`
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`__________________________
`
`MICROSTRATEGY, INC., )
`
` Petitioner, )
`
` vs. ) Case No. IPR2013-00034 (JL)
`
`ZILLOW, INC., ) Patent No. 7,970,674
`
` Patent Owner. )
`
`__________________________
`
` Deposition of JOHN A. KILPATRICK, Ph.D.,
`
` MAI, FRICS, taken at 1201 Third Avenue,
`
` Suite 4900, Seattle, Washington, commencing
`
` at 9:36 a.m., Thursday, August 8, 2013,
`
` before Donald W. McKay, RMR, CRR, CCR 3237.
`
`JOB No. 1712507
`
`PAGES 1 - 222
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 1
`
`1
`
`2
`
`3 4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 1
`
`MICROSTRATEGY 1022
`Microstrategy, Inc. v. Zillow, Inc.
`IPR2013-00034
`
`
`
`APPEARANCES OF COUNSEL:
`
`FOR THE PETITIONER: BY: LAUREN A. DEGNAN, ESQ.
`
` Fish & Richardson P.C.
`
` 1425 K Street, N.W.
`
` 11th Floor
`
` Washington, D.C. 20005
`
` 202.626.6392
`
` degnan@fr.com
`
`FOR THE PATENT BY: RAMSEY M. AL-SALAM, ESQ.
`
`OWNER: Perkins Coie LLP
`
` 1201 Third Avenue
`
` Suite 4900
`
` Seattle, Washington 98101
`
` 206.359.6385
`
` RAlSalam@perkinscoie.com
`
`ALSO PRESENT: BY: JORDAN TALGE, ESQ.
`
` Susman Godfrey LLP
`
` 1201 Third Avenue
`
` Suite 3800
`
` Seattle, Washington 98101
`
` 206.505.3842
`
` jtalge@susmangodfrey.com
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 2
`
`1
`
`2 3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 2
`
`
`
` I N D E X
`
`EXAMINATION BY PAGE
`
`MS. DEGNAN.......................................... 4
`
`MR. AL-SALAM........................................ 213
`
`MS. DEGNAN.......................................... 214
`
` E X H I B I T S
`
`NUMBER DESCRIPTION PAGE
`
`Exhibit 1001 U.S. Patent No. 7,970,674 B2, dated 61
`
` June 28, 2011
`
`Exhibit 1003 U.S. Patent No. 5,857,174, dated 163
`
` January 5, 1999
`
`Exhibit 1004 U.S. Patent No. US2005/0154657 A1, 155
`
` dated July 14, 2005
`
`Exhibit 1013 Declaration of John Kilpatrick 39
`
` Under 37 CFR 1.132
`
`Exhibit 1014 Uniform Standards of Professional 28
`
` Appraisal Practice and Advisory
`
` Opinions, 2005 Edition, effective
`
` January 1, 2005
`
`Exhibit 1018 The Appraisal Foundation, Appraisal 27
`
` Standards Board, USPAP Q&A
`
`Exhibit 1019 Standard on Automated Valuation 24
`
` Models (AVMs), Approved September
`
` 2003
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 3
`
`1
`
`2 3
`
`4
`
`5
`
`6 7 8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 3
`
`
`
` Seattle, Washington; Thursday, August 8, 2013
`
` 9:36 a.m.
`
` JOHN A. KILPATRICK, Ph.D., MAI, FRICS
`
` called as a witness in the
`
` above-entitled cause, being
`
` first duly sworn, testified
`
` as follows:
`
` EXAMINATION
`
`BY MS. DEGNAN:
`
` Q. Good morning.
`
` A. Good morning.
`
` Q. State your name for the record.
`
` A. I'm Dr. John Aaron Kilpatrick.
`
` Q. Can you give me your address, please.
`
` A. Home or business?
`
` Q. Whatever you prefer.
`
` A. Business is Suite 240, 2101 Fourth Avenue,
`
`Seattle, Washington, 98121.
`
` Q. Do you understand you've taken an oath to tell
`
`the truth and it's exactly the same as if you were in a
`
`court?
`
` A. Yes.
