1. Institutional Perspectives on Real Estate Investing: the Role of Risk and Uncertainty Ravi Dhar William N. Goetzmann Yale School of Management

2. Problem Why is the allocation to Real Estate so low? (1) It’s the right number (2) costs and trouble (3) conceptual issues

3. Approach Collect current institutional investor views about real estate as an asset class. Goal: to understand determinant factors of institutional allocation to real estate. We survey a large sample of major institutional investors via a web questionnaire.

4. Findings Long-term estimates of risk and return are the driving factors A strong trend towards increasing their real estate allocation. Trend more pronounced for managers who felt relatively comfortable in relying on historical statistics about real estate returns. Conjectures about the role of uncertainty.

5. Data Collection Focus groups and interviews Greenwich Associates questionnaire E-mail administered. 202 responses, 173completed/1,500 mailed Focus on CIOs

6. Sample Size, Composition and Assets Under Management RespsComp.Assets $MM Endow/Found656094,184 Corp Pension8368274,556 Public Pension52431,114,916 Union221,310 Total2021731,484,966

8. Cross-Tabulation: Years Invested 0 to 23 to 45 to 1011 to 2020+No Endow/Foun0.3280.0690.2070.1720.0520.172 Corp Pension0.1210.0450.1060.1820.1360.409 Public Pension0.0880.059 0.4410.176

9. Table 3: Cross-Tabulation of Plans to Increase Allocation vs. Type of Institution Note: rows sum to 100% IncreaseDecreaseStay the SameUncertain Endow/Found0.5520.0000.2760.172 Corporate Pension0.2880.0450.4700.197 Public Pension0.2940.0290.5000.176

16. Views on Real Estate Expected Return EquityFixedHedgePEEM ER of RE significantly less6.4%1.7%4.0%30.1%24.3% ER of RE Little less42.2%4.6%16.8%31.8%34.1% ER of RE same22.5%5.8%24.3%7.5%6.9% ER of RE little more15.0%55.5%13.3%1.7%4.6% ER of RE significantly more1.7%20.2%2.3%0.6% Don't know12.1% 39.3%28.3%29.5%

17. Relative Risk EquityFixedHedgePEEM ER of RE significantly less11.6%2.3%9.8%42.8%41.0% ER of RE Little less49.1%11.0%30.6% 31.2% ER of RE same21.4%14.5%18.5%5.8%5.2% ER of RE little more11.0%54.3%13.3%1.2% ER of RE significantly more0.6%11.0%3.5%2.3%1.7% Don't know6.4%6.9%24.3%17.3%19.7%

18. Relative Cost EquityFixedHedgePEEM ER of RE significantly less4.6%6.4%15.0%21.4%4.6% ER of RE Little less7.5%4.6%20.8%23.7%15.6% ER of RE same12.7%5.8%15.0%18.5%22.0% ER of RE little more38.7%23.1%4.6%0.0%19.7% ER of RE significantly more17.3%41.0%0.0%1.7%2.9% Don't know19.1% 44.5%34.7%35.3%

19. Uncertainty and Decision-Making Frank Knight’s 1921 study Risk, Uncertainty and Profit Uncertainty: you don’t know the distribution Risk: you don’t know the Std. Bewley (1986) –investors were prone to inertia –the tendency to remain with the status quo choice. – might even explain the gross under—diversification observed among individual investors.

20. Institution vs. Individual Hirshleifer and Welch (2002) model the organization as distinct from the manager. As the institutional memory for why a decision was originally made fades Remaining managers have a bias towards the status quo in a steady economic environment. In a volatile environment they might exhibit impulsiveness.

21. Does Uncertainty Affect Choice? Factors to look for: –Extrapolation Comfort –Efficiency of Market Price

22. Extrapolation Comfort Least com234Most comDon’t know RE9.8%15.0%30.1%30.6%4.0%10.4% EQ10.4%15.0%26.0%29.5%14.5%4.6% FEQ11.6%26.6%27.2%24.3%4.0%6.4% EM22.5%35.3%19.7%11.6%0.0%11.0% Debt8.1%11.0%27.7%31.2%16.8%5.2% HY13.9%20.2%33.5%21.4%1.2%9.8% Com27.2%21.4%15.6%7.5%0.0%28.3% HF34.7%21.4%15.0%4.0%0.0%24.9% VC26.0%27.2%21.4%4.6%0.0%20.8%

23. A strong relationship between confidence of extrapolation and target allocation! Alloca tion Least com234 Most com Don’t know 0-118.4% 20.4%14.3%0.0%28.6% 1-58.0%20.0%40.0%24.0%2.0%6.0% 5-95.9% 29.4%52.9%2.9% >95.0%12.5%30.0%40.0%12.5%0.0%

24. A strong relationship between length of time invested in RE, however it is non-linear. Years Least.c om234 Most.co m Don’t know 0-218.2%9.1%29.1%20.0%0.0%23.6% 3-45.7%24.5%34.0%26.4%3.8%5.7% 5-103.7%14.8%25.9%44.4%7.4%3.7% 11-209.7% 25.8%45.2%6.5%3.2% 20+0.0%14.3%42.9%28.6%14.3%0.0%

25. Crash Probability in next 10 years A Crash is:Real EstateU.S. EquitiesFixed Income Not at all likely2.9%1.7%13.9% Not too likely45.7%30.6%50.9% Somewhat28.3%43.9%20.2% Very10.4%13.9%5.2% Extremely1.2%0.6%0.0% Don’t know11.6%9.2%9.8%

26. Relative Efficiency Real estate is:EQFEQEMDebtHYCom Substantially less Efficient22.5%11.0%4.6%28.3%5.8%11.6% Somewhat less30.6%34.7%18.5%33.5%31.2%20.8% About as17.9%16.8%16.2%21.4%30.1%14.5% Somewhat more15.0%20.2%29.5%2.9%13.9%15.6% Substantially more Efficient4.0%4.6%13.9%2.9%2.3%3.5% Don’t know9.8%12.7%17.3%11.0%16.8%34.1%

27. Conclusions MPT statistical model as base of decision. Other costs, issues and risks pertinent. Uncertainty (as opposed to risk) plays a role in allocation. Strong recent “herding” into real estate. Is this herding irrational? Is it informational?