01: The Landscape and the Agency Problem

Institutional Investors: Theory and Evidence

Author

Affiliation

Calvin J. Chiou (邱健嘉)

Department of Finance · National Chengchi University

Published

September 17, 2026

Welcome

Hi — a bit about me

Assistant Professor, Department of Finance, NCCU
Research: mutual funds, institutional ownership, corporate governance, ESG
I use R for almost everything — data, analysis, this slide deck
Office hours: by appointment — please reach out, I respond quickly

. . .

What I hope you take away from this course:

A clear picture of how institutional investors behave, why it matters for markets and firms, and the practical skills to study it yourself.

A question to start

. . .

You have NT$5 million to invest.

. . .

Do you pick individual stocks yourself?

. . .

Probably not. You hand it to a professional fund manager.

. . .

Now multiply that by millions of people doing the same thing.

. . .

That is what this course is about — what happens when enormous pools of capital are managed by a relatively small group of professional institutions.

What share of the US stock market do institutions own today?

. . .

1950

Mostly individual retail investors.

Institutions held just 6.1% of outstanding US corporate equity (Tonello & Rabimov, 2010).

2016

Institutions held 63% of outstanding US public corporate equity (Federal Reserve Flow of Funds, 2016).

In Taiwan, foreign institutional investors alone account for a significant fraction of daily trading volume.

. . .

The shift from retail to institutional ownership is one of the defining structural changes in modern financial markets — and it is still accelerating.

Part 1 · The Landscape

Who counts as an institutional investor?

Any organization that pools money from individuals or other entities and invests it on their behalf.

. . .

The big categories

Mutual funds (共同基金)
Pension funds (退休基金)
Insurance companies (壽險)
Sovereign wealth funds (主權基金)
Endowments (大學基金)
ETFs (exchange-traded funds)

What they share

Manage other people’s money
Subject to regulation and disclosure rules
Have mandates, constraints, and benchmarks
Their decisions move markets

Scale: just how big are we talking?

The five largest asset managers in the world:

Manager	AUM (approx.)
BlackRock	~$10 trillion
Vanguard	~$8 trillion
Fidelity	~$4.5 trillion
State Street Global Advisors	~$4 trillion
JPMorgan Asset Management	~$3 trillion

. . .

Taiwan’s GDP is roughly $800 billion.

BlackRock alone manages more than 12 times Taiwan’s annual GDP.

The cast of players, briefly

This course’s focus

Mutual funds — best available data (CRSP Mutual Fund Database), the richest literature, and the clearest agency problems
The “Big Three” (BlackRock, Vanguard, State Street) — dominate index investing and are simultaneously the largest shareholders in most S&P 500 firms

Also on stage

Pension funds — long horizons, tend to be more engaged (CalPERS, GPIF, Bureau of Labor Funds)
Insurance companies — fixed-income tilt, regulated equity holdings
In Taiwan — foreign institutional investors (外資) dominate large-cap trading; the Bureau of Labor Funds (勞動基金) is the major domestic player

. . .

We will dig into ownership patterns and the evidence on how these types differ next week.

Part 2 · Why Does It Matter?

Three reasons this is not a boring topic

. . .

① Institutions affect stock prices

When large institutions buy or sell, prices move. Their portfolio decisions create and destroy the price signals that guide capital allocation in the economy.

. . .

② Institutions govern corporations

As major shareholders, they vote on boards, executive pay, and mergers. Whether they use this power effectively — or not — shapes how companies behave.

. . .

③ Institutions face their own agency problems

The fund manager is paid to act in your interest. But they have their own career concerns, fee incentives, and performance pressures. Do these distort their decisions?

The agency problem — the course’s central tension

You hire a fund manager to maximize your wealth.

. . .

But the manager cares about:

Their fee (often a percentage of AUM, not performance)
Their career (avoiding underperformance that gets them fired)
Their bonus (calendar-year performance)
Looking good at quarter-end (window dressing)

. . .

These incentives do not perfectly align with yours.

. . .

This tension — between the principal (you) and the agent (the manager) — runs through almost every paper we read this semester.

Today’s motivating paper

Bebchuk, Cohen & Hirst (2017)
“The Agency Problems of Institutional Investors”
Journal of Economic Perspectives

. . .

This is an accessible, well-written piece — a good entry point.

The core argument:

Institutional investors are not simply “investor-principals” who solve agency problems at the firm level. They are themselves agents — with their own principals (their investors) and their own incentive distortions.

. . .

We will return to specific arguments from this paper throughout the semester.

Two agency problems, not one

. . .

Problem 1 — the classic one (Berle & Means, 1932)

Ownership is dispersed across many small shareholders. Each is “rationally apathetic” — too small to bother monitoring — so managers are left relatively unconstrained: “management control.”