`
` Q. Is there any reason why you can't give full and
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 4
`
`1
`
`2
`
`3 4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 4
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`complete testimony today?
`
` A. I can't think of one.
`
` Q. Are you on any medication that might impair your
`
`ability to do so?
`
` A. I can't think of any.
`
` Q. I understand you've been deposed before. I'm
`
`sure you are somewhat familiar with the process. But
`
`just so we both have an understanding between us, if you
`
`don't understand my questions, please feel free to ask
`
`me for clarification.
`
` A. I'll do that.
`
` Q. The rules are actually such that if there is a
`
`question pending, I would prefer to get an answer before
`
`we take a break.
`
` A. Sure.
`
` Q. But I will try to take breaks on a regular
`
`basis. If you need one, just let me know.
`
` A. Thank you, very much.
`
` Q. Have you ever provided any expert opinion in a
`
`patent case other than this one?
`
` A. Yes.
`
` Q. How many?
`
` A. I can't recall. At least one in which I've been
`
`deposed a couple of times, but not testified in court
`
`yet. There may be others, but I can only think of one
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 5
`
` 5
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`offhand.
`
` Q. Who were the parties in that case?
`
` A. There were multiple parties on both sides. The
`
`case was captioned Move, M-O-V-E, versus Real, R-E-A-L.
`
` Q. For which party did you provide testimony?
`
` A. This is going to be embarrassing. I can't
`
`remember. It's been so long.
`
` Q. When was it?
`
` A. I was deposed once probably five or six years
`
`ago, and then once again probably about three years ago.
`
` Q. Do you remember where the case was pending?
`
` A. No. I mean, both my depositions were in
`
`New York City, but I don't know that it was pending in
`
`New York.
`
` Q. Do you recall whether you were testifying on
`
`behalf of the party that owned the patent or on behalf
`
`of someone accused of infringement?
`
` A. Again, it's been so long, I can't recall.
`
` Q. Generally, what was the general subject matter
`
`of your testimony?
`
` A. The real estate market. What ordinary
`
`participants in the real estate market would understand
`
`this patent to do. People of ordinary knowledge and
`
`experience, how they would interpret this patent, what
`
`it would do. The nature of the real estate market,
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 6
`
` 6
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`particularly the nature of the computerization of the
`
`real estate market -- and I use market in an overly
`
`broad sense -- over the course of the last say 15, 20
`
`years.
`
` Q. Are you named as an inventor on any patents,
`
`yourself?
`
` A. I don't believe so, no.
`
` Q. Have you ever worked as an appraiser?
`
` A. Yes.
`
` Q. More than once?
`
` A. More than once?
`
` Q. Yes.
`
` A. I have practiced appraisal off and on since
`
`1984, and I'm currently licensed in 49 of the 50 states,
`
`with Montana pending.
`
` Q. Would you say you've done hundreds of
`
`appraisals?
`
` A. Tens of thousands.
`
` Q. As part of your appraisal work, have you made an
`
`effort to learn about the particular properties you were
`
`appraising?
`
` A. Yes.
`
` Q. Would you say that's standard practice, to learn
`
`about a property that's being appraised?
`
` A. Yes.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 7
`
` 7
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` Q. Are there any standards that govern your line of
`
`work that say that one should make an effort to learn
`
`about the property being appraised?
`
` A. There are a host of different standards, some
`
`core, some supplemental, and different standards apply
`
`in different circumstances; but they all do have, to
`
`varying degrees, certain implications about
`
`familiarization in a general sense with the subject
`
`property. They're not really phrased that way.
`
` Q. Are there other predominant or key standards
`
`relating to conducting appraisals?
`
` A. It depends on who you are. If you are a
`
`real estate appraiser, practicing in the United States,
`
`then you are generally -- not always, but generally
`
`required to adhere to, among others, the Uniform
`
`Standards of Professional Appraisal Practice. Depending
`
`on the situation, again, if you are a real estate
`
`appraiser practicing in the United States, you might be
`
`required to adhere to any one of a number of
`
`supplemental standards.
