Home About Us Reports Research Paper 2002 Leveraging Knowledge Assets Page 7

About Us

Reports

Research Paper

Leveraging Knowledge Assets

6.1              Introduction

Under a federal approach reformed federal substantive law would govern all significant issues relating to federal IPRs, including the requirements for the creation and validity of a security right, its registration and priority status, the appropriate filing venue and the mode and manner of enforcement on default of the debtor. The key difference relates to registration and priority: under a provincial approach, an interest registered federally would be subject to any prior provincially registered interest. In contrast, under a federal approach, any interest registered federally would not be subject to any interest registered elsewhere.

The advantages and disadvantages of a federal approach are in some ways the converse of those of the provincial approach. The main disadvantage is that significant reform of the federal registry system is needed. (Though, as we have seen, significant technical reform to the federal registry is required even under the provincial approach.) A major advantage is that the unified federal jurisdiction eliminates the choice of law and chain of title search problems which dominated our analysis of the provincial approach. The advantage here is not simply that searching would be simplified because only one registry would be searched. More importantly, because federal law, not the law of the debtor’s location, would govern, the problem of difficult to search or undiscoverable interests granted by foreign debtor’s would be eliminated.

Three questions need to be dealt with in the context of a federal approach. First, will the system be a pure federal approach in which federal registration is the exclusive means of publicizing a security interest, or a mixed approach in which provincial registration retains some role? Secondly, what reforms are needed to the federal registration system? Finally, what is the scope of the federal system: for example, are security interests in unregistered copyright or royalty payments to be governed by federal or provincial law? The remainder of this Part deals with these questions in turn.

6.2              Pure or Mixed Approach ?

6.2.1        Recommended Approach

Under a provincial approach, an interest registered federally would be subject to any prior provincially registered interest. In contrast, under a federal approach, any interest registered federally would not be subject to any interest registered provincially. Thus to achieve the maximum degree of protection it would be necessary to register federally.

From this starting point, there are a number of variants. The main issue is whether the federal system would be the exclusive means of perfecting an interest in IPRs. In a pure federal approach, such as that established in the U.S in respect of copyright according to the notorious Peregrine decision, any security interest which is not registered federally would be entirely ineffective against third parties. In what is sometimes referred to as a “mixed system” provincial registration would be effective against any party who had not registered federally, including any judgment creditor and the trustee in bankruptcy.

In our view there is no advantage to a pure federal approach. The argument made by Judge Kozinski in Peregrine was that a pure federal registration system would facilitate searching as all relevant interests would be registered in a single venue. But this is also true in a mixed system. Since federal registration trumps any prior provincial registration, any person intending to register federally need only search federally. Any interests which were only registered provincially could be safely ignored. Accordingly, we recommend that the federal approach adopt this “mixed” system. We will assume a mixed system in the discussion which follows.

We noted earlier that qualification based on actual knowledge in the federal registry system should be abolished no matter what approach is adopted. It bears emphasizing here that it is essential to the “mixed” federal approach that the holder of a federal interest will have priority over any security interest which is only registered provincially, even if the federally registered interest holder has prior actual knowledge of the provincial registration. Actual knowledge might be very common as a result of a search of the provincial system, and if this knowledge were to affect priority the mixed federal approach would be fatally undermined.

