Agency relationships in appraising for institutional asset managers.
The focus of this article is open-end equity CREFs that are typically managed by insurance companies, bank holding companies, and real estate investment companies on behalf of their pension fund clients. Investors in open-end commingled funds buy and sell CREF units on the basis of appraised values of the properties within each fund. The incentive and compensation schemes for parties either involved in or affected by valuations of externally managed properties are addressed here. Left unchecked, these agency relationships and resulting agency problems may cause non-random errors(2) in individual valuations of properties held in institutional portfolios.
First, agency theory is reviewed, the importance of appraisal accuracy is discussed, and historical evidence regarding appraisal accuracy is briefly examined. The various sources of potential agency problems facing institutional property managers are then examined, and specific methods for reducing potential agency problems are briefly explained.
AGENCY THEORY AND APPRAISAL ACCURACY
Agency relationships occur when one party, the principal, hires another party, the agent, to perform a service on behalf of the principal. The most frequently cited agency relationship is that between stockholders and managers. Prior to the development of agency theory, it was typically assumed that agents acted in a strictly professional manner in performing their services, without regard for the economic consequences of their actions. Now, however, greater attention is being paid to possible conflicts of interest between parties, known as agency problems, and means of resolving such conflicts, which may involve agency costs.
Agency problems arise because of conflicting economic incentives between a principal and an agent. For stockholders and managers, the conflict is usually maximizing the value of the firm versus maximizing manager wealth. A potential solution to this dilemma is to create the common goal of maximizing the value of the firm by allowing managers to become stockholders. Costs of making agents' behavior consistent with achieving the common goal by compensating managers in the form of stock options or some other form of compensation are referred to as agency COSTS.(3)
The importance of appraisal accuracy
An ongoing agency problem exists between CREF investors and asset managers because the fee each manager receives from investors for managing a portfolio is typically calculated as a percentage of the appraised values of the assets under management (because actual selling prices cannot be observed).(4) A consequence of this type of management fee structure is that there is an economic incentive for CREF management to maximize valuations of their fund's properties.
Further, another agency problem exists between asset managers and appraisers because they too maintain an ongoing relationship. Specifically, because CREF managers hire appraisers to undertake valuations at least annually for each individual property within a fund, an economic incentive exists for these appraisers to furnish valuations beneficial to asset managers.
Currently, many CREF managers repeatedly use the same appraisal firm, or firms, each year for their independent appraisals. Managers can merely discontinue hiring a particular appraisal firm if the results of that firm's annual valuations do not satisfy management's expectations. Obviously, the loss of a lucrative CREF appraisal contract could be costly to any independent appraisal firm.
Clearly, the heavy reliance on appraisals as proxies for market values of properties in institutional portfolios makes their accuracy critically important to market participants and observers. Of greatest concern to CREF investors is that wealth actually changes hands on the basis of appraised values of the properties within each CREF. In addition, measurement of investment performance as well as compensation of asset managers is based on aggregated appraisals of individual CREF properties. As a result, the unit trading prices of CREF investments may be subject to the influence of these and other agency problems.
The accuracy of CREF property appraisals is important to market observers as well as to market participants because valuations of individual properties held in CREFs are used to create aggregated equity real estate returns series such as the Russell-NCREIF Property Index. These returns series are then used in empirical studies of commercial real estate and mixed-asset portfolios that include equity real estate.
Recent studies that examine the inflation-hedging and diversification benefits of including equity real estate in a traditional pension fund portfolio of stocks and bonds have concluded that equity real estate should be an integral part of the institutional asset mix. In fact, empirical results of historical returns series in these studies indicate that from 15% to 60% of a total mixed-asset pension fund portfolio should be invested in equity real estate if risk-adjusted portfolio return is to be maximized.
