Docstoc

Equity portfolio management strategies

Document Sample
Equity portfolio management strategies Powered By Docstoc
					Equity portfolio management strategies
Objective
Outline
               Portfolio management style



Passive
Buy and hold strategy, often known as indexing



Active
Continuos rebalancing
                          Passive management


Objective                          Techniques
Match the return of a benchmark
                                   Full replication
                                   Issues: Transaction costs
Approach
Replicate the benchmark            Sampling
                                   Issues: Tracking error



                                   Quadratic optimization
                                   Issues: Programming



                                   Completeness funds
                                   Issues: Special benchmark to complement active
                                   portfolio management
                         Active management



Objective
Outperform a passive benchmark portfolio on a risk adjusted basis
Portfolio return > Benchmark return + transaction costs


Issues
Benchmark
Measuring returns on a risk adjusted basis
          Themes in active portfolio management


Sector rotation

Value vs. growth

Earnings & price momentum

Factors models
Identify stocks that are sensitive to _________factors


Long-short approach
Screen & rank; buy the top, sell the bottom
                         Style analysis


Compare manager’s return to that of different styles of indices

• Grid style
• Regression analysis
                Style analysis: Grid style

Large                         S&P 5000
cap
                             Russel 1000


                            Wilshire 5000

                                    TSE300

                            Russel midcap


                                             Nasdaq

                 Joe B.   Russel 2000



Small
cap

        Value                                Growth
            Style analysis: Regression analysis




R = b1F1 + b2F2 + ….+ bjFj + …. + e


Where:

R = return on the portfolio under analysis

bj = sensitivity of portfolio to style factor j


Fj = return on a factor j style portfolio
        Style analysis: Regression interpretation



Look for (bj)s that are large and significant


They reveal which factor style portfolios are similar to the portfolio
under analysis
          Asset allocation strategies



Integrated asset allocation

Strategic asset allocation

Tactical asset allocation

Insured asset allocation
                Integrated asset allocation



Evaluate and integrate:

• Capital market conditions
• Investor’s objectives & constraints
                      Integrated asset allocation

                                           Investor’s assets, liabilities, and net
                                                          worth
   Capital market conditions




                                                 Investor’s risk tolerance
           Predictions                                   function




    Expected returns, risk,                        Investor’s objectives
        correlations




           Optimized portfolio: asset allocation & security selection


Feedback                                                           Feedback

                         Return evaluation & feedback
                   Strategic asset allocation



Classical optimization: It results in a constant asset allocation mix

Similar to integrated asset allocation, without a feedback loop

Exemplification:
• Pension plans
                Tactical asset allocation


Assumption
Mean reversion

Aka. timing the market

It’s a contrarian strategy:
“Buy low, sell high”
                   Insured asset allocation

Assumption
Returns & risks constant over time, but investors change

• Switch between equity & cash to accommodate investor’s risk
  tolerance

• Similar to integrated asset allocation without feedback on the
  capital market side
          Evaluation of portfolio performance



Requirements of a good portfolio manager:

• Derive no less than average returns for a given risk-class
  (timing & security selection skills)

• Diversify away all non-systematic risk
         Approaches to measuring performance




•   Peer-group comparisons
•   Treynor’s composite measure
•   Shapre’s measure
•   Jensen’s measure
•   Fama’s approach
•   Attribution analysis
•   Market timing skills measurement
                 Peer-group comparisons



Ranking can be random

Most data tracks funds, not individual portfolio managers

See also textbook
               Treynor’s composite measure

Comparison of risk premium per unit of relative risk

Measure
Ti = (Ri - Rf)/bi

Benchmark
Tm = (Rm - Rf)bm


Issues
• Looks at performance only
• Uses realized returns
                      Sharpe’s measure
Comparison of risk premium per unit of absolute risk

Measure
Si = (Ri - Rf)/si

Benchmark
Sm = (Rm - Rf)sm


Issues
• Looks at performance and diversification
• Uses realized returns
                           Jensen’s measure

Measures excess return (above and beyond that required by the market)


