R15 - Application of Economic Analysis to Portfolio Management Flashcards
Steps to develop capital market expectations?
- determine relevant factors for investor: TH, allowable asset classes, other
- Determine historical perf and driving factors of asset classes
- Identify valuation model and methods to use
- Identify best sources of info
- Interpret current market conditions
- Document current capital market expectations
- Monitor results and refine
Capital market forecasts should be:
consistent unbiased objective well supported accurate (min forecasting errors)
What are some limitations in economic data?
- data available in time lag and subject to revision
- inconsistent definitions and calculation among sources
- indexed and rebased
What are some errors and biases that can exist in data?
- transcript error: numbers wrong
- survivorship bias: overstate return, understate risk
- smoothed data on illiquid assets
How smoothing works and consequences?
- illiquid assets infrequently traded/priced
- analysis implicitly assumes continuous price change between two points
- risk calculations unrelated and correlation closer to 0
What are some challenges to forecasting?
-Ex post understate ex ante risk: what actually happened may understate what can happen in future
- history-based estimates have limitations:
- –future diff from past
- –regime change may have happened b/c fundamental factors changed
When selecting time periods:
- longer time periods + more frequent observations increase data quality
- shorter periods less likely to include regime changes
- use longer periods unless:
- –reason to believe fundamentals have changed
- –stat analysis of sub-periods reveals nonstationarity
Analyst forecasting biases?
- data mining - keep analyzing until find pattern even if not real
- time period bias - relationship holds in one period but not another
- failure to condition data - e.g. using past nominal return without considering inflation
- mistaking correlation for causation
- psychological biases
- –anchoring trap
- –status quo
- –confirmation bias
- –overconfidence
- –prudence
- –availability
- model risk
- –selecting wrong model
- –input uncertainty
Using statistical tools for setting capital market expectations:
- If have stationary data, use historical return, variance, and correlations
- –use arithmetic avg for one period estimates of return
- –use geometric avg for multiperiod - Apply shrinkage estimate to historical data
- –i.e. use weighted avg of historical return and model estimated return - Apply time series analysis
- –sometimes see volatility clustering (shortrun volatility can persist) - Use multifactor models when have more than one driver of returns
Benefits of using multifactor models to unify statistical modeling?
- relates returns to common factors
- reduces noise
- factors can generate internally consistent covariance matrix
Gordon growth model:
E(R) = (Div1/P) + g_nominal
Grinold Kroner model:
Return = Div1/P + I + g - (% change in shares outstanding) + (% change in P/E ratio)
risk premium approach with bonds: what factors used?
risk free rate inflation premium default risk liquidity maturity taxes
risk premium approach for equity
long term bond yield + equity risk premium
Financial equilibrium formula for market ERP
ERPi = correlation with global portfolio x standard deviation of I x (ERP_m/std dev_m)
What is full integration of a market?
can fully diversify
What is full segmentation of market?
cannot diversify at all
Calculate ERP considering degree of integration/segmentation:
Steps:
1. Calculate ERP for the asset under full integration (corr x std dev x sharpe ratio of market) and full segmentation each (std dev x sharpe ratio of market)
- Weight ERP based on degree of integration and segmentation
- Add risk free rate to get expected return
- may have to add liquidity premium if in a problem
Beta formula:
beta = Covariance (ri,rm )/Variance of Market
Covariance formula:
Covariance.i,j = B.i x B.j x std dev ^ 2 of market
How inflation behaves at diff points in business cycle:
expansion phase - inflation increasing
declining phase - inflation decreasing
What happens during business cycle: Initial Recovery
Economic effects: exit recession -gov stimulates through policy -consumer confidence increasing -inflation declining initially
Capital market effects:
ST rates low or declining
LT rates bottoming and bond prices peaking
Stocks do well anticipating economic recovery
What happens during business cycle: Early Expansion
Economic effects:
- economy growing faster than trend
- output gap shrinks
- policy less stimulative
- increasing confidence
- inflation low
Capital market effects
- ST rates increasing
- LT rates bottoming/increasing, bond prices start declining
- Stocks improving
What happens during business cycle: Late Expansion
Economic effects:
- GDP growth above trend, but slowing
- policy becoming restrictive
- consumer confidence excessive
- inflation increasing
Capital market effects:
- ST rates increasing
- LT rates increasing with bond prices declining
- Stock prices volatile/topping out
What happens during business cycle: Slowdown
Economic effects:
- GDP still above trend but growth below trend and turning negative
- policy turning neutral
- consumer confidence peaking
- inflation still increasing
Capital market effects:
- ST rates peaking then declining
- LT rates high and the declining, bond returns favorable
- Stock prices declining in anticipation of recession
What happens during business cycle: Recession
Economic effects:
- 2 consecutive quarters of negative real growth
- output gap increasing
- policy easing
- consumer confidence weak
- inflation peaking
Capital market effects:
- ST/LT rates declining
- Bonds doing well
- Stock prices turn up later in recession