6.2.2        Alternative Mixed Approach

In principle a mixed system in which the law governing issues other than priority, such as validity, attachment and enforcement, would be left to the law of the debtor’s location. (This approach would have sub-variants depending again on whether a provincially registered security interests was ineffective for all purposes or only in a competition with a federally registered interest.) Our view is that this type of mixed approach should be rejected in favour of the holistic approach described above in which all issues would be dealt with by federal law. The holistic approach would not be significantly more difficult to implement. Most of the major system reforms are required to accommodate the basic priority rules, and addressing validity, enforcement and similar issues would be primarily a matter of drafting. There would also be some economies in addressing the non-priority matters federally. For example, specific enforcement procedures might be useful for IPRs and federal law would avoid the need to amend the provincial laws in this respect. It would also allow more effective integration of enforcement and title issues, as it would be straightforward to provide for transfer of ownership as an enforcement mechanism. This is particularly relevant in the case of foreign debtors. If enforcement mechanisms were left to foreign law, then what would be the effect of a foreign judgment ordering the transfer of ownership in a Canadian IPR? In the Canadian context uncertainty on this matter could be resolved through federal legislation recognizing the validity of the provincial enforcement process, but a solution along these lines in respect of foreign debtors would be much more difficult. Perhaps none of these objections is overwhelming, but once priority issues are allocated to federal jurisdiction, there is no positive reason in favour of leaving validity, enforcement and other such matters to the law of the debtor’s jurisdiction.

6.3              Reforms to the Federal Registry System

6.3.1        Basic Reforms

We have seen that even in a provincial system the federal register would need to be reformed to allow legally reliable, up to date, on-line searching. Debtor name rules would also need to be specified to enable reliable searching. The same basic reforms would needed in a federal system.

6.3.2        Notice Filing

There is some ambiguity in usage as to whether notice filing is simply opposed to document filing or automatically implies perfection of after-acquired property. We use it in the former sense. The superiority of a notice registry for security interests generally is a principle that has been widely accepted in Canadian law. All the provincial movables registries and the federal Bank Act registries are notice registries and the same model has been proposed for security interests in land.⁸⁰ There is no reason why it would be any less advantageous in respect of a registry for security in intellectual property.⁸¹ Accordingly, we recommend and will assume in the remainder of this Report that notice filing would be adopted in any federal registry system that might be established for security in intellectual property rights.

6.3.3        Integrated or Separate Charge and Ownership Registries

The federal approach has several possible implementations which vary according to degree of unification of the federal IP registries. Perhaps the most obvious approach would use the existing separate federal IPR registries for registration of security interests as well as ownership interests; that is, all copyright related interests would be filed in the copyright register, etc. Alternatively, it is sometimes suggested that any federal system for registration of security interests should be a unified security interest registry for all types of IPRs.⁸² This alternative itself can be further broken down. The current separate federal IP registries might be maintained as ownership registries only and a new separate federal IP security interest registry would be established. Another option would be to integrate all existing and proposed registries for the various IPRs into a single “grand unified” IP registry which would provide a single venue for registration of all types of interest in any type of IP. This option would avoid some of the legal problems with coordination of separate registries, but it might well present significant technical hurdles, particularly in view of the different search fields which must be maintained for the different kinds of IPRs.

In our view this issue is not central. The advantage to either type of unified federal registry is that it would provide a ‘one-stop’ search and registration venue. But this can equally well be achieved using a “gateway” approach to searching multiple registries in which a single search at an electronic “meta-site” would automatically be routed to all relevant registries and with the results returned as a single report. Thus to the user the system would appear to have only one registry, no matter how the registries are configured technically. As we have seen in the discussion of the provincial system, one impediment to the effective implementation of the gateway concept is the need for uniform debtor name rules. This requirement should not present any difficulty in respect of the various federal IP registries given that all are within the federal jurisdiction. There is no substantive reason for different name criteria for in these registries and if they are modernized it should be very straightforward to ensure that the same criteria are implemented in respect of each registry. Once this is accomplished the legal problem is solved and the issue of whether the federal IP registries should be unified either wholly or in part turns into a purely technical question of whether it is easier to build a unified gateway or an entirely new registry.

Since the resolution of this issue is not crucial, for convenience in characterizing the federal position in our framework we will adopt the simplest approach, which is that separate registries would be maintained for different types of IPRs and both security interests and ownership interests would be registered in the appropriate IP registry. We emphasize that we are also recommending and assuming that the separate federal registries would have uniform debtor name rules so that ‘one-stop’ gateway searching and registration could be easily implemented.