(Ri - Rf) = a + (Rm - Rf)bi + e


If a > 0
Portfolio earned more than the required rate

If a < 0
Portfolio earned less than the required rate


Issues
• Uses realized returns
                       Fama’s approach




Excess return = Portfolio risk + Selectivity

See also textbook
                          Attribution analysis
Attribute performance to:
• Selection
• Tactical asset allocation (market timing)



Allocation effect:
[(wp - wb)stocks(Rbstocks - Rb)] + [(wp - wb) bonds(Rbbonds - Rb)] + …


Selection effect:
[wp(Rp - Rb)]stocks + [wp(Rp - Rb)]bonds + …
              Attribution analysis: Exemplification



Asset class   Portfolio weights   Benchmark weights   Difference
Stock                 0.5               0.6               -0.1
Bonds                0.38               0.3               0.08
Cash                 0.12               0.1               0.02
              Attribution analysis: Exemplification



Asset class   Portfolio weights   Benchmark weights   Difference
Stock                 0.5               0.6               -0.1
Bonds                0.38               0.3               0.08
Cash                 0.12               0.1               0.02
              Attribution analysis: Exemplification



Asset class   Portfolio weights   Benchmark weights    Difference
Stock                 0.5               0.6                -0.1
Bonds                0.38               0.3                0.08
Cash                 0.12               0.1                0.02




Asset class   Portfolio return    Benchmark return    Difference
Stock               9.7%                8.6%              1.1%
Bonds               9.1%                9.2%             -0.1%
Cash                5.6%                5.4%              0.2%
              Attribution analysis: Exemplification



Asset class   Portfolio weights   Benchmark weights   Difference
Stock                 0.5               0.6               -0.1
Bonds                0.38               0.3               0.08
Cash                 0.12               0.1               0.02




Asset class   Portfolio return    Benchmark return    Difference
Stock               9.7%                8.6%             1.1%
Bonds               9.1%                9.2%             -0.1%
Cash                5.6%                5.4%             0.2%
              Attribution analysis: Exemplification



Asset class   Portfolio weights   Benchmark weights   Difference
Stock                 0.5               0.6               -0.1
Bonds                0.38               0.3               0.08
Cash                 0.12               0.1               0.02




Asset class   Portfolio return    Benchmark return    Difference
Stock               9.7%                8.6%             1.1%
Bonds               9.1%                9.2%             -0.1%
Cash                5.6%                5.4%             0.2%
                    Attribution analysis: Exemplification



Portfolio return = (0.5)(9.7%) + (0.38)(9.1%) + (0.12)(5.65) = 8.98%
Benchmark return = (0.6)(8.6%) + (0.3)(9.2%) + (0.1)(5.4) = 8.46%

Allocation effect
(-0.1)(8.6% -8.46%) + (0.08)(9.2%-8.46%) + (0.2)(5.4% - 8.465) = -0.02%

Selection effect
(0.5)(9.7% - 8.6%) + (0.38)(9.1% - 9.2%) + (0.12)(5.6% - 5.4%) = 0.54%

Allocation effect + Selection effect = -0.02% + 0.54% = 8.98% - 8.46%
              Attribution analysis: Interpretation




Manager underperformed benchmark by 0.02% due to deviations
from benchmark’s weight

Manager outperformed the benchmark by 0.54%, due to its
superior selection skills
                           Measuring timing skills


Measure the effectiveness of switching between asset classes
Having perfect timing skills (hindsight 20/20) is equivalent to owning a lookback option:
At expiration, it pays the return of the best-performing asset class.

Ri = Rf + max[(Rb- Rf), (Rst- Rf)]


Regression measure:
(Ri - Rf) = a + (Rb- Rf)bib + (Rst - Rf)bist + y max[(Rb- Rf), (Rst- Rf)] + e

a = excess return
y = proportion of the lookback option captured by manager
            Factors that affect performance measures



Knowing what is the true return generating process
All the above measures are based on CAPM



Finding the real market portfolio
Changing the proxy for the market portfolio completely changes the ranking



Accounting for manager’s style
                    A question of benchmark




Portfolio managers have different objectives and styles.

Wee need customized benchmarks.
         The making of a good benchmark



•   Unambiguous
•   Investable
•   Measurable
•   Appropriate: Consistent with manager’s style
•   Reflective of manager’s current investment opinions
•   Specified in advance