Benchmarking real estate investment performance: the application of real estate indices.
Nearly all of the real estate newsletters and conferences that were targeted to pension funds in 1997 had benchmarking and performance measurement as major topics. Why was this topic so popular? The relatively poor performance experienced by pension funds with their real estate investments in the late '80s and throughout the '90s is a big factor. Another contributing factor is the perception that real estate performance measurement and benchmarking practices significantly lag behind benchmarking techniques for other asset classes, such as stocks and bonds.
When you can pick up the daily financial newspapers and find at least eleven stock indices, four bond indices, and four mutual fund indices reflecting the previous day's market close, real estate does appear to be lagging behind. The most widely used index to benchmark real estate, the National Council of Real Estate Investment Fiduciaries (NCREIF) Property Index, tracks quarterly total return performance and is published at least 90 days after the close of the previous quarter.
While this disparity may cause you to shake your head, comparing real estate to stock benchmarks like the S&P 500 is like comparing apples to oranges. Stocks are publicly-traded investments, while most real estate is still owned and traded privately [ILLUSTRATION FOR FIGURE 1 OMITTED]. However, when you compare real estate benchmarking efforts to other private asset classes (e.g. private equity, venture capital, and hedge funds), then the relative efforts are commendable.
To understand the value and the limitations of real estate benchmark practices, it is useful to review the more widely-used real estate benchmarks and how they are employed by pension funds and other real estate investors to evaluate real estate performance. It only includes private real estate performance and excludes publicly-traded real estate investments, such as REIT stocks. There are four principal methods for measuring the performance of private asset classes - indices, statistical universe/peer group comparisons, normal portfolio/portfolio opportunity distributions, and risk-adjusted performance measures.
Any discussion of institutional real estate benchmarks necessarily begins with the NCREIF indices. Today, NCREIF maintains several indices, including composite total return and property type/geographic sub-indices.
Key features of the NCREIF Property Index include: * Index series dates back to 1978.
* The index tracks the property-level performance of over 2,500 properties valued at over $60 billion.
* Calculations are based on quarterly returns of individual properties before deduction of management fees.
* Each property's return is weighted by its market value
* All properties have been acquired on behalf of tax-exempt institutions and held in a fiduciary environment.
* All properties are existing, investment-grade commercial properties.
* Each property's market value is determined by real estate appraisal methodology, consistently applied.
The NCREIF Property Index accounts for approximately 25 percent of the investment-grade real estate owned directly by pension funds and other institutional investors. In contrast, the S&P 500 accounts for 70 percent of the Stock market capitalization.
The NCREIF Property Index is a property-level index, which makes it a good benchmark for individual property performance, but not for evaluating fund manager performance. Nevertheless, many pension plans compare their real estate managers' performance to either the entire NCREIF Property Index or one or more its sub-indices. For example, if a real estate advisor manages a private investment fund of regional malls in the East, a pension fund may be inclined to compare the overall fund performance to the NCREIF property sub-index of regional malls in the East.
Understandably, the NCREIF Performance Index is widely used because it provides much more information then is available from any other real estate index. For example, the third-quarter 1997 NCREIF "Metropolitan Statistical Area Detailed Quarterly Performance Report" tracked the performance of 22 office buildings in the Houston MSA alone, a level of detail unavailable from other indices.
To properly select and apply an index for performance benchmarking purposes, the analyst must understand the index design and computation methodology. One important design feature is whether the index member returns are weighed on a market value or equal unit weight basis. The popular Dow Jones Industrial Average is calculated on an equal value basis. The NCREIF Property Index is a market value-weighted index similar to the S&P 500 Index; this means that the return on a $100 million property will carry more weight in the index compared to the returns generated by a $10 million property. It would not be appropriate to use an equally-weighted index for real estate benchmarking purposes since it is unrealistic to acquire and own real estate assets in equally sized units.
Another factor to consider in benchmarking is to ensure that both the portfolio returns and index returns are computed on a time-weighted basis. Time-weighted returns, as contrasted with dollar-weighted returns (also referred to as Internal Rate of Return), ignore the return impacts due to the timing of cash flows. Manager performance standards promulgated by the Association of Investment Management Research require that stock, bond, and real estate investment managers use time-weighted returns for reporting investment performance. It may not be unusual for some real estate managers to report their performance on the basis of cash flows, which would not be comparable to any index in which the underlying methodology resulted in time-weighted returns. The computation of Internal Rate of Return based on cash flow is equivalent to the computation of dollar-weighted returns.
A second technique for evaluating investment managers' performance is peer evaluations, that is, comparing the manager's performance results with a group of similar real estate managers. This is referred to as a statistical universe, or peer group, comparison. Figure 2 provides a listing of real estate manager and/or fund performance indices that may be employed in this technique. These manager/fund indices are usually published by investment consulting firms. The consultants compile the manager/fund performance data, produce time-weighted indicies, and also have the capability to calculate quartile performance results. For example, a specific fund featuring a regional mali investment strategy may be charted relative to the quartile performance for 25 funds that featured a similar strategy over the same period.
Figure 3 is a hypothetical report that displays the annual performance for all funds that are grouped into a statistical universe report. The table in Figure 3 discloses the highest return, the first-quartile breakpoint, the median, the third-quartile breakpoint, and the lowest return as compared to the subject fund labeled "Your Performance." In 1996, the peer group returns ranged from a low of 2.6 percent to a high of 14.1 percent, with a median of 8.1 percent. In comparison, the subject fund's performance of 5.9 percent placed it in the third quartile of relative performance.
