The Money Management Industry

0
Outline of the lecture

On the topic of money management we will discuss:

• The money management industry
• Differences between active and passive management
• Index funds and ETFs

• Then in the next lecture we will talk about Portfolio Performance

Evaluation where we will evaluate the performance of the fund managers.

1
Introduction
Institutional Investors

• What is an Institutional Investor?

• It is a financial intermediary that:
• Pools funds of participants
• Participants get shares in pool proportionate to their contribution
• Makes buying and selling decisions

investors: institutional investors, retailor investors

regulations

2
Instiitutional Investors

• The main institutional investors are:

• Mutual Funds most of it are retail investors
• Hedge Funds
• Pension Funds
• Others: banks, insurance companies, brokerage firms, university endowments, etc
• Differences between Mutual Fund and Hedge Fund:
• Mutual funds must register shares with SEC and must file-annual reports but hedge
funds are legally exempt management fees
• Hedge funds thus have less restriction on the investment strategies they can pursue
(e.g. riskier) management fees + incentive fee
2 - 20 structure (2%management fee + 20% on the profit incentive)
• See SEC website for regulation:
https://www.sec.gov/Archives/edgar

3
Why do we have Instiitutional Investors

Potential Benefits of Institutional Investors

• Economies of scale allow each investor to benefit from:
• Professional Management (as long as the manager is able to add value –
more on this on the performance evaluation section)
• More diversification

Institutional Investors versus Individual investors

• Mutual funds manage over $8 trillion as of Feb 2005
• The fraction of U.S. equity held by institutions rose from 16% in 1965 to
61% in 2001.
• Institutions accounted for 96% of the trading volume in the NYSE in Nov
2002 (Jones and Lipson, 2002)

US stock mkt: institutional investors that are rational investors

CHINA stock mkt: retail investors are irrational investors
4
Total Net Assets of Open-End Funds (Worldwide, in trillions of US dollars)

5
Total Net Assets by Fund Types (US)

6
Total Net Assets by Fund Types (US)

Index Funds, percent

Total Net Assets grew from $9.5tn to $19.2
while actively managed funds shrank
2007 2017

Index Mutual Funds

Index ETFs

Actively managed Mutual Funds

Actively managed ETFs

0% 20% 40% 60% 80% 0% 20% 40% 60% 80%

Source: ICI Factbook

7
Active vs. Passive Management
The Players

• Many institutional investors follow actively managed strategies (e.g.

hedge funds) while others follow passive strategies.
• Well known fact: active money managers as a whole (or on average)
underperform the market (we’ll see the evidence later)
• Does this mean that no active manager add value?
• No. Some individual money managers might be able to outperform the market
• What is necessary for some managers to add value?
• Good traders can outperform the market only if there are some bad traders who
underperform the market
• Are there any losers?
• Yes. Mostly individual investors.
Some people are simply confused and some managers can take advantage
of that….

8
Who is on the Short End

• Example of confused investors

• Tele-Communications Inc. – tcoma
• Transcontinental Realty Investors – tci
• On October 13, 1993, news came out of a potential merger between Bell
Atlantic Corporation and Tele-Communications Inc., tcoma
• Unsophisticated investors incorrectly bought shares of Transcontinental
Realty Investors (tci), a small and thinly traded real estate stock
• The price of tci rose from $15.50 to $17.875 in just a few minutes
• 56 trades in tci stock, representing 55,000 shares occurred before the
NYSE halted trading in the stock at 10:27am
• the day before, the total trading volume in tci was just 100 shares...
• Trading later resumed and the stock ended at $15.25

9
Passive Managers

• In general, the return from investing on a fund consists of:

Fund Return = Expected Passive Return

+ Return From Active Trading
− Trading Costs
− Management Fees

• Compared with actively managed funds, passive funds have very small (or
zero) returns from active trading. But they also have very low trading
costs and low management fees.