`
` Now, there are also people who appraise
`
`real estate who are not licensed and certified
`
`appraisers. One might be a, quote, unquote, valuer
`
`practicing before the IRS, in which case the salient
`
`standards are those contained in IRS Reg 170-H as
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 8
`
` 8
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`opposed to the Uniform Standards.
`
` If you're an accountant appraising real estate
`
`before the IRS, then 170-H might apply; but USPAP, as we
`
`know it in the profession, might not.
`
` If you are a non-licensed tax assessor, then the
`
`standards come from the International Association of
`
`Assessing Officers, not USPAP, again.
`
` Then finally, you may be a non-licensee, a
`
`homeowner, valuing your own home. I do a lot of
`
`litigation in a lot of states. A homeowner may value
`
`his or her own home and may offer that up in testimony.
`
`Homeowners are considered, in many states, to be valid
`
`authorities on the value of their home, in which case,
`
`there are no standards which apply to that valuation
`
`exercise.
`
` Q. So this IRS Regulation 170(h) would apply to --
`
`just so I wrote it down correctly -- both a valuer for
`
`the IRS and an accountant for the IRS?
`
` A. It's 170-H. The IRS is, for a lot of reasons,
`
`very peculiar; but in this particular sense, they use
`
`the word valuer, or actually valuator, V-A-L-U-A-T-O-R.
`
`And 170-H would apply to anyone who is allowed to submit
`
`a value of real estate to the IRS.
`
` Now, there are certain other criteria which
`
`govern who might do that. Certainly a real estate
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 9
`
` 9
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`appraiser who is licensed under state law may do that;
`
`but also, under certain circumstances, an attorney might
`
`do that, or an accountant might do that.
`
`BY MS. DEGNAN:
`
` Q. For non-licensed tax assessors, you mentioned
`
`the International Association of --
`
` A. Assessing Officers.
`
` Q. Thank you.
`
` A. IAAO as it's commonly known.
`
` Q. Do real estate appraisers in the U.S. also
`
`comply with these standards put out by this IAAO?
`
` A. If one is doing work which is governed by those
`
`standards. I frequently do work in what's called the
`
`automated valuation model space, in which case I will
`
`not adhere to, but make reference to the IAAO standards
`
`as what we might call supplemental standards in the
`
`appraisal profession.
`
` If one is operating an automated valuation
`
`model, but one is not an appraisal licensee and one is
`
`not doing ad valorem tax work, then the IAAO standards
`
`are good guidance, but they're not required adherence.
`
` Q. Would you say that they're widely accepted as
`
`authoritative for that subject matter of automated
`
`valuation space?
`
` A. No. No. They are widely accepted when an
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 10
`
`
`10
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`automated valuation model is used in a property tax
`
`situation, but not -- I mean, we use automated valuation
`
`models for a whole host of things, and only one of the
`
`utilizations of an AVM might be -- I say might be for
`
`property tax purposes, because there are a lot of other
`
`ways you can do property tax valuation that would still
`
`fall under the rubric of those IAAO standards.
`
` Q. So are there other standards that are widely
`
`accepted to be followed when someone is doing an
`
`automated valuation?
`
` A. Not really. I mean, if you are a real estate
`
`appraiser, using any sort of methodology, be it a
`
`traditional appraisal methodology or an AVM or something
`
`else, then there are certain standards which apply as a
`
`matter of law and regulation. However, if you are
`
`someone else using an AVM -- and I would dare say most
`
`of the people who use AVMs are not licensed real estate
`
`appraisers -- then there aren't any regulatorily or
`
`widely accepted authoritative standards.
`
` Q. You mentioned AVM. Just so the record is clear,
`
`does that stand for automated valuation model?
`
` A. Yes.
`
` Q. So today, if we slip into AVM, we'll understand
`
`it means automated valuation model?
`
` A. Yes.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 11
`
`
`11
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` Q. So to the extent there are accepted standards at
`
`all for the use of AVMs, the IAAO standard is the one
`
`you're aware of?
`
` A. Well, there is also, for example, under the
`
`Uniform Standards -- which are called USPAP -- there is
`
`USPAP standard Rule 6. But both the IAAO standards and
`
`USPAP standard Rule 6 apply to a whole host of mass
`
`appraisal situations, only one of which might be an AVM,
`
`and then only under those circumstances in which an AVM
`
`is being used within the confines of what those
`
`standards envision.
