
· temporary market impact cost - temporary, short-lived impact on security price from trading
- usually price reversion after the order is complete
· permanent component of price change associated with trading is the market impact caused by the information content of the trade
· execution risk lowers with faster trading
- trade too fast ➝ increased market impact
- trade too slow ➝ increased execution risk/market risk
Types of Reference Prices
1/ pre-trade benchmarks
2/ Intraday benchmarks
3/ Post-trade benchmarks
4/ Price-target benchmarks
Pre-Trade Benchmark
a/ opening price
b/ arrival trade
c/ decision price
d/ previous close
Decision price
security price at the time the PM makes the decision to buy and sell the security
Previous Close
often specified by quantitative PMs
Opening Price
- used by PMs who trade from the open
- used as the ‘decision price’ by fundamental PMs
- opening price does not have overnight risk
Arrival Price
- price of the security at the time the order is entered into the market for execution
- typically used by short-term alpha traders
- goal is to transact at or close to current
market prices
Intraday benchmarks
- for funds that trade passively over the pay, see liquidity or rebalancing
- do not expect the security to exhibit any ST price momentum
1/ VWAP
2/ TWAP
VWAP (Volume weighted average price)
- volume-weighted average price of all the trades executed over the day or trading horizon
- usually when PM wants to participate with volume
patterns - works well when PM is executing buy and sell orders
TWAP
- time-weighted average price - the average price of trades executed over the trading day or trading horizon
- used when outlier trades make VWAP unreliable
i. e. large buy orders at the day’s low or large sell orders at the day’s high
Post-Trade Benchmarks
- most common is closing price
- funds valued at end of day (NAV) will ‘buy on close’ to minimize tracking error
- typically used by index & mutual funds
Price-target benchmarks
- PMs seeking short-term alpha
- purchase shares at or below some target price
Large Block Trades + Urgent (Trade Execution)
- principal trade or broker risk trades
- dealer is the counterparty
- buy/sell from their own inventory
- if the security is less active, the dealer will quote a wider bid-ask spread
- called quote-driven, OTC or off-exchange markets
Large block trades ➝ non-urgent or very illiquid (Trade Execution)
➝ agency trade
- dealer attempt to arrange trades as a broker
- attempt a cross ➝ matching with other client orders
- then open market ➝ split order up
RFQ - Request for quote ➝ PM requests quotes from dealers
Liquid, standardized securities with order-driven markets
- trading done electronically w/ multiple venues
- for trades other than large orders
Large orders, liquid, standardized (Trade Execution)
➝ algorithmic trading (DMA) - high-touch generally inefficient
- subject to front-running
Algorithmic Trading
slice orders into smaller pieces and
trade across venues and overtime to reduce the
the price impact of an order
- 2 purposes
➝ execution - PM determines what to buy/sell
➝ profit-seeking - algorithm determines what to buy/sell
Scheduled Execution algorithm classifications
1/ Scheduled:
a) POV - percent of volume (a.k.a. participation algorithms)
b) VWAP/TWAP
POV - percent of volume (a.k.a. participation algorithms)
- as volume increases, the algorithm trades more
- trader specifies a participation rate
e. g./ 10%, for every 10K shares traded, algorithm trades 1000
VWAP/TWAP
- VWAP ➝ follows a time slicing schedule based on historical volume profiles
- higher percentage traded at the open and close
- may not be optimal for illiquid stocks (may not complete order)
- TWAP ➝ follows an equal-weighted time schedule - same number of shares per time period
Scheduled algos. appropriate for:
- PM has no expectations of momentum (adverse price moves)
- PM has greater risk tolerance for longer execution time periods - minimizing market impact
- relatively small order size (5%-10% of expected volume)
- relatively liquid or balanced buy/sell orders
Liquidity seeking (opportunistic algorithms)
- trade faster when liquidity shows up across multiple venues
- will use exchanges, ATS, and dark pools (for larger quantities)
· appropriate for
· large orders with high trade urgency while avoiding market impact
· printing limit orders would reveal information (minimizes information leakage)
· less liquid, thinly-traded, or when liquidity is episodic
Arrival price (Execution algorithm)
- trade more aggressively at beginning of order - called front-loaded strategy
- used when momentum is expected in prices
- appropriate for · PM who are risk adverse and
wish to trade more quickly to reduce execution risk - security is relatively liquid, order is not out-sized (i.e. < 15% expected volume)
Dark Strategies/liquidity aggregators (Execution algorithm)
- used to avoid any information leakage volume
- order size is relatively large and may have a significant market impact
appropriate for:
· illiquid securities
· wide bid-ask spreads
· PM does not require full execution (low trade urgency)
Smart order routers
- SOR determines the destination with the highest probability of executing a limit order
- used with small market or limit orders that will have no market impact
- best used for orders that require immediate execution or for
limit orders for which printing the order will not leak information - appropriate for · securities traded on multiple markets
Trade Executions for Equities
- exchanges and dark pools (ATS or MTF – multilateral
trading facilities) - most trades are electronic, algo. trading is common
large & urgent trades ➝ high touch (especially illiquid securities)
large & non-urgent ➝ trading algorithms
small trades - electronic trading
Trade Executions for Fixed Income
- primarily OTC, quote-driven
- sourcing liquidity relies heavily on dealers
- except for on-the-run U.S. Treasuries, bond/interest rate futures
- small trades and large urgent trades ➝ principal trades
- large, non-urgent ➝ agency trades
Trade Executions for Exchange-Traded Derivatives
- electronic trading widespread
- algo.-trading ➝ mostly futures
- large, urgent trades ➝ liquidity seeking algos.
