FINA 3103: Intermediate Investments

Lecture 9. Alpha

Don Noh

Hong Kong University of Science and Technology

Spring 2025
Overview
Types of Registered Investment Companies

• Open-end mutual funds ($18,746 billion in 2017)

◦ Number of shares adjust at the end of day, depending on buy and sell orders.
◦ All orders are executed at net asset value (NAV; closing price) at 4pm EST.
◦ Holdings are disclosed quarterly.
• Closed-end mutual funds ($275 billion in 2017)
◦ Fixed number of shares trade on exchanges.
◦ Orders can be executed at any time during the trading day.
• Exchange-traded funds (ETFs; $3,401 billion in 2017)
◦ Fixed number of shares trade on exchanges.
◦ Shares can be shorted.
◦ Holdings are disclosed daily.

FINA 3103 Lecture 9. Alpha Spring 2025 1 / 32

Types of Mutual Funds and ETFs

• Equity mutual funds

◦ Domestic vs international
◦ Value vs growth
◦ Large, mid, or small cap
◦ Sector
• Bond mutual funds
◦ Government bonds
◦ Municipal bonds (tax-exempt)
◦ Investment-grade bonds (rated BBB− and above by S&P or Baa and above by Moody’s)
◦ High-yield bonds
• Money market funds
◦ Government bonds
◦ Municipal bonds (tax-exempt)

FINA 3103 Lecture 9. Alpha Spring 2025 2 / 32

Composition of US Mutual Funds and ETFs

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 3 / 32

Actively Managed vs Index Funds

• Actively managed funds

◦ Investment mandate puts limits on types of securities that can be held (e.g., a technology fund can only invest
in technology stocks).
◦ Within the mandate, portfolio choice is up to the investment manager.
◦ A skilled manager can add value (but also charge higher fees).
• Index funds
◦ Tracks published indices such as S&P 500 or MSCI World Value Index
◦ Or passively managed by following mechanical rules
◦ Could generate alpha relative to CAPM (e.g., value, low volatility, or momentum)
◦ But nearly zero alpha relative to appropriate benchmark

FINA 3103 Lecture 9. Alpha Spring 2025 4 / 32

Growth of Index Funds and ETFs

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 5 / 32

Fees
Types of Fees

• Expense ratio: A percentage of account value is deducted to cover costs of portfolio management,
administration, marketing, and distribution.
◦ Net return = Gross return − Expense ratio
• Loads:
◦ Purchase fee is a percentage of account value at purchase.
◦ Redemption fee is a percentage of account value at sale.
◦ There are no-load funds with no purchase and redemption fees (Vanguard and Fidelity).

FINA 3103 Lecture 9. Alpha Spring 2025 6 / 32

Mutual Fund Expense Ratios

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 7 / 32

Expense Ratios of Actively Managed vs Index Funds

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 8 / 32

Distribution of Mutual Fund Expense Ratios

• S&P 500 index funds: Expense ratios in 2000 varied from 0.09% (min) to 2.68% (max) with
asset-weighted mean of 0.32%.
• Another failure of law of one price!

Percentile Average
Asset-
Type 10th 50th 90th Simple
weighted
Equity 0.66 1.18 2.00 0.59 1.25
Index equity 0.06 0.33 1.53 0.09 0.61
Hybrid 0.65 1.15 1.98 0.70 1.26
Bond 0.45 0.81 1.61 0.48 0.93
Money market 0.17 0.40 0.66 0.25 0.40

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 9 / 32

Mutual Funds
Zero-Sum Game

• The market overall earns zero alpha.

• Index funds or passive strategies must earn zero alpha relative to the appropriate benchmark.
• Some actively managed funds may earn positive alpha (before fees) at the expense of other actively
managed funds.
• Or all actively managed funds may earn positive alpha (before fees) at the expense of retail investors
or other institutions (pension funds, insurance companies, etc.).
• Even if gross alpha (before fees) is positive, net alpha (after fees) could be negative.

