Effects of asbestos on commercial real estate: a survey of MAI appraisers.

Considerable evidence exists that asbestos-containing materials (ACMs) in buildings do not pose a significant health hazard to building occupants.(1) Nonetheless, the health risks associated with exposure to asbestos fibers give rise to legal (e.g., regulatory and tort) risks for an operator or anyone with an ownership interest in a building. The legal risks that arise from potential health risks from exposure to asbestos have economic implications for owners and operators of property with asbestos.

Under Superfund legislation,(2) the classes of parties who may be declared "potentially responsible parties" include tenants and lenders as well as owners. As operators of the spaces they lease, tenants can become liable for asbestos-related damages. A number of court decisions in the 1980s have held that a lender qualifies as an owner or operator of a property that contains hazardous or toxic substances when, as the result of a foreclosure or a loan workout, the lender becomes the owner or operator of the property.(3) Insurers face potentially huge claims as a result of judgments awarded by the courts to parties found to be injured by exposure to asbestos. Because of the potential problems ACMs pose for lenders and insurers, property owners and tenants of buildings with ACMs may find it difficult to obtain loans and insurance coverage for asbestos-related costs.

In spite of the publicity asbestos has received, there is an apparent lack of consensus about how its presence affects the market value of commercial property. For example, a recent study by Koehn, MacAvoy, and De Silva(4) on the effects of asbestos on commercial property values in Los Angeles found that the presence of asbestos in buildings had no significant effect on values. On the other hand, a recent ruling by a New York trial judge reduced the assessment of a New York City office building because of the presence of asbestos.(5)

Estimating the impact of asbestos on property values is complicated by the fact that in the current overbuilt real estate markets there is no shortage of asbestos-free new buildings. Tenant rejection of older buildings with ACMs may be caused as much by the plentiful supply of relatively cheap new space as by the presence of asbestos.

Most of the published information about the effects of asbestos on the value and marketability of commercial property is in the form of case studies.(6) Little general information exists about the effects of asbestos on property values. One reason for the lack of generalized information is that the extent and types of ACMs present in buildings--and thus the costs of removing, controlling, or containing those ACMs--vary considerably from building to building. Further, the exposure to legal risks, the costs of dealing with asbestos, and the real estate market conditions tend to vary by location.

The purpose of the study presented in this article is to obtain from commercial real estate appraisers an assessment of how asbestos affects the value of commercial properties. Probing the collective wisdom of real estate appraisers can increase understanding about how the market value of properties is affected by the presence of asbestos. Selected members of the American Institute of Real Estate Appraisers (AIREA) were thus surveyed during June and July of 1990.

THE WAYS ASBESTOS AFFECTS PROPERTY VALUES

The value of a property with asbestos (|V.sub.A~) can be expressed generally in a present value (PV) framework as follows:

|V.sub.A~ = PV of expected net operating income (NOI) + PV of expected net sales price (NSP) - PV of expected containment and/or removal costs. (1)

To the extent the presence of asbestos reduces a property's expected net operating income (NOI) and expected net sale price (NSP), increases a property's discount rate, and causes a property owner to incur costs of containment or removal, the value of a property with ACMs is depressed below that of a comparable clean property (i.e., a property without ACMs). The loss in property value (LV) caused by the presence of asbestos can be expressed generally as:

LV = PV of property without asbestos - PV of property with asbestos (|V.sub.A~). (2)

Analyzing the adverse impact of asbestos on a property's NOI involves estimating how the presence of asbestos may affect revenue and expense items such as occupancy rates, rental rates, rental concessions, and operating expenses. The property's discount rate may be higher than that of a comparable property without asbestos if equity investors or mortgage lenders perceive the cash flows of the property to be subject to greater economic and legal risks. If ACMs remain in a building at the time of sale, the NSP may be lower because of the reduced NOI and the increased risk anticipated by prospective buyers. The NSP may also be adversely affected by the increase in the transaction costs of the sale, notably the need for environmental audits and surveys of various degrees of complexity.