For example, studies such as those by Webb, Curcio, and Rubens(5) and Firstenburg, Ross, and Zisler,(6) which specifically examine and identify several mean/variance-efficient combinations of stocks, bonds, and equity real estate assets imply that equity real estate should be at least one-third of the total value of a mixed-asset class institutional portfolio. In making their findings, the studies each use historical returns on CREF investments as well as returns on other types of real estate in combination with historical returns on stocks and bonds.
The use of appraised values for real estate returns and market values for stock and bond returns in the construction of mean/variance-efficient portfolios makes the results of such methods suspect at best. Unreflected differences in liquidity between commercial real estate and stocks and bonds make this approach questionable as well. Further, agency problems that influence CREF appraisals may directly affect the findings of these studies. The problem for analysts remains, however, that broadly diversified market value databases on individual commercial real estate properties simply do not exist.
Historical evidence on the accuracy of appraisals
In practice, it is difficult to assess the accuracy of an appraisal without the sale of the property. Therefore, it is virtually impossible to draw conclusions as to the accuracy of most CREF property appraisals in light of the fact that they are typically made on an interim basis in the time between the purchase and sale of a property. Moreover, motives may exist for selling CREF properties that would bring into question straight comparisons of sale prices to appraised values.(7) Roberts and Roberts have further argued that it is unrealistic to expect that commercial property appraisals should be within 10% of selling prices because highly different circumstances are often involved in valuing versus selling commercial properties.(8)
To no one's surprise, published studies that examine the differences between appraisals of CREF properties and their actual selling prices have found significant differences between some of the appraised values and the selling prices of individual properties. In fact, in analyses of hundreds of properties taken from over a dozen different CREFs, more than 25% of the most recent independent appraisals prior to their being sold differed from actual selling prices by more than 10%.(9) Further examination of all studies published to date, however, shows no clear evidence of systematic bias, either negative or positive, in the aggregate.
In terms of the reliability of CREF appraisals, acceptable levels of accuracy have never been established within the CREF industry. In light of the large number of factors that affect valuations and selling motives differently, it may be unreasonable to expect that the differences between independent CREF appraisals and actual selling prices should always be lower than some arbitrarily selected rate such as 10%.
In any case, investors are concerned not only with the appraisal accuracy of sold properties,(10) but with the accuracy of interim appraisals of unsold properties. In fact, large errors in interim appraisals would be of monumental concern to investors, particularly if they believed the source of the errors to be primarily nonrandom.(11) Of course, if appraisal errors were entirely random, they would tend to cancel each other out in a portfolio of properties of equal value, but CREF properties range in size from less than $1 million to more than $100 million. Thus, nonrandom errors, whether positive or negative, would tend to adversely affect investor confidence in commingled funds.
The fact that both unit prices of CREFs and asset management fees are determined by appraised values of the properties within them certainly leads to investor concerns over the accuracy of each appraisal performed.(12) A detailed discussion of these and other investor concerns with regard to the accuracy of appraisals follows.
SOURCES OF POTENTIAL AGENCY PROBLEMS IN APPRAISALS
Management fee incentives
As was previously mentioned, a recurring concern of investors in open-end CREFs has been that CREF unit prices and asset management fees are determined on the basis of appraised values of the underlying properties rather than on actual selling prices. The major cause of investor disapproval over this type of trading and compensation scheme is neither the appraisal process nor appraisers or CREF managers, but the fact that an agency relationship exists such that asset managers have an economic incentive to provide information to an appraiser that may improperly influence the appraisal process.
In practice, most of the information used in appraising individual CREF properties is provided to independent appraisers by each commingled fund's manager. It may be difficult, if not impossible, for an appraiser to verify all of the data from the CREF manager because the manager is both the client and the source of much of the needed input data. Clearly, an unsuspecting appraiser or analyst may unknowingly incorporate into a report data from a client that are inaccurate or incomplete. It is also plausible that an appraiser may arrive at a value estimate not achievable absent such deficient information.
