Video Summary - 9. Guest Lecture by David Swensen

Finance-Focused Summary (Markets, Investing Strategy, Portfolio Construction, Risk/Performance)

Core Thesis

For investors with long time horizons, portfolio management is “simple” in the sense that outcomes are driven primarily by:

Equity bias (stocks tend to outperform over the long run)
Diversification (“free lunch” per Markowitz)
Avoiding market timing
Treating active security selection cautiously because (net of costs) it tends to be negative-sum

“Yale Model” vs. Traditional Endowment Allocation (Early–Mid 1980s)

In the mid-1980s, endowments were roughly:

50% in U.S. common stocks
40% in U.S. bonds and cash
10% in “alternatives”

Yale concluded that this was insufficient diversification, and began shifting toward a more equity-oriented and diversified approach across asset classes, including:

Real assets
Absolute return / hedge-fund-like strategies
Private equity

Three Return Drivers / Framework (Explicit Tools)

Asset allocation: asset classes + portfolio weights
Market timing: short-term deviations from long-term target weights
Security selection: active management within each asset class (vs. indexing / market-matching)

Key Numbers, Examples, and Performance Metrics

Yale Size / Context

Yale endowment grew from a “billion dollar portfolio” (1985) to about $22.5 billion (estimate).

Relative Importance of Asset Allocation

Using Roger Ibbotson’s work (as described):

>90% of variability in institutional portfolio returns is attributed to asset allocation
“More than 100% of returns” attributed to asset allocation because security selection and market timing reduce aggregate returns after:
- trading costs
- fees/commissions
- taxes
- other “leakage”

Equity vs. Bonds/Bills Over Long Horizons (Inflation-Adjusted Discussion)

Over an ~81-year window (1925–end of 2006), real outcome multiples described:

Treasury bills: ~19x nominal (less attractive after inflation)
Bond market: ~72x nominal (better after inflation)
Broad U.S. stocks: ~3,077x
Small stocks: ~15,922x

Implication: riskier equities can massively outperform over the long run, but investors must survive drawdowns.

Crash Risk Illustration (Oct 1929 → Early 1930s)

For small stocks at peak:

Lost 54% by end of 1929
Lost another 38% by end of 1930
Lost another 50% by end of 1931
Lost another 32% by June 1932

Implied drawdown: about $0.10 left per $1 invested at the trough.

Takeaway: diversification and allocation discipline are required to hold risk assets through extreme stress.

Evidence Against Market Timing (Behavior + Dollar-Weighted Returns)

Internet Bubble Case (1997–2002)

For mutual fund behavior:

Time-weighted return (top 10 internet funds): +1.5%/yr
Dollar-weighted experience:
- investors lost $9.9B out of $13.7B (72% lost)
Additional capital gains distributions:
- $3.3B (~24% of invested amount)

Morningstar Study

In 17 equity mutual fund categories:
- in all 17, dollar-weighted < time-weighted
Timing gaps:
- “Conservative allocation fund”: 0.3%/yr
- Tech fund category: 13.4% per annum over 10 years (dollar-weighted vs time-weighted)

Institutional Timing Example (Oct 1987 Crash)

Oct 19, 1987: described as an extreme crash (one-day drop ~21–22% for major U.S. indices; globally similar).

Investment committee / accounting behavior:

June 1987: equity allocation at a 15-year high, bond allocation at a 15-year low
By June 30, 1988: equities fell and bonds rose beyond what price moves alone implied

Stated conclusion: endowments/foundations sold stocks in Nov/Dec/Jan and bought bonds—i.e., emotion-driven timing.

Cost: it took until 1993 (about 6 years) to return bond allocations toward pre-crash levels.

Security Selection Evidence (Active Management Net of Fees/Costs)

Example: Frank Russell Benchmark (Wilshire 5000)

10 years ended June 30, 2005:

Benchmark return (Wilshire 5000): 9.9%/yr
Average actively managed equity funds: 9.6%/yr
Net underperformance: about 0.3%/yr (30 bps)

Bias problem: survivorship + backfill bias can worsen inferred performance:

1996: 307 managers
2005: 177 remained (130 disappeared)

Bias-corrected claim:

deficit may be ~2% vs the 0.3% implied by reported averages

Hedge Funds: Bias Magnitude and “Net” Expectation

Malkiel (Hedge Funds)

Starting cohort: 331 funds (1996)
By 2004: 75% disappeared
Survivorship bias: 4.4%/yr
Backfill bias: 7.3%/yr
Combined bias: 11.7%/yr explained

Translation: reported hedge fund returns may overstate what investors actually experienced; “real” net may be mid-single digits depending on corrections.