The three terms on the right-hand side of equation 1 are interrelated insofar as decisions about how to deal with asbestos affect the PVNOI and PVNSP. Alternatives for dealing with asbestos thus must be considered before the impact of asbestos on the market value of the property can be estimated. The primary alternatives for dealing with asbestos are 1) immediate removal; 2) no removal until demolition of the building; and 3) postponement of removal until a time between immediate removal and demolition of the building.(7)

All alternatives other than immediate complete removal of asbestos require the development and implementation of an operations and maintenance (O&M) plan for the building. O&M procedures are intended to maintain the asbestos in a safe (i.e., nonfriable and nonairborne) condition as long as the building is operated while the asbestos is present.(8) Implementation of these procedures increases operating expenses. Additional costs may be incurred to encapsulate or enclose to safely contain the asbestos in lieu of removal, but the cost of either action is normally significantly less than the cost of removal.

Broadly speaking, efforts to remove ACMs from a building involve both direct costs and indirect costs. The direct costs of removal include all of the expenditures for removing and disposing of the asbestos. The direct costs, in brief, can be regarded as the costs associated with working on the asbestos itself. The indirect costs are the costs associated with the disruption of building operations and the dislocation of a building's occupants while removal takes place. These include the costs of moving and temporarily relocating tenants and the loss of rental revenue resulting from building "down time" while the asbestos is being removed.

If after asbestos is removed from a building the property's NOI and discount rate quickly adjust to levels appropriate for comparable buildings without asbestos, then the total cost of removing the asbestos immediately following acquisition of the property should be the ceiling of the loss in value caused by the presence of asbestos. In some situations, however, a building may continue to experience asbestos-related problems that linger past completion of total removal. Lingering costs can result from outstanding contracts that produce lags in the adjustment of rental income and operating expenses to levels appropriate for comparable buildings without asbestos, from overhanging legal liability, or from a taint a building continues to bear for once having contained asbestos. The effect of these lingering costs is to reduce the anticipated benefits from removal.

THE ASBESTOS SURVEY

The sample of appraisers to whom the survey was distributed consisted of 200 MAI-designated members of AIREA. To maximize the chance of reaching appraisers who had experience appraising properties containing asbestos, the sample was limited to MAI-designated appraisers who belonged to chapters whose memberships were exclusively located in metropolitan areas. Consequently, the sample of appraisers was drawn from five chapters representing AIREA's largest exclusively metropolitan chapters: New York, Philadelphia, Washington, D.C., San Diego, and Houston. Through a form of stratified sampling, 40 MAIs were randomly selected from each of these chapters. A total of 78 members returned questionnaires that were at least partially completed, for a response rate of 39%. Thirteen of the 78, however, were eliminated from the sample because the respondents indicated little or no experience with buildings containing ACMs or because the questionnaires were not sufficiently complete to process. The usable response rate was thus 33%.

The survey instrument

The survey questionnaire consisted of 24 questions containing a total of 115 response items. Table 1 through Table 5 contain the text of all of the questions in the survey as well as a statistical summary of the responses to each item. The questions can be divided into three broad sets. The first set (questions 1-9) contains questions designed to obtain qualitative assessments of how the market is likely to react to buildings with asbestos (questions 2-4, 6-9) and of how anticipated marketplace reactions are likely to affect the strategies pursued by owners of such buildings (questions 1 and 5). The second set of questions (questions 10-16) requests that appraisers provide quantitative estimates of adjustments for commercial property in general. The third set of questions (questions 17-24) asks for estimates of adjustments for different types of commercial property.

REPORT OF SURVEY RESULTS

The statistical summary next to each of the question items listed in Table 1 through Table 5 reports the number of responses to the item, the mean, the median, and the mode responses to the item, and the standard deviation of the response scores for questions 1 through 9 and 17 through 24. For questions 10 through 16, which required open-ended rather than structured responses, the modal response has been replaced by a column indicating the range of responses.

The multiple-item questions attempt to determine whether some factors are considered by the respondents to be more important than others. To analyze the responses, two levels of statistical tests were performed. The one-way analysis of variance (ANOVA) test was performed on questions 1, 4, 5, 6, 7, 9, and 17 through 24.

The null hypothesis,

|H.sub.0~: ||Mu~.sub.1~ = ||Mu~.sub.2~ = ... = ||Mu~.sub.K~.

is tested against

|H.sub.a~: the mean responses are not equal, where

K = the number of items in the question

Conditional on the null hypothesis, the appropriate test is an F-test with (K - 1) and (|n.sub.k~ - K) degrees of freedom, where |n.sub.k~ is the number of responses to the kth item in the question. Table 6 presents the ANOVA test results for the previously mentioned questions.