. . .

Problem 2 — Bebchuk, Cohen & Hirst’s contribution

Institutional investors were supposed to fix Problem 1 by concentrating ownership. But an investment manager invests other people’s money. So a second, nested agency problem appears: the fund manager (agent) vs. the fund’s own beneficial investors (principals).

The concentration numbers behind the paper

Largest 20 US corporations by market cap, June 2016 — average ownership by:

Holder group	Mean	Median
Largest 5 institutional holders	20.8%	19.8%
Largest 20 institutional holders	33.4%	32.9%
Largest 50 institutional holders	44.2%	44.2%

. . .

Compare this to Berle & Means (1932): the largest 20 shareholders of the 200 largest US corporations in 1930 held a mean of just 10.55%.

. . .

Today’s largest 20 institutional holders alone own more than three times that. Shareholders are no longer atomistic — the question is no longer can institutions influence firms, but will they, and in whose interest?

Three structural drivers of underinvestment in stewardship

(Bebchuk, Cohen & Hirst, 2017)

① Capture only a small fraction of the benefit

Investment managers bear the full cost of stewardship (monitoring, engagement, proxy fights) but earn fees calculated as a small percentage of assets under management — so they keep only a sliver of any value increase they help create.

② Limits of competition — especially for index funds

Performance is judged relative to a benchmark or to rival funds tracking the same index. Raising a portfolio company’s value also lifts the benchmark, so it buys no competitive edge. A classic collective-action problem: everyone in the index fund would benefit from more stewardship spending, but no single fund can raise fees to pay for it without investors defecting to a cheaper rival tracking the same index.

③ Private costs from opposing corporate managers

Many fund families also sell services — 401(k) administration, cash management — to the very corporations they invest in, creating pressure to vote with management. Crossing the 5% ownership threshold with intent to influence control also triggers costly Schedule 13D disclosure, discouraging active engagement by the largest holders.

The arithmetic of underinvestment

A stewardship activity costs $C$ and raises portfolio value by $\Delta V$.

. . .

No-agency benchmark: undertake it whenever $C < \Delta V$
Investment manager who captures a fee share $\alpha$ of the gain: undertakes it only when $C < \alpha \times \Delta V$

. . .

Everything in between — $\alpha \Delta V < C < \Delta V$ — is value-creating stewardship that never happens.

The arithmetic, with real numbers

A $1 billion equity stake; stewardship could raise its value by 0.1%, or $1,000,000.

. . .

Manager type	Fee share captured	Willing to spend up to
No-agency benchmark	100%	$1,000,000
Index fund (avg. fee ≈ 0.12%)	0.12%	~$1,200
Active mutual fund (avg. fee ≈ 0.79%)	0.79%	~$7,900
Activist hedge fund (“2 and 20”)	~20%	~$200,000

. . .

Same $1 million of value on the table. Wildly different willingness to go get it.

Governance passivity, in staff counts

Vanguard: ~15 staff for voting and stewardship, covering 13,000 portfolio companies
BlackRock: ~24 staff, 14,000 portfolio companies
State Street Global Advisors: fewer than 10 staff, 9,000 portfolio companies

(Krouse, Benoit & McGinty, 2016)

. . .

Less than one person-workday per company per year, on average — even though each of these managers holds a meaningful stake in most of them.

The “giant three” problem

Bebchuk & Hirst (2019) extend this to a specific concern:

. . .

BlackRock, Vanguard, and State Street together hold 20–25% of S&P 500 companies.

. . .

Their fee income is proportional to AUM, not to how well they monitor portfolio companies.

The return on investing in better corporate governance is tiny relative to the cost for a fund that already holds every company in the index.

. . .

Result: the institutions with the most power to discipline corporate management may have the least incentive to do so.

Why activist hedge funds are different

“2 and 20” fee structure → managers capture ~20% of any value increase they create, an order of magnitude more than a mutual fund’s fee-based cut
Concentrated portfolios (sometimes as few as 10 positions) → one successful engagement can meaningfully move the fund’s own performance
Not registered investment companies, and don’t sell 401(k) or cash-management services to corporations → far less exposure to the private costs that make mutual funds reluctant to antagonize management
Willing to run costly proxy contests (~$10 million on average, Gantchev 2013) that mutual fund families almost never initiate

…but hedge fund activism is not a full solution

Only pursue targets where the expected value increase is large — engagements are typically associated with average abnormal returns above 5%, just to clear the cost and fee hurdle
Need the support of other institutional investors (mutual funds, index funds) to actually win — or credibly threaten — a proxy fight
If mutual funds are expected to side with management anyway, activist hedge funds have little leverage — the two problems are linked, not independent