Trade Executions for OTC Derivatives
– dealer market, trade sizes usually large
- large, urgent trades ➝ principal trades
- non-urgent ➝ agency trades
Trade Executions for Spot forex (currency)
- primarily electronic trading
- large, urgent trades ➝ RFQ
IS decomposes total cost into
- measures total costs associated with implementing the investment decision IS = Paper return - Actual Return where: Paper return = (𝐏𝐧−𝐏𝐝) 𝐒 Actual return = ∑𝐬𝐣(𝐏𝐧) − ∑𝐬𝐣𝐩𝐣 − 𝐅𝐞𝐞𝐬 - IS decomposes total cost into 1/ Execution cost ∑𝐬𝐣𝐩𝐣 − ∑𝐬𝐣𝐩𝐝 - shares that were transacted - market impact + price drift 2/ Opportunity cost ^𝐒 − ∑𝐬𝐣a(𝐏𝐧−𝐏𝐝) – unexecuted shares 3/ Fixed Fees – explicit fees
Exp
➝ execution cost divided into delay and trading cost
- delay cost = (p0-pd)Sj
p0 = price when the order is placed
- trading cost = Sj(pj-Po)
- the time from receiving the order from the PM to actually placing the order
- often resulting from the time required to determine the trading strategy (broker or algorithm choice)
Improving Execution Performance
· have a process in place that provides traders with broker performance metrics (reduce delay)
· determine proper order size for the market
- rather than typing up funds/time trying to fill a difficult trade, reduce the size and concurrently trade something else (reduce lost opportunity)
- typically the result of adverse price movement and/or illiquidity
Evaluating Trade Execution
- measure execution quality of a trade
- measure overall performance of the trader/broker/algo.
𝐂 𝐨 𝐬 𝐭 ( $ ) = 𝐒 𝐢 𝐝 𝐞 × ( 𝐏1 − 𝐏 ∗ ) × 𝐒 𝐡 𝐚 𝐫 𝐞 𝐬
𝐂𝐨𝐬𝐭 ($/𝐬𝐡) = 𝐒𝐢𝐝𝐞 × (𝐏1 − 𝐏∗)
𝐂𝐨𝐬𝐭(𝐛𝐩𝐬)=𝐒𝐢𝐝𝐞 ×9 𝐏∗ : 𝟏𝟎,𝟎𝟎𝟎
𝐏 = average execution price of order
𝐏∗= reference price can be: arrival price, VWAP, TWAP, market-on-close
Mark Adjusted Cost
Adjusts for general market movement
MAC = Arrival cost (bps) – 𝛃 · Index Cost (bps)
𝐈𝐧𝐝𝐞𝐱 𝐜𝐨𝐬𝐭 (𝐛𝐩𝐬) = 𝐒𝐢𝐝𝐞 × ( 𝐈𝐧𝐝𝐞𝐱 𝐕𝐖𝐀𝐏 - 𝐈𝐧𝐝𝐞𝐱 𝐚𝐫𝐫𝐢𝐯𝐚𝐥 𝐩𝐫𝐢𝐜𝐞) / 𝐈𝐧𝐝𝐞𝐱 𝐚𝐫𝐫𝐢𝐯𝐚𝐥 𝐩𝐫𝐢𝐜𝐞 × 𝟏𝟎, 𝟎𝟎𝟎
Added Value
- compare arrival cost with estimated pre-trade cost
- Added Valued (BPS) = Arrival cost (BPS) - estimated pre-trade cost (BPS)
Trade governance
- all asset managers should have a trade policy
a document that articulates the firms trading policies
(mandated by major market regulators, e.g. SEC)
➝ objective of a trade policy ➝ ensure execution and order-handling procedures are in line with the duty of best execution owed to clients
Meaning of best execution
· not just best price at lowest cost
- involves identifying the most appropriate trade-off between:
· execution price
· trading costs
· speed of execution
· likelihood of execution and settlement · order size
· nature of trade
Factors determining the optimal execution approach
· urgency of order
· characteristics of the securities traded
· rationale for a trade
· characteristics of the execution venue used
· investment strategy objectives
List of eligible brokers and execution venues
· quality of service · reputation · speed of execution · financial stability · settlement capabilities · cost competitiveness · willing to do principal trades
Process used to monitor execution arrangements
- all brokers and execution venues used should be subject to ongoing monitoring for reputational risk, irregularities, criminal actions and financial stability
- use of a ‘Best Execution Monitoring Committee’
- firm-wide responsibility for trade execution monitoring
- trading records stored and accessible for several years
Performance measurement
- absolute return ➝ what the portfolio achieved over a specific period
- excess return ➝ portfolio return - benchmark return
