Trading and Performance Evaluation Flashcards
Motivations/Rationals
- Profit Seeking
- Risk Management Hedging Needs
- Cash Flows Needs
- Corporate Actions, Margin Calls, and Index Reconstitutions
Profit Seeking
- Active portfolio managers are trying to outperform a benchmark by generating Alpha.
- Done by trading securities that they believe are under/over-priced
- Management needs to act on their insights before the rest of the market does
- Key part is Alpha decay
Alpha Decay
- Deterioration of Alpha once an investment decision has been made.
- Management trades on the flow of daily news, and has a high urgency of trading
Dark Pools
- Trading systems with low pre-trade transparency; trades cannot be seen until the trading occurs (post trade)
- Disadvantages: Traders cannot see orders on the other side they don’t have any of the pre-traded likelihood that a trade will be executed
Lit Venue
High pre-trade transparency, you can see both sides of a trade when an order is entered “stock exchange”
Cash Flows Needs
- Trades are primary caused by subscriptions in the fund or redemptions out of the fund
- If a fund has a lot of illiquid holdings it will be difficult to invest new client funds in a short-term period
Cash Drag
- Cash doesn’t earn a return which will reduce a portfolios return
- To avoid cash drag a portfolio can use a cash equitization strategy (ETFs or derivatives)
Index Reconstitution
Overtime a index adds and remove securities
Factors that Dictated the Appropriate Trading Strategy
- Order Characteristics
- Security Characteristics
- Market Conditions
- Individual Risk Aversion
Order Characteristics
- Side of the Trade: The direction of the order (buy, sell, short sell)
- Absolute Size: The number of the securities being trade
- Relative Size: Orders that make up a higher percentage of the ADV will have a significant impact
Market Impact
- Impact on the price of the shares by a large order “buying liquidity”
- Market Impact costs and timing cost are inverse in nature
- Upward pressure: act of buying a lot
- Downward pressure: act of selling a lot
- Large block order: Tries to trade overtime with less urgency to reduce the impact.
Security Characteristics
- Security Type: Different types, trade in different markets, and have different costs
- Short-Term Alpha:
- Price Volatility: If prices are fluctuating widely implies high execution risk.
- Security Liquidity: More liquidity the narrower the bid-ask spread.
Short-term Alpha
A high urgency trade due to capturing mispricing there will be a high rate of alpha decay for an active management
Impact and Execution Risk
These are Inverse if one goes up, the other goes down
* Market Impact costs: Trading to quickly information leakage
* Execution Risk: Trading too slowly; adverse price movement
Implicit Costs
Hidden, largest part to a trade’s costs
* Market (Price) Impact Cost: (Cost of buying liquidity) Actual execution price less the price when the order is presented to the broker (M - B)
* Opportunity Cost: (Cost of the investment idea) The difference in price when the order was cancelled and when the order was placed “missed trade”
* Execution Risks: (Timing, Delay, or Slippage Cost) The price when the broker presents the order to the market less the price when the original order was presented to the trader (B- T)
* Bid-Ask Spread
Explicated Costs
The tangible visible cost you can see:
* Commissions,
* Fees
* Stamp duties
* Taxes
Reference Price
Are a key impact for the actual costs of trading for post trade evaluation
* Pre-Trade
* Intra-Trade
* Post-Trade
* Price-Target
Pre-Trade Benchmarks
Know before the start of the trading there are 4-types
* Decisions “Benchmark” Price
* Previous Close: Closing price on the previous day
* Open Price: Opening price of the day
* Arrival Price
Decisions “Benchmark” Price
“Original/Initial” The original price when the investment was order initiation “conceive”.
A lot of orders are initiated when the market is closed to the previous day’s closing price is used.
Arrival Price
“Revised Benchmark Price”
* Market price when it was sent to the market for execution if the order was fulfilled
* If an order wasn’t fulfilled at a limit order price, then the limit price needs to change or the price of close of previous day.