Examples of types of information used in an appraisal assignment that may be difficult to obtain include the status of present leases with renewal options and the terms of leases currently under negotiation. In recent years, the oversupply of office space in many markets around the country has caused building owners to offer various types and levels of rent concessions. An asset manager may be reluctant to fully disclose the magnitude of rent concessions to an appraiser or anyone else for fear that such concessions may become publicly known.(13) In such instances, the appraiser may wish to inform the client of Canon 4 of the Code of Professional Ethics of the Appraisal Institute, which states that an appraiser "must not violate the confidential nature of the appraiser-client relationship."(14)
Certainly it is an independent appraiser's responsibility to gather and interpret relevant information concerning market conditions, including the use of rent concessions. An appraiser, however, cannot review all leases and pending leases within a market. In many instances, then, an appraiser must first gain the trust of a client by stressing the utmost respect for the confidentiality of the client's lease data. Then, the lease information provided by a client may be assumed to be complete and accurate, unless data obtained by the appraiser from alternative sources prove otherwise.
The complex agency relationships among investors, fund managers, and appraisers result in agency problems among the parties for which there are no simple solutions. Clearly, however, the mere possibility of "sweetheart appraisals" must be eliminated to fully resolve these agency problems. Existing constraints to these problems are discussed next.
One investor constraint on the behavior of managers of open-end funds is the ability of investors to withdraw from a CREF if it is believed that overall CREF portfolio value is overstated. There may be a considerable lag, however, between the time a redemption request is initially made and the time it is fully completed. In fact, complete withdrawals may take a year or more to fully execute.
There are, of course, other less drastic measures that can be taken by investors to reduce the likelihood of non-random errors in appraisals. For example, although investors have limited means of knowing the actual extent to which CREF appraisals are influenced by agency problems, they can and often do hire their own independent appraisers to verify asset managers' appraisals. Investors may also examine the scope of CREF management's appraisals for clues as to their accuracy.
In some instances, investors even hire independent appraisers to perform separate narrative appraisals that produce alternate value estimates of CREF properties. The agency costs for this type of monitoring procedure are quite prohibitive, however, and may only be warranted if investors have serious doubts as to the accuracy of particular valuations of relatively large properties. Appraisals funded by investors may or may not increase their confidence in overall CREF portfolio values, depending on the findings. In any case, however, the ability of market participants' appraisers to consistently produce more accurate value estimates than those generated by asset managers' appraisers would be highly unlikely in light of information asymmetries between CREF managers and investors' appraisers. Further, agency problems related to compensation as well as asymmetric information exist between investors and their appraisers.
Other constraints in place to inhibit parties involved in CREF valuations from attempting to influence the appraisal process are the professional and regulatory standards. For example, a fund manager is regulated by federal fiduciary responsibilities outlined in the Employee Retirement Income Security Act of 1974, and an appraiser is constrained by the rigorous professional and ethical standards of such professional associations as the Appraisal Institute, which, if violated, could result in a member's censure, suspension, or expulsion.(15) The loss of business to an appraisal firm that violates its own strictly enforced professional standards clearly acts as a powerful economic disincentive to provide clients with beneficial value estimates.
In sum, current agency relationships create an economic incentive for an asset manager to maximize the value of the commingled fund, and for an independent appraiser to comply in this endeavor. Acting as constraints on these agency problems are the ability of an investor to monitor an asset manager and withdraw from a fund as well as the integrity, credibility, and professional reputation of both a fund manager and an appraiser.
The combination of these disincentives with existing economic incentives does not, however, eliminate the potential for agency problems to influence value estimates. For example, because it is possible for an asset manager to provide an asymmetric information set to an appraiser that results in a value estimate higher than otherwise could have been reached, it is also entirely possible that all parties to an appraisal, with the exception of the manager, will believe that any appraisal error is purely random.(16) Even if investors believe that there is systematic non-random error in managers' appraisals, they may not be willing to incur the monitoring costs associated with identifying the magnitude of any errors unless they believe the benefits of monitoring will outweigh the costs.