Ibbotson (Larger Set)

3,500 funds over 10 years
Survivorship bias: 2.9%/yr
Backfill bias: 4.6%/yr

Dispersion / Market Efficiency (Where Active Management May Have a Better Chance)

First vs third quartile dispersion in active returns (10-year period ended June 30, 2005):

Bonds (high-quality): ~0.5%/yr
Large-cap equities: ~2.0–2.0+%/yr
Small-cap equities: ~4.7%/yr
Hedge funds: ~7.1%/yr
Real estate: ~9.3%/yr
Leveraged buyouts: ~13.7%/yr
Venture capital: ~43.2% differential (top vs bottom quartile)

Conclusion: if pursuing active management, target areas with most inefficiency/dispersion (e.g., VC/LBO/real estate) rather than highly efficient markets (e.g., high-quality bonds).

Yale Current Portfolio Allocation (Construction Numbers)

Traditional marketable securities: (30% total)

11% domestic equities
15% foreign equities
4% bonds

Alternatives / nontraditional: (70% total)

23% absolute return (hedge-fund-like, uncorrelated goals)
28% real assets (timber, oil & gas, real estate)
19% private equity (venture capital, leveraged buyouts)

Equity-like exposure

96% of assets expected to generate equity-like returns over long periods.

Diversification range

Asset-class weights roughly 4% to 28% each.

Yale Performance and Risk Stability Metrics

Past 20 years: 15.6% per annum (headline)
No down years since 1987
- 1987 negative return was < 1%
- Through the Internet bubble bust, Yale had positive returns every year since 1987
- Institutional investors were negative on average in 2001–2002

Spending / real economy link

June 1985 operating budget distribution: $45M (10% of revenues; lowest in the century described)
Year ending June 30, 2008 spending: $843M (37% of revenues)
Projected following year: $1.15B

Value-added comparison

Yale vs average college/university returns: +$14.4B added to coffers (over two decades)

Explicit Recommendations / Cautions

Equity bias is sensible for long-horizon investors, but:
- Diversify to sustain risky allocations through severe drawdowns.
Avoid market timing
- Timing is typically driven by fear/greed/chasing performance, harming dollar-weighted outcomes.
- Timing also increases costs and often requires more leverage to emotional decisions.
Active security selection
- Treat as potentially negative-sum after fees, commissions, and market impact.
- Pursue active strategies only if you have a genuine edge and consider market efficiency/dispersion.
- Beware survivorship bias and backfill bias when judging managers/funds.
Value/contrarian investing (“falling knives”)
- Works best with a long time horizon, a strong fundamental thesis, and sufficient dry powder to add on declines.

Instruments / Sectors / Assets Mentioned

Stocks
- U.S. common stocks (equities), small stocks
- Internet stocks (1999–2000 short bet; hypothetical context)
Fixed income
- U.S. bonds, Treasury bills, Treasury bonds
- High-quality bonds (as an “efficient” example)
- Bond futures (hypothetical day-trading example)
Real assets
- Timber, oil & gas, real estate
Alternatives / private markets
- Absolute return strategies (hedge-fund-like, uncorrelated goals)
- Private equity: venture capital, leveraged buyouts
- Venture capital, distressed securities, bank loans
- Real estate
Credit / structured credit examples
- Subprime mortgage-backed securities
- Credit spreads (mortgages and corporates)
Hypothetical company example
- Google (putting Yale-like allocation into a single stock; illustrative)
Benchmarks / data sources
- Wilshire 5000
- Frank Russell Corporation (data cited)

Methodology / Step-by-Step Framework (Portfolio Construction)

Set long-term equity orientation + diversification targets
- Use asset allocation as the primary driver.
- Maintain stable policy weights; review infrequently.
Define “market timing” boundaries
- Allow only deviations that are not wholesale shifts from policy targets.
- Avoid short-term tactical changes; keep targets stable (e.g., modest updates, often yearly).
Decide on active vs. passive by market efficiency
- Assess whether markets are efficiently priced.
- Prefer active effort where active-return dispersion is high (e.g., VC/LBO/real estate).
Evaluate managers with bias awareness
- Correct historical performance interpretation for survivorship and backfill biases.

Disclosures / Disclaimers

No explicit “not financial advice” disclaimer appears in the provided subtitles.

Presenters / Sources Mentioned

Professor David Swensen (Yale)
Roger Ibbotson
Bob Shiller
Jeremy Siegel (Stocks for the Long Run)
Harry Markowitz (“diversification is a free lunch”)
John Maynard Keynes (quoted re: timing)
Seth Klarman (Baupost quote about top-down vs bottom-up)
Morningstar (time-weighted vs dollar-weighted study)
Frank Russell Corporation
Burt Malkiel (A Random Walk Down Wall Street)
Wilshire 5000

9. Guest Lecture by David Swensen

Key takeaways