10
Passive Managers 1: Index Funds

Typical approach in passive management: Index Funds

• Example: create a fund to track an index like the S&P 500
• You can do this by creating a portfolio of all the firms (and weights) as in
the index, or just focus on a subset of firms
• Why would you want to hold just a subset of firms? Minimize transaction
costs.
• Tracking Error (TEp ) of Fund P relative to Benchmark B:

TEp = σ(Rp − RB )

• Objective of passive management is to find the weights of portfolio p to

minimize tracking error and transaction costs.
• Manager needs to find the optimal tradeoff between the two

11
Passive Managers 1: Index Funds

Index funds have been growing fast

• The first equity index portfolio was created in July 1971 by Wells and
Fargo Investment Advisors to track the NYSE Composite Index
• The first equity index mutual fund was created in 1976 by Vanguard to
track the S&P 500 index
• Index funds have performed better than active funds on average

12
Passive Managers 1: Index Funds

12
Index Funds: Expense ratios (percent) have declined

13
Index Funds: Total Net Assets and Number have increased (Billions of dollars)

14
Index Funds: Expense ratio of all funds have fallen (percent)

15
The Buffet Challenge

All sources on
https://seekingalpha.com/article/

4107853-warren-buffett-wins-1m-bet-made-decade-ago-s-and-p-500-stock-index-outperform-hedge-funds

16
The Buffet Challenge

• A few years later, Buffett offered to wager $500,000 that no investment

pro could choose a set of at least five hedge funds that would beat the
performance of a low-cost S&P 500 index fund of Buffett’s choosing, over a
10-year period.
• Only one fund manager, Ted Seides, took the bet and chose five “funds of
funds” as his picks. The first year tracked for the purposes of the wager
was 2008, so 2016 represented the ninth full year. Buffett detailed the
year-to-year performance of each investment thus far, and the results
aren’t even close.

16
The Buffet Challenge

2008–2016 Cumulative Returns of Funds in the Buffett Challenge

Hedge funds Index Fund

Fund A 8.7%
Fund B 28.3%
Fund C 62.8%
Fund D 2.9%
Fund E 7.5%
Average 22.0% Index 85.4%

16
The Buffet Challenge

16
The Buffet Challenge

16
Passive Managers 1: Index Funds

Example: Vanguard Group, Inc.

• Passive Management Industry Leader
• Founded in September 24, 1974 by John Bogle
• Main objective: minimize costs while providing high-quality service
• Approach is similar in spirit to the logic of the CAPM: hold the market

17
Passive Managers 1: Index Funds

Example: Vanguard Group, Inc.

17
Passive Managers 2: Exchange Traded Funds (ETFs)

• Exchange-traded Funds (ETFs) have become quite popular recently. They

are closely related to index funds, but they are slightly different.
• What is an ETF?
• Shares in a Trust that holds a basket of stocks
• Essentially, these are index funds that trade like stocks on major stock exchanges
• Examples: SPY (S&P 500), QQQQ (Nasdaq 100), DIA (Dow Jones), IWM (Russel
2000), VTI (Wilshire 5000), EFA (MSCI EAFE), TIP (TIPS), GLD (Gold)
• In January 1993, AMEX began trading Standard and Poor’s Depositary Receipts, or
“SPDR’s”, aka, “spiders”
• Since then, ETFs have grown dramatically…

18
Total Net Assets and Number of ETFs (billions of dollars)

19
How do ETFs work?

20
Passive Managers: ETFs or Index?

What makes ETFs different from passive index funds (IF)?

• Advantages of ETF:
• Can trade ETFs during the day like any common share. In contrast, IF can only be
traded at the end of the day
• Like a common share, you can short ETFs, as well as buy it on margin.
• Advantages of Index Funds
• Dividend policy: in an IF any dividend received is immediately re-invested. The nature
of ETFs is such that they need to hold dividends in cash and distribute to their
shareholders at the end-of-quarter.
• Since ETF are purchased from brokers, you pay a fee. In IF there is very small or no
front-load fees.
• Other subtle differences in tax efficiency etc

21
Active Management
Passive Managers: ETFs or Index?

• In contrast with Passive strategies, many mutual funds (and certainly hedge
funds), follow actively managed strategies
• Active funds have considerably higher fees than passive funds:
• Vanguard S&P 500 index fund has expenses of only 0.2% per year
• Fidelity Megallan Fund has initial load of 3%, expenses of 0.95% per year
• Running a mutual fund is expensive, as there are lot of parties that need to
be compensated…
• From the investors’ perspective: it is very important to investigate if the
fees justify the performance.
• That will be the topic of the next lecture(s): Performance Evaluation
Measures

22
Conclusion
Teaching Points

In this class you learned...

• Understand the general structure of the money management industry
• The difference between passive versus active management
• Index Funds and ETFs

Next Lecture(s): Talk about different performance evaluation measures

that we can use to evaluate the performance of actively managed funds
(mutual funds).

23
 Portfolio Performance Evaluation
Standard Measures of Performance

0
Outline of the lecture

Portfolio Performance Evaluation

We will learn how to evaluate the performance of any portfolio (Mutual Funds,
etc...)