Other issues related to appraisals of CREF properties are: 1) the types of manager-initiated adjustments or updates made to independent appraisals; 2) the treatment of anticipated future closing costs in appraisals of unsold CREF properties; 3) the sensitivity of appraisals to the critical assumptions and particular methods used by an appraisal firm in the valuation process;(17) 4) the frequency and timing of both the in-house CREF valuation updates and the independent appraisals; and 5) the stability and smoothness of overall CREF returns as opposed to returns on other asset classes such as stocks and bonds. These additional concerns are briefly discussed next.
Properties in open-end CREF portfolios are typically appraised by an independent appraisal firm at least once each year. The largest properties within a pooled fund may be independently evaluated on a more frequent basis at the CREF manager's discretion, because their contribution to overall fund value is most significant.
During the year-long intervals between independent appraisals, CREF managers also undertake in-house staff appraisals of the properties in their fund, usually on a quarterly basis. The regularity and degree of thoroughness of in-house appraisals, however, varies from fund to fund. They are typically updates of the most recent independent appraisals rather than full narrative reports, but may also include the most recent sale data on comparables as well as updated discounted cash flow (DCF) analyses. These interim updates may of course either increase or decrease the appraised values of the most recent independent appraisals. Reasons for increasing previous independent appraised values of individual properties include: 1) general adjustments made to reflect favorable macroeconomic changes in the particular real estate market; and 2) specific adjustments made to capture rental increases or "value-added" capital improvements made to properties. The latter type of upward adjustment is easiest for a staff appraiser to justify, although the potential exists for dispute as to whether certain capital improvements truly add value.
Downward adjustments, or "writedowns," to property values are usually made on the basis of local, market-related phenomena such as a suddenly weakening local economy or an oversupply of space that was not evident during the course of the most recent independent appraisal. Normal depreciation and deterioration of quality, on the other hand, are not routine reasons for making in-house adjustments to value.
Other investor concerns focus on differences in assumptions and methods in the valuation process that are a result of asset manager influence or repeated error. For example, closing costs may or may not be considered by the asset manager, or they may be estimated inaccurately in the appraisal report. Further, other errors or omissions could be made repeatedly by the same appraiser, or critical assumptions, whether or not they are influenced by asset managers, could be unrealistic.
It is conceivable that an independent appraiser might inadvertently make the same error repeatedly on a particular property each time it is evaluated. It is also conceivable that an appraiser who has not remained current as to consensus forecasts of local market conditions could make significant errors on critical assumptions related to the future strength of the local economy.
Conservative assumptions may result in the undervaluation of a property just as bold and aggressive assumptions could result in overstated values. In addition, in some cases minute differences in appraisal assumptions can have a large impact on valuations. The most frequently cited example of the sensitivity of appraisals to small changes is based on the assumed discount rate used in DCF analysis. A small percentage change in the discount rate typically results in a much larger percentage change in the value estimate. For pooled funds, investment performance measurements based on appraised values may be even more sensitive to changes in the discount rate than the value estimates themselves.(18)
A remaining concern among investors in open-end commingled funds is that their unit prices will "jump" on the basis of the frequency and timing of both independent and in-house staff appraisals, particularly in the quarter that independent appraisals of the largest properties in a fund are reported. Currently, there is little uniformity within the CREF industry with the exception that independent appraisals are usually performed once a year. Across funds, inconsistencies abound in terms of the quarter in which independent valuations are made. Some CREF managers have most or all of these appraisals done in the same quarter each year while other asset managers spread them out evenly throughout the year.
In terms of the frequency and timing of in-house appraisals, some fund managers undertake them each quarter (other than the quarter of the independent appraisal) on all of the properties in their fund. Other managers perform staff appraisals regularly only for large properties, and on a sporadic basis, if at all, for small properties. This inconsistency across and within funds may result in either upward or downward errors in overall CREF values depending on the direction of unreflected changes in the market. Further, the use of "stale" comparable sale data may cause staff appraisals to lag behind market values, which, in turn, would cause CREF values to lag behind the true market values.