`
` Q. So just so I have the list, we've got the USPAP
`
`Rule 6 standard and the IAAO standard that are used in
`
`connection with AVMs, although they also can be used as
`
`other things. Is there any other standard that comes to
`
`mind that is applicable to the use of AVMs?
`
` A. Insofar as you and I are using the word
`
`"standards" to mean the same thing -- because I -- as a
`
`real estate appraiser, I think of standards as something
`
`that's regulatorily established or established by
`
`convention by a professional organization. Certainly
`
`the International Valuation Standards, which are
`
`proffered by the International Valuation Standards
`
`Committee, pays homage to AVMs. There are some
`
`standards I can't recite or give you the exact citation
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 12
`
`
`12
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`within International Financial Reporting Standards,
`
`within the -- there are some allusions to AVMs within
`
`the Financial Accounting Standards Board's
`
`promulgations, allusions. I don't know if they actually
`
`specify AVM, but they certainly provide some guidance
`
`that one would want to pay attention to.
`
` I can't think of any others offhand, but that
`
`doesn't mean there aren't any. In other words, that
`
`list is neither exhaustive, nor wholly inclusive.
`
` Q. But those are the ones that come to mind right
`
`now?
`
` A. Off the top of my head.
`
` Q. Sure. So we got off on a little bit -- I was
`
`asking about your experience as an appraiser.
`
` A. Yes.
`
` Q. Tens of thousands. Impressive.
`
` A. Probably more than that, but go ahead.
`
` Q. I understand.
`
` We were talking about in that work, you would
`
`typically become familiar with the property that's being
`
`appraised.
`
` A. To varying degrees. Yes.
`
` Q. What sorts of things did you typically become
`
`knowledgeable about?
`
` A. Well, it really varies. I've used both
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 13
`
`
`13
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`individual appraisal models, as well as automated
`
`valuation models. When one is doing a more heuristic
`
`individual appraisal model, one might go out and tape
`
`measure, photograph the property, measure its condition,
`
`measure its square footage, number of bathrooms, number
`
`of bedrooms, those sorts of things. There are also
`
`geographic models which simply utilize the locational
`
`characteristics. In other words, they recognize that
`
`all houses in a given neighborhood tend to be about the
`
`same price. There are models which leverage off of tax
`
`assessment data. In other words, we go in and we test
`
`the tax assessment data against sales prices. If we
`
`find it to be effective and an efficient measure, then
`
`we simply utilize the time adjustment to the tax
`
`assessment within the confines of an automation
`
`valuation model. Goolsby, in 1997, writing in the
`
`Journal of Real Estate Research, discussed that sort of
`
`methodology.
`
` Case & Shiller use what's, in essence, a repeat
`
`sales model. I wrote about that in the Journal of
`
`Housing Research in 19 -- no, 2005, 2006, somewhere in
`
`that time frame, where we don't worry about the
`
`specifics of the house; we simply look at repeat sales,
`
`and we're dependent entirely on time measure. So I say
`
`that because depending on the valuation model one uses,
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 14
`
`
`14
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`the conditions may or may not be important. In some
`
`heuristic cross-sectional models, they're very
`
`important. In some longitudinal models, not important
`
`at all.
`
` Q. What is the difference between a heuristic model
`
`and a longitudinal model, just so the record is clear,
`
`for those of us who haven't been doing this for 20 years?
`
` A. Individualized appraisals tend to be fairly
`
`heuristic, even though they come with a very hardened
`
`set of standards and a body of knowledge which dictates
`
`methodology. And in the case of residential financing
`
`appraisals, they're usually done on a
`
`fill-out-the-blanks form, Fannie Mae and Freddie Mac
`
`proffer forms which one might use for this purpose. But
`
`that having been said, the adjustments, the size and
`
`scope of the adjustments, the choice of comparables, the
`
`weightings, all tend to be fairly heuristic; that is to
`
`say, made up by the appraiser him or herself. So, given
`
`that, those sorts of tools are best utilized by somebody
`
`trained and licensed as a real estate appraiser, so that
`
`one is familiar with the way one's peers in the industry
`
`would do that sort of work. That's actually how we
`
`specify it in USPAP.