FINA 3103 Lecture 9. Alpha Spring 2025 10 / 32

Performance of Actively Managed Funds

Alpha (% annual) Beta

Model Net Gross Market SMB HML WML
CAPM − 1.13 − 0.18 0.99
( − 3.03) ( − 0.49) ( − 2.10)
3-factor − 0.81 0.13 0.96 0.07 − 0.03
( − 2.50) (0.40) ( − 5.42) (7.96) ( − 3.22)
4-factor − 1.00 − 0.05 0.97 0.07 − 0.03 0.02
( − 3.02) ( − 0.15) ( − 5.03) (7.78) ( − 3.03) (2.60)

Source: Fama and French (2010)

FINA 3103 Lecture 9. Alpha Spring 2025 11 / 32

Constructing Portfolios Sorted by Prior Returns

• Although actively managed funds collectively do not earn alpha, some funds may do so.
• Skill or luck?
• Skill should persist.
• Evaluate funds relative to 3-factor model plus an additional WML (Winner Minus Loser) factor to
account for momentum.
1. Sort mutual funds into 10 portfolios based on prior annual return (i.e., cutoff at each 10th percentile).
2. Compute returns on these portfolios over the subsequent year.
3. Rebalance annually.

FINA 3103 Lecture 9. Alpha Spring 2025 12 / 32

Mutual Funds Sorted by Prior Returns

Portfolio
Statistics Low 2 3 4 5 6 7 8 9 High
Mean (%) 0.01 0.23 0.34 0.36 0.40 0.38 0.45 0.43 0.59 0.68
SD (%) 4.90 4.60 4.48 4.30 4.36 4.35 4.41 4.56 4.72 5.04
Market beta 0.93 0.93 0.93 0.90 0.90 0.90 0.90 0.90 0.89 0.88
SMB beta 0.32 0.22 0.20 0.21 0.22 0.22 0.27 0.34 0.46 0.62
HML beta − 0.08 − 0.10 − 0.06 − 0.03 − 0.04 − 0.05 − 0.05 − 0.07 − 0.05 − 0.05
WML beta − 0.09 − 0.02 0.01 0.04 0.08 0.07 0.11 0.16 0.20 0.29
Alpha (%) − 0.40 − 0.20 − 0.13 − 0.14 − 0.12 − 0.14 − 0.12 − 0.18 − 0.10 − 0.12
Alpha (t-stat) − 4.33 − 3.11 − 2.52 − 3.09 − 2.82 − 3.31 − 2.81 − 3.65 − 1.78 − 1.60

Source: Carhart (1997)

FINA 3103 Lecture 9. Alpha Spring 2025 13 / 32

Alpha on Mutual Funds Sorted by Prior Returns

Source: Ang (2014) Figure 16.4

FINA 3103 Lecture 9. Alpha Spring 2025 14 / 32

Persistence of Mutual Fund Performance
Note: x-axis is flipped so that 1 is the portfolio of mutual funds with the highest prior return.

Source: Bodie, Kane, and Marcus (2018) Figure 11.8

FINA 3103 Lecture 9. Alpha Spring 2025 15 / 32

Explaining Under-Performance

• We see some persistence in performance relative to the CAPM, but not relative to the 4-factor model
that includes the WML factor.
• Lucky funds that happen to hold winner stocks (i.e., high returns over the prior year) continue to do
well because of momentum.
• Alpha is uniformly negative because of fees.
• Instead, poor performance (i.e., funds in portfolio 1) is especially persistent.
• Cahart (1997) shows that poorly performing funds charge fees that are too high and trade too much.

FINA 3103 Lecture 9. Alpha Spring 2025 16 / 32

Performance-Flow Sensitivity

• Given that good performance does not persist, investors should not react to past performance when
deciding which fund to invest in.
• However, investor flows react strongly to past performance in the cross-section of mutual funds.
• In the aggregate, money flows into (out of) the mutual fund sector when returns are high (low).
• This is the opposite of rebalancing and also the market-timing strategy from the lecture on Stocks!
◦ Optimal portfolio choice requires that investors go into mutual funds when the stock market falls.
• Which behavioral bias is this consistent with?

