CFA Book 3 Flashcards
Bond management | active vs passive
- passive: manager agrees with mkt >> mimics mkt
- active: manager does not completely agree with mkt >> invests diff from mkt index >> alpha
Bond management | managing against liabilities
ALM with math
Bond management | 3 things change as go from passive to active
- active management increases
- expected return increases
- tracking error increases
◎ portfolio - index return = alpha
◎ std dev of alpha = tracking error
Bond management | 5 degrees of management activity
- pure bond indexing (least)
- enhanced indexing by matching primary risk factors
- enhanced indexing by small risk factor mismatches
- active management by larger risk factor mismatches
- full-blown active management
Bond management | pure bond indexing
- mirror specific securities and weights
- expensive to implement >> lower return than index and rarely implemented
- low fees, risk, tracking error
Bond management | enhanced indexing by matching primary risk factors
- sample of index (fewer securities) to mirror index’s primary risk factors
- lower transaction costs to implement, but high corr to index returns >> beats pure bond index strat, but still often less than index
Bond management | enhanced indexing by small risk factor mismatches
- beginning of active management; reduced manager restrictions
- mirror exposure to large risk factors; same duration as index
- deviate on small risk factors >> increase return >> cover admin costs
- eg. of mismatch: relative value strat
- goal: slightly beat index
- increased risk, tracking error (volatility), management fees
Bond management | active management by larger risk factor mismatches
- increase mismatches more than in the previous level (diff duration vs index) >> higher return >> cover admin and increased transaction costs
- reduced manager restrictions
- eg. mismatch: alter duration of portfolio vs index
- increased risk, tracking error (volatility), management fees
Bond management | full-blown active management
- no restrictions, very aggressive
- tilting/mismatching (risk factors), duration, relative value
- increased risk, tracking error (volatility), management fees
Bond management | 4 criteria for portfolio vs benchmark
- mkt value risk
◎ low risk = short duration - income risk
◎ low risk = long duration - credit risk
◎ portfolio should equal benchmark risk - liability framework risk
◎ managing against liabilities
Bond management | portfolio or index risk profile
- duration
- key rate durations
- duration contributions
- spread durations
- sector weights
- cash flow distributions
- diversification
Bond management | stratified sampling
• goal: mirror index risk/return with fewer securities
• process:
◎ create matrix of risk factors
◎ calc weight of each cell in matrix
◎ mirror risk factors and weights of index
• does NOT need to be same securities as in index, just same risk factor exposure
Bond management | risk profile: duration
- aka effective duration, option adjusted or adjusted duration
- measures change in bond value given small parallel shift in interest rates
- duration is tangent >> underestimate increases/overestimate decreases >> add convexity component
Bond management | risk profile: key duration
• measures bond value given ‘twists’ in yield curve
Bond management | risk profile: present value distribution of cash flows
• aka PVD
• all numbers are PV’d
• process:
◎ calc PV of all cash flows
◎ divide term into equal time periods
◎ find PV of each periods CF
◎ period CF PV / total CF PV = proportion of total duration attributable to that period
◎ duration contribution = period prop of duration * period duration (taken as end of period)
◎ duration contribution / total duration = piece of PVD
◎ aggregate of all pieces = PVD
• PVD shows how duration is distributed across its maturity
• mirroring PVD in a portfolio >> same sensitivities to interest rates changes as benchmark (both parallel and twists)
Bond management | risk profile: sector and quality percent
match sector weights and qualities of benchmark
Bond management | risk profile: sector duration contributions
match proportion of benchmark duration attributable to each sector
Bond management | risk profile: quality spread duration contribution
match proportion of benchmark duration attributable to each quality (ie credit ratings)
Bond managment | risk profile: sector, coupon, maturity cell weights
if callability is an issue, match sector, coupon, maturity of benchmark
Bond management | risk profile: issuer exposure
- aka event exposure
- diversify to reduce exposure to any one issuer
Bond managment | risk profile for MBS: 3 factors (???)