What the paper recommends

More disclosure: how investment managers vote, and their business ties to portfolio companies
Rethink mutual fund fee rules that block managers from charging for dedicated stewardship
The trade-off from the rise of index investing: lower costs for investors, but weaker incentives to monitor
Their conclusion: modern corporations suffer not from too much shareholder intervention, but from too little — little basis for weakening shareholder rights out of fear of “activist” institutions

Some big open questions

These are the kinds of questions this course prepares you to investigate:

Do active fund managers actually beat passive benchmarks, after fees?
Does institutional ownership improve firm performance — or does it just reflect good firms attracting institutional attention?
When institutions become large shareholders simultaneously across competitors, do they soften competitive pressure?
Can ESG-focused funds actually influence firm behavior, or is it mostly marketing?
Why do fund managers disclose certain trades at quarter-end that look very different from the rest of the year?
Would more disclosure — of proxy votes, of business ties to portfolio companies — actually change how institutions behave?

Part 3 · Getting Set Up in R

Why R?

You might already know Python, Excel, Stata, or SAS.

We use R because:

The academic finance literature increasingly publishes in R (and Python)
tidyverse + tidyquant + frenchdata gives you a clean pipeline from raw data to publication-quality output
Quarto (.qmd) lets you write your report and run your analysis in the same document
The CRSP Mutual Fund Holdings data we use in later labs is easiest to work with in R
I can actually help you debug it

The R stack for this course

# Run this before Week 2
install.packages(c(
  "tidyverse",    # data wrangling and visualization
  "tidyquant",    # financial data download and analysis
  "frenchdata",   # Kenneth French's factor library
  "broom",        # tidy model output
  "gt",           # publication-quality tables
  "scales",       # number formatting
  "slider",       # rolling window calculations
  "fixest"        # fast panel regressions with fixed effects
))

. . .

A setup guide and package check script are on Moodle. Run the check script and email me if anything fails — before Week 2.

Public data vs. CRSP data

Public data only — no accounts needed

ETF prices via tidyquant::tq_get()
Factor returns via frenchdata

You can run these at home immediately.

CRSP Mutual Fund Holdings

I will distribute the data files via Moodle before the relevant labs.

These are the real datasets used in published papers — the same ones you would use in your own research.

Lecture notes format

All lecture notes — including this slide deck — are written in R Quarto (.qmd).

. . .

What this means for you:

Every .qmd file I distribute is fully runnable — open it in RStudio, click Render, and it reproduces all analysis and output
You can modify my code and run your own versions

. . .

If you have never used Quarto before, don’t worry. The learning curve is gentle, and the payoff is high.

Wrapping Up Week 1

What we covered today

Who institutional investors are, and how concentrated their ownership has become
Two nested agency problems: managers vs. dispersed shareholders (Berle & Means), and investment managers vs. their own beneficial investors (Bebchuk, Cohen & Hirst 2017)
Three structural drivers of underinvestment in stewardship — and why index funds are especially exposed
Why activist hedge funds behave differently, and why they can’t fully substitute for institutional stewardship

Before Week 2 — your to-do list

Install R ≥ 4.3 and RStudio (or Positron)
Run the package check script from Moodle — email me if anything fails
Read Bebchuk, Cohen & Hirst (2017) — full paper, not just the abstract. It’s in JEP so it’s written to be accessible. Should take about an hour.

Week 2 preview

Topic: Types of Institutional Investors and Asset Concentration Trends

Readings:

Gompers & Metrick (2001, QJE) — how institutional demand affects stock prices
Ferreira & Matos (2008, JFE) — institutional ownership around the world

We will also talk about the Taiwan institutional investor landscape in more detail.

References

Bebchuk, L. A., Cohen, A., & Hirst, S. (2017). The agency problems of institutional investors. Journal of Economic Perspectives, 31(3), 89–112.
Bebchuk, L., & Hirst, S. (2019). The specter of the giant three. Boston University Law Review, 99(3), 721–741.
Berle, A. A., & Means, G. C. (1932). The Modern Corporation and Private Property. Macmillan.
Board of Governors of the Federal Reserve System. (2016). Financial Accounts of the United States: Flow of Funds, Balance Sheets, and Integrated Macroeconomic Accounts, Z.1.
Gantchev, N. (2013). The costs of shareholder activism: Evidence from a sequential decision model. Journal of Financial Economics, 107(3), 610–631.
Krouse, S., Benoit, D., & McGinty, T. (2016, October 24). Meet the new corporate power brokers: Passive investors. Wall Street Journal.
Tonello, M., & Rabimov, S. R. (2010). The 2010 Institutional Investment Report: Trends in Asset Allocation and Portfolio Composition. The Conference Board.

Questions?

Calvin J. Chiou
jjchiou@nccu.edu.tw
Office: Commerce Building 261233

See you next Thursday.