- also involves measuring the risk incurred to achieve that return
Performance Attribution
- how that performance was achieved or how the risk was incurred
- explain absolute or relative return
- what portion was driven by active mgr. decisions - decompose excess return into component sources - decompose risk
Performance Evaluation
- draw conclusions regarding the quality of performance - distinguish manager luck from skill
An effective performance process must
1) account for all the portfolio’s return or risk exposure 2) reflect the investment decision-making process
3) quantify the active decisions of the PM
4) provide a complete understanding of the excess
return/risk of the portfolio
Return attribution
– analyzes the impact of active investment decisions on returns
Risk attribution
- analyzes the risk consequences of those decisions (absolute or benchmark relative terms)
Micro attribution
- understanding the drivers of a manager’s returns and whether those drivers are consistent with the stated investment process
Macro attribution
- measures the effect of the asset owner’s (sponsor’s) choice to deviate from the SAA
- also measures the effect of the manager selection and
timing decisions
Returns-based attribution
- uses only total portfolio return
- most appropriate when underlying portfolio information is not available at the required frequency or detail
- easiest to implement
- least accurate
- vulnerable to data manipulation
Holdings-based attribution
- used actual holdings (beginning of period)
- all transactions are assumed to occur at end of the day
- accuracy improves when data has shorter time intervals
- most appropriate for investment strategies with little turnover
Transactions-based attribution
- uses both the holdings and the transactions during the evaluation period
- most accurate, but most difficult and time-consuming to implement
Brinson Model Equity Attribution
➝ Allocation (𝑷 −𝑩 )(𝑹 −𝑹o )➝∆𝐰(𝐑 −𝐑o )
➝ Security (𝑹𝑷−𝑹𝑩)𝑩𝒘 ➝ 𝑹𝑨𝒊 · 𝑩𝒘𝒊
➝ Interaction (𝑷𝒘 − 𝑩𝒘)^𝐑𝐩 − 𝐑𝐁a➝ ∆𝐰𝐢 ⋅ 𝑹𝑨𝒊
Factor-Based Equity Attribution
- fundamental factor models
- decompose contributions to excess returns from factors
- market, size, value, momentum
- size (<0 = Lg. Cap)
- value (>0 = value)
RMRF = 1 = broad-based market index
Exposure decomposition – top-down approach (under Fixed Income Attribution)
Benchmark
➝ Duration
➝ Yield curve positioning
➝ Sectors (i.e. gov’t., corporate)
Portfolio
- active decisions to deviate from benchmark exposures
e.g./ active duration bets ➝ increase duration relative to the benchmark in anticipation of falling rates
yield curve positioning ➝ barbell for a flattening curve
sectors ➝ overweight credits in anticipation of spreads narrowing
· how well, relative to the benchmark, did these active decisions work out, what contribution to active return
- used primarily for client reports, easy to understand
Fixed Income Attribution Approaches
1/ Exposure decomposition – top-down approach
2/ Yield curve decomposition - duration based
3/ Yield curve decomposition - full repricing
Yield curve decomposition - duration based (under Fixed Income Attribution)
- can be either top-down or bottom-up
➝ estimates the returns based on duration
% total return = % Income return + % Price change - ModDur × ∆𝐲𝐢𝐞𝐥𝐝
- applied to both the portfolio and the benchmark
- difference = effect of active PM decisions
- requires more data points than exposure decomp.
- typically used in reports for analysts & PMs
Yield curve decomposition - full repricing (under Fixed Income Attribution)
- instead of using estimates
- prices out each security
- most complex attribution of the three