Pair-wise comparison of the means using Tukey's studentized range test was also performed on all item means in multiple-item questions to explore how the items could be grouped based on significantly different mean levels.

Owner strategies for dealing with asbestos

The mean response scores to question 1 listed in Table 1 indicate that the most significant motive for removal is that desired financing is difficult to obtain for properties with asbestos. The least important motivation is that ACMs are difficult to safely control. All of the motivations had a mean response score lower than 3, indicating that they are considered to be more than "somewhat important" motivations for removal.

The ANOVA F-statistic shown in Table TABULAR DATA OMITTED 6 indicates that the 11 problems are not all equally important. In accord with Tukey's studentized range test, the problems are arranged into the following three groups representing three statistically distinct levels of motivations:

Most Important

Group I

* Desired financing difficult to get

* Potential lawsuits

* Property difficult to sell

Group II

* Bad publicity and bad image

* Potential costly regulations

* Increase in cost of financing

* Tenants react adversely

Least Important

Group III 

* Insurance coverage difficult to get

* Health threat to occupants

* Increase in cost of insurance

* ACMs difficult to safely control

The mean response scores to question TABULAR DATA OMITTED TABULAR DATA OMITTED 5 in Table 1 show that the respondents rated the development of a plan for removal in stages as the strategy most frequently pursued by building owners. Containment of ACMs by encapsulation or enclosure is the second most frequently pursued strategy.

The ANOVA result indicates that the six strategies are not all equally likely to be undertaken by new property owners. Tukey's studentized range test allows the separation of four of the six items into the following three statistically distinct groups:

Most Frequent

Group I

* Develop plan to remove asbestos in stages

* Encapsulation or enclosure of ACMs

Group II

* No removal of ACMs until demolition

Least Frequent

Group III

* No removal of ACMs until just prior to sale of property

Market responses to buildings with asbestos

Table 2 summarizes the responses to questions 2 through 4 and 6 through 8. The results for question 2 imply that potential purchasers do not rule out buying buildings with asbestos.(9) Question 3 indicates that prospective buyers may or may not require the seller to pay for both TABULAR DATA OMITTED preliminary and comprehensive environmental audits.(10)

Questions 4 and 6 concern whether adverse consequences are expected to linger after ACMs have been completely removed. Question 4 indirectly addresses this issue by asking how the cost of removal is reflected in the negotiations of the sale. If buyers generally accept a price discount approximately equal to the estimated total cost of removal, the implication is that investors are confident that removal of the asbestos will eliminate all associated problems. On the other hand, if buyers demand a price discount greater than the total cost removal, the implication is that they expect adverse consequences to linger after removal is completed. Responses to question 4, however, do not indicate that the latter expectation prevails. The mean response scores for the three buyer demands are similar. The ANOVA F-statistic implies that these demands are equally likely. These results are supported by the responses to question 6, which indicate that appraisers do not TABULAR DATA OMITTED expect the adverse consequences listed to occur frequently.

Responses to question 7 indicate that indemnification guarantees are the most frequent of the listed loan requirements imposed by lenders on property with ACMs. The ANOVA F-statistic implies that the four loan requirements are not equally likely. The Tukey's studentized range test allows the arrangement of the four items into two distinct groups.

Most Frequent

Group I

* Requirement of indemnification guarantee

Least Frequent

Group II

* Lower loan-to-value ratio 

* Higher "points" or origination fees

* Higher interest rate

Lenders may require indemnification guarantees more frequently than higher loan costs or lower loan amounts because they are unable to accurately price the risks posed to them by asbestos.(11) Responses TABULAR DATA OMITTED to question 8 indicate that refusal to lend is fairly common, confirming that many lenders tend to be cautious about lending on buildings with ACMs.(12)

Appraiser adjustments to properties with ACMs

Table 3 summarizes the responses to question 9. The responses indicate that the adjustments most frequently made by appraisers are to reduce the property value by the combined direct and indirect costs of asbestos removal; that adjustments are more likely to be made to the operating expenses of the property than to the rental income; that the most frequent of the operating expense adjustments is to increase repair and maintenance expenses (when O&M expenses are likely to be included); and that adjustments are made to the cap rate and the discount rate with almost equal frequency.