Intraday Benchmarks
Based during the trading period
Normally used by the portfolio manager who trades passively over a day the ones included are:
* VWAP
* TWAP
Post-Trade Benchmarks
Determine after trading has been completed (most common is the closing price)
Disadvantages: The closing price is not known until after the trade has been completed; cannot access performance during the trading horizon
Price Target Benchmark
- High urgency for with high alpha decay
- For profit seeking portfolio managers trying to earn a short-term alphas
- Is quick strategy that would purchase a security below a predetermined price
High-Touch Approach
High level of human interaction, for large block trades, and/or less liquid markets
1. Principal Trade: Agency
2. Agency Trade: Broker
Principal trades
- “Agency” they can act as the buyer to any seller or the seller to any buyer
- Done using the dealer’s own inventory
- Dealer/market makers assume all the risk in executing the order, priced into their spread “bid-ask”
- Other markets: Quoted driven, OTC, and RFQ markets
Agency Trades
- Where the broker finds the other side of the trade
- Risk for the execution remains with the portfolio manager or trader.
Electronic Trading
- Trading via computers, done in more liquid markets
- The trading is order driven cause it allows the buyer and sellers to advertise their limit orders in a central limit order book
- Allows direct market access (DMA)
- Algorithmic trading
- Provides anonymity
Direct Market Access (DMA)
Allows a person on the buy side to access the order book of the exchange directly through the structure the broker has.
Used for:
* Small currency trades
* Buy-side traders for exchange-traded derivatives (particularly smaller trades)
Algorithmic Trading
Use of programs rules to electronically trade order for trade execution or profit seeking
Trade Execution Algorithms
Execution algorithms that trade by specified rules set by the portfolio manager to meet their
Specific objectives they have:
* Scheduled algorithms
* Liquidity seeking
* Arrival Price
* Dark Strategies or liquidity aggregators
* Smart order routers
Schedule Algorithms
- Percentage of volume participating algorithm will send order on a volume participation schedule
- They execute trades using rules driven based on a historical time period or historical volume
- Are most appropriate for smaller orders in liquid markets that have less urgency to execute the trade while, trying to minimize the market impact
- Market Impact cost decreased; execution risk increased
Disadvantages:
* They continue to trade at any price can be adverse
* May not fill the order in the specified time in there is a lack of trading.
Volume-Weighted Average Price (VWAP)
- Schedule Algorithm
- Weighted average of execution price during the day where the weight is applied is the proportion of the days trading volume
- Breaks the trade down and sending the order based on historical intraday volumes
- Trades more at the opening and the closing not the trading day and not so much in the middle of the day
Disadvantages:
* Not useful if a trader is a significantly part of the trading volume
* Unethical; Can simply trade at the end of the day to execute a trade order
* If prices are moving Down = only execute buy orders
* If prices are moving Up = only execute sale which would be above the VWAP
Time-Weighted Average Price (TWAP)
- Schedule Algorithm
- Used for thinly traded stocks
- Is the equal weighted average price of all trades over a specific trading horizon
- Ignores actual volume good for highly fluctuating volume throughout the day.
- Excludes Potential Trade Outliers
- Equal number of shares are being traded over the period of the day.
Liquidity Seeking Algorithms
- Opportunistic Algorithms
- Try to take advantage of favorable liquidity conditions
- Will use both Lit and Dark venues to get their trades done
- For large orders in less liquid markets
- Have a high urgency to mitigate market impact costs
- Concerns for when management displays their limit orders that can lead to information leakages.
Arrival Price Algorithms
- Try to trade close to market prices, prevailing at the time the trade is entered
- Trades more aggressively than other algorithms
- For small orders in more liquid markets
- For managers that believe the price will move adversely and has a high urgency to trade
- Market Impact cost increased; execution risk decreased
- For profit seeking management with high alpha decay.
Dark Strategies Liquidity Aggregators
- Execute strategies in the dark pools venues
- Aggregators trying to optimize the trade across multiple dark venues
- For large orders in illiquid markets
- For management that does not have to execute the whole order immediately
Smart Order Routers (SOR)
- Algorithmics that try to determine the best destination (Dark or Lit venues) to route an electronic order
- Tries to find the best market pricing
- Appropriate for small market orders, low market impact, where the price/market moves very quickly
- For small limit orders with low information leakage
Crossing Network Order
Electronic trading order with anonymity (not urgent trading)
Execution Price
- The actual price a security was bought or sold at
- Relates to shares transacted in the market and corresponds to price drift from buying (which can increase the market price) or selling (which can decrease the market price).