An additional problem related to the frequency and timing of appraisals focuses on the trading method for open-end CREF funds. Transaction prices of units are blind; that is, neither a buyer nor a seller knows what the actual unit price will be when purchase or redemption transactions are authorized. Typically, transactions occur only at the end of each quarter, and notice of purchases or withdrawals must be given by an investor one full quarter prior to the determination of the effective unit price.(19) Because trades usually occur quarterly, unit prices are only calculated once each quarter, and the value of the CREF in the interim period between quarters is not even estimated.
As a result of this type of trading scheme, both investors and asset managers may prefer that changes in unit price be smooth over time so that they will not be surprised by changes in the effective trading value. If appraisals of CREF properties are only undertaken annually, then CREF unit prices will indeed adjust at a slower rate than actual portfolio market values, regardless of whether the appraisals are spread out evenly throughout the year.
Any smoothing errors in overall CREF unit prices would tend to be positive during dramatic decreases in actual selling prices and negative during market surges. Also, smoothing errors would tend to reduce the volatility of CREF values over time, making it difficult to make straight comparisons of CREF returns to returns on other assets such as stocks and bonds. In the following section, suggestions are offered for reducing these and other potential sources of non-random error.
METHODS OF REDUCING POTENTIAL AGENCY PROBLEMS
Changing the asset management fee structure
There are several potential methods for further reducing agency problems caused by management fees. For example, one improvement might be to change the fee structure so that any fees based on property appreciation must be directly tied to the actual selling prices of the properties upon sale rather than the appraised values prior to sale. A potential problem with this type of performance-based incentive structure is that fund managers' actions regarding the timing and motivation for sales may be influenced by their desire to generate fees. Further, the Department of Labor has not officially approved this type of fee structure, and it may be interpreted as violating Employee Retirement Income Security Act requirements that specify that fees be based on fair market value of assets.
Another improvement might be for fund managers to purchase unsold redemption requests from investors at the stated unit price. This alternative, however, would generate additional agency and regulatory problems caused by the conflicting roles of an asset manager as both fiduciary of, and investor in, a pooled fund.
Yet another means of changing incentives would be to require CREF managers to reimburse investors for excess fees collected if properties sell below appraised values. If this were done, managers would not have economic incentives for sweetheart appraisals, and investors would not be inclined to suspect that CREF managers were attempting to obtain favorable appraisals.
A balanced approach to the issue of fees would be to pay them initially on the basis of appraised values, with fee adjustments made based on the actual selling prices of individual properties within a fund. Upon sale of a property, the fees that would have been paid if the value of the property had increased or decreased on a straight-line basis from the date of purchase can be calculated. Then, the differential between calculated fees and those that have actually been paid by investors can be reimbursed, credited, or deducted by the fund manager (on a pro-rata basis covering only the time that the investor held ownership interest in the fund). According to this type of fee arrangement, all parties would benefit from appraisal accuracy, and the agency costs of adopting such a fee structure would be minimal. In addition, current agency costs incurred by investors to monitor CREF appraisal accuracy would be reduced.
Other partial solutions to potential agency problems would be to require, or at least recommend, that uniform appraisal guidelines be followed by each CREF fund manager. Examples of requirements that might improve the accuracy and quality of CREF appraisals include: 1) consistently estimating and reflecting closing costs in the appraisals; 2) periodically rotating independent appraisers; 3) consulting more extensively with local or specialist appraisers and analysts each time an appraisal is undertaken; and 4) updating CREF appraisals more frequently while fully disclosing to an investor the costs and methods of updating and calculating effective CREF unit prices. These suggestions are briefly discussed in the following section.