The Tukey's studentized range test allows the arrangement of the adjustments into the following three groups:

Most Frequent

Group I

* Reduction of property value by indirect costs

* Reduction of property value by direct costs

* Increase in maintenance and repair expenses

Group II

* Increase in insurance costs

* Increase in cap rate

* Increase in discount rate

* Increase in workletter/buildout allowances

* Increase in legal expenses

Least Frequent

Group III

* Increase in the amount of rental concessions to tenants

* Decrease in rental rate

* Increase in vacancy rate

Magnitudes of selected asbestos effects

Responses to questions 10 through 16 are presented in Table 4. The mean and median estimates to the questions have the following implications.

* The discount in the market value of properties with ACMs typically ranges from about one-tenth to about one-third of the value of comparable properties without asbestos.

* Properties with ACMs take, on average, 8.2 months longer to sell than comparable properties without asbestos,(13) and rental properties with ACMs take, on average, 4.88 months longer to obtain tenants to fill vacancies than properties without asbestos.(14)

* The average increase in the discount rate for properties with ACMs (i.e., the size of the additional risk premium) is 1.99%.(15)

* The average range in the lump sum cost for an environmental audit to determine whether ACMs are present is $4,091 to $13,432 (the median range is $2,250 to $5,500), and the average range in the lump sum cost for a comprehensive engineering study to determine the extent as well as the specific characteristics of the asbestos present in a building is $5,638 to $20,844 (the median range is $3,750 to $10,000).

* The average increase in the effective interest rate for mortgage loans made on properties with ACMs is 73 basis points (.73%) and the median increase is 50 basis points.

Income and value effects of asbestos on different property types

Table 5 contains summaries of the responses to questions 17 through 24.(16) With the exception of question 23, a consistent pattern of responses is apparent. This pattern is that commercial properties (e.g., office buildings and shopping centers) are generally perceived to be more adversely affected by ACMs than industrial properties (e.g., warehouses and factories). Further, among the commercial property types office buildings are perceived to be more adversely affected by ACMs than retail properties. Heavy industrial buildings are generally perceived to be the most severely affected of the industrial property categories.(17) Responses to question 23, the exception, indicate that the value reduction for heavy industrial buildings as a result of the direct costs of removal may be slightly higher than for retail space. This deviation may be explained by the commonly accepted notion that heavy industrial buildings contain a large amount of asbestos.

The mean response scores for questions 17 through 24 imply the following approximate percentage adjustments to properties with ACMs.

* An average decrease in rental rates of 11% to 12% for office buildings, 8% to 10% for retail buildings, and 6% to 7% for industrial properties

* An average increase in rental concessions of 6% to 7% for office buildings, and of less than 6% for both retail and industrial properties

* An average increase in operating expenses of 7% to 8% for office properties, 5% to 7% for retail properties, and 3% to 4% for industrial properties

* An average increase in vacancies of 9% to 10% for office buildings, 8% for shopping centers, and 6% to 7% for industrial buildings

* An average decrease in NOI of 10% to 11% for office buildings, 7% to 8% for retail properties, and 6% to 7% for industrial buildings

* An average reduction in the loan-to-value ratio of 9% to 10% for office buildings, 8% to 9% for retail buildings, and 6% to 7% for industrial buildings(18)

* An average reduction in value caused by direct removal costs of 17% to 19% for office buildings, 15% to 16% for retail buildings, and 13% to 15% for industrial buildings

* An average reduction in value caused by indirect removal costs of 11% to 13% for office buildings, 11% to 12% for retail buildings, and 8% to 9% for industrial buildings

* An average loss in value caused by the total cost of removal (i.e., combined direct and indirect costs) of 28% to 32% for office buildings, 26% to 28% for retail buildings, and 21% to 24% for industrial buildings

Apparent congruence can be observed between the results summarized above and the estimates of value reduction provided in question 10. The upper end of the range of the estimates of value reduction provided in question 10 closely corresponds to the total cost of removal for high-rise office buildings indicated. On the other hand, the lower end of the range of estimates may represent buildings in which the value loss associated with the presence of asbestos is less than the total cost to remove it.

CONCLUSION

Asbestos in buildings clearly remains a problem for many building owners, and thus it must be dealt with in appraisal. The survey of MAI appraisers reported in this article is intended to contribute toward understanding of the costs imposed on real property by environmental problems such as asbestos, and also about how the market and the institutional players in the market are likely to respond to contaminated properties. This questionnaire also can be easily modified to deal with most types of hazardous and toxic valuation problems currently confronted by appraisers.