Implementation Shortfall
- Another way to calculate trade cost but is more comprehensive
- Takes a portfolio wide perspective
- Most enact approach to cost measurement
- It captures all elements of transaction costs
- Can be decomposed into all different costs
- Is a frontloaded strategy that can be adjusted to aggressively execute an order when the order has a high urgency.
Paper Portfolio vs Actual Portfolio Return
- Paper return: hypothetical return, at the original decision price with not cost
- Actual return: net of all the cost
Difference Between is the total cost of executing the trade can be presented
* A dollar amount
* Percentage amount
* Per share amount of the initial order
* Basis points amount
Mid-Quote
The average of the best bid and best offer.
* Uses a benchmark to see if it’s too over/under valued
* Can be used to calculate an effective spread
* If the mid-quote is not be available, cannot use a mid-quote as a benchmark
* To measure your implicit cost another alternative to use is the VWAP
Cancellation Price
Market price of a security is order is not fully executed on the remaining portion of the traded is canceled
Keys Areas of the Trade Policy for the Trade Management
- Best Execution: Meaning of
- Optimal Trading Approach: Factors that determine
- Eligible Brokers and Venues: List of the approved
- Monitoring: Any details of the process that are used by the asset management
Best Execution Trade-Offs
- Nature of the trade and order size
- Execution price
- Trading costs
- Execution speed
- Execution risk
- Settlement risk
Performance Measurement
- Componet of Portfolio Evaluation
- Calculating both a return and the risk over a time period(s)
- Can be looked at it on a relative or absolute benchmark (target, specify in advance)
Performance Attribution
- Componet of Portfolio Evaluation
- Determine what are the key factors generated the performance of the account.
- Explains how the return was achieved given the level of risk taken
Performance Appraisal
- Componet of Portfolio Evaluation
- Looks at different ratios to see if the performance was by effected by investment decision, the market price, or by chance
- Renders a professional judgement of the fund
- Design to asses if the portfolio manager where the investment results are due to skill or luck
- Can the portfolio manager outperform their benchmark on a risk adjusted basis consistently
Return Attribution
Evaluates the impact that active management decisions have on the funds returns
Risk Attribution
- Done in sync with return attribution and analyzes the active management decisions and impact on portfolio risk
- Risk is generally compared to portfolio appropriate benchmark and also calculated on absolute terms
Micro Attribution Analysis
- Done at the individual Portfolio management level (the specific manager)
- Tries to determine or verify what the portfolio managers actual did, what they said they would do, and to understand what are the actual factors/drivers of the portfolio return
Macro Attribution
- Done on a fund sponsor level
- Quantifying the fund sponsor decisions to deviate from the SAA or TAA and the timing when they made their decisions
Equity Attribution Method’s
- Brinson-Hood-Beebower (BHB)
- Factor-based Attribution Model
Brinson-Hood-Beebower (BHB) Model
Quantifies a portfolio return into 3 attribution effects this shows the value added
1. Allocation effect: Pure selector allocation
2. Security selection effect: security selection; within sector selection
3. Interaction Effect: allocation selection interaction
Allocation Effect
- Decision to under/over-weight specific sector weightings in the portfolio vs the benchmark
- Purely capture the portfolio managers ability to select outperforming sectors and avoid underperforming ones.
Selection Effect: Within Sector Selection
- Was the decision to under/over-weight within certain sectors vs a benchmark a good or bad decision
- Portfolio managers stock picking skills
Interaction Effect
- “Plug” residual amount of the benchmark
- It considered to be a really good if it captures the style and the effect it should be zero
- It must sum up the total return with the others effects (allocation and selcetion) to 100%
Fundamental Factor Model
Used where a portfolio’s sensitively to additional factor can be tested and example is the Carhart Model.
