A Macro-Economic Model Providing Patent Valuation and Patent Based Company Financial Indicators[1]


I.              Introduction


                A.            Summary


                I present a new and useful macro economic model for valuing patents.  The model provides equations predicting the market share for products and services covered by a patent.  One benefit of the macro economic model is that it enables inexpensive automated determination of the value of any patent.  One drawback of the macro economic model is its reliance on a nominal determination of market share.  However, the reliance upon a nominal determination of market share obviates data defining the actual market and actual sales covered by the patent, and that twist is what makes the model feasible to implement! 

                Applying income valuation theory to the annual net earnings provided by the patent for the remaining term of the patent results in a valuation for the patent.  Extensions of the model to the time dependence of company patent portfolios provides patent based company financial predictors.

                I have implemented the macro economic model in a programmed computer system, and I have used that system to automatically generate valuations for all enforceable non-expired U.S. patents.  Valuations resulting from the model should  available by the time this article is in print at www.patentvaluepredictor.com.   Patent values obtained in January 2001 (for three patents issued in each of the last 10 years) appear in the following table.












Invalid's bathtub

Joycelyn Forbes





Ventilated welding shield

Arnold E. Crowson





Welding protected coveralls

Barry Worton





Closed cell foam ground pad and methods for making same

James M. Lea





Garment waistband construction

Crown Textile Company





Water filled cushion

Peter A. Smith





Beach towel and pillow removably contained within carrying bag

Cambridge Products





Clathrate composition including essential oils and method of using same

Kurita Water Industries Ltd.





Ultrasonic measuring apparatus using a high-damping probe

Shimizu Construction Co., Ltd.





Magnification changing lens

Canon Kabushiki Kaisha





Wringing device in particular for fringed strips for cleaning floors

VDM S.r.l.





Cover for hospital bed rails

John J. Marra, Jr.





Sculptured, stretchable waterbed mattress with aesthetic appearance

Advanced Sleep Products





Portable hygenic apparatus

Les Placements Jean-Claude Lemyre Inc.





Bath and shower device with a bath tub and a related shower partition

Ideal-Standard GmbH





Shower soap system

Gerald W. Berry





Mobile surface cleaning machine

Advance Machine Company





Bottom structure of a bed

Paramount Bed Company Limited





Hospital bed with collapsing wing

Hill-Rom Company, Inc.





Foot egress chair bed

Hill-Rom Company, Inc.





Revolving suntan bed

Richard W. Cooper, Jr.





Face mask and face mask cover

Daniel S. Harris





Electrophysiology table

General Electric Company





Gel filled deformable cushion and composition contained therein

Joel L. Sereboff





Stretcher frame clamp

Kevin R. O'Connell





Occipital retention strap for cyclist headgear

9001-6262 Quebec Inc.





Toilet gas suction vent

Pavel Baldea





Stay-dry toilet seat

Mindy Machanic





Crib with infant hammock

Wade Swift, Jr.






Kimberley D. Field





Inflatable beach bed

Marinus Anthonius Maria Van Tol





Head and neck support for racing

Robert P. Hubbard





Inline sanitary conditioning system

Waterbury Companies, Inc.





Opening-closing device of western style toilet seat and seat cover

Katoh Electrical Machinery Co., Ltd.


                The model predicts much larger values for some patents.  The largest patent value determined using the model is $6.2 billion.

                In sequence below, I discuss the importance of patent valuation, define valuation, discuss conventional valuation of patents, define the macro economic model, discuss the reasonableness of the axioms defining the macro economic model, extend the model to provide patent based company financial predictors, and then discuss methods for validating the results of the model.


                B.            The Significance of Patent Valuation


                Valuing patents is important for many purposes including determining business values, capital allocations, taxes, licensing rates, and patent infringement damages.  There is a growing interest in valuing patents because our economy is shifting from a tangible assets based economy to an intangible assets based economy.[2]  The business world has recognized that the intangible assets of many companies exceed the value of their tangible assets,[3] and that patents are part of these intangible assets.[4]


                C.            Conventional Patent Valuation


                Valuation is an accounting term which means a lump sum of money payable to receive the future benefits of an asset at a particular time.[5]  There are three generally accepted accounting theories for valuing assets: market, cost, and income.  Market theory values an asset as the present value ascribed to similar assets in an active public market.  Cost theory values an asset by the cost of replacing the asset.  Income theory values an asset by the present worth of the net anticipated economic benefit of the asset.

