FIN9200: Portfolio Management

Group Project #1

Asset Allocation Decisions at University of Iowa

(The data and information discussed in this document are based on public disclosures by the University of Iowa and the Iowa Board of Regents)

In this assignment, you are asked to assist University of Iowa to review its asset allocation policies. The investment team at the universitys Office of Treasury Operations periodically analyzes the policy portfolio weights to ensure the portfolios it manages are consistent with the ongoing market conditions. Consider yourself a member of the investment team tasked with such an analysis. Based on long-term the return and risk expectations on a set of asset classes, you are asked to first generate a series of feasible optimal portfolios subject to various constraints. You will then compare the current policy weights with the optimal ones. Based on the analysis, you will provide a recommendation to the Treasurer on whether the current asset allocation policies require revision.

I. Background

I.A Investment Management at UI

The treasurers office of the University of Iowa currently manages about $3 billion of funds in financial investments. Depending on their purposes, the funds are divided into mainly three portfolios. The operating portfolio houses money intended for short-term use, i.e., the daily financial operations. The intermediate portfolio manages funds intended for use in the intermediate horizon, e.g., five to ten years. An example of such funds is the money allocated for the expansion plan of university-affiliated hospitals. Finally, the endowment portfolio manages the universitys endowment and quasi-endowment money. The endowment portfolio is managed to maintain and preserve the long-term value of the funds. At the end of September 2023, the three portfolios have assets of $2,065 million, $504 million, and $657 million respectively.¹

The majority of the universitys investments are delegated to external fund managers. Any internal investment decisions, such as changing asset allocations, balancing investment portfolios, the selection and interaction with external fund managers, are handled by the treasurers office. The university also hires an external investment consulting firm to provide advice and assistance on asset allocation and investment manager selection. The current consultant is Marquette Associates, a Chicago-based boutique firm that provides investment advice for public pensions and non-profit organizations. Previously, Wilshire Associates served as the investment consultant from May 1996 to October 2015.

I.B Strategic and Tactical Asset Allocation Decisions

The Iowa Board of Regents has determined that all the three state universities should invest in the following assets class categories: US equities, International (non-US) equities, fixed-income securities, private equities, real assets, and cash. These assets differ in their characteristics of expected return, risk, and liquidity. Their returns tend to be imperfectly correlated.

One of the most important investment decisions the University of Iowa has to make is the portfolio allocation policies across the asset classes permitted by the Board of Regents. This is also known as strategic asset allocation in the investment management practice. The strategic asset allocation decisions are based on the long-run risk and return characteristics of asset classes (e.g., what will happen to each asset class in the next ten years), as well as the level of risk and liquidity desirable for each of the three portfolios the operating portfolio, the intermediate portfolio, and the endowment portfolio. The outcomes of this decision are the target portfolio weights on various asset classes, known as the policy weights. Once set, the policy weights are expected to remain relatively stable, and are revised only every a few years. At any given time, the actual portfolio weights are allowed to deviate away from the policy weights as long as such deviations are within a reasonable range, typically 2.5% to 5%.

Since the three investment portfolios have different risk and liquidity requirements, they have different policy weights on the same set of asset classes. For example, since the funds in the operating portfolio are expected to be withdrawn within a year, this portfolio should hold mostly low-risk and liquid investments. As a consequence, cash is typically the largest asset class in this portfolio, followed by fixed income. Real assets and equities have smaller weights in the operating portfolio, and private equities are considered not appropriate for the operating portfolio due to the highly illiquid nature of such investments. By comparison, the intermediate portfolio can tolerate a higher level of risk and can afford to hold more illiquid investments, and therefore has a higher weight on fixed income and a lower weight on cash. The endowment portfolio has the longest investment horizon by far, and therefore holds more risky and illiquid assets such as public equities (US and international), real estate, and private equities.

Another important investment decision made internally is tactical asset allocation. Relative to the strategic allocation decision, the tactical decision focuses on the risk and return of asset classes at shorter horizons, e.g., one to three years.

The risk and return characteristics at short horizons can deviate substantially from the long-run. Normally, equities are more risky and have higher returns than bonds. However, during the five- year period that spans the great financial crisis (2008 to 2012), bonds significantly outperformed equities. For example, the annualized returns to high-yield bonds and core bonds during this period are 10.3% and 5.9% respectively, while the returns to US equities, developed market equities, and emerging market equities are 2.0%, -3.2%, and -0.6% respectively. Commodities are the worst- performing asset class during this period, with an average annual return of -5.2%.