• If a portfolio manager (fund) systematically outperforms the market

• Detect if portfolio managers have market timing ability
• Detect if portfolio managers have stock picking ability
• Detect if portfolio managers really follow the advertised strategy. For
example: is a “small-growth fund” really buying small stocks?

We will study
• Standard risk adjusted measures of performance often used in the industry
and will discuss its limitations
• Apply these measures in the real world and discuss some EMPIRICAL
EVIDENCE on the performance of mutual funds

1
Introduction
Performance Evaluation

Why is this topic important? Why do we need to learn HOW to evaluate

the performance of a Portfolio?
• If you are the manager of a portfolio, you need to know if you need to
make changes in your portfolio
• If someone else manages your portfolio:
• Active funds have considerable high fees and expenses!
• You need to know if portfolio managers are “adding value” for their fee. If not they
should:
• Manage a smaller amount of your money or be replaced or have additional constraint
on what they should do
• Also there is a performance evaluation industry out there:
• Morningstar, Lipper, industry publications, etc...
• These firms provide information that can is used by investors to decide how to allocate
funds between different portfolio managers

2
How to choose a Measure

As you will see, evaluating the performance of a portfolio or a fund is a

complicated subject
• Different measures have been proposed and all are appropriate for
different types of investment assumptions: the fact that there is no “one
measure fits all” complicates the analysis!
• Because of this you must always interpret with caution the stellar
performances advertised by some funds! Obviously, funds will always
report the performance measure for which they look better!
• Funds tend to cherry pick both the timing and the type of performance
they report in their prospectus.

3
Which measure not to Choose

IMPORTANT: to evaluate the performance of a fund it is NOT appropriate

to look ONLY at the average return of a fund
• What you should NEVER do in evaluating the performance of a fund:
compare two funds by just looking at their average returns.
• E.g.: It’s wrong to say: “Fund A is better than Fund B because Fund A had higher
average returns than fund B”
• WHY IS THIS WRONG? You’re not controlling for risk! If fund A follows
a risky strategy and Fund B follows a safe strategy (e.g. T-Bill) then of
course A should have higher returns on average. But that does not reflect
that Fund A is doing a good job, i.e. it does not reflect skill! This is just a
compensation for risk!
• THUS, in evaluating the performance of a fund, we ALWAYS NEED to
adjust its returns for the level of risk of the fund.

4
Benchmarks
The CAPM

The performance measures that we will discuss use the CAPM AS THE
BENCHMARK:
• Thus we will compare the performance of a Fund with the predicted
performance of that fund according to the CAPM
• Thus, to implement this method, the first thing that we need to obtain is a
measure of the risk of the portfolio and then adjust the returns for this
risk. As you know, there are two different types of risk:
1. Portfolio’s systematic risk, measured by its beta
2. Portfolio’s total risk, measured by its standard deviation
• To determine WHICH measure of risk to use, the key is to determine the
impact of a given fund on the investor’s GLOBAL portfolio:
• If a client invests in MANY funds: then systematic risk of the fund provides the relevant
measure of the fund’s impact on the risk of the investor’s GLOBAL portfolio
• If a client invests only in that fund: then total risk of the fund provides the relevant
measure of the fund’s impact on the risk of the investor’s GLOBAL portfolio (Why?
The fund IS the GLOBAL portfolio!)

5
Four Benchmarks

Measures of Systematic Risk: benchmark is SML

• Jensen’s Alpha
• Treynor’s Reward to Volatility Ratio (RVOL)

Measures of performance based on Total risk: benchmark is the CML

• Sharpe Ratio
• Modigliani and Modigliani M2

6
Benchmarks

Jensen’s alpha
Jensen’s Alpha

• Key idea: Compare with ex-post SML

E{Rp } = avg(Rf ) + β̂p · (avg(Rm ) − avg(Rf ))

• where avg(Rf ) and avg(Rm ) are sample averages (that’s why is called ex-post)

• Alpha measures the average return on the portfolio over and above that
predicted by the CAPM, given the (estimated) portfolio’s beta and the
average market return and average riskfree rate

7
Finding Jensen’s Alpha

Four step procedure

1. Determine the time period to evaluate performance
2. Find an index that broadly represents the stock market, including the
segment that the manager invests in
• If manager invest mainly in large cap stocks, then may use the S&P 500 or Russell 1000
• If manager invests mainly in mid and small cap stocks then you may use the S&P 1500
or Russell 3000