                Market theory valuation of patents has little or no utility because no two patents are similar enough for the sales price of one to define the value of another.  Moreover, until recently, there was no public market.[6]  Cost theory is generally inapplicable since a patent cannot be replaced. 

                Conventional methods for using income theory to value a patent analyze micro economic data to determine the anticipated economic benefit of owning the patent.  This micro economic data includes market data indicating the gross sales and net income derived from the sale of products covered by the patent, and the revenue derived from licensing the patent.

                Applying income theory to micro economic data to value a patent is labor intensive, costly, and complex.  This method requires an analysis to determine the meaning of the claims of the patent, a comparison of products to the claims of the patent to determine what products are actually covered by the patent, a determination of the size of the market covered by the patent, and a determination of the cost advantage of the patented technology compared to alternative technologies for that market.  A micro economic analysis may be necessary to prove damages in patent infringement litigation.  However, a micro economic analysis of a patent is often cost prohibitive for purposes of business valuation, capital allocation, taxes, and licensing.

                Moreover, the data necessary for members of the public to perform micro economic analysis of patents is simply not available.  This is because that data includes relationships between patents, product lines, product line specific costs and earnings information, and licensing royalty rates and terms, and companies rarely release that type of  information to the public.  Thus, micro economic analysis of patents is often not feasible.  The model presented below fulfills the need for an economic analysis of patents.


II.            The Macro Economic Model For Valuing Patents


                A first axiom of the model is that each enforceable patent covers a fraction of Gross Domestic Product (GDP) of the country in which the patent is enforceable.  I name this fraction  CFj where j represents the jth enforceable patent.

                The sum of the fractions of GDP covered by all of the enforceable patents equals a fraction, K, of GDP.  The first axiom provides the equation:


                                                                                K * GDP  =  G CFj                                                                (1)


where G represents the sum over j = 1 to n, and where n is the number of enforceable patents.

                A second axiom is that the CFj is a function of certain formal characteristics of the patent such that the function correlates to the strength and breadth of the claims of the patent.  These formal characteristics include, for example, measures of the length of the independent claims, the statutory classes of the independent claims, the number of independent claims, the number of claims, the length of the specification, the number of figures, the number of examples, the number of embodiments, the number of references cited, etc.  I name the variables which are the measures of these formal characteristics, v1, v2, v3, ...,  I name this function the RPN (relative patent number) function.  The second axiom provides the equation:


                                                                                CFj =  C * RPNj(v1, v2, v3...)                              (2)


where C is a proportionality constant.

                 Substituting (2) into (1) yields the equation:

                                                                K * GDP  =  C * G RPNj(v1, v2, v3...)                                (3)


                I measure the values of v1, v2, v3 ... for the jth patent and evaluate RPNj(v1, v2, v3...) for the jth patent to obtain a value RPNj for RPNj(v1, v2, v3...) so that:


                                                                                K * GDP  =  C * G RPNj                                      (4)


                I know the values for GDP and G RPNj.  I select K to be unity[7] and solve for C.   Substituting the value for C and the value RPNj for RPNj(v1, v2, v3...) into (2) solves for CFj.  I now have values for all of the CFjs.

                A third axiom of the model is that there is a profit margin, Mj, associated with the sale of goods and services covered by the jth patent.

                Combining the third axiom with (2) provides:


                                                                                Pj = Mj * CFj =  Mj * C * RPNj                          (5)


where Pj is the annual profit provided by the sale of goods and services covered by the jth patent.

                A fourth axiom of the model is that the jth enforceable patent is the asset that provides the profit Pj.   A result of the fourth axiom is that the value of the patent can be determined using conventional income theory by calculating the present value of Pj annual income attributable to the patent over the enforceable life of the patent. 

