Patents have become a direct source of revenue, apart from providing legal protections for inventions. A patent generates revenue in its patented form through licensing or sell-out. Three challenges confront this shift in use of patents as direct revenue generators especially in the context of patent sale. First, the challenge of estimating the price a patent fetches in a sale (referred as ‗patent value‘) remains scarcely addressed owing to the intangible nature of patents and lack of patent sale data in public domain. Secondly, the variations in price and the method used to estimate the price while selling a patent individually (called as a singleton) or in a group along with other patents (called as portfolio) are little understood. Thirdly, literature focus so far has been on the static dimensions of patent value determinants leaving scope for researching the time varying dynamic nature of potential value determinants. This thesis aims to systematically address the above challenges and research gaps through five specific research objectives pertaining to the following four aspects of patent sale: (1) Successful sale aspect of patent lots in an auction (Research objective 1) Research objective 1: To explore the determinants of successful sale of singletons and portfolios. (2) Selling strategy aspect of the sold patents (Research objective 2) Research objective 2: To characterize the singleton and portfolio patent lots successfully sold in an auction. (3) Time dynamic aspect of factors influencing price (Research objectives 3 & 4) Research objective 3: To investigate the effect of time dynamic nature of patent age on the selling prices of singletons and portfolios. Research objective 4: To understand the effect of time dynamic trend of technological importance (TI) and patent class trend on selling prices of singletons and portfolios. (4) Valuation aspect (Research objective 5) Research objective 5: To identify the determinants of selling prices of singletons and portfolios. In this thesis, we use a combination of patent bibliometric and market based approach to study various aspects of singleton and portfolio patent sale incorporating the time dynamic aspect. A set of US patents auctioned by an US auction firm, called Ocean Tomo, is used as sample. The sample was dominated by patents from Computer & Communication field. The research findings showed significant differences in the characteristics of sold and unsold lots both in the case of singletons and portfolios. Amongst the sold lots, singletons were found to include more novel (lesser patent backward citations) and technology specific (higher C&C technology scope) patents than the portfolios sold. Further investigation on the variation in selling prices fetched by singletons and portfolios showed portfolios fetching significantly higher prices than singletons at the lot level. Interestingly, at a granular level, the equation reversed with singletons showing significantly higher price compared to the average price per individual US patent member within the portfolio. Along the time dynamic aspect, the existence of an optimal age to sell patents for higher prices is evidenced especially in the case of singletons from C&C technology field. The optimal age was found to be around 10 year 2 months of age from grant of singletons. No such optimal age was identified for portfolios. The growth trend analysis of the forward citations (technological importance of patent) and the patent applications filed within the specific patent class of a patent (attractiveness of patent class) illustrated the significance of the patent class trend in explaining variation in selling price of singletons. Finally, the valuation models of singletons and portfolios showed singletons being valued more on the patent related characteristics unlike portfolios which were valued based on the all three aspects – patent, seller and bundling. In summary, the novelty of the thesis resides in the (1) incorporation of new set of variables namely forward citation diversity, forward citations trend, patent class trend, technology scope and portfolio type in understanding patent price (2) treatment of time dynamic variables in understanding patent value and (3) focus on portfolio analysis through independent analysis of singleton and portfolio selling prices.