To the extent that short-term risk and return changes are predictable, it may be desirable to adjust portfolio weights according to predicted market conditions of each asset classes. However, predicting asset returns at short horizon is no easy task; it requires substantial financial expertise and efforts.

The UI investment team does not consider itself to have any competitive advantages in making tactical decisions. For this reason, the university does not aggressively pursue tactical asset allocation in its investments. When its actual portfolio weights deviate from the policy weights, it is more likely for reasons other than tactical allocation. For example, the actual weight of an asset class may be higher (lower) than the policy weight after the asset class has generated high (low) returns.

Nonetheless, some policy weight changes in recent years could be viewed as response to changes in the interest rate environment, and therefore have the flavor of tactical decisions. For example, in light of low yields of fixed income assets, the operating and intermediate portfolios started to invest in an equity subclass, Global Low Volatility, which features fixed-income like risk but equity-like expected returns. Out of the same concern, part of the fixed income allocation was dedicated to the subclasses of Senior Bank Loans and High Yield Corporate Bonds.

I.CChanges in Asset Allocation Policies Over Time

The Board of Regents and the University review investment policies periodically. As a result of these reviews, the policy weights on various asset classes change over time. Major asset allocation policy changes took place in 2008, 2013, 2014, and 2017. Appendix A provides asset allocation policies at the end of 2012, 2016, and 2020. A major shift in the endowment portfolio is the increase of policy weight on private equity. The weight was 10% in 2012 and 2016. After the 2017 review, as the board members and investment staff became more comfortable with alternative asset classes, the policy weight to private equity was increased to 25%.

II. Asset Allocation Analysis, February 2025

As a member of the investment team at the Treasurers office, you are asked to perform an analysis on whether the policy portfolio weights on major asset classes for the endowment portfolio are within a reasonable range of what might be considered optimal, based on a set of long-term return and risk projections. If there are significant differences between the policy portfolio and the optimal portfolio, you will need to alert the Treasurer for possible revision of the endowment portfolio.

II.A An Example of Analysis

Appendix B provides an example of what (part of) the analysis might involve. In 2013, at the request of UI investment team, Wilshire Associates (investment consultant at that time) provided an analysis similar to what you are expected to perform. The appendix includes forecasts of expected returns, volatility and correlations of various asset classes, provided by the consultant. The expected return forecasts are for a long-term horizon, i.e., 10 year or above. The volatility (expected risk) and correlation forecasts are based on historical estimates but adjusted to incorporate recent trends. The consultant further produced what is known as the efficient frontier based on the portfolio constraints the university sets. The efficient frontier plot in the appendix shows that the endowment portfolio at that time is pretty close to be on the efficient frontier.

Note that when making short-horizon forecasts, it is straightforward to identify a risk-free rate. For example, the yield on Treasury securities with one-year maturity is considered the riskfree rate at short-horizon. When making long-horizon forecasts, however, it is much harder to identify an appropriate riskfree rate. In particular, because returns to cash and short-term Treasury securities fluctuate year to year, they are no longer considered risk-free at long horizons. Therefore, in Wilshires long-run forecasts, cash is considered a risky asset class with an expected annual return of 1.5% and an expected risk (annual standard deviation) of 1.25%.

II.B Data for Analysis

Appendix C provides the policy weights of University of Iowas endowment portfolio on various asset classes.

Appendix D, a spreadsheet file titled LTCMA202501.xlsx, provides the long-term forecasts for expected returns, volatilities, and correlations of the asset classes. The forecasts are obtained from several major brokerage firms and asset management firms, such as BlackRock, INVESCO, J.P. Morgan Asset Management, Wilshire Associates, Research Affiliates, State Street Global Advisors, Callan Associates, PGIM, Amundi, BNY Mellon, and INVESCO; see sheet Sources for a list of institutions. The forecasts are made by these firms during the recent period and have a forward-looking horizon of 5 to 10 years.

For the convenience of analysis, UIs policy weights and portfolio constraints are also provided in Appendix D.

II.C Your Assignment:

You are asked to perform the following specific analyses on the endowment portfolio.

Step 1: select a source of forecasts from Appendix D for your analysis. Provide your selection in the sign-up sheet on ICON. Ideally, each group should choose a different source of forecast. Do not choose a source that is marked as Stale or Do Not Use in the sheet Sources.

Step 2: produce the portfolio efficient frontier subject to the following portfolio constraints and plot the efficient frontier.

