14  Session 5: Term Structure of Private Capital Returns

Unit 1 — Why DCF Fails
Book Chapter 3 (all sections)
Track Common core (both tracks)
Assessment milestone Track declaration deadline (end of session)

14.1 Learning Objectives

By the end of this session, students will be able to:

  1. Describe the empirical term structure of private capital returns by asset class (PE buyout, growth equity, infrastructure, real estate, private credit).
  2. Explain why long-horizon assets are disproportionately exposed to the liquidity illusion.
  3. Compute the Jensen bias \(B(T)\) for a given asset class and horizon using the affine formula.
  4. Compare the GE-LAV liquidity-adjusted term structure to the standard DLOM (Discount for Lack of Marketability) methodology.
  5. State and justify their Track selection by the end of class.

14.2 Pre-Class Assignment

  • Read: Book Chapter 3, all sections (~25 pages)
  • Optional: Section 1 of Phalippou & Gottschalg (2009), “The Performance of Private Equity Funds,” Review of Financial Studies

14.3 In-Class Outline (75 minutes)

Time Segment Format
0:00–0:05 Recap: OU dynamics → today: how dynamics interact with horizon Lecture
0:05–0:20 Empirical term structure of private capital returns Lecture + chart deep-dive
0:20–0:40 The liquidity risk term structure Lecture + worked formula
0:40–0:55 Asset class implications: which assets are most exposed Lecture
0:55–1:05 DLOM vs. GE-LAV — comparative framework Lecture
1:05–1:15 Track declaration · Q&A · final track guidance Activity

14.4 Discussion Questions

  1. The Jensen bias for a 30-year PPP asset is ~5.4%. The DLOM convention for infrastructure is 10–20%. Are these incompatible? How should they be reconciled?
  2. Pension funds and insurers hold most of the long-duration private assets where Jensen bias is largest. If GE-LAV becomes adopted methodology, what happens to reported PE returns at large pension funds?
  3. Some private credit assets have short effective duration (1–3 years for direct lending). Jensen bias is small there. Does that mean private credit is “GE-LAV-immune”? What other GE-LAV channels might still matter?

14.5 Worked Numerical Example: Choosing Between DLOM and GE-LAV

Setup: Insurance company holds a 15-year core infrastructure asset (regulated water utility). DCF mark = $500M.

Method 1: DLOM at 15% - Adjusted value: $500M × 0.85 = $425M - Applied uniformly across all market conditions - “Always 15% off DCF”

Method 2: GE-LAV (current state, normalization regime) - Jensen bias for T = 15: \(A \cdot 15 + C = 0.18\% \cdot 15 = 2.7\%\) - LAV value (just Jensen): $500M × (1 + 0.027) = \(513.5M\) - GE-LAV (normal regime, small equilibrium correction): ~$495M - “About flat with DCF in normal markets”

Method 3: GE-LAV (current state, stressed regime, L = −0.8) - Equilibrium rate: \(r_\text{GE-LAV}(L = -0.8) \approx 12\%\) (vs. DCF 7.5%) - GE-LAV value: $500M × e^{-(12% - 7.5%) } $250M - “50% off DCF in stress”

Method 4: GE-LAV (GFC depth, L = −1.5) - Equilibrium rate: ~32% - GE-LAV value: ~$150M - “70% off DCF at GFC depth”

Comparison: DLOM gives one answer ($425M). GE-LAV gives a menu of answers conditional on regime. The insurance regulator should care about the GFC-depth number ($150M) for solvency purposes.

14.6 What to Expect Next Session

Session 6 begins Unit 2 (Measurement and Theory) with a deep look at IRR. Why IRR is the dominant performance metric in private markets, what its two structural biases are, and why a constant-rate metric cannot correctly measure performance when rates are stochastic.

Reading: Book Chapter 4, sections 4.2–4.4 (~10 pages).

For Track 2 students who declared today: Begin reading Karatzas-Shreve Ch. 5 (Brownian motion and stochastic calculus). You’ll need this background by Session 25.


← Session 4 | Schedule | Next: Session 6 →