- sector
- prepayment
- convexity risk
Bond managment | non-MBS risk profile summary: issue, metric
- yld curve shift : duration
- yld curve twists : PVD, key rate duration
- spread changes : spread duration
- credit changes : duration contribution by credit rating
- call/put exposure : delta
Bond management | scenario analysis
- poor man’s monte carlo
- providing distribution of possibilities vs a single point estimate
- three levers to change: bond price, coupons, interest on coupons
Bond management | two key factors in bond risk
duration and convexity
Bond immunization | goal
minimize interest rate risk:
◎ price (mkt value) risk
◎ reinvestment risk
• applied to a single or multiple liabilities
Bond immunization | classical immunization
- price risk and reinvestment risk (neg corr) offset each other
- process: set effective duration = liability horizon >> interest rate risk elminated (price risk offsets reinv risk)
- immunized for instantaneous, small parallel shift in yld curve
- often needs to be extended to be sufficient
Bond immunization | process
- select bond w/ effective duration = liability duration
- set bond PV = PV of liability >> DONE
• unmatched duration:
◎ if bond duration < liablity duration >> reinv loss > price gain if rates go down
◎ if bond duration > liability duration >> price loss > reinv gain if rates rise
Bond immunization | duration changes over what two parameters
- time
- interest rates
• unlikely bonds and liabilities change in identical ways >> periodic rebalancing of immunization
Bond immunization | bond characteristics to consider
- credit rating
2 embedded options - liquidity (neccesary for rebalancing)
Bond immunization | immunization risk
- portfolio shortfall risk vs liability
- corr with reinv risk (higher reinv risk >> higher immune risk)
- zero coupon have zero immunization risk
Bond immunization | effective duration
- %change in bond dollar value given 1% change in interest rates (100bp increase)
- weighted avg (mkt val) of effective durations of bonds in portfolio
Bonds | duration contribution
contribution of bond (individual or by some grouping) to portfolio duration = wD
w = bond (mkt val) weight in portfolio = bond mkt val / portfolio mkt val
D = effective duration of bond
• sum of contributions = portfolio duration
Bonds | bond vs grouping of bonds when calcing duration
effective duration (weighted by mkt val), duration contribution (weighted by mkt val), dollar duration (not weighted, sum of each bond’s DD) can all be calced on an individual bond or a group of bonds
Bonds | dollar duration
DD = -1 * 0.01 * (modified or effective duration) * mkt val
Bonds | dollar duration: rebalancing
• process:
◎ calc new DD
◎ rebalancing ratio = DDold / DDnew
◎ (rebalancing ratio - 1) * new portfolio mkt value = new dollar val of bonds to be bot/sold (in same proportion as original portfolio)
• alt: buy/sell bond (the controlling position) with greatest duration >> lower dollar amount req to rebalance
Bonds | 3 type of spread duration
- nominal spread: %chg price / 100bp chg in nominal spread
- zero-volatility spread (aka static spread) = %chg price / 100bp chg zero-vol spread
- option adjusted spread (OAS) = %chg price / 100bp change in OAS
Bonds | sum of mkt val weighted sector spread durations =
portfolio spread duration
Bonds | spread duration measures
100bps parallel shift in spread over Treasuries
• spread measures risk aversion
Bond immunization | 4 immunization extensions
- multifunctional duration (aka key rate duration)
- multiple-liability immunization (multiple horizons)
- relaxation of min risk requirement (allow more risk >> higher returns)
- contingent immunization (mix of active and passive strats)
Bond immunization | contingent immunization: goal and process
• goal: return maximization
• process:
1. set safety net return
2. pursue higher returns with active management 3. if hit safety net return >> immunize to lock in safety net return
• must have procedures to monitor for safety net trigger
Bond immunization | contingent immunization: dollar safety margin
- = Current mkt val - liability PV discounted at immunized rate (current obtainable rate)
- current assets PV - liability PV
Bond immunization | contingent immunization: safety net trigger
If PV of asset CF < liability PV discounted at immunized rate (NEW current obtainable rate) >> safety net is triggered >> passive management
Bond immunization | contingnent immunization: what can go wrong
- mkt can move to fast and no chance to lock in immunization rate
- immunization rate might not be achieved once triggered (???)