                The generic formula for valuing a stream of income using conventional income theory is:


                                                                                V = G I/(1+R)k                                                        (6)


where G represents a summation over the k = 1 to N time periods in which the income I is received, and where R represents the internal rate of return (which is the discounted present value of the right to future income).  Substituting (5) into (6) for the jth patent provides the current value Vj  of the jth patent:


Vj = G Pj/(1 + R)k                                                   (7)


where G represents a summation over the k = 1 to N years that the patent will be enforceable.[8] 

                I select R to be 0.14 (14 percent), since this is the often quoted historical rate of return on common stocks, and because I perceive a primary utility of this model to be for decisions on allocation of capital.[9]  I select Mj for all j to equal MA, where MA is the average corporate profit margin across the entire economy.[10]  I evaluate Vj using these selections in (6) to obtain a value for each currently enforceable patent.[11]


III.           Reasonableness of the Axioms and Selected Values


                I consider an axiom to be reasonable if the axiom tends to result in patent valuations correlating to valuations of the patents that would be obtained using income theory analysis of micro economic data.

                Selecting the fraction of GDP covered by all of the enforceable patents to be unity is arbitrary.  I believe (based upon anecdotal evidence only) that most GDP is covered by patent rights.  I know of no economic data defining the fraction of GDP covered by enforceable patent rights.

                Axiom 2 assumes a correlation between the gross sales of products covered by patents and the strength and breadth of the patents.  This is the fundamental assumption of the macro economic model since it provides that the CF equals the gross sales of products covered by the patent.  If the CFs of patents correlate to the actual gross sales of products covered by the patents, then axiom 2 is reasonable.

                The value selected for the R has a significant effect on the value the model determines for patents (1) because patents are enforceable for multiple years and (2) because of the compounding factor (1+R)k in the multi year income theory analysis.  However, that compounding factor effect on valuation is present in any income theory based valuation. 

                Axiom 4 explicitly excludes the contribution to profits due to the assets involved in manufacturing, sales, other IP, and contracts associated with the sales of products covered by the patent.  Those other assets may dilute the patent's contribution to profit.  However, it may be that those other assets, absent the existence of the patent, would only result in a competition-driven low or negligible profit margin, and that all substantial profit margin is due to the exclusivity provided by the existence of the patent.  This axiom reflects the well known difference in profit margins between a monopolized market and a free competition market.  Hence, this axiom is not unreasonable.

                It is useful to compare the results obtained using the macro economic model to results obtained using micro economic data, for a limiting case.  Consider the limiting case evaluating the sum of all enforceable patents.  A micro economic analysis would determine that a fraction of GDP was covered by enforceable patent rights.  I assume for this example that that fraction is unity, so that the micro economic analysis shows that the fraction of GDP covered by enforceable patent rights is unity.  In the macro economic model, summing up the fractions of GDP covered by each enforceable patent also results in the GDP.  Hence, when the sum of all enforceable patents are considered, the macro economic model and the micro economic model will provide the same results.  Therefore, the correlation between the predictions of the macro economic model and a micro economic analysis will converge as the number of patents under consideration increases.  Hence, the model should be more accurate when applied to portfolios of patents, such as those owned by artificial legal entities, which is the subject of the next section.


IV.           Extensions of the Macro Economic Model to Company Patent Portfolios: Patent   Based Company Financial Indicators and Their Time Dependence


                Aperiodically, patents owned by a company expire, and new patents owned by the company issue.  Each patent is enforceable for a discrete amount of time.  At least for those reasons, the value of a company's patent portfolio, VT, changes with time.  Using the valuations obtained by this model and the term of enforceability of each patent, I can calculate the time dependence of VT.

                For the reasons explained below, the sum of CFj (equation 2) for all patents owned by a company, CFT, should correlate to a company's gross sales, and the sum of  Pj (equation 5) for all patents owned by a company, PT, should correlate to the company's net earnings.  In the macro economic model, the sum of CFT for a company provides a value which is the nominal fraction of GDP covered by all of the company's sales.  The company's gross sales define the actual fraction of GDP produced by the company.  I have normalized the sum of CF for all patents to GDP and the sum of goods and services produced by all companies is in fact GDP.[12]  Therefore, the  average over all companies, of the difference between CFT and the actual fraction of GDP produced by a company, is zero.  This limitation statistically constrains CFT to correlate to the actual of sales of a company.  As discussed above in the limiting example at the end of the last section, the CFT tends to converge on the actual gross sales of the company when the number of patents evaluated increases.  Moreover, a company's patents are more likely to cover the actual goods and services produced by that company than another company, since (1) the company cannot rightfully produce goods and services covered by patents of another entity, (2) no other entity has the right to produce goods and services covered by the company's patents, and (3) entities primarily attempt to get patents covering their products in order to protect their markets.   For all of these reasons, CFT should approximate a company's gross sales.  Based upon the same reasoning, PT should approximate a company's actual net earnings.