6.3.4        After-acquired Property and Asset Indexing

Perhaps the single most trenchant criticism of the federal approach is that, because the federal registries are asset based, the federal approach would prevent lenders from effectively taking security in after-acquired property. We agree that this is a serious concern, but we believe it can be addressed with proper registry design.

Asset indexing minimizes the search burden on subsequent grantees, since a single search reveals all current interests in that asset. The difficulty with asset indexing is that it impedes granting security interests in after-acquired property. But a technical solution to this problem is possible in an appropriately designed unified ownership/security interest registry system. It should be possible to provide that when a new asset is registered in the debtor’s name the system would automatically search for security interests also registered against that name and any such interest would automatically be registered against the specific asset.⁸³ We can refer to this as a system with automatic cross-registration. To implement such a system debtor name regulations would have to apply to ownership interests registered in the ownership registry as well as to security interests. Further, a specific check box on the financing statement would be needed to reflect whether the security agreement has an after-acquired property clause. The system must not register a security interest which does not have an after-acquired property clause against the debtor’s after-acquired property. A specific field for this purpose on the financing statement, similar in principle to the separate field for serial number descriptions, would be necessary to allow this approach to be implemented by a computerized system.

If implemented, such a system would greatly reduce the burden of providing specific asset descriptions in the initial registration as well as allowing automatic perfection of interests in after-acquired property. The demands on the system are more severe in the context of automatic perfection of after-acquired property than in the case of initial registration. On initial registration the automatic cross-indexing capacity would serve as an aid to the registrant in identifying property belonging to the debtor and in ensuring that no errors are made in the asset descriptions. Inexact match problems could be dealt with directly by the registrant.

Inexact matches would pose a more significant problem in the context of after-acquired property. There are two concerns. First, what would happen if ownership of the newly acquired asset (e.g. a patent for a new invention) was registered in a name which was not the correct name (as prescribed by regulation) of the newly registered owner? Secondly, what would the system do with security interests registered against names which are inexact matches to the name under which the new asset is registered?

The first problem is significant because a debtor might intentionally register a newly acquired property under a name other than the debtor’s correct legal name precisely in order to avoid having prior security interests attach to the new property.⁸⁴ We should distinguish between the consequences of an incorrect name registration as between the secured party and a subsequent secured party taking a security interest in the after-acquired property specifically, and as between the secured party and a debtor. Suppose SP1 takes a security interest in all the present and after-acquired property of D. D then acquires new property but registers its ownership under an incorrect name. The system does not discover the prior security interest registered by the correct name, and so the prior interest is not specifically attached to that property. It is nonetheless straightforward to provide that the security interest in that after-acquired asset is enforceable against the debtor, since the point of perfection is to protect third parties. To deal with the problem of a competition with a third party, an ownership registration under any name other than the correct name would be invalid, so that a subsequent transferee from the same transferor who registered under their correct name would have priority. This provides D with a strong incentive to register under the correct name in order to protect its ownership interest. Further, one of the main advantages of avoiding the application of the after-acquired property clause is that D would then be in a position to grant a security interest in an unencumbered asset. But a subsequent secured party (SP2) taking an interest in the property would have a strong incentive to ensure that the property was registered under the correct debtor name, since otherwise SP2 would be exposed to subsequent interest holders who did register by the correct name. The debtor who had registered by an incorrect name would be entitled to amend the registration, but on doing so SP1's prior interest would be revealed.

The next problem is how to deal with security interests registered against names which are inexact matches to the name of the transferee. One approach would be to automatically register all such inexact matches against the newly acquired specific asset and rely on the procedures allowing the debtor to discharge financing statements which are erroneously registered. This approach has the disadvantage of burdening ‘innocent’ transferees who have the misfortune of having a name similar to that of a debtor with a security interest in after-acquired property registered against them. An alternative approach would be to use more stringent match criteria for automatic attachment to after-acquired property than is used in searches generally. This puts a higher onus on secured parties to use the correct name when registering their security interest. This is not unreasonable. In a sense it is a quid pro quo for the ability to automatically capture after-acquired property. And it is better to but the burden on the secured party, who benefits from the transaction, than on the innocent transferee with a name similar to that of the debtor. Note that inexact matches under the PPSAs burden the secured party dealing with the debtor, who must verify whether the inexact matches are the same person as their debtor. They do not burden the innocent parties whose similar names resulted in the inexact match.