Obviously, informed opinions (i.e., best judgments) are not the same as real knowledge. In addition, the respondents to this survey may not have dealt with the asbestos problem in sufficient numbers and depth to be able to answer many of the questions knowledgeably. Notwithstanding the limits of the survey, disseminating the informed opinions of practicing real estate appraisers can contribute to a better understanding of how asbestos affects the value and marketability of properties with ACMs.

1. Harvard University, Summary of Symposium on Health Aspects of Exposure to Asbestos in Buildings (Cambridge, Mass.: Energy and Environmental Policy Center, John F. Kennedy School of Government, Harvard University, 1989). See also B. T. Mossman, J. Bignon, M. Corn, A. Seaton, and J. B. L. Gee, "Asbestos: Scientific Developments and Implications for Public Policy," Science, v. 247 (January 19, 1990): 294-301; letters in discussion of this article, "Asbestos, Carcinogenicity, and Public Policy," Science, v. 248 (May 18, 1990): 795-802; and "Asbestos and Carcinogenicity," Science, v. 249 (August 24, 1990): 844.

2. The original Superfund legislation, called the Comprehensive Environmental Response, Compensation and Liability Act of 1980 (CERCLA), Public Law 96-510, was amended by the Superfund Amendments and Reauthorization Act of 1986 (SARA), Public Law 99-499. The statute, as amended, is located at 42 U.S.C. Section 9601 et seq.

3. Cases that establish how a lender can face financial liability as an "operator" of property if it acts to protect its security interest in the property include United States v. Mirable, No. 84-2280 (E.D. {a/. Sept/4. 1985), 15 Envtl. L. Rep. (Envtl. L. Inst.) 20994 (1985), In re T. P. Long Chemical Co., Inc., 45 B.R. 278 (Bkrtcy. 1985), U.S. v. Maryland Bank and Trust Co., 632 F. Supp 573, Guidice v. BFG Electroplating, 30 Env't Rep. Cas. 1665 (BNA), and U.S. v. Fleet Factors Corp., CA 11, No. 89-8094, May 23, 1990. However, the economic interests of lenders can be affected even if they do not become owners or operators insofar as the solvency of the borrower and the value of the collateral property are adversely affected by the ACMs. See Johnine J. Brown, "Superfunds and Superliens: Super Problems for Secured Lenders," BNA's Banking Report, v. 50, no. 14 (April 4, 1988); Jay Gwin and Judon Fambrough, The Effect of Environmental Law in Real Estate Transactions, a technical report of The Real Estate Center (College Station, Texas: Texas A&M University, January 1990); and Walter D. James, III, "Financial Institutions and Hazardous Waste Litigation: Limiting the Exposure to Superfund Liability," Natural Resources Journal (Spring 1988): 329-355.

4. Michael F. Koehn, Paul W. MacAvoy, and Harindra De Silva, "Market Responses to Asbestos in Buildings," Journal of Real Estate Finance and Economics (September 1990): 213-232.

5. Neil Barsky, "Assessment Cut on Office Building Due to Asbestos," The Wall Street Journal (January 23, 1991): B2.

6. Douglas S. Bible, Marshall F. Graham, and Michael T. Newman, "Travis Square: A Study of ACM Removal in a Class A Office Building," Environmental Watch (Chicago: Appraisal Institute, Winter 1991); Will McIntosh, "Study of the Southern State Office Building: The Effect of ACMs on Rents and Marketability in a Healthy Market," Environmental Watch (Chicago: Appraisal Institute, Winter 1991); Wade R. Ragas and R. Dunbar Argote, "Valuation of Office Buildings with ACMs Using DCF Analysis," Environmental Watch (Chicago: Appraisal Institute, Winter 1991); and Albert Wilson, "Probable Financial Effect of Asbestos Removal on Real Estate," The Appraisal Journal (July 1989): 378-391. Two major exceptions are the recent article by Koehn, MacAvoy, and De Silva and the economic analysis of the effects of asbestos in buildings by Donald N. Dewees, Controlling Asbestos in Buildings: An Economic Investigation (Washington, DC: Resources for the Future, 1986).

