Advanced Applied Finance Javier Estrada Winter, 2014 Assessment Background: Mean Returns AM v . GM v . DWM Javier Estrada Three different questions, three different answers IESE Business • What has been the return in a ‘typical’ period? School The AM ( Not as widely used as typically believed) Barcelona Spain • What has been the periodic rate at which a passively‐ invested capital evolved over time, compounded ? The GM ( This is what we typically call ‘mean return’) Remember: AM ≥ GM and AM‒GM = f + (Volatility) • What has been the periodic return of an active investor? The DWM (The investor’s IRR) Remember: DWM ≷ GM The difference between GM and DWM goes to the heart of the active/passive management debate ADFIN Winter/2014 1
Long ‐ Term Trends – Returns Motivation Javier Estrada Long‐term trends provide useful perspective IESE Business Finance models are long ‐term (equilibrium) models School Evidence: In the long‐term … Barcelona Spain S outperformed B in every country • The compounding power of S is much higher The ERP varies substantially across countries (Remember this when using the CAPM) World‐market ERP ≈ 5% B reduced purchasing power in many countries Q : Can the difference in returns between S and B be explained by differences in risk? ADFIN Winter/2014 Long ‐ Term Trends – Risk First and foremost Javier Estrada Risk just cannot be assessed independently from IESE Business the holding period School What is risk? Barcelona Spain In the short term • Volatility, spreads, worst‐case scenarios, … The evidence suggests that S are riskier than B In the long term • Shortfall probability (Key benchmark: Inflation) The evidence suggests that S are less risky than B Time diversification The longer the holding period, the more likely is a higher exposure to risk to turn into higher return • This idea is at the heart of lifecycle strategies and ADFIN standard advice about asset allocation Winter/2014 2
Long ‐ Term Trends – Forecasting Predictability tends to increase with … Javier Estrada the aggregation of the portfolio IESE Business the length of the holding period School E + ∆ P/E 1 ) The RDM ( R 1 ≈ DY 0 + g 1 Barcelona Spain Observed and expected returns must be the sum of these three components Returns have investment/speculative components • Long‐term returns basically are investment returns Valuation plays a critical role in long ‐term forecasts • But valuation does not give timing signals Usefulness of the forecasting matrix • What are the conditions that support a prediction? How plausible are those conditions? ADFIN Winter/2014 Risk Revisited – Downside Risk An alternative to the standard framework Javier Estrada Focuses on the way most investors think about risk IESE Business • Aims to isolate ‘bad’ outcomes (downside potential) School Measures of downside risk Barcelona Spain Semideviation • Measures volatility below a chosen benchmark • Introduces a distinction between good/bad volatility VaR (Value at Risk) • Measures ‘really bad’ observed/expected outcomes • Easy to interpret and communicate • Trivial to calculate only under normality Most of the time normality may be badly misleading Downside beta and Morningstar risk ADFIN Winter/2014 3
Risk Revisited – The 3 ‐ Factor Model An alternative way to estimate required returns Javier Estrada Accounts for the size and value effects IESE Business Has CF and PM applications School Keep in mind: The 3FM … Barcelona Spain is based on pervasive and global evidence assumes that value and small‐cap stocks are riskier • But it is controversial whether this is really the case argues that higher exposure to the size/value effects (higher risk) calls for higher required return • Argues that size/value tilts increase expected return is very widely used in PM • Estimation of a manager’s alpha ADFIN Winter/2014 Risk & Return – Risk ‐ Adjusted Return Performance can be the result of luck, risk taking, Javier Estrada or skill (or, typically, a mix of them) IESE Business No reason to reward being lucky or taking risk School • The only reason to pay a manager is for his ‘skill’ Barcelona Spain Most of the popular rankings are flawed • Based on short‐term returns • Skill can be assessed by ranking managers by their risk ‐ adjusted , long ‐term performance Measures of risk‐adjusted performance Jensen and Treynor: Based on beta Sharpe and RAP: Based on volatility Sortino: Based on downside volatility Information ratio: Based on alpha and its volatility ADFIN Winter/2014 4
Risk & Return – Stars and Costs Stars (Morningstar) Javier Estrada Returns should be measured net of all costs IESE Business • Hence cost ‐ and ‐ risk ‐adjusted return School • This is what the Morningstar stars measure Barcelona Spain Proprietary, objective, relative, and easy to explain Costs Explicit (Public) • Loads, management fee, incentive fee ‘Hidden’ (Not known ex‐ante) • Trading, bid‐ask spreads, taxes Critical factor determining/ predicting performance This is the big edge of passive management • The less you pay, the more you get • Keep a very close eye on the fees you pay ADFIN Winter/2014 Portfolio Optimization – Excel Optimization tool Javier Estrada Can easily handle … IESE Business • any goal School • any number of assets Barcelona Spain • any number and type of restrictions Inputs • ERs and Var‐Cov matrix (and sometimes R f ) Remember GIGO! Remember observed short/long‐term patterns Remember the portfolio’s intended holding period Outputs • Weights ⇒ E p / SD p / S p / GM p ADFIN Winter/2014 5
Portfolio Optimization – GMM GMM is an attractive alternative to SRM Javier Estrada Designed to deal with a multiperiod horizon and IESE Business the reinvestment of capital School Maximizes the probability than GM p and W T will be Barcelona Spain higher than with any other strategy Typically yields concentrated and volatile portfolios Particularly plausible for … • aggressive investors • long‐term investors • investors unlikely to have to bail out along the way Simple to implement • Requires the same information than SRM ADFIN Winter/2014 Emerging Markets – Performance During 1988‐2013 the evidence shows … Javier Estrada EMs equity … IESE Business • Produced annualized returns of 12.1% with volatility School of 23.6% Barcelona Spain • Outperformed the US/Europe/World markets (Not so over the past five years / Still below ‘07 peak) • Still provide substantial diversification benefits • Call for a 10‐15% allocation of equity portfolio EMs debt … • Produced annualized returns of 9.8% with volatility of 13.9% (‘DM‐equity‐like’ performance) • Were clearly re‐rated in the last few years • Had wildly ‐fluctuating spreads ADFIN Winter/2014 6
Emerging Markets – Risk The popular perception is that EMs are ‘risky’ Javier Estrada This is largely due to a focus on … IESE Business • the volatility of individual (equity) markets School • political risk Barcelona Spain Big question 1: Should it be priced? (Is it diversifiable?) Big question 2: How to measure it? (Spreads, CCRs, …) But EMs are much less risky when considered appropriately from a portfolio perspective • Correlations across EMs are relatively low A large part of the risk gets diversified away Critical issue: Where to account for risk? In the CFs (scenarios)? In the DR (risk premium)? • These two alternatives are interconnected • Remember the implications of an arbitrarily‐high DR ADFIN Winter/2014 Emerging Markets – Cost of Capital Models Javier Estrada L: Systematic country and industry risk IESE Business GE: Yield spread / Total country risk School GS: Improves GE’s double‐counting adjustment Barcelona Spain SSB: Detailed political risk adjustment / Subjective Relevant issues to keep in mind There is a wide variety of proposed methodologies, none of which is currently widely accepted Risk adjustments can be made through the DR or through scenarios in CFs Avoid an arbitrary estimation of the DR • Can you defend your approach and DR estimate? ADFIN Winter/2014 7
Recommend
More recommend