FINA 3103 Lecture 9. Alpha Spring 2025 17 / 32

Cross-Section of Equity Mutual Fund Flows

Source: Sirri and Tufano (1988)

FINA 3103 Lecture 9. Alpha Spring 2025 18 / 32

Aggregate Equity Mutual Fund Flows

Source: 2018 Investment Company Fact Book

FINA 3103 Lecture 9. Alpha Spring 2025 19 / 32

Performance Evaluation
Does Anyone Have Skill?

• Evidence presented thus far suggests that active management does not benefit investors.
• Any alpha that active management generates is more than offset by fees.
• But there are always exceptional performers in any industry.
1. Fidelity Growth Company Fund
2. Berkshire Hathaway
3. Princeton University Investment Company (PRINCO)

FINA 3103 Lecture 9. Alpha Spring 2025 20 / 32

Benchmark Portfolio

• “Grading on a curve”: Compare scores across students taking the course.

• Similarly, compare managers who have the same benchmark portfolio.
• Sometimes benchmarks are explicit.
◦ S&P 500 index fund
◦ Cambridge Associates’ benchmark of university endowments (PRINCO’s benchmark).
• When not explicit, evaluate relative to the 3-factor model.
◦ Betas in the 3-factor regression defines “what course”?
◦ Alpha determines “what grade”?

FINA 3103 Lecture 9. Alpha Spring 2025 21 / 32

Fidelity Growth Company Fund (FDGRX)

• Large-cap growth fund with expense ratio of 0.8%

• Managed by Steven Wymer since 1997
• Wymer was named by Morningstar as the 2017 US Fund Manager of the Year.

FINA 3103 Lecture 9. Alpha Spring 2025 22 / 32

Example 1: Evaluating the Fidelity Growth Company Fund

• 3-factor regression:

Ri,t − Rb,t = 0.30% + 1.07 × ( Rm,t − Rb,t ) + 0.33 × SMBt − 0.47 × HMLt + 𝜀 i,t
| {z }
t-stat=3.20

◦ 1.07 × ( Rm,t − Rb,t ) : long $1.07 in market and short $0.07 in riskless
◦ 0.33 × SMBt : long $0.33 in Small and short $0.33 in Big
◦ −0.47 × HMLt : short $0.47 in High and long $0.47 in Low

FINA 3103 Lecture 9. Alpha Spring 2025 23 / 32

Information Ratio

• Breaking down performance:

Ri,t − Rb,t = 0.30%

| {z }
𝛼
+ 1.07 × ( Rm,t − Rb,t ) + 0.33 × SMBt − 0.47 × HMLt
| {z }
benchmark
+ 𝜀 i,t
|{z}
tracking error

• Skilled managers can generate alpha without deviating too much from the benchmark.

FINA 3103 Lecture 9. Alpha Spring 2025 24 / 32

Information Ratio

• Information ratio is defined as:

𝛼 0.30%
= = 0.16.
𝜎 (𝜀 i,t ) 1.90%

◦ 𝛼 : average performance relative to benchmark

◦ 𝜎 (𝜀 i,t ) : deviation from benchmark
• It measures how much alpha you can generate without deviating too much.
• Proof is not provided, but information ratio measures how much you gain in terms of Sharpe ratio
when you mix this investment with a passive portfolio.

FINA 3103 Lecture 9. Alpha Spring 2025 25 / 32

Berkshire Hathaway (BRK)

Warren Buffett (1930–): CEO of Berkshire Hathaway and the 7th

wealthiest ($118.6 billion in Jan 2024).