Reflecting closing costs in the appraisals
Closing costs can be substantial, perhaps as much as 5% or more, especially if potential sale commissions are taken into consideration. Existing appraisal guidelines from the Appraisal Institute do not indicate whether closing costs should be subtracted prior to calculating the present value of the reversion. Also, few guidelines exist to explain the appropriate method of estimating closing costs. The inconsistency of treatment may result in large distortions in the value estimate, particularly for appraisals of large properties held by institutional asset managers.
A solution to this potential problem would be to establish detailed, consistent guidelines specifically for the purpose of valuing externally managed properties. Good appraisal practice, as suggested by the Appraisal Institute, calls for an appraiser to identify the purpose of the value estimate. The purpose is distinct from the objective in that estimating the market value may be the objective while the purpose could be to establish the adequacy of collateral for a mortgage, to assist in a merger of two firms, to settle an estate, or in this case, to price a CREF.
If an appraiser is told the purpose of an endeavor and is furnished with detailed valuation guidelines to be used for that specific purpose, then discrepancies caused by various treatments of the same issue (e.g., the treatment of closing costs) can be minimized. The National Council of Real Estate Investment Fiduciaries (NCREIF) has suggested guidelines to be used for the CREF appraisal engagement letter; however, these guidelines are general in nature. Guidelines for appraising for institutional asset managers would certainly be more beneficial to independent appraisers if they were written in more specific terms than those currently covered in a typical engagement letter.
Rotating independent appraisers
Several CREF managers publish information that describes their policies concerning the hiring of independent fee appraisers. In some cases, the fund literature lists the name of a single appraisal firm that handles all of a fund's valuation needs. Fund managers should be encouraged to deal with several different independent appraisal firms so that valuation assignments are routinely rotated and the same appraiser does not repeatedly provide a value estimate for any particular holding. Periodic rotations may serve to mitigate any errors, intentional or otherwise, that an appraiser brings to an assignment. They may also serve to increase investor confidence in the appraisal process by reducing the agency problems associated with an ongoing relationship between asset managers and appraisers.
One potentially negative result of a frequent rotation policy is a likely increase in the long-run costs of CREF appraisal reports. If the same appraisal firm provides the value estimate each year, it can generally afford to provide a report for a lower fee than if it were starting from scratch. Of course, the appraisal firm may also carry forward any prior errors in updating its work. Nonetheless, a first-time narrative appraisal of a large commercial property can be quite expensive. Collectively, these fees can amount to a substantial sum for an individual CREF over a relatively short period of time, and the costs versus the potential benefits, in terms of gains in accuracy, must be carefully considered by a CREF manager in determining both the appraisal firm and the scope of the appraisal assignment.
Consulting with other analysts
Local and specialist appraisers may or may not be used in an appraisal assignment depending on the property or the asset manager. Of course, ignorance of the local economy and lack of experience in appraising a certain type of property are not valid excuses for inferior appraisal work. These issues are addressed by the Appraisal Institute's "Competency Provision" of the Uniform Standards of Professional Appraisal Practice (USPAP). In a nutshell, input from local and specialist appraisers and analysts should be used by outside independent appraisers as a matter of standard practice whether appraising for institutional asset managers or any other client.
McIntosh suggests that the Appraisal Institute should periodically publish survey data on expectations of appraisers and analysts.(20) If this were done frequently for various property types and local markets, it would not only serve as a useful tool for appraisers and market participants in their analyses of market conditions, but it would also reduce the potential for asset managers to unduly influence the information used by appraisers in valuation assignments. This latter benefit would be further enhanced if detailed historical as well as forecast data regarding information such as lease terms and vacancy rates were included in the survey.(21)
Calculating unit prices more frequently
As discussed earlier, some funds reappraise properties each quarter and others just once a year. Also, CREF unit prices are only calculated at predetermined trading intervals (quarterly for most CREFs) rather than on a more frequent basis as new information is received. As a result, heavy emphasis is often placed on "stale" appraisals to calculate CREF unit prices, and their trading values may be artificially slow in adjusting to equilibrium market values. Further, as a result of the current trading scheme for open-end CREFs, both investors and asset managers tend to prefer that changes in CREF unit values occur gradually.
