CFA III SS 14 Risk Management

• centralized: identifies, quantifies, reports and monitors firm-wide risks; preferred 
• decentralized: each unit manages own risk, does not consider correlations

• interest rate
• exchange rate
• equity prices
• commodity prices
• credit
• liquidity

first four are market risks

• operational
• model
• settlement
• regulatory
• legal/contract
• tax
• accounting
• sovereign

• estimate of loss at a specified probability over a specified time
• ex. 5% VAR: E(r) - 1.645*stnd dev = VAR (amount below 0)
• ignore E(r) for 1 day VAR
• 1% is 2.33

• 1. Analytical: variance and covariance; easy to apply; relies on normality (which is probably not true)
• 2. Historical: uses actual historic returns; ex if have 100 returns, take the 5th lowest and thats the 95% VAR
• 3. Monte Carlo: same as historic, but use simulation to determine set of returns

• factor push: push factors in direction that will hurt the firm
• maximum loss optimization: optimize mathematically to find the risk variables that will cause the maximum loss
• worst case scenario: guess at what the worst case would be

• current: payment is currently due
• potential: payment may be due or is due in the future

• trading price of option * number of shares (usually number of contracts*100)
• not dependent on amount in the money

• 1. Sharpe
• 2. RAROC
• 3. ROMAD
• 4. Sortino
• 1. 
• 2.  - VAR can be any capital at risk measure
• 3.  where maximum drawdown = max return - min return
• 4.  where MAR = minimum acceptable return (never less than Rf) and downside deviation is deviation of negative returns

• Return = Rdc = (1+Rfc)(1+Rfx) - 1
• Rfc = return of asset in foreign currency
• Rfx = relative change in foreign currency value

Var(Rdc) = Var(Rfc) + Var(Rfx) + 2*std(Rfc)*std(Rfx)*corr(Rfc,Rfx)

If asset is risk free, std(Rdc) = std(Rfx)(1+Rfc)

• passive: match benchmark's exposure
• discretionary: small deviations, but primary goal is still risk reduction
• active: try to earn alpha on currency 
• overlay: separate currency exposure management from asset management

• 1. economic fundamentals
• 2. technical rules
• 3. carry trade
• 4. volatility trading
• 1. assume purchasing power parity holds in long-run and short run deviations can be exploited. Increase in relative currency value associated with low value relative to long-term, low inflation, higher real rates, decreasing currency risk premium
• 2. assumes past prices predict future prices; not trusted by CFAI
• 3. borrow at low interest rate and lend at high interest rate; assumes currency lent in will not depreciate as dictated by IRP; usually works, but very bad when goes wrong
• 4. if volatility will be high, buy put and call; if low, sell put and call. ATM called straddle, OTM called strangle

• (Ft - Fo) - (St - So)
• Since Ft = St at expiration, roll yield at expiration = (So - Fo)/So

hedge currency exposure with another, highly correlated currency

• ex. portfolio is long multiple currencies, hedge with basket instead of each one individually
• cheaper, but less perfect hedge

• MVHR uses regression to determine hedge ratio that will minimize risk
• regress asset returns against currency returns, and multiple the slope coefficient and amount invested to determine amount of currency hedge to obtain

Forward contract in an illiquid currency, so calculate gain/loss and settle in another currency