The above solutions to the problem of inexact matches highlight a disadvantage of our proposed system, namely that it requires registration by both name and asset description. In comparison, an advantage of traditional land titles systems which are parcel indexed is that a secured party can be entirely secure in its registration if it registers by the correct asset description alone, and there is no need to prescribe a strict form for debtor name. But registration by debtor name is not unknown in parcel indexed land systems. New Brunswick applies the same debtor name rule under its Land Titles Act as it does under the PPSA out of a concern for improving the ability of subsequent judgment creditors to discover and register against the assets of a debtor.⁸⁵ Anecdotal evidence indicates that, as would be predicted, this has indeed imposed noticeable additional inconvenience and costs in mortgage lending. At the same time, the New Brunswick experience shows that it is not infeasible to apply naming conventions in the context of asset based registration. Note that the rule requiring a transferee to register by its correct name is not as onerous as the same rule in respect of a secured party, since the transferee doesn’t need proof from itself of its own correct name.

The system could be implemented either an archival system or a title guarantee system, though it would be more certain under a title guarantee system. The system operates by matching the debtor name against which the security interest is registered, with the name of the owner of any newly registered property. In an archival system the most recently registered transferee is owner is not definitively the owner, since ownership depends on the efficacy of the instrument itself, which is not assessed on registration. An invalid transfer in the chain of title will invalidate the security interest. As a practical matter this should be a relatively minor problem.

A further constraint is that the system must be electronic in order to implement the automatic perfection feature efficiently. In principle functionally the same method could be implemented manually. A rule could be implemented that whenever a party seeks to have title to an asset registered in their name, they must present a recent search of their name in the debtor index. If the search discloses a security interest in collateral including after-acquired property, the registry staff will then register the interest against the specific asset. The manual method has obvious disadvantages. It would be more costly because of the extra paperwork and searching required on the part of both the registrant and the registry staff. These costs would be incurred in respect of each registration, even when there was in fact no prior security interest with an after-acquired property clause to be disclosed. Nor could it provide the same level of security as an automated system, since there would necessarily be some window depending on how a “recent” search result would be defined. In contrast, if designed in from the outset, this approach should add relatively little cost to a computerized registry system. It would probably not be worthwhile to implement this approach in a paper based system.

6.3.5        Copyright and Asset Based Searching

Asset based searching is not feasible for copyrights. Asset based searching requires a unique identifier for each property such as a motor vehicle serial number or a geographical parcel identifier in the case of land. This criterion is clearly satisfied for patents and trade-marks which are assigned unique identifiers on registration. Copyrights, however, are probably incapable of being sufficiently reliably identified to form the basis for asset based searching. Asset indexing under the current copyright registration system is by title. But the title is probably not an adequate identifier for the purposes of asset based searching.⁸⁶ First, unlike a motor vehicle serial number or a patent number, the title of a copyrighted work is not unique. Different works with the same or similar titles are common. Titles are more like debtor names in this respect. Defining the registration and search requirements for debtor name has proven to be a source of considerable difficultly in the operation of the PPSA. But the fact that debtor name searches are workable does not imply that copyright title searches would also be workable. Some ambiguity in debtor name is acceptable because a party dealing directly with the debtor can take steps to determine whether close matches or proximate name turned up in a search do refer to the debtor. In contrast, reliance on the asset description is most important when dealing with the collateral in the hands of someone other than the debtor who granted the security interest. With nothing more than a title, the secured party cannot verify that the song “Loving You” which has a security interest registered against it is different from the song of the same title which their debtor wants to use as collateral.⁸⁷ This is why an asset description, unlike a debtor name description, must be truly unique if it is to be used for asset based searching.