`
` On the other hand, in, for example, automated
`
`valuation models, which are something quite different,
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 15
`
`
`15
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`we -- the model, itself, captures all of that in a
`
`statistically valid fashion. Therefore, an automated
`
`valuation model could be utilized by someone who is not
`
`trained and experienced or licensed, because the
`
`methodology is already captured in the model. It's not
`
`heuristic; it's statistically validated.
`
` Q. You mentioned the term "longitudinal." What did
`
`you mean by that?
`
` A. Time series. Like Case-Shiller or like my paper
`
`in the Journal of Housing Research where the value of a
`
`house today is simply a function of its value yesterday.
`
` Q. So in terms of things that you looked at
`
`typically, I think you mentioned location of the
`
`property, maybe things like the number of bedrooms, the
`
`number of baths, in your work as an appraiser. Is that
`
`the kind of thing that you also looked at when you used
`
`automated valuation models?
`
` A. Not necessarily. There are some AVMs that look
`
`at those things. There are many AVMs that do not look
`
`at those things. But even when an AVM looks at those
`
`things, it looks at them within the confines of a
`
`statistical validation. In other words, rather than the
`
`appraiser making judgment calls about those things or
`
`ranking or scoring those things, the computer, itself,
`
`does it, utilizing any one of a number of fairly
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 16
`
`
`16
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`time-honored algorithms, weighted or ordinary least
`
`squares being more of the two common ones.
`
` Q. In terms of just isolating one of these things,
`
`like location -- I mean, would it be the case that
`
`location is something that is always used in performing
`
`an appraisal?
`
` A. It depends on the kind of appraisal.
`
` Q. How would it depend? Let me ask it a different
`
`way. What kind of appraisal would location be
`
`irrelevant?
`
` A. Well, let's assume for a moment that I'm asked
`
`to appraise your house in a subdivision in which there
`
`are plenty of sales of comparable houses. In that
`
`context, the location of your house is irrelevant.
`
` Q. I mean, aren't you looking at only the
`
`subdivision? Isn't that a geographic location?
`
` A. Sure, but I'm not adjusting for the location.
`
`In other words, I'm not considering location, because
`
`location is implicitly and endogenously embedded in the
`
`heuristic model, itself.
`
` Q. But if you look at comparables only in the
`
`subdivision, aren't you necessarily including location
`
`as a factor in your analysis?
`
` A. Endogenously, but not exogenously. In other
`
`words, I'm not factoring or adjusting for location.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 17
`
`
`17
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`See, in appraisal speak, when one considers location, it
`
`means one has to exogenously think about and adjust for
`
`differences in location. But if I can go into a
`
`subdivision and only pick comparables from that same
`
`subdivision in which location is already endogenously
`
`accounted for, then in appraisal speak, I'm not
`
`considering the location, because I don't have to.
`
` Q. What does endogenously mean in your vernacular?
`
` A. It's already built in. It's part of -- a more
`
`technical definition of endogeneity would mean that it's
`
`of the model, itself. So the appraiser is not thinking
`
`about it, because it's already there.
`
` Q. So when you've performed appraisals, is it fair
`
`to say that you have always considered location either
`
`endogenously or exogenously?
`
` A. Hmm. Always considered location -- you know,
`
`I'm trying to reconcile what I know as a real estate
`
`appraiser to a question which I sense is asking me about
`
`something outside of the narrow box of real estate
`
`appraisal. So I hadn't really thought about that.
`
` Q. Would it be unusual to take no account
`
`whatsoever, either endogenously or exogenously, of
`
`location?
`
` A. Of course not. Because -- well, I talked about
`
`Case-Shiller a minute ago, and I talked about my own
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 18
`
`
`18
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`repeat sales paper. And I point out mine just to brag,
`
`because there are a whole lot of repeat sales papers out
`
`there in the scholarly literature. They don't ever take
`
`into account location, because a repeat sales model
`
`doesn't need to.