To maximize the benefits from peer group analysis, investors employing this method need to be fully informed about the universe design methodology and peer group selection criteria and limitations of this approach. Otherwise eligible funds may be excluded for various reasons by the universe creators and publishers (e.g. ambiguous strategy classifications, incomplete information, and shareholder classifications). Additionally, some fund manager may not fully cooperate with the regular reporting process. Again, time-weighted return calculations for both the subject fund and the peer universe will ensure a consistent comparison.
A third benchmarking technique is known as normal portfolio, or portfolio opportunity, distribution. The goal of this approach is to measure the added value produced by the investment manager relative to a randomly generated investment portfolio with the same investment strategy. This technique first identifies the specific investment objectives, establishes the opportunity set in which the manager may invest, and then randomly generates portfolios that are created, measured, and weighted into a benchmark return series. This type of analysis is quantitative and requires extensive data to be valid. Additionally, it is quite difficult to apply this technique to private real estate given both the data requirement, the ambiguity that is frequently associated with investment objectives, and the fact that all real estate investment opportunities are not equally accessible to all market participants:
Risk-adjusted performance measures are based on the underlying expectation that high (low) risk portfolios should generate high (low) returns. Figure 4 presents the calculation of the Sharpe Ratio, which represents one method to evaluate risk-adjusted returns. The portfolio performance premium (return in excess of the risk-free rate) is divided by the measured volatility (standard deviation) of the portfolio. The resulting ratio expresses the amount of premium return per unit of risk. High ratios represent superior performance since the ratios adjust the returns for the risks incurred. This technique may be applied to the real estate asset class provided all portfolios use a comparable basis of determining periodic values and measuring volatility.
Finding the Best Benchmark
Given the complexities associated with the four benchmarking methods presented in this article, some people may try to keep the process simple and just compare manager performance to public real estate benchmarks such as the National Association of Real Estate Investment Trusts (NAREIT) indices consisting of publicly-traded REITs. After all, stock managers compare themselves to the S&P 500 Index.
However, these indices have changed significantly in recent years with the rapid growth of both the market capitalization and the index members (number of companies). In addition, public real estate security indexes still represent a small portion of the real estate universe. Finally, the underlying financial and operating characteristics of associated real estate companies may be quite different from the investment portfolio of properties being reviewed. Benchmarking analysis with real estate security indexes are to be encouraged, but with the caveat that other benchmarking techniques should also be employed and that extensive disclosures should emphasize the differences between the subject portfolio and the benchmark index employed.
Over time, practitioners will continue to improve upon the most basic technique for benchmarking real estate management performance, which is called "the herbal tea and granola approach" by investment consultant Jeffrey Slocum. This method basically boils down to a three step process of evaluating the manager against:
* what he was told to do,
* the strategy he said he was going to apply, and
* his ability to stay within established risk guidelines.
As the name implies, this approach is more hands-on and statistically less rigorous than the other benchmarking techniques we have described. But given the complex nature of the other approaches, it may give the investor the best qualitative measure of whether the manager met the original expectations established.
FIGURE 2: STATISTICAL UNIVERSE/PEER GROUP COMPARISONS
* IPC Portfolio Index
* Evaluation Associates Index
* Callan REal Estate Composite Index
* Russell Performance Index
* Wilshire Real Estate Fund Index
FIGURE 4: RISK ADJUSTED PERFORMANCE MEASURES
Sharpe Ratio = Return (i) - Return (TBill)/Volatility (i)
(i) = investment
(TBill) = Rosk Free Treasury Bill
RELATED ARTICLE: How Pension Funds Are Measured
Pension fund administrators strive to maintain the right balance between risk and reward at a reasonable cost. Through active management, they try to create value for the fund's beneficiaries.
As shown in Figure 5, one pension fund consultant, Keith Ambachtsheer of Cost Effective Management, describes a process that pension funds may employ to evaluate their own performance in communications that are directed to the plan beneficiaries. In determining how you will be evaluated on real estate performance, it is useful to understand how real estate money managers are scored. At the end of any investment period, the portfolio gross return is calculated, in this case 10.7 percent. This represents the total return for the year for all of the funds' assets. From this, we can subtract what the pension plan assets would have generated if they were all invested in risk-free assets (such as U.S. Government Bonds, shown at 8.4 percent). This results in a gross risk premium earned, calculated as 2.3 percent.
The pension plan manager endeavors to produce this premium by doing two things: selecting an asset mix beyond just investing in Treasuries and by actively managing those assets versus just buying index funds. Subtraction of the pension fund's costs to manage these assets (0.3 percent) results in a net risk premium earned, calculated as 2 percent.
Finally, a risk premium for investing in risky assets is deducted, which is 1.6 percent in the example. This represents the "charge" for taking on increased return volatility, which is calculated by applying a 15-percent hurdle rate to an increase in the standard deviation of return times two (5.3% x 2 x .15). The result is a risk-adjusted, value-added return that, hopefully, is positive. In this example, the amount is 0.4 percent, showing the benefit to the plan's beneficiaries of having the plan administrator diversify and actively manage their assets.
Also note that while 0.4 percent may not seem like a great deal, it can represent a significant amount of money when a fund manager is overseeing millions in assets.
Kevin J. Maxwell and Paul S. Saint-Pierre
Kevin J. Maxwell is senior vice president of The O'Connor Group, a real estate investment company in New York City.
Paul S. Saint-Pierre is the senior equity analyst at Putnam, Lovel, and Thornton in San Francisco.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||part two; includes related article on pension-fund measurement|
|Author:||Maxwell, Kevin J.; Saint-Pierre, Paul S.|
|Publication:||Journal of Property Management|
|Date:||May 1, 1998|
|Previous Article:||Between the lines: profitable garage management.|
|Next Article:||Telephone terror.|