Exposure Decompositions
- Duration based
- Top-Down approach, process the segments risk by same specific attributed or characteristics
- Tries to quantify what arises from interest rate risk
- Segments the portfolio by its market value weights
- Assigns securities to the difference duration bucket, by the nature of each security
Yield Curve Decomposition - Duration Based
- Can be top down or bottom up
- Uses both YTM and duration to calculate price return
- Looks what factors that drive the return when the YTM changes
- It requires a lot more data because it has to performed on the portfolio and the benchmark
Yield Curve Decomposition – Full Pricing Based
- Alternative to the Yield Curve Decomposition
- Securities can be repriced by using spot rates
- The use of spot rates is known as the full repricing method
- It is the most accurate measure of reprising securities as its more in depth
- It will bemore difficult and costly.
Liability-Based Benchmark
- Used when a specific liability needs to be paid in the future
- Will focus on the cash flows when they are needed as the liabilities come due
- It might limit the number and type of investment choices such as inflation-adjusted bonds, nominal bonds, high quality stock
What are the qualities of a good benchmark
S – Specified in Advance: The benchmark is known by the portfolio manager and fund sponsor before the investment period
A – Appropriate: the benchmark has a consistent style or approach
M – Measurable: The benchmark’s value or calculated on a reasonable basis
U – Unambiguous: Clearly identities and weights of securities that make up the benchmark
R – Reflective of Current Investment Options: Whatever is in the benchmark or expertise of those securities in the benchmark
A – Accountable: portfolio manager accepts the applicability of the benchmark in advance and agrees to it any difference is only their active decisions make
I – Investable: We can replicate the benchmark passively and can forgo active management if desired.
Absolute Benchmark
- A return objective that tries to get a desired return that exceeds a minimum target return, or return of a specific percentage
- Disadvantage: Will fail the investability test, not an investable benchmark
Broad-Based Market Index Benchmark
- Measures performance of a broad category of assets (like the S&P 500)
- Disadvantage: May be inappropriate if the management style deviates from the style reflected in the index
Style Index Benchmark
- Represents specific portions of an asset category based on natural characteristics (value or growth)
Disadvantage:
* Some of the style indexes can contain weights on security and sectors that might not be appropriate or prudent
* Different definitions of investment style
Factor-Model Benchmark
- Specify a set of a factor exposures to a return of an account measures the sensitivity to risk exposures
Disadvantage:
* Focusing on factor exposure might not be easily explained
* The data and modeling are not always available and might be expensive
* May be ambiguous which can lead to misspecifications
Return-Based Benchmark
- Looks at portfolio return and tries to explain the return
- Based on sensitivity to style indexes/approaches
Disadvantage:
* Might not reflect what is currently owned or is willing to own
* Not good at detecting (Style drift)
* For any meaningful analysis we need enough data to come up with a statical significant pattern of style exposures
Manager Universe Benchmark
- Median Manager, the fund at the middle, where funds are ranked from highest to lowest
Disadvantage:
* Doesn’t fit any of the criteria but being measurable
* Subject to the survivorship bias (not unbiased)
Customized Benchmark
- Satisfies all of the criteria
- Designed to reflect the management security allocation and investment process
- Also, knowns as a “normal portfolio” the universe of all the potential securities that are investable the management can select from
Disadvantage:
* If you allow a manager to create their own benchmark they will use one that is easy to beat
* Expensive to contract and maintain them
Sharpe Ratio
- Assumes returns are normally distributed risk adjusted measure of performance
- Considers the risk relative return per unit of risk
Disadvantage:
* It doesn’t differentiate between volatility that is upward or downward; it takes all volatility into account.
* There is a penalty for all volatility even good volatility
Treynor Measure
- Similar to the Sharpe-ratio but uses beta only considers systematic risk (by using beta)
- Considers the risk-relative return per unit of systematic risk
- Disadvantage: Only useful to evaluate portfolios that are well diversified.
Information Ratio
- Ex-ante future returns expected it’s a standard deviation of error
- Used to measure a portfolios performance with a benchmark, but accounts for differences in risks.