7. In Jeffrey D. Fisher, George H. Lentz, and K. S. Maurice Tse, "Valuation of the Effects of Asbestos on Commercial Real Estate," The Journal of Real Estate Research (Summer 1992): 331-350, the authors show that an optimal time may exist for removing the asbestos such that the loss in value resulting from ACMs is minimized.

8. Operations and maintenance activities include recording the location and condition of any ACMs in the building, cleaning up any asbestos that may have been released from the ACMs, periodically inspecting and following other procedures for monitoring the condition of the ACMs to determine whether repairs or other corrective action may be warranted, developing procedures for responding to asbestos emergencies, and informing building employees about the location of the ACMs and instructing them about safe methods of working in the vicinity of ACMs.

9. Comments to this question indicated that a prospective purchaser's reaction depends on who the purchaser is. Institutional investors and government were identified as being the most reluctant to purchase such buildings.

10. Comments to this question indicated that practice varies among buyers and sellers. A few respondents commented that some sellers routinely provide a preliminary type of survey for older buildings (the class of buildings most likely to contain asbestos). Several other responses, however, indicated that it is common for a potential buyer to pay for the preliminary audit. One possible explanation of why buyers pay for the preliminary audit is that a typical buyer may want to be able to show, in accordance with the "innocent party" defense provided in CERCLA, that it exercised due diligence in attempting to detect the presence of ACMs.

11. Comments indicated that other loan costs (not listed in the question) that may be imposed by lenders are required environmental surveys and periodic air sampling. The requirement that surveys, air sampling, and so on be completed before loan origination extends the loan processing time for buildings with ACMs.

12. Comments indicated that lender actions depend on the type of lender making the loan as well as on the amount and condition of the ACMs in the building. It was noted that some institutions (individual insurance companies were specifically cited as examples) have a policy of refusing to lend on properties with ACMs. Yet the general feeling of the respondents was that many lenders will make a loan if the ACMs are adequately controlled.

13. Some comments indicated that at least a portion of the additional time to sell is caused by the need for environmental surveys and engineering studies, providing further evidence that the transaction costs imposed by the need for asbestos (environmental) audits/surveys/studies include delay costs as well as actual expenditures.

14. Some comments indicated that some of the additional time to re-lease might be attributable to the down time needed to remove asbestos from the leased space.

15. An impression obtained from comments is that some respondents felt that adjustments to the discount rate to reflect the presence of asbestos are difficult to support.

16. The results summarized in Table 5 for questions 17 through 21 must be interpreted with caution. In particular, the average percentage change in the individual components of NOI taken together, as indicated by the mean response scores to questions 17 through 20 do not reconcile with the percentage change indicated in question 21, the overall percentage change in NOI. The reason they do not reconcile is that the percentage calculations are applied to different bases. Questions 17 through 20 are nonetheless useful because they contain information about the effects of asbestos on the individual components of NOI.

17. There are several possible explanations of why commercial property types might be more adversely affected by asbestos than industrial property types. One is that commercial properties are more likely than industrial properties to cater to the public. A second reason is that industrial properties are perhaps more likely to be owner-occupied properties, and thus experience less pressure from the marketplace. A third possibility is that industrial properties are more likely to be special-purpose properties used by clientele not as prone to be alarmed by asbestos as clientele of commercial buildings.

18. The less adverse lender reactions to industrial properties than to the other property types may be because loans on industrial properties are based on the credit of the company occupying the space rather than on the value of the property. Moreover, because industrial properties are more likely to be owner occupied, the firm occupying the space may try harder to make a loan work out because of the value "in use" of that space to the company.

Jeffrey D. Fisher, PhD, is associate professor of finance and director of the Center of Real Estate Studies in the School of Business at Indiana University, Bloomington, Indiana. He obtained a BSIM from Purdue University, an MBA from Wright State University, and a PhD from Ohio State University. Mr. Fisher has published extensively.

George H. Lentz, PhD, is an associate professor of finance and real estate at California State Polytechnic University in Pomma, California. Mr. Lentz obtained a BA from the University of Santa Clara, an MBA from Indiana University, and a PhD from the University of Arizona, and has written several real estate-related articles.

K. S. Maurice Tse, PhD, is assistant professor of finance, School of Business, Indiana University. Mr. Tse has authored several articles, and received a BSc from Hong Kong University as well as an MS, an MBA, and a PhD from Michigan State University.