8
Finding Jensen’s Alpha

Four step procedure

3. Run the regression

Rp − Rf = αp + βp · (Rm − Rf ) + up

• where intercept is α and the slope is β

4. If αp > 0 the manager did beat the market; αp < 0 the manager lost to the
market

• Portfolios can be ranked based on their αp

• Also you should always ask: but did the manager beat/lost to the market in
a statistically meaningful way?
• i.e. is αp statistically different from zero or not? Answer: check if t-statistics of αp is
small than −1.96 or greather than 1.96 (remember your stats classes)

9
Jensen’s Alpha: Example

EXAMPLE: based on monthly performance over the last five years

• Food Fund (FF) had average return of 9% and beta of .5
• Peer Fund (PF) had average return of 12% and beta of 2.0
• Average return on market was 10%; average riskfree rate was 7%

• FF: 9% − [7% + (10% − 7%) · 0.5] = 0.5% > 0, so FF beat the market
• PF: 12% − [7% + (10% − 7%) · 2] = −1% < 0, so PF lost to market and FF beat PF
• Note: alpha corresponds to vertical distance that p plots from SML

10
Jensen’s Alpha: Example

Some large funds

11
Benchmarks

Treynor RVOL
RVOL

The RVOLp corresponds to the slope of a line from avg(Rf ) to FF

• Compare to slope of peers of SML avg(Rm ) − avg(Rf ) = RVOLm

• FF: (9% − 7%)/0.5 = 4; PF : (12% − 7%)/2 = 2.5; Mkt: (10% − 7%) = 3
• FF beat the market but PF did not

12
Systematic Risk Adjusted Indicators RVOL or Jensen’s α

1. Alpha and RVOL always give the same sign for p relative to market
• This means that if a given portfolio p beat the market according to alpha, then this
portfolio also beats the market according to Treynor
2. But Alpha and RVOL can rank two portfolios differently.
• This means that if portfolio p beat portfolio z according to alpha, it is possible that
portfolio p loses to portfolio z according to Treynor
• Why? The approach used in each measure to adjust return for risk is different
• see class for an ilustration of this result

13
Benchmarks

Sharpe Ratio
Sharpe Ratio

• Key idea: Compare portfolio returns with ex-post CML

ˆ m · σi
E{Ri } = avg(Rf ) + SR

• So you compute the Sharpe ratio of the portfolio:

avg(Rp ) − avf(Rf )
SRp =
σp
• And compare SRp with SRm
• If SRp > SRm then p beat the market
• Can use the SR to rank the different funds in the market

14
Sharpe Ratio: picture

• The Sharpe ratio corresponds to the slope of a line from avg(Rf ) to FF

• Compare to slope of peers of CML, avg(Rm ) − avg(Rf )/σm = SRm

• FF: (9% − 7%)/5.5% = 0.36; PF : (12% − 7%)/20% = 0.25; Mkt:
(10% − 7%)/11% = 0.273
• FF beat the market but PF did not

15
Sharpe Ratio: numbers

16
Benchmarks

Modigliani & Modigliani

Modigliani & Modigliani: M2 Measure

• Determines what p’s average return would have been if it had the same
standard deviation as the market
• IDEA: “If I combine this portfolio with the riskfree asset such that the standard
deviation of the total portfolio is the same as the market, how does the average
returns of this global portfolio compares with those from the market?”
• WHY THIS IS USEFUL?: values of Sharpe ratio are hard to interpret. M2 gives the
same information but with a nice interpretation (in terms of returns)

17
M2: the measure

• Equation of the line in the mean-std space that goes through avg(Rf ) and p:
avg(Rp ) − avg(Rf )
avg(Ri ) = avg(Rf ) + · σi
σp
It is the CAL if the p is the market portfolio
• Thus, the average return that p would have earned if it had the same
standard deviation as the market, σm , is the M2:
avg(Rp ) − avg(Rf )
M2p = avg(Rf ) + · σm
σp
• Compare M2p with avg(Rm ):
• If M2p > avg(Rm ) then p is above CML, and hence beat the market
• If M2p < avg(Rm ) then p is bellow CML, and hence lost to the market
• Can also rank portfolios based on the values of M2p

18
M2: picture

• Equation of line from avg(Rf ) thru FF:

avg(Ri ) = 7 + [(9 − 7)/5.5]σi = 7 + 0.3636σi
• So M2FF = 7 + 0.3636 · 11 = 11%, since M2FF = 11% > avg(Rm ) = 10%,
FF beat the market
• For PF: M2PF = 7 + [(12 − 7)/20] · 11 = 9.75%, since
M2PF = 9.75% < avg(Rm ) = 10%, PF lost to market and rank bellow FF
19
Sharpe Ratio vs. M2

1. Basically, these measures provide the SAME information

2. SR and M2 always give the SAME sign for p relative to market and SAME
rankings for portfolios
3. Units of M2 are easier to interpret (because they are in percents %)

20
Benchmarks

Is there a best measure?