- Rewards per incremental unit of risk created by deviating from the benchmark holdings
Appraisal Ratio
- Raito of returns to risks from active management
- The active return per unit for a unit of active risks
- Considers the managers value per unit of non-systemic risk
Sortino Raito
- Very Similar to the Sharpe ratio
- Only considers the standard deviation of the downside risk
- Clients are more concerned with risks of negative returns
- Penalizes bad volatility by considering returns only below the minimum accepted return
- For investments that have non-systematical or skew return distributions
- Considers target-relative return per unit of downside risks
Capture Ratio
- Determines the management relative performance when a market is either up or down
- This both would be the opposite in an upward market
- Downside capture vs upside capture
If the capture ratio is:
* >1 positively asymmetrical convex return profile more upside than downside
* <1 negative asymmetrical concave return profile more downside vs upside
Drawdown
- Represents the peak to trough loss in portfolio value during a continues period of negative returns
- Measures the total time from the start of the drawdown until recovery to the previous peak
Can be broken-down into:
* Drawdown Phase
* Recovery Phase
Drawdown Duration
Total time that is full recover form a drawdown when a drawdown begins to when up to a cumulative drawdown reverse cand becomes zero
Maximum Drawdown
Occurs at the very end of the drawdown phase and right before it hits the recovery phase it’s the point at which the cumulative drawdown is at its highest
Time Weighted Rate of Return
is used to evaluate the external managers returns because they ignore factors they have no discretion.
Type 1 Error
- Rejecting the HO when in fact it is true
- Believe the manager adds value when they do not
- Are easier to determine because we can measure a manager’s vs a benchmark
- Results in costs associated in retaining management is weak
Type 2 Error
- Accepting the HO when in fact it is false, Beta Error
- Believe the manager does not add value, when they do
- Errors are harder to determine as how do we compare them (there are opportunity costs)
- If there is an excessive number of type-2 errors indicates that there is a problem of hiring and firing
- Bad decision result in cost associated with not retaining these managers
Solutions to Prevent Type 2 Errors:
* Reduce by tracking the subsequent performance of those managers that were not hired or fired
* When there is any type of personal changes should not be done by short-term performance
AUM fees
- Fees based on a percentage of assets under management.
- The assets will grow due to manager skill and effective marketing
- Over the short term, market cycles can affect AUM for reasons not linked to the manager.
- The principal-agent problem can occur
- This potential conflict is addressed through performance-based fees.
Performance-based fees
- Fees designed to reward manager performance.
- May be based on absolute returns, or benchmark-relative returns.
- Symmetrical exposure: Manager earns a base fee and shares in the gains/losses of the fund
- Capped symmetrical exposure: Manager shares in gains above the base fee up to a limit, but not losses
Asymmetrical (upside) exposure:
* The manager shares in gains above the base fee, but not losses
* Offers the manager unlimited upside and capped downside, similar to a call option.
Issue with Performance Fees
- Create tension when base fees must be paid in periods of poor performance.
- Reduce variability in net returns
- Cause managers to unethically favor clients
- Cause delay in realizing investments until performance fees can be earned
- Cause hedge fund managers to return capital to investors when significantly below the high-water mark
Transactions-based Attribution Method
- Most effective for active stock selection portfolios
- Captures both the holdings and the transactions (purchases/sales) completed within the period
- Allow the entire excess return to be quantified and explained
Returns-based Attribution Method
Most appropriate when the underlying portfolio holdings are not readily available with sufficient frequency at the required level of detail (hedge funds).
Holdings-based Attribution Method
- Most appropriate for investment strategies with little turnover (passive strategies)
- Only references the beginning-of-period and end-of-period holdings and ignores individual transactions.
Quality of Investment Personnel
- Sufficient expertise and experience
- Sufficient depth to execute the strategy
- Exposure to key-person risk
- Incentives to reduce personnel turnover
- Consistent investment strategy
Trade Error Logs
- Trade policy document should describe the factors used in determining how an order can be executed in an optimal manner for a given scenario.
- Any resulting gains/losses need to be disclosed to the compliance department and documented in a trade error log.
- The priority is to ensure errors are resolved in a way that prevents adverse impact for the client, not to ensure complete disclosure.