                CFT can be viewed as what a company's sales should be if the company's patents cover the company's products and services and no other products and services.  PT can be viewed as what a company's net earnings should be if the company's patents cover the company's products and services and no other products and services.  The corollaries to these conclusions are that corporate gross sales exceeding CFT indicate that the company has weak patent protection in the sense that its patents do not cover all of its products and services and that other entities' patents may cover some of the company's products and services.  Conversely, CFT exceeding corporate gross sales indicates that the company has strong patent protection in the sense that its patents cover its own products and services and may cover other entities products and services.  A company whose patents extend to markets other than the markets in which the company is currently active can be expected to expand into those markets.  A company whose patents do not cover its existing markets can be expected to contract in those markets in response to the exclusive  rights of others.  In addition, weak patent protection indicates low profit margins and strong patent protection indicates high profit margins.  Hence, the ratio of CFT to the company's gross sales is an indicator of whether the company's sales and profit margins should be increasing or decreasing.  Likewise, the ratio of  PT to the company's actual net earnings is an indicator of whether the company's profit margins should be increasing or decreasing.[13]

                It typically takes a few years from the time a company changes the amount of capital allocated to patent generating activities until there will be a significant change in the rate at which the company obtains patents. This is because research activities typically have invention payoffs a few years after the research is initially funded, and it takes a few years, on average, from the time an application for patent is filed until a patent is granted.  Moreover, the expiration date of existing patents owned by a company can be extrapolated seventeen years into the future.  Therefore, extrapolations of CFT, PT and  Vc a few years into the future may be statistically significant predictors of a company's future of CFT, PT and  Vc, and therefore significant predictors of the company's future sales and earnings.  Since historical values for CFT, PT, and VT can be derived from historical data, I can extrapolate the time dependence for these quantities to future times using trend line analysis.  Hence, the extensions of the basic model provide company financial analysis tools.

                Several alternatives and refinements to the basic model are feasible.

                Instead of using GDP as the macro economic datum, the model is applicable to a direct macro economic measure of corporate profits in which case the step of multiplying the CF of a patent by a profit margin to determine net income is bypassed.

                Instead of relying upon axiom 1, the macro economic model could distinguish between economic sectors.  The alternative model could equate the CF of all enforceable patents associated with the economic sector with the fraction of GDP associated with the economic sector.  Data for the fraction of GDP associated with each economic sector defined by SIC code is available.  Patents may be associated with economic sector by associating the USPCS code for the patent with an SIC or NAISC code.

                The profit margin associated with the patent may adjusted down by an amount equal to a  profit margin associated with a free market.  The profit margin associated with a free market, such as a commodities market, is a profit margin that exists for products and services produced without patent protection.  A commodities market's profit margin is an estimate of the profit margin provided by non-patent assets.


V.            Validation of the Model


                Actual patent valuations are not a useful measure for validating the model because of the significant compounding effect of the selected value of R. Of course, anecdotal patent sales values could be assembled and compared with the model's valuations from which statistical deviations could be calculated.  However, actual patent sale data is scarce and anecdotal evidence would not be rigorous.

                Better measures for validating the model would be CFT and  to a lesser extent PT, since CFT and  PT do not depend upon multi year extrapolations.  CFT should correlate to company's gross sales.  PT should correlate to the company's net earnings.  Historical values of CFT and PT can be calculated from recorded assignment data for the company's patents.  Historical sales and earnings data is available for publicly traded companies.  Hence, it is feasible to determine the correlation of CFT to sales, and of PT to earnings, as a function of time for publicly traded companies.

                Another method for validating the model is to look at the effect of the issuance of a single patent on the sales, earnings, and market capitalization of the assignee company.  If the value of the patent is large relative to the sales, earnings, and market capitalization of the company prior to the patent's issuance, the company's sales, earnings, and market capitalization should significantly increase as a result of the issuance of the patent.

                I have not yet attempted a systematic validation of the model.  I intend to report on validation in a subsequent article.