Comparing all measures

1. SR vs. M2 can indicate that p lost to the market while Alpha and Treynor
RVOL indicate p beat the market
• WHY? Presence of firm-specific risk in p:
• a portfolio can have a low beta but high σp
2. For the same reason, SR and M2 can rank the different portfolios
differently than Alpha and Treynor.
3. ALWAYS REMEMBER: In SR and M2 the benchmark is the CML while in
alpha and Treynor the benchmark is the SML

21
Which Measure is Appropriate

It depends on Investment Assumption

• If the portfolio/fund represents the entire investment for an individual,
then use Sharpe ratio or M2 . Here, risk is measured by the total risk of
the portfolio (WHY? since you only hold one portfolio you care about the
total risk of that portfolio).
• If the investor invests in many different portfolios/fund, then use the Jensen
α or Treynor. Here, risk of a portfolio should be measured by its
systematic risk
• WHY? Because this systematic risk measures the contribution of this portfolio to the
risk of your GLOBAL portfolio (recall discussion on CAPM/Market portfolio).

22
Limitations of Single Performance Measure

1. CAPM as benchmark? Is this the right asset pricing model?

• Exception: Sharpe ratio cannot be considered inappropriate on these grounds because
it uses standard deviation as a measure of risk and does not rely on the validity of the
CAPM.
• Maybe should be using the APT (Fama-French three factors)
2. These measures try to infer value added of the manager of the fund, i.e. its
stock picking ability (skill):
• This interpretation is correct ONLY if the beta is constant over time
• This is NOT true if the portfolio is being actively managed
• We will analyze time variation in beta later (market timing)
3. Which market index? Remember Roll’s critique: we don’t observe the
true market portfolio thus using a proxy such as the S&P500 or other
might not be correct!
• This is important: in practice, making slight changes in the benchmark portfolio may
completely change performance rankings
4. All the measures may not be meaningful if avg(Rm ) < avg(Rf )
• Why? Both the ex post SML and CML will be downward sloping

23
APT based measures

• We can create measures of performance that use the APT as the

benchmark model.
• This one is analogous to Jensen’s alpha:
• For example, if we use the Fama-French (1993) three factor model, we can
run a regression of the fund’s returns (Rp) on the Fama-French factors:

Rp,t = αp + βmkt · Rmkt

t + βsmb · Rsmb
t + βhml · Rhml
t

• We then rank funds according to their alphas (intercept) which is a

measure of abnormal performance of the fund.

24
Benchmarks

Applying our measures

Mutual Funds Jensen’s alpha

• Roughly half are positive, half negative, BEFORE costs

• Here are the MEDIAN net alphas (after costs and fees) for 2,609 U.S.
equity mutual funds between 7/1963-12/1998 (from Pastor and Stambaugh,
2002)

Investment Objective CAPM α

Small + growth −8.45
Other aggressive growth −5.41
Growth −2.17
Income −0.39
Growth + Income −0.51
Maximum Capital Gains −2.29
Sector funds −1.06
All funds −2.13

• The median fund is a looser! Of course, there are winners out there
25
Mutual Funds Jensen’s alpha

Evidence shows that average fund after transaction costs generates a

negative alpha
• Underperforms the S&P 500 index, i.e. plots below the SML

• CAPM tells us:

Rp − Rf = βp · (Rm − Rf )
• CAPM test using a regression

Rp − Rf = αp + βp · (Rm − Rf ) + εi,t

• Test evidence: average αp ≤ 0

26
Mutual Funds Jensen’s alpha

Evidence shows that average fund after transaction costs generates a

negative alpha
• Underperforms the S&P 500 index, i.e. plots below the SML

26
Application: Evidence of Persistence in Performance? (Carhart 97)

27
Conclusion
Teaching Points

In this class you learned...

• Compute risk-adjusted measures of performance of a fund (Jensen’s alpha,
Treynor RVOL, Sharpe ratio and M2)
• Understand the differences between the different performance evaluation
measures
• Understand the limitation of these measures

28