When assessing the turbulence on the financial markets of the last few years, one of the main outcomes for investors was a heightened awareness of cash and security arrangements. In particular, it has become en vogue among asset owners to ask their custodians to segregate client assets from the custodians’ proprietary assets. While asset segregation has thus become a global topic, the legal effects of such segregation may differ from jurisdiction to jurisdiction. This article provides a summary overview of the patterns of asset segregation and their (possible) effects under Swiss law.

By Renato Costantini (Reference: CapLaw-2012-40)

1) Introduction

It is fair to say that the function of custody in the world of securities has, hitherto, been seen as rather dull. There was a perception that custody was a commodity and was viewed as the poor relation compared to other services purchased by investors. Today, we are seeing a much greater interest in and awareness of cash and asset safety. After all, the smartest allocation decisions and the appointment of the finest investment managers may all be in vain if the ownership of cash and securities is ever called into question.

As a result, among other measures, such as contractual limits on set-off and rehypothecation, strengthened liability for (sub)-custodians, and the introduction of OTC clearing regimes, investors are increasingly demanding for segregation of their cash and securities deposits. But what does that exactly mean and what are the legal effects of asset segregation? Is it possible to assess this on a general basis or can it only be based on knowledge and experience of the particular circumstances? In order to answer these questions, in particular the one on the legal effects of asset segregation, it is important to first understand the patterns and motives of asset segregation.

2) The Omnibus Account Structure—The World as We Know It

Normally, investors appoint a custodian bank to safeguard their securities and to hold their cash balances. The custodian banks, in turn, have securities and cash accounts with other custodians or with the Central Securities Depository (CSD), where securities are ultimately held. For cross-border securities holdings, the chain typically involves multiple custodians. In fact, in many non-domestic markets, custodians engage local sub-custodians, who, in turn, have accounts with the domestic CSD.

To maximize the efficiencies of the multi-tiered structure, most financial intermediaries, including CSDs, have adopted the practice of holding securities in fungible pools in their own name with their upper intermediary, CSD or issuer (as the case may be). Accordingly, in such omnibus account structure the ultimate investor’s ownership interest is not visible on the books of the issuer, the CSD or the other intermediaries, except, of course, for the custodian of the ultimate investor.

While the trend towards asset segregation in the aftermath of the financial crisis has apparently led to a general questioning of the omnibus account structure, it would be far too simplistic to just adopt such set-up universally. Rather does appropriate and efficient asset protection require an in-depth analysis of the asset holding structures at stake, in particular in light of all relevant jurisdictions involved. When deciding on asset segregation, one must first of all be aware that the sole fact that securities are normally held in the name of the intermediaries, and not in the name of the ultimate investor (omnibus account structure), does not allow any definite conclusion as to the entitlement of the investor to the securities held through such structure. In fact, to put it simplistic, two different ownership models have to be distinguished in modern securities holding:

In Model 1, the holding of the securities through and in the name of intermediaries does not alter the contractual or corporate relationship between the ultimate investors and the issuers of the securities. Absent agreement to the contrary, investors are the creditors and the shareholders of the issuers and, thus, the legal owners of the securities. Model 1 is normally adopted in most civil-law jurisdictions, such as Switzerland. In contrast, in Model 2, the holding of securities through and in the name of intermediaries leads to a transfer of title in the securities to the respective intermediaries. Accordingly, the investor does not have any direct legal relationship with the issuer. Rather does the investor (and any upper intermediary, except for the ultimate intermediary facing the issuer) have a mere claim against its upper intermediary. Model 2 is, in some instances, adopted in certain common-law jurisdictions, such as the UK and the US.

Further, and importantly, any cash balances, do, as a rule, only confer entitlements against the respective intermediary (but no upper-tier claims whatsoever). This basically applies to both, Model 1 and Model 2.

3) Motives and Effects of Asset Segregation

Given the different ownership models for intermediated securities holding, it is apparent that the motives and effects of asset segregation may differ from jurisdiction to jurisdiction:

a) Proof of Claim

In Model 1 jurisdictions, the segregation of client assets from the custodian’s proprietary assets does normally have no direct influence on the entitlement to the assets booked on a specific account. This, because, here, a person in whose name assets are held in custody is not automatically regarded as being their legal owner. This would only be the case if the person depositing the assets in its own name is also the legal owner of such assets and, therefore, not only acts in its own name, but also on its own behalf. If, by contrast, the person holding the asset in its own name acts on another person’s behalf, it does normally not have any ownership interest over the respective holdings. This typically applies to intermediaries who hold assets on behalf of their clients. Or, to put it the other way round, in Model 1 jurisdictions, the entitlement to assets in a securities account is not primarily determined by the name on that account. Consequently, asset segregation (or the change of the account name in favour of the client) does basically have no direct influence on the entitlements to the assets booked in that specific account.