An even more significant hurdle is that while debtors do have legal names which can be well defined (even though it may be inconvenient to verify that legal name), there is no single legal title of a copyrighted work. The title of a work for the purposes of registration is whatever the copyright owner says it is. How can a searcher know whether the book “Revolt in the Desert” which is unencumbered is different from the book “The Seven Pillars of Wisdom” which has a security interest registered against it?⁸⁸ The current provincial systems for serial numbered goods are premised on the notion that the serial number is in fact a reliable unique identifier, and the dispute is largely as to when the burden of registration by serial number should be imposed. In contrast, if asset based searching were implemented for copyright, it would be necessary to provide special rules to determine priorities when the security interest was not discovered because the title of the work it was registered against was not the title which the third party used to search, recognizing that neither title can be considered the true title.

It might be suggested that the CIPO could assign a unique number to each federally registered copyrighted work, but in order to ensure that the same work was not registered twice under different titles and so assigned different numbers, the copyright office staff would have to examine the works themselves and compare them to already registered works. The administrative burden and consequent cost makes this impractical. Thus, though asset based searching could be mandated for copyright, it would be so unreliable as to be scarcely be worthwhile.

6.3.6        Search and Registration Burden

One advantage sometimes claimed for a provincial approach is that it would allow a creditor to take a security interests in all the debtor’s personal property with a single registration. But just as federal/provincial systems coordination integration can be used to reduce the search burden in a provincial system, so it can be used to reduce the registration burden in a federal system. We have seen that the federal system should use notice financing and should be automatically cross-indexed so that it would be possible to register a security interest in IPRs by debtor name rather than by specific asset. If this is done, then a “gateway” approach to registration of security interests can be adopted in which a centralized request to register a security interest in all of the debtor’s personal property would automatically register the interest federally against the IPR as well as provincially against the traditional personal property. This functionality could either be provided through the provincial personal property security systems or through a commercial service provider. Admittedly this approach would not provide a truly transparent “one-stop” registration, since if provincial debtor name rules were not uniform the registration, in at least some jurisdiction (where federal and provincial name rules differed) would have to provide two different debtor names in different registration fields. In this sense the underlying dual registration would be seen by the user and the system would not be truly seamless. Nonetheless, the necessary dual registration could be made very convenient.

Conversely, a federal system would require dual searching in any case in which the creditor wished to take a security interest in all the debtor’s property, not just the IPRs. Even so the search burden would be less than in a provincial system. In a federal system the creditor would need to search the provincial system only by the names of its own debtor, rather than by the names of all predecessors in title. Further, the results in a federal system would be more reliable. Because the search on the IPR side would be conducted in a single registry rather than back and forth between the federal ownership and provincial security registries, the uncertainty stemming from inconsistent debtor names would be eliminated. Nor would there be a need to search any foreign system, or accept the risk of undiscoverable foreign interests. At the same time, a gateway approach could also be used to further reduce the search burden in the federal system by providing a one-stop search venue. As with registration, the dual search would not be completely transparent because of the potential need to enter two different debtor names to accommodate federal and provincial name rules.

6.3.7        Title Guarantee System

As noted above, the fact of a ownership registry marks a fundamental difference between federal IPRs and other types of personal property. This immediately raises a question which, curiously, seems not to have been raised in previous commentary. This question is whether the federal system should be reformed to implement a title guarantee system analogous to a land titles (or Torrens) system for land.

Existing federal registration systems are, at most, archival systems. That is to say, they operate in the same fashion as a traditional instrument registry system in the land context: unregistered documents are void as against registered documents, but registration does not itself determine ownership. It is necessary to review the chain of title to ensure that the person currently shown as owner is in fact the owner. The alternative to an archival system is a title guarantee system analogous to a land titles system for land, whereby the registered owner is statutorily deemed to be the true owner. Such a system could, if desired, be implemented in respect of uniquely identifiable types of intellectual property such as patents and registered trade-marks.