`
` Q. What is a repeat sales model?
`
` A. The value of your house as a function of the
`
`last couple of sales of the house.
`
` Q. Oh, of that particular house.
`
` A. Yes.
`
` Q. That sits in the same location the entire time?
`
` A. Well, again, I'm thinking in terms of how an
`
`appraiser would think about things. So the fact is we
`
`don't consider location, because it's not part of the
`
`model.
`
` Q. But the location remains constant in that sort
`
`of analysis?
`
` A. But because it's constant, it's not part of the
`
`analysis.
`
` Q. Right. But it does -- just in this particular
`
`question -- it does, in fact, remain constant the entire
`
`time during the analysis.
`
` A. Yes. That's an almost metaphysical question.
`
`It's just not something that an appraiser or somebody
`
`who is used to -- to use the vernacular of patents,
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 19
`
`
`19
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`which I'm becoming familiar with -- somebody of ordinary
`
`knowledge in the real estate appraisal profession -- not
`
`something that they would have thought about, therefore,
`
`it's not something that I've really ever thought about.
`
` Q. We were focused on the appraisals and the use of
`
`location. Do AVMs typically take into consideration the
`
`location of property?
`
` A. May or may not. A longitudinal AVM would not.
`
` Q. That's the value of the property over time?
`
` A. Right.
`
` Q. But such a valuation would -- I can't remember
`
`the word -- endonically (sic) or inherently take into
`
`consideration location, because it's the same property
`
`value over time. Right?
`
` A. Well, it's just not relevant. It's just not
`
`relevant. It's just not a relevant consideration.
`
`Therefore, appraisers don't think about things, don't
`
`take the time to worry about things that aren't relevant.
`
` Q. In AVMs that you have used in the past, have
`
`they ever taken into consideration location information?
`
` A. Some do; some don't.
`
` Q. Would you say there are important inputs into an
`
`AVM -- important attributes that a good AVM should
`
`consider?
`
` MR. AL-SALAM: Object to the form.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 20
`
`
`20
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` MS. DEGNAN: I think it's compound. Let me
`
`clear it up.
`
`BY MS. DEGNAN:
`
` Q. Do you have an opinion as to whether there are
`
`important attributes that a valid AVM should have?
`
` A. Well, I certainly have an opinion on that, yeah.
`
` Q. So what is your opinion?
`
` A. That there is no universal truth. In other
`
`words, there are a lot of different kinds of AVMs. So
`
`different AVMs may have -- depending on the AVM, itself,
`
`may have different input sets.
`
` Q. Say for a residential AVM, are there, in your
`
`opinion, important attributes that you could consider?
`
` A. There is no universal truth to that. In other
`
`words, there are a lot of different AVMs that measure
`
`residential values, and different AVMs attack the
`
`problem in different ways.
`
` Q. You said there is a lot of different AVMs. Are
`
`there different categories of AVMs?
`
` A. I don't know that there is a taxonomy for
`
`categorization.
`
` Q. So there is no bucket -- there is no big-picture
`
`buckets you can kind of classify an AVM into, this kind
`
`of AVM versus that kind of AVM?
`
` A. Right.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 21
`
`
`21
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` Q. So every AVM has to be named individually,
`
`because it doesn't fall into a class?
`
` A. Right. We tend to trademark them.
`
` Q. How would you define an AVM?
`
` A. Well, the name sort of says it all. It's an
`
`automated valuation model. It utilizes some statistical
`
`algorithm to produce value.
`
` Q. So any statistical algorithm that produces value
`
`can be considered an AVM?
`
` A. I would say so. Sure.
`
` Q. Does an AVM have to be computer generated?
`
` A. Hard to imagine not doing one, but there was a
`
`fellow named Chiang, C-H-I-A-N-G, who wrote a text, the
`
`Mathematics of Econometrics, back in 19, I want to say
`
`24; and he showed that if you tried to solve a matrix --
`
`he said in a footnote that if you try to solve the
`
`determinant of a matrix greater than five by five, you
`
`would, in his opinion, go insane. So I would suggest
`
`that attempting to do an AVM without a computer would be
`
`certainly mathematically possible, but perhaps taxing on
`
`one's soul.