VI.           Conclusion

                I present a new macro economic model that enables patents to be valued without relying upon data for actual sales of products covered by the patents or actual costs associated with generating those sales.  The model is readily implemented for all patents because it does not require sales and costs of sales data for actual products to be associated with specific patents.  Extensions of the model provide financial indicators based upon company patent portfolios.  The patent valuation model has been implemented, and patent valuations based upon it are commercially available.  Work is in progress to implement extensions of the model that provide patent based financial indicators, and to validate the results of the model.


[1]Richard A. Neifeld, Ph.D., Patent Attorney.  I wish to thank my colleague, Martin Goffman, Ph.D. for useful discussions, and I credit  him with the conception of the sector dependence alternative of the model.  The implementation of the model is the subject of pending patent applications.  I can be contacted at .  See www.patentvaluepredictor.com for implementation of the model and various financial products derived therefrom.  First publication at  83 JPTOS 211 (March, 2001). 

[2]For example, the AIPLA  recently formed a committee entitled "Management of IP Assets."  One subcommittee of that committee is devoted to exploring patent performance metrics.  Also see the discussion of patents to economies as a whole in  Roy et al., "Global Assessment of Patents, R & D Investment and Economic Output: Part 1 - Macro Economic Comparisons at the Country Level" 79 JPTOS 110 (February 1997).

[3]See Smith et al., "Valuation of Intellectual Property and Intangible Assets," published by John Wiley & Sons, Inc., New York, NY, Copyright 1994,  ISBN 0-471-30412-3.

[4]The amount at risk in patent suits, and hence the value of patents in suit appears to have risen dramatically over the last twenty years.  Coolley, "Overview and Statistical Study of the Law on Patent Damages" 75 JPTOS 515 (July 1993), reports on patent damage awards during 1982-1992.  Coolley shows that there were only three damage awards over one hundred million dollars in 1982-1992, and seventeen awards in the ten to one hundred million dollar range.  In contrast, the AIPLA "Report of Economic Survey 1997" page 70 indicated forty practitioners reporting patent infringement suits with amounts at risk of over one hundred million dollars, and one hundred and eighty six practitioners reporting patent infringement suites with estimated amounts at risk of ten to one hundred million dollars.  Another measure of the increasing value placed on patents is the number of patents requested and the number granted versus time.  The number of issued patents grew from sixty five thousand in 1982 to one hundred and twenty three thousand in 1997.   See the "Fiscal Year 1997 Patent and Trademark Review," Table 6, at page 87.

[5]See Henry A. Babcock, Ph.D., FASA, "Appraisal Principles and Procedures," Chapter 6, p. 95, (1995), published by the American Society of Appraisers, Washington, DC.

[6]A few companies have launched online patent exchanges for the sale and licensing of patents, which may eventually evolve into an active public market in patent sales and licensing, including "The Patent and License Exchange, Inc." and its web site at www.pl-x.com.

[7]If there is macro economic data indicating what fraction of GDP is covered by patent rights, that fraction of GDP may be used.

[8]Alternatively, anticipated annual variations in sales and in Pj over time may be included in the equations where industry wide variations due to product life cycle are anticipated.  For example, for pharmaceutical patents where market data may indicate expected life cycles for sales of patented drugs.

[9]Alternatively, R may be economic sector dependent, depending upon the different perceived risks and rewards for patents in different economic sectors.

[10]Various alternative profit margins may be selected.  For example, the profit margin for each patent may be selected to be the average profit margin in the economic sector most closely associated with the claims of that patent.  Economic sectors of a patent can be automatically determined, for example by correlating the United States Patent Classification System (USPCS) codes for the patent to the Standard Industrial Codes (SICs) or NAISCs.  NAISCs are industrial classification codes replacing SIC codes.  Both GDP and corporate profit margin broken down by SIC and NAISC codes are publicly available.

[11]Without selecting values, the general formula is:


Vi = Gk (( K * GDP * Mi * RPNi(v1, v2, v3...) / (Gj RPNj(v1, v2, v3...))) /(1 + R)k)

[12]By companies, here, I refer to all entities that contribute to GDP.

[13]Note that, when Pj is a function of economic sector of the jth patent, PT and CFT do not track one another.

Return to www.neifeld.com