In the land context a title guarantee (land titles) system is widely considered to be superior to a instrument registry system as it provides increased certainty and reduces duplication of effort in searching title. On its face, a title guarantee system would be desirable for IPRs for the same reason that it is desirable in the land context, namely the elimination of duplicative searching. Admittedly, there are differences between IPRs and land which make the analogy imperfect. The term of most IPRs is limited and IPRs become less valuable near the end of their term and so are increasingly less likely to be transferred. For both these reasons searching a chain of title to an IPR may be less onerous than searching a chain of title to land.⁸⁹ But the question is not whether a title guarantee system for IPRs is as desirable as such a system for land, but whether it is more efficient than an archival system for IPRs. The disadvantage of a title guarantee system is the additional burden it shifts onto the registry office. The advantage is that title documents are reviewed only once, rather than each time there is a transfer. This suggests that so long as IPRs are, on average, transferred more than once,⁹⁰ a title guarantee system is preferable to an archival system.

If a title guarantee system is desirable, this almost necessitates a federal approach to security interests in IPRs. This is because under a provincial system the full chain of title must be available to allow parties dealing with an IPR to search for the name of prior debtors in order to search the provincial security interest registry for security interests given by predecessors in title. In a title guarantee system prior owners are hidden behind the registry “curtain.”

A title guarantee system probably could not be implemented in respect of copyrights, primarily because of the difficulty of uniquely identifying the copyrighted work. A title guarantee system requires that the registry staff examine the documents of transfer and verify the true ownership once and for all at the time of registration of a transfer. The degree of certainty required means that the registry staff would not be able to rely on the title alone. Examination and comparison of the work itself with previous works would be required and this would present a prohibitively costly administrative burden. Further, where the first owner of the copyright is the author, ownership reverts to the heirs of the author 25 years after the death of the author notwithstanding an previous assignment.⁹¹ Since this is a generally unknown group it would not be possible for a title register to automatically show the heirs as the owner.

6.4              Scope

6.4.1        Security in Provincial IPRs

As we have seen, the major problems with the provincial approach stem from federal authority over the ownership and the transfer of ownership of federal IPRs in combination with the existence of federal ownership registries for all six categories of federal IPRs. Because there is no ownership registry, either provincial or federal, for provincial IPRs, there is no apparent advantage in including these types of collateral within the sweep of a substantive federal secured transactions regime even assuming this were constitutionally permissible. Rather they should continue to be governed, as they are now under the provincial choice of law rules, by the law of the jurisdiction where the debtor is located. This would include the protection accorded to unregistered trade-marks under s.7(b) of the Trade-marks Act. Though the protection is accorded by federal law, no ownership registry is required.

6.4.2        Security in Unregistered Copyrights

We have seen that the principal advantage of the federal substantive solution is that it permits coordination of the law governing security with the rules governing registration of ownership in the federal IPRs. The difficult case is copyright, the one form of federal IPR that is not dependent on federal registration for its existence. A particular problem arises in respect of works under development, software being a prime example. Under the provincial approach a security interest in such a work can be readily obtained with a single registration under the debtor name. This is often regarded as a significant advantage of the provincial approach.

Under a federal approach, one option to unregistered copyright is to provide that the general secured transactions law of the debtor’s location would govern. Unfortunately, this does not solve the problem of works under development. In order to maintain the integrity and reliability of the federal record, any secured party taking an interest in an unregistered right would have to be subordinated in the event that the debtor subsequently registered the copyright federally and then granted a competing federal security interest in (or assigned) the registered copyright. Thus a secured creditor would be unable to secure its priority ranking by registration federally in cases where registration of the copyright itself would be premature at the time security is taken or where the product is under continuous evolution. The secured creditor could protect itself by requiring the debtor to register the copyright federally before a security interest is taken, but this would face practical problems when the work is under development. The other option is to provide that federal registration would be required to perfect a security interest in all copyrights, even where the copyright itself has not yet been registered at the time security is granted. The objection to this approach which is usually raised is that it appears to be even worse than the first alternative with respect to work under development, in particular software. Under the first alternative continually updated federal registration is required to avoid the possibility of subordination; in this second alternative continual registration is required to achieve any protection at all.