Bond immunization | immunization risk: ALM related
- interest rate risk
◎ match duration and convexity
◎ neg convexity hard to match - contingent claim risk (call, prepay)
- cap risk
• convexity (non-linear chg) key issue
Bond immunization | minimizing reinvestment risk when managing to a liability
- minimize distribution of cash flow around liability horizon
- bullet strategy best, barbell worst
- if maturity too early >> reinv risk; if too late >> interest rate risk (sensitive to int rate chgs)
Bond immunization | maturity variance
- aka M^2
- var of asset maturities around liability maturity date
- lower >> better
Bond immunization | cash flows used for immunization that happen at liability maturity
- treat as a zero coupon bond in the immunization portfolio
- relates to expected cash flow that arrive around or at the time the liability matures/is due
Bond immunization | cash flow matching
• process:
◎ select bond that matures on last liability maturity date
◎ combine that bond’s extra cash flows with another bond to meet 2nd to last liability maturity
◎ and so on
• more costly than duration matching
• pros:
◎ reduces non-parallel shift risk
◎ no duration rebalancing necessary
• always have assets mature before liability
• all cash flows from bonds are included >> reinv risk is important
Bond immunization | duration vs cash flow matching
- duration requires duration rebalancing; cash flow does not
- cash flow reduces non-parallel yld curve movement risk
- duration matching is easier and less expensive
- zero-coupon bonds >> eliminate all immunization risk
Bond immunization | combination matching
- aka horizon matching
- combination of multiple liability matching (duration matching) and cash flow matching
- usually entails matching cash flow for near term liabilities
- cash flow matching reduces non-parallel yld curve movement risk and eliminates need to rebalance
- more expensive than multi-liability matching
Bond relative value | bond relative value analysis
• relative value of bonds is compared across:
◎ sector
◎ issuer
◎ duration
◎ structure
Bond relative value | 2 general approaches to bond relative value
- top-down: macroeconomic
- bottom-up: specific undervalued issues
• same as with equities
Bond relative value | classic relative value analysis
- use both top-down and bottom-up
- look at total return
Bonds | secular
a long term time frame (eg. 10 yrs)
Bonds | cyclical changes
- chg in supply >> chg in demand
- increase supply >>good credit signal to demand >> lower credit spreads & higher prices
- reduced supply >> bad credit signal to demand >> higher credit spread & lower prices
Bonds | secular changes
• corp bond mkt is dominated by intermediate term, bullet structures
• high yld mkt dominated by callable issues
◎ this is expected to decline as credit quality improves >> easier credit
• 3 results:
◎ bonds with embedded options trade at premium due to scarcity
◎ longer durations trade at premium due to scarcity
◎ credit-based derivatives will increase in use for diversification and tailoring structure
Bonds | liquidity vs risk
- greater liquidity >> lower risk premium
- bonds becoming more liquid due to cheaper, faster transactions due to technology and other factors
Bonds | secondary market purpose
- yld spread/pickup trades: more yld given credit rating (ignores total return)
- credit upside trades: buy before credit upgrade
- credit defense trades: sell bonds in risk of downgrade
- new issues swap: swap into newer issues for liquidity
- sector rotation: move into outperforming sectors; out of underperforming
- yld curve adjustment: adjust portfolio given yld curve forecast
- structure trades: adjust portfolio given volatility and yld curve forecast
- cash flow reinvestment
Bonds | 3 yield spread measures
- nominal spread
◎ corp yld - govt yld of similar maturities
◎ basic unit of price and relative value analysis for global mkt - swap spreads:rate paid by fixed-rate payer over on-the-run T’s with same maturity
◎ widely used in Europe for indication of credit spreads - option adjust spread
◎ good for comparing corp vs MBS
◎ declining due to reduction in bonds w/ optionality
Bond relative value | 3 spread analysis methods
- spread mean reversion
◎ most common reason - quality spread analysis: high credit quality spreads vs low credit quality spreads
- percentage yld spread analysis: corp yld / T ylds with simiar duration
◎ not commonly used
Bond relative value | bond structures: bullet
• short term: 1 - 5 yrs
• medium term: 5 - 12 yrs
◎ most common (esp in europe)
• long term: 12 - 30 yrs
Bond relative value | bond structures: early retirement provision
• callable bonds:
◎ underperform with rates rise
◎ outperform when rates fall
• sinking funds
◎ portion of issue is retired on a schedule
◎ don’t fall as much when rates rise due to buy back provision