“You will notice that our major equity holdings are relatively few. We select such investments
on a long-term basis, weighing the same factors as would be involved in the purchase of 100%
of an operating business: (1) favorable long-term economic characteristics; (2) competent and
honest management; (3) purchase price attractive when measured against the yardstick of
value to a private owner; and (4) an industry with which we are familiar and whose long-term
business characteristics we feel competent to judge.”

FINA 3103 Lecture 9. Alpha Spring 2025 26 / 32

Example 1 (continued): Evaluating Berkshire Hathaway

• We take the monthly returns of Berkshire and the monthly factors.

• 3-factor regression:

Ri,t − Rb,t = 0.96% + 0.82 × ( Rm,t − Rb,t ) − 0.29 × SMBt + 0.44 × HMLt + 𝜀 i,t
| {z }
t-stat=3.61

◦ 0.82 × ( Rm,t − Rb,t ) : long $0.82 in market and short $0.18 in riskless
◦ −0.29 × SMBt : short $0.29 in Small and long $0.33 in Big
◦ 0.44 × HMLt : long $0.44 in High and short $0.44 in Low
• Information ratio is
𝛼 0.96%
= = 0.17.
𝜎 (𝜀 i,t ) 5.82%

FINA 3103 Lecture 9. Alpha Spring 2025 27 / 32

Princeton University Investment Company (PRINCO)

• PRINCO managed $25.9 billion (as of June 30, 2018)

• Policy portfolio establishes guidelines for diversification across asset classes.

Asset class Policy (%) Actual (%)

Domestic equity 9 4.6
International equity
Developed markets 6 4.7
Emerging markets 10 8.7
Hedge funds 25 23.3
Private equity 27 36.2
Real assets 18 17.1
Fixed income & cash 5 5.3

FINA 3103 Lecture 9. Alpha Spring 2025 28 / 32

Example 2: Evaluating PRINCO

• We take the annual returns of PRINCO and the annual factors.

1. Policy portfolio:
Ri,t − Rb,t = 2.35% + 1.00 × ( Policyt − Rb,t ) + 𝜀 i,t
|{z}
t-stat=2.30

◦ Same risk as the policy portfolio, but added value both within and across asset classes
2. Benchmark of 65% in S&P 500 and 35% in Barclays Government/Credit Bond Index:

Ri,t − Rb,t = 6.30% + 0.73 × ( Benchmarkt − Rb,t ) + 𝜀 i,t

| {z }
t-stat=3.27

◦ Lower risk and much higher return

FINA 3103 Lecture 9. Alpha Spring 2025 29 / 32

Example 2: Evaluating PRINCO

3. Cambridge Associates’ benchmark of 160 university endowments:

Ri,t − Rb,t = 4.63% + 0.94 × ( CABt − Rb,t ) + 𝜀 i,t

|{z}
t-stat=3.50

◦ Similar risk and much higher return

4. 3-factor model:

Ri,t − Rb,t = 6.41% + 0.51 × ( Rm,t − Rb,t ) + 0.24 × SMBt − 0.06 × HMLt + 𝜀 i,t
|{z}
t-stat=3.71

◦ Low beta, small-cap strategy

FINA 3103 Lecture 9. Alpha Spring 2025 30 / 32

Summary of PRINCO’s Performance

• Ability to enter alternative asset classes (hedge funds, private equity, and real assets) at the right time
• Ability to identify alpha within asset classes
• Beats most other university endowments
• What about “fees”?
◦ In 2016, Andrew Golden (president of PRINCO) earned $5.784 million.
◦ Collectively, Golden and 4 managing directors earned $14.745 million.
◦ Total executive compensation of PRINCO as a share of the $25.9 billion endowment is only 0.06%.

FINA 3103 Lecture 9. Alpha Spring 2025 31 / 32

Summary
Summary

• Shop around because fees vary across funds.

• It’s hard to find actively managed funds that outperform benchmarks, especially after fees.
• For most of us, it’s better to stick to index funds with low fees.
• Don’t chase past performance, and don’t panic (and sell out) when the market falls.

FINA 3103 Lecture 9. Alpha Spring 2025 32 / 32