政府透過法律制度將原先具有非排他性及非敵對性等公共財性質之專利財產化，藉此鼓勵發明人進行研發，間接為整體社會帶來促進創新與阻礙創新等不同面向之影響，而專利之私有化使擁有專利之所有人取得排他獨佔權限，如此累積創新所帶來之結果便是形成專利叢林現象，使單一產品生產時須取得眾多專利，增加未取得專利之侵權風險，該專利叢林現象後續更促使訴訟成為新興專利聚集型態獲取利益之手段。 從專利取得之角度觀察，其取得專利之方式可系統化分為自行研發、併購等內化形式或取得專利授權等外部形式，而專利取得之動機則從單純產業利用轉變為企業策略性考量；現今專利市場中，更經常將取得之專利以授權或出售方式商品化，甚或以專利作為商業談判、利益交換之籌碼，使專利有貨幣化現象。從經濟學之角度思考，專利貨幣化或有交易、預防及投機等動機，並得從專利本質上與後續發展上之特性觀察到專利貨幣化之因素。 由於本文將專利市場分為專利聚集、專利交易平台、專利資訊提供者及一般個人、實際從事生產公司與研究機構等不同類型參與者，並將研究對象著重在專利聚集與專利交易平台二者，因此特於本文中討論專利聚集形成之階段與交易模式，且因應專利交易模式而衍生探討專利交易價金之計算。從本文可知，專利聚集得區分為攻擊型、防禦型、以授權營利型及混合型四大類，各類型專利聚集及專利交易平台皆有其創造價值之價值鏈活動安排、整體產業競爭作用力之五力分析與支援核心競爭優勢之相對應策略活動系統，並從各市場參與者之運作模式中亦得觀察可能存在之缺失。 攻擊型專利聚集表面上雖提供授權服務，然手段上具有強制性，以訴訟等手段迫使下游實際從事生產等公司支付授權金以取得專利授權，從模式上觀察，其存在未經挑選專利、擁有專利過多及組織體系過於龐雜等缺失；而防禦型專利聚集則從下游買方之迫切需求角度出發，以防禦目的之專利授權或出售為訴求，藉由廣泛涵蓋不同領域之專利組合滿足下游買方對於專利侵權風險之控管，然其現有模式可能因支付有限年費導致難以確保購買大量專利之品質，並有為符合眾多會員需求使專利相關性降低以及整體產業難以避免搭便車現象等缺失；至於以授權營利型專利聚集透過技術或產品標準定義之專利組合，包裹式將下游實際從事生產廠商需要之專利一次性授權，雖不見得有授權之急迫需求，但顯然降低下游買方取得授權之交易成本，惟其缺失為權利金分配制度之公平性與專利組合之區別標準。混合型專利聚集則涵蓋上述三種專利聚集之優勢，並以特殊之智財資本市場重新定位無形資產能夠創造之價值，對於下游買方及上游專利供應者而言，皆帶來產業結構變化之衝擊，同時以專利組合授權來滿足下游專利被授權人之需求，不過，混合型專利聚集管理複雜度高、經濟利潤有下降可能，並且長期將有價格扭曲之應變風險。 專利交易平台與專利聚集不同之處在於其不直接取得專利或專利授權，僅以專利供給方與需求方之交易中介者自居，然從廣義概念上，其亦為專利聚集之一種形態，有助於專利買賣雙方取得資訊，降低搜尋、談判或執行等交易成本，並透過其他業務互補專利交易平台業務可能之不足，惟其仍舊有所缺失，亦即價格決定機制之困難、服務範疇過大導致成本控管之不經濟與交易公開之接受程度等問題。從本文各章節之分析中，可比較各市場參與者之不同，亦得觀察彼此間之互動與缺失，從而提出未來可能之研究方向，並針對專利市場之管制與開放給予建議。 / The propertization of patents which intrinsically show the traits of public goods including non-excludable and non-rival was established to encourage inventors and therefore leads to positive and negative influence to the society indirectly. The privatization of patents protects the exclusive rights of the owners, whereas what we called the accumulated innovation set up the patent thickets that enhance the risks of infringement and promote the strategic litigations raised by new patent aggregators. The methods of acquiring patents could systematically divided into internalization which includes R&D and M&A, and externalization which includes licensing and so on. While the motivations of acquiring patents transform from industrial applications to strategic considerations, and the patents gradually become commercialized and even express its monetization. This paper classifies different players in the patent market, such as patent aggregators, patent transaction platforms, patent information providers and operating companies, while with the focus on patent aggregators and patent transaction platform. Besides, the patent aggregators could be categorized into “offensive patent aggregators”, “defensive patent aggregators”, “running by licensing patent aggregators” and “hybrid patent aggregators”. Moreover, the findings of this paper stand on what each patent aggregator and patent transaction platform has its scheduled activities of the value chain, the five forces model to the analysis of its industry and the strategic activities system for supporting its core competence, and even the demerits of its operating model. The offensive patent aggregators provide licensing services, while its services apparently contain compulsive licensing model by raising claims against the operating companies. Instead, defensive patent aggregators help the downstream buyers in need to fight against the offensive patent aggregators by licensing or selling patent portfolios which cover broad technological areas and benefit risk management. Next, the running by licensing patent aggregators package their patents by standards of technologies or products and provide “one-stop-shop licensing solutions” to the downstream buyers to save the transaction costs. Furthermore, the hybrid patent aggregators embrace the advantages of three kinds of patent aggregators mentioned above and try to create the IP capital market leading to the transformation of the industry structure. On the other hand, the patent transaction platforms are different from the patent aggregators in the ownerships of patents. The patent transaction platforms will never become the owner of the patents or acquire the rights of patent licensing, it just named themselves intermediaries of transaction that reduce the transaction cost and enhance the transparency of information. Nevertheless, the diversified operating models of different patent aggregators and patent transaction platforms exist its improvable or inevitable drawbacks. The analysis of each chapter in this paper could help to compare the players in patent market and contribute to observe the shortages and interactions between the ones. What’s more, this paper gives some suggestions for further researches in the future as the conclusion.