In practice, the costs of continually updating each appraisal for each CREF property would be ridiculously prohibitive and would far outweigh the benefits of more accurate CREF unit prices. There is, however, some optimal level of updating appraisals at which the marginal costs would be equal to the marginal benefits. As a matter of current policy, asset managers already carefully document total appraisal costs for having their CREF units valued. If they were to further break out and document internal and external appraisal expenses for each individual property and periodically inform investors of these costs, managers and investors would be better able to determine the most desirable frequency and scope of valuation for each property in the CREF portfolio.
In regard to the pricing of overall CREF portfolios, one potential method of reducing market participants' desires for small and steady changes in fund values would be to change the frequency of reporting unit prices of CREF investments. Unit price calculations should be adjusted between trading dates to reflect reported changes in income or appraised values of the underlying properties (or any other new information) as soon as the information becomes available to the asset manager. As a matter of standard CREF policy, the most recent valuations of individual properties should be reflected immediately in overall CREF values rather than at the end of each quarter, so that investors can better anticipate changes in quarterly unit prices. If this were done, the actual CREF unit prices to be used for purchase and redemption transactions would be less likely to present unexpected problems.
Specific guidelines set forth for independent appraisals of institutionally managed properties could serve to temper the problems that these and other inconsistencies present, yet virtually none exist. The Appraisal Institute, The Appraisal Foundation, the U.S. Department of Labor, NCREIF, and other concerned parties should coordinate their efforts to adopt a detailed set of guidelines for asset managers, appraisers, and others involved in the appraisal process to use when appraising institutionally managed properties.
The clear purpose of such guidelines should be to reduce agency problems, whether real or perceived, in an effort to improve both appraisal accuracy and investor confidence. In any case, investor confidence in appraisals of externally managed properties would certainly improve if the industry took formal action as a group to address the issue of agency problems. Investors' perceptions of the appraisal process would improve, and asset managers, appraisers, investors, and regulators would all benefit from the exchange. Ultimately, the reduction of agency problems facing market participants may even serve to permanently increase investor demand for commercial real estate.
1. One of the most widely used commercial property indexes is the Russell-NCREIF Property Index, which includes appraised values of more than 1,200 commercial properties by type and location held by more than 30 different CREF managers who are voting members of the National Council of Real Estate Investment Fiduciaries.
2. Previous studies focus on the statistical problems that occur when appraised values of individual properties are pooled to calculate portfolio returns. See, for example, S. Michael Giliberto, "A Note on the Use of Appraisal Data in Indexes of Performance," The Appraisal Journal (April 1988): 77-83; and David Geltner, "Bias in Appraisal-Based Returns," AREUEA Journal (Fall 1989): 338-352.
3. For a thorough discussion of the development of agency theory and agency costs, see the seminal work on the topic by Michael C. Jensen and William H. Meckling, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics (October 1976): 305-360.
4. The Employee Retirement Income Security Act of 1974 requires that external investment management compensation be based on fair market value of assets under management (as opposed to historical value).
5. James R. Webb, Richard J. Curcio, and Jack H. Rubens, "Diversifications Gains from Including Real Estate in Mixed-Asset Portfolios," Decision Sciences (Spring 1988): 434-452.
6. Paul B. Firstenburg, Stephen A. Ross, and Randall C. Zisler, "Real Estate: The Whole Story," Journal of Portfolio Management (Spring 1988): 22-34.
7. David Guilkey, Mike Miles, and Rebel Cole, "The Motivation for Institutional Real Estate Sales and Implications for Asset Class Returns," AREUEA Journal (5989): 70-86.