The problem here is not unique to copyright. Rather, it is the problem of after-acquired property. We have discussed above an approach to the federal registry which would accommodate security interests in after-acquired property. If this approach to registry design is adopted federal registration is no longer an impediment to publicizing interests in unregistered copyright. The first approach outlined above should be adopted. Provincial law would govern copyright in unregistered works, so that a secured creditor who had registered provincially would be fully protected, unless the work were subsequently registered. To protect itself against a subsequent registration followed by a grant of a security interest/assignment, the secured creditor could, if it wished, register a security interest federally against the debtor’s name. If the unregistered work were then registered, it would automatically attract the security interest.

6.4.3        Security in Royalties and Debtor/Licensor’s Contract Rights

A debtor/licensor’s right to collect royalty payments from its licensees can be used as security or assigned outright in the same way as any other chose in action or claim. Should a security interest in royalty payments owing to a debtor/licensor be governed by federal law, as being a form of a security interest in an IPR, or by the law of the debtor’s location, as being a form of account? A lender taking a security interest in the IPR will normally wish to take a security interest in the royalty stream at the same time. This suggests that a security interest in royalties in the same manner as the underlying IPR to avoid the need for dual registration. But dual registration itself should not be a serious burden, if a gateway approach to registration is adopted, or when the lender is taking a security interest in all of the debtor assets. And provincial registration of security interests in royalties does not raise the same search problems as we saw in the provincial approach to security interests in IPRs generally. The royalty payments in question will normally be based on licences granted by the immediate debtor. Long chains of assignments of the royalty payments themselves will not be common. For example, if the debtor is an exclusive licensee with the power to grant end-user licences in a particular geographic area, the lender will need a chain of title search to ensure the debtors exclusive licence is not subject to prior encumbrances, but the lender is concerned only with the royalty payments derived from the sub-licences, not payments owing to the original owner of the IPR who granted the exclusive licence to the debtor. For this reason the relevant security interests in royalties would be readily discoverable even if registered only the debtor’s jurisdiction, and not in the federal registry. Thus federal registration of security interests in royalty payments is not essential in a federal system.

The advantage of allowing such security interests to be governed by the law of the debtor’s location is that all accounts, whether based on IPRs or otherwise, would be governed by the same law. This avoid characterization issues which might otherwise arise. (For example, would accounts due for technical support services provided by the licensor in respect of the licenced software be considered as royalty payments or as a separate account? Would it matter if the service agreement were in the original licence or in a separate contract?)

On the whole we recommend that security interests in the licensor’s rights under a licence, including the right to royalty payments and any non-monetary rights, should be treated as a separate form of collateral governed by the general secured transactions law of each province or territory of residence of the debtor.

6.4.4        Security in Licensees Interest in a Licence

Exclusive licences which must be registered in the federal ownership registry in order to be binding against subsequent parties are tantamount to assignments. Security interests in such exclusive licences, like security interests in an ownership interest, would be governed by federal law under the federal approach.

In contrast, security interests in non-exclusive licences, which do not have to be registered in the federal registry in order to be effective against subsequent parties, should be perfected under provincial law (or, more precisely, the law of the debtor’s location). Since the non-exclusive licence itself cannot be discovered by a federal search, leaving the security interests in the non-exclusive licence to the provincial system does not impair the integrity of the federal register any further. Provincial registration is also more convenient, particularly in cases where the licence in question is an end-user licence needed to run the business and the debtor does not have other significant IP assets, so that lender might have no other reason to register federally.

Note that this does not imply that a security interest registered provincially will necessarily be enforceable against the licensor. This issue is discussed below.⁹²