• putable bonds: few bonds have puts
Bond relative value | credit analysis factors
- corp: ability to pay
- ABS: collateral & servicer
- Municipal: ability to assess and collect taxes
- govt: ability (econ) and willingness (political) to pay
Bond relative value | methodolgies
- total return analysis
- primary mkt analysis (supply/demand effects)
- liquidity and trading analysis (liquidity effects)
- secondary trading rationales
- secondary trading constraints
- spread analysis (volatility effects; mean rev, quality spread, % yld)
- structural analysis (derivatives + macro econ)
- credit curve analysis (yld curve + macro econ)
- credit analysis (up/down grades)
- asset allocation / sector analysis (econ: sectors/firm performance)
Bond relative value | secondary trading constraints
- portfolio constraints
◎ quality
◎ structure
◎ foreign bonds
◎ floating rate for commercial banks
◎ high yld limit for insurance firms
◎ structure, quality for european investors
cause of global bond mkt inefficiency - ‘story’ disagreement: buy/sell side disagreement
- buy and hold
◎ desire not to recoganize loss - seasonality
◎ trading slows at end of mon, qtr, yr due to paper work
• cause of global bond mkt inefficiency
Leverage effect on p/l
- magnifies outcome, good or bad
- return generated on assets, but return rate is off of equity
Levered return calculation
Re = Ri + (D/E * (Ri - Rd)
Re = return on equity
Ri = return on assets
D/E = debt/equity ratio
Rd = cost of debt
• as leverage OR investment return increase, return variability increases
Levered duration calculation
De = (Di \* I - Db \* B) / E De = equity duration Di = asset duration Db = borrowed funds duration I = assets B = debt E = equity
repurchase agreement
• repo agreement: collateralized loan
• repo rate = loan rate
• process:
◎ A sells B a security with a promise to repurchase with a specified date and price
◎ A buys the security back from B
◎ diff in price is the cost of borrowing (at the repo rate)
• 360 days/yr
repo agreement | 4 collateral delivery methods
- none: small credit risk or short loan period
- physical delivery (most expensive)
- custodial acct at borrower’s clearing firm
- electronic bank transfer (expensive)
repo agreement | factors affecting repo rate
- credit risk & delivery
◎ high credit risk & no delivery >> high repo rate - quality of collateral
- repo term: longer >> higher repo rate
- collateral scarcity: if scarce and lender needs it >> lower repo rate
- seasonal factors
• fed funds rate (bank to bank) is repo rate benchmark
seasonal factors
not Christmas, but end of period (mon, qtr, yr) where trading slows due to paper work
bonds | risk measures: standard deviation: cons
- bond returns often not norm dist
- many inputs into calc: N * (N + 1) / 2, N = # bonds
- estimates for inputs is hard
bonds | risk measures: semi-variance
• only measures variance below the mean
• cons:
◎ computation is hard
◎ if returns symmetric >> same as var, but var is better understood
◎ if returns not symmetric, semi-var may not be the best
◎ estimated with 1/2 the dist >> smaller sample >> less statistifcally accurate
bonds | risk measures: shortfall risk
- prob return will be below target
- cons: no info as to size of shortfall
- similar to VAR
bonds | risk measures: value at risk (VAR)
- prob, given a period of time, return will be less than target
- con: no info as to size magnitude of shortfall
- similar to shortfall
bonds | futures contracts
- adjust portfolio without as big capital outlay • contracts not securities
- 30 days - 30 years
- react the same as bonds, but exactly the same
- exchange: CBOT
bonds | futures contracts: cheapest to deliver bond
- CTD
- exchange allows short position to deliver various bonds to fulfill contract
- options: quality or swap, timing, wild card
- use conversion factor (from exchange) to calc correct price for bonds delivered
bonds | 3 advantages of futures contracts
- more liquid
- less expensive
- easier to short
bonds | futures contracts: duration
- same as bonds
- buy futures >> increase duration
- sell futures >> decrease duration
bonds | duration vs dollar duration
dollar duration = duration * 0.01 * mkt value
bonds | futures contracts: hedging bonds
DDt = DDp + DDfut
DDt = total DD
DDp = portfolio DD
DDf = futures DD
• if DDfut > 0, then but futures; if
bonds | futures contracts: calcing # of futures to buy/sell
• # contracts = (DDt - DDp) / DDf
DDt = total DD
DDp = portfolio DD, DDf = ONE futures DD
• OR # contracts = (Dt - Dp) * Pp * CTD conversion factor / (Dctd * Pctd)
D = duration
P = price
trade horizon date
date when the trade is complete, taken off, matures, etc
bonds | futures contracts: basis risk
- risk the price basis will not be the same at the horizon date
- price basis = spot price - futures delivery price
- risk of using futures rather than identical bonds; more risky with ‘cross hedge’