Changes
Jump to navigation
Jump to search
Entezarkheir (2010) - Patent Thickets And Market Value An Empirical Analysis (view source)
Revision as of 18:51, 25 March 2013
, 18:51, 25 March 2013no edit summary
*This page is referenced in the [[Patent Thicket Literature Review]]
*This page is listed on the [[PTLR Core Papers]] page
The pro-patent shift of the United States has created a patent thicket. This has made the use of other firms innovations more costly, due to higher transaction costs and the possibility of hold up. Using a panel data on publicly traded US manufacturing firms from 1979 to 1996, this study finds a negative impact from the patent thicket on the market value of the firm. I also find that firms with larger patent portfolios experience a smaller effect, likely because stronger bargaining position lowers the occurrence of the hold-up problem for these firms. The advantage of larger firms is even more prominent following the pro-patent shift. My results also capture heterogeneity in the impact of the patent thicket across industries.
==Review==
===Measures of thicket===
Following Ziedonis, patent thicket is measured by the fragmentation of patent ownership faced by each firm based on the citations of that firm's patents to patents of other firms.
*USPTO data identifies patented invention, assignee, citations or references to other patents or non-patented inventions;
*Fragmentation is measured using a normalized Herfindahl index that gives more weight to firms with firms faced by fragmentation since citation shares are squared.
**[insert equation 4 HHI] 1 minus the sum of squared shares citations to other firms patents in each firms patent portfolio in each year.
===Sample===
Size, data source, industries, geography:
*1,975 publicly traded US manufacturing firms (SIC=2000-3999) with at least one patent, for 10,273 observations from 1979 to 1996.
*NBER patent data files are used to construct the thicket index using patents granted between 1963-2002 and citations from 1976 to 2002;
*Compustat North American Annual Industrial file (with company identifier) is used for company financial information from 1979 to 2002;
*Observed citations are truncated (citations data is unavailble prior to 1976 and are observed only with a lag), so:
**data is restricted to 1979 to 1996 period for analysis, and **values are corrected based on distribution of fraction of citations each year since grant date
===Results===
:''"...results show that firms experience a significant decline in their market value when the technology market is fragmented. The results also show that the pro-patent shift in the 1980s has increased the size of this impact. I also find that firms with larger patent portfolios experience a smaller negative premium in their market value. This is likely because firms with larger patent portfolios face fewer problems in their cross-licensing negotiations with external entities as the larger portfolio size increases their bargaining power in the licensing negotiations and lowers the risk of being held-up by their rivals. This advantage of firms with larger patent portfolios is even more prominent following the pro-patent shift in the 1980s.
*Calculating semi-elasticities, market value declines by 1.1% for a 10% increase in the (log) fragmentation index;
*Analyzing data before and after USPTO change and establishment of CAFC in 1980s (for 1979-1989 and for 1990-1996), market value declines by 0.98% and 1.29% respectively for a 10% increase in the log fragmentation index;
**Firms with larger portfolios of patents are less impacted, as reflected in a significantly positive interaction term between the fragmentation index and patent portfolio size, although this effect shrinks in the later period.
**There is a varying negative effect of patent thicket across industries, which is largest and most significant in chemical and mechanical industries.
===Social Welfare Consequences===
:''"The pro-patent shift of the United States has created a patent thicket. This has made the use of other firms’ innovations more costly, due to higher transaction costs and the possibility of hold up."''
===Dependent Variable and Model===
*Tobin's Q, the log of the ratio of the market value of the firm divided by its book value at a point in time. Non-linear least squares equations are estimated that controls for the following:
**A patent thicket measure which is the logarithm of the Ziedonis fragmentation index [see equation 4 model, pg.6];
**Market value, which is sum of common and preferred stock, long-term debt adjusted for inflation, and short-term debt;
**Tangible assets, which is the book value of the firm or net plant and equipment, inventories, investments in unconsolidated subsidiaries and intangibles;
**Intangible assets, which following Hall's definition is R&D intensity (R&D expenditure/tangible assets), patent intensity (number ot patents per R&D expenditure), citation yield per patent; this stock is depreciated at a rate of 15%.
**The model also controls for shadow value of firm assets, which is captured with firm, time and random effects in one specification.