1. Eight asset (sub-)classes to be included in the endowment portfolio: US equities, non-US equities, core fixed income, high-yield fixed income, senior bank loans, global fixed income, real estate, and private markets.


2. No short-sale rule and no leverage rule: the university does not engage in short-selling, Thus, the portfolio weight on any asset (sub-)class must be non-negative. Nor does it use financial leverage (i.e., borrowing money to invest). Thus the portfolio weights on the eight asset classes must sum up to exactly 100%.


3. Due to liquidity concerns and other considerations, the university keeps policy weights on a few less-liquid or speculative asset (sub-)classes on or below the following maximum:

Max weight for high yield fixed income: 5% Max weight for bank loans: 5%

Max weight for private markets: 25% Max weight for real estate: 10%

Step 3: analyze the optimality/efficiency of the current endowment policy portfolio.

After producing the feasible efficient frontier, your next step is to compare the current endowment portfolio with the frontier. You only need to do this for the policy portfolio (i.e., the portfolio with the policy weights, not the actual weights). The quick and intuitive way to perform this analysis is to add the point representing the expected return and volatility of the policy endowment portfolio to the plot of the efficient frontier. An example for reference is the plot produced by Wilshire Associates in Appendix B.

Please let the Treasurer know if you find the policy portfolio to be very inefficient and thus it requires immediate revision. It is likely that the policy portfolio wont be exactly on the efficient frontier. However, if the policy portfolio is not too far below the efficient frontier on the plot, you wont be too concerned.

Step 4: identify the optimal portfolio based on the targeted portfolio volatility and compare the optimal portfolio weights with the policy weights.

In general, there is one unique optimal portfolio on the efficient frontier that suits the university the best based the universitys risk preference for the endowment portfolio. However, identify this optimal portfolio requires additional assumptions. The textbook approach is to define a utility function and then identify the portfolio that maximizes the expected utility. In reality, the utility function of an investor is hard to define.

A popular alternative approach of identifying the optimal portfolio is called risk targeting. In this approach, the optimal portfolio is the one that maximizes the expected return of the portfolio subject to that the volatility (standard deviation) of the portfolio is set to a target level.

In this step of the analysis, you decide to take the volatility of the policy endowment portfolio as the risk target. Historically, the targeted volatility of the endowment portfolio is between 11% and 12%, but the specific number is not provided. You can estimate this targeted volatility using the policy weights and the provided forecasts of volatilities and correlations of various asset classes. Note that depending on the source of forecasts you use, the estimated volatility target may not necessarily fall within the range of 11% to 12%. Just take your estimate as the target even if it falls outside that range.

Use the estimated volatility (call it X) as the volatility target. Based on this target, you are to find the optimal portfolio that maximizes the expected return subject to that the volatility of the portfolio equals X. In the efficient frontier plot, this amounts to find the point on the frontier that has the same x-axis value (volatility) as the policy portfolio.

After obtaining the optimal portfolio, compare its weights with those of the policy portfolio. Do you see dramatic differences? If so, what might explain/cause the big differences? And finally, based on your analysis in this step, do you recommend changes to the policy weights? Please provide a discussion.

Step 5: perform further analysis to address some of the following questions:

1. Effect of portfolio constraints. How different are the optimal portfolio weights in Step 4 if you do not impose any portfolio constraints?


2. Sensitivity to assumptions. Identify asset (sub)classes that have high correlations with others. Change the expected return assumptions of these assets slightly and see whether the optimal portfolio weights in Step 4 are sensitive to the Try this for both with and without portfolio constraints.


3. Importance of private equity. Remove the private equity asset class. Suppose you still would like to keep the targeted volatility identified in Step 4, what expected return would the optimal portfolio delivers?


4. Effect of riskfree How different will the efficient frontier and the optimal portfolio look like if a riskfree asset is included? You can use the yield on the 10-year US Treasury as the riskfree asset.


5. Any other questions that you believe are interesting and

You can pick any (one or two) of the above questions for analysis. The instructions for the analyses may appear vague. This is intentional; you will need to first have an understanding of the possible issues of interest, before developing your own procedures or assumptions to complete the analysis.

Submission Instructions

Perform the required analyses specified in the three steps. Write a memo explaining the results of your analyses. Put your memo in a WORD file with no more than 5 pages (no page restriction on plots or tables if you would like to include as appendices). In addition, put all your spreadsheet analysis in an EXCEL file. Make sure to put the names of all team members in both files. No other requirements on format or content.

Submit both files to ICON. Since this is a group assignment, each group can select one member to submit the files.