`
` Q. So it's not required to be computer generated,
`
`but it would be helpful, in your opinion.
`
` A. Right. And the earliest works on which ordinary
`
`and weighted least squares depend were by the two
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 22
`
`
`22
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`mathematicians, Gauss and Markov, who were writing in
`
`the early and late 1800s, respectively. Gauss and
`
`Markov never conceived of the notion of a mechanical
`
`computer. But the AVMs that we have today are
`
`universally computer driven. I think that's axiomatic.
`
` Q. I understand you have some teaching experience
`
`in the field of appraisals and valuations. Is that
`
`correct?
`
` A. A little bit.
`
` Q. Would you say it's in the field of appraisals or
`
`the field of valuation?
`
` A. All the above. I mean, I've taught valuation to
`
`non-appraisers, and I've taught non-appraisal stuff to
`
`appraisers.
`
` Q. I think your C.V. indicates you've also done
`
`some teaching at some chapters of the International
`
`Association of Assessing Officials?
`
` A. Among other practitioner groups, yes.
`
` Q. How much teaching have you done with that
`
`organization?
`
` A. Not very much. Just a couple of lectures in
`
`front of a couple of their chapters.
`
` Q. And you're a member of that organization?
`
` A. I think so. I've got a woman at the office that
`
`keeps track of all that.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 23
`
`
`23
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` Q. Are there dues you have to pay?
`
` A. I'm sure she does. Then she sends a piece of
`
`paper over to a different woman in a different office
`
`who writes checks. They all do these things, and every
`
`year, they tell me what they did.
`
` Q. Your C.V. also indicates you're a nationally
`
`certified instructor of the Uniform Standards of
`
`Professional Appraisal Practice. That's the USPAP we
`
`talked about earlier?
`
` A. Yes.
`
` Q. That's another sort of teaching position you
`
`have?
`
` A. Yes. Sort of. It's a little more complicated
`
`than that. That's probably a simple way of putting it.
`
` Q. Was there more vigorous training to become a
`
`nationally certified instructor for USPAP than it was to
`
`be able to teach chapters of the IAAO?
`
` A. Yes. Quite a bit more rigorous.
`
` MS. DEGNAN: We talked earlier about some
`
`publications from these organizations or some standards
`
`at least. Just so we can maybe try to put some names to
`
`some standards that might be relevant --
`
` THE WITNESS: Sure.
`
` MS. DEGNAN: -- I have a couple of exhibits.
`
` (Exhibit 1019 premarked for identification.)
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 24
`
`
`24
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` MS. DEGNAN: I'm going to hand you what we have
`
`pre-marked as MicroStrategy Exhibit 1019.
`
` I don't have four of everything, but I have an
`
`extra of this one.
`
` MR. TALGE: Thank you.
`
` MR. AL-SALAM: This is the first exhibit in the
`
`case. Correct?
`
` MS. DEGNAN: I don't think so. Weren't there
`
`exhibits attached to the petition?
`
` MR. AL-SALAM: I see. I was just wondering
`
`where the 1019 came from. There doesn't have to be a
`
`reason.
`
` MS. DEGNAN: I think we can probably put in the
`
`list that says not used, not used, just as you would in
`
`district court, if it turns out there is lots that
`
`aren't used.
`
` MR. AL-SALAM: Okay.
`
`BY MS. DEGNAN:
`
` Q. Dr. Kilpatrick, have you ever seen this standard
`
`before or this document that is entitled, "Standard on
`
`Automated Valuation Models (AVMs)"?
`
` A. Yes.
`
` Q. Are you familiar with it?
`
` A. Yes. I will say that I'm not altogether sure
`
`that that is the most current edition of this document.
`
`Veritext National Deposition & Litigation Services
`866 299-5127
`
`Page 25
`
`
`25
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`This is dated about ten years ago. I think there may be
`
`an update on it. But I am familiar with this one, yes.
`
` Q. You're familiar with the one dated September
`
`2003?
`
` A. Yes.
`
`