8. Joe R. Roberts and Eric Roberts, "The Myth About Appraisals," The Appraisal Journal (April 1991): 212-220.
9. Rebel Cole, David Guilkey, and Mike Miles, "Pension Fund Investment Managers' Unit Values Deserve Confidence," Real Estate Review (Spring 1987): 84-89; and Rebel Cole, David Guilkey, and Mike Miles, "Toward an Assessment of the Reliability of Commercial Appraisals," The Appraisal Journal (July 1986): 422-432.
10. The event of a CREF property sale ultimately negates the importance of an appraisal to market participants because the market price can then be used to replace the appraised value of the property in the calculation of unit price and overall CREF value.
11. In theory, the appraised value of a property is equal to the unobserved market value, plus random error. As a result of agency problems between parties, however, investors may believe that the appraised value of a commingled fund property is equal to the unobserved market value, plus non-random error, plus random error.
12. A recent article in the Wall Street Journal confirms that many pension fund investors believe that agency problems may result in positive non-random errors in asset managers' valuations of properties. See James A. White, "Pension Funds Take a Beating From Appraisals on Real Estate," Wall Street Journal (March 5, 1991): C1, C21.
13. Arguably, public knowledge of current and pending rent concessions may unduly limit the ability of asset managers to negotiate leases with other tenants.
14. Appraisal Institute, Code of Professional Ethics of the Appraisal Institute (Chicago: Appraisal Institute, 1990), A-6 and A-7.
15. The Appraisal Institute employs an MAI-designated appraiser in a full-time capacity as Director of Screening, and the director is responsible for enforcing the Appraisal Institute's strict code of ethics and standards of practice. The costs borne by members of the Appraisal Institute of setting and maintaining professional designations and standards and monitoring compliance serve as further examples of agency costs.
16. The maximum level of positive non-random error by CREF managers' appraisers that will not trigger withdrawals by investors is bounded by the point at which it can be detected by market participants. Thus, the optimal appraised value of a CREF property for asset managers who receive compensation based on valuations of all properties under management will be equal to unobserved market value, plus an undetectable but bounded level of positive systematic error, plus random error.
17. In a recent survey of CREF asset managers, major differences in valuations prepared for CREF asset managers are reported, and evidence of CREF manager influence on specific critical assumptions as well as on the appraisal process in general are documented. See Robert A. Gibson, "Asset Managers' View of the Appraisal of Real Estate Assets," The Appraisal Journal (January 1989): 65-78.
18. Mike Miles, "Commercial Appraisals for Institutional Clients," The Appraisal Journal (October 1984): 550-564.
19. As previously mentioned, a redemption request often requires several quarters to complete; thus, only a fraction of the withdrawal will be priced at the initial effective trading value. The remaining portions will be redeemed at the unit prices that are in effect at the time that they actually occur.
20. Willard McIntosh, "Forecasting Cash Flows: Evidence from the Financial Literature," The Appraisal Journal (April 1990): 221-229.
21. MarketSource, the recently introduced research quarterly published by the Research Department of the Appraisal Institute, represents an important step toward the development of quality trend analysis and survey data for real estate markets.
Stuart Fletcher, PhD, is Director of Operations for C.F. Parks & Co., Inc., an investment management firm. He received his PhD from Florida State University, and has previously contributed to professional and academic journals.
Barry A. Diskin, MAI, PhD, is professor in the Department of Risk Management/Insurance, Real Estate & Business Law at Florida State University, Tallahassee, Florida. He received his PhD from Georgia State University. Mr. Diskin has previously contributed to The Appraisal Journal and other professional and academic journals.
|Printer friendly Cite/link Email Feedback|
|Author:||Fletcher, Stuart; Diskin, Barry A.|
|Date:||Jan 1, 1994|
|Previous Article:||Appraisal assignments involving endangered species.|
|Next Article:||1993 survey of appraisal clients.|