This does, however, not mean that asset segregation is of no use whatsoever in Model 1 jurisdictions: Any co-mingling of client assets in omnibus accounts may complicate proprietary claims. Co-mingling and difficulties with proving ownership in an insolvency are exacerbated when assets are held through chains of intermediaries. In fact, for successful recovery of assets from the financial institution holding the assets, will require full proof of ownership. As custodians often use sub-custodians, which themselves may pool assets with those of persons who are unconnected to the client or its custodian, it may prove essential for the client in order to establish its ownership that assets where duly designated and segregated from third-party assets. Therefore, though not legally effective in the first instance, asset segregation may at least help to properly prove ownership interests in Model 1 jurisdictions.

Further, and in this context, asset segregation may, but need not, help to accelerate the recovery process in a custodian’s insolvency. Liquidators may take time to return assets in the possession of an insolvent intermediary. This is particularly true for assets held as security, given that liquidators will want to be sure that the security cannot be applied before making a distribution. However, and contrary to what is sometimes suggested, asset segregation is no guarantee for an accelerated recovery in insolvency.

b) Establishment of Claim

In Model 2 jurisdictions, the segregation of client assets from the custodian’s proprietary assets may, in fact, have a direct influence on the entitlements to the assetsbooked on a specific account. This, because such jurisdictions depart from the principle that the person in whose name the assets are held is to be regarded as their legal owner, regardless of whether that person acts on its own behalf or not. For instance, the fact that the assets of the custodian’s client are duly segregated and categorized appropriately in the books and records of the custodian may lead to the establishment of a client asset trust. Consequently, if the custodian becomes insolvent, its clients will have a proprietary claim over the trust assets. Where this is the case, the clients are entitled to full recovery of the relevant assets. This applies to both, cash and securities. In fact, in case of a client trust, the clients will also have a proprietary right over any cash amount received by the financial institution from a bank at which the financial institution held an account containing client money. In contrast, if the relevant assets are not properly segregated and therefore not subject to a trust, the client is an unsecured creditor and would be entitled only to a share of any assets of the insolvent financial institution after all secured creditors and the costs of the insolvency have been paid for. This example illustrates that, in Model 2 jurisdictions, where title to assets held with intermediaries normally vests in the intermediaries (and not in the client), the segregation of assets (cash and securities) may well have a direct legal effect, e.g. the creation of a proprietary interest in client trust assets.

c) Regulation

Regardless of whether asset segregation has a direct effect on ownership under civil law (Model 2 jurisdiction) or not (Model 1 jurisdiction), the obligation to segregate client assets may also be driven by regulation. This out of two, basically different, rationales: First, for asset protection purposes and, second, for reporting purposes. While in the first instance the motives of the regulator are closely linked to civil law asset protection, the regulation in the latter case is driven by reporting duties of any kind, such as tax and qualified shareholder reporting. Over the last years, both types of regulation have increased and, thus, on their part, also contributed to the lifting of the omnibus-account-veil.

In some instances, regulation does not impose asset segregation, but triggers certain obligations, such as increased capital charges, if not put in place. For instance, under the current Basel III framework, lower capital charges apply for direct participants in central clearing systems only if client portfolios are fully segregated from the portfolios of the client’s clearing member (segregation) and duly portable to another clearing member in the event of the default of the client’s clearing member (portability). Further to such scenarios, where regulatory obligations are triggered by the fact that asset segregation is not put in place, it is also possible that regulatory obligations are triggered by asset segregation which was implemented on a voluntary basis. Accordingly, also a segregation done on a voluntary basis should be thoroughly assessed prior to implementation.

4) Client Asset Segregation under Swiss Law

As Model 1 jurisdiction, Swiss civil law does not alter the ownership interest in securities just because they are held through and in the name of intermediaries. In fact, absent agreement to the contrary, such as fiduciary arrangements, a custodian holding client securities with a sub-custodian is not regarded as being the legal owner of its holdings. This irrespective of whether the custodian is holding the securities in its own name (omnibus account) or not. In order to have a valid ownership interest over securities held in (sub)-custody, it is therefore not necessary to book such securities in the client’s name. Accordingly, under Swiss civil law, asset segregation does basically have no direct influence on the ownership interest in the securities held in a securities account (establishment of claim). As a consequence of this, the protection of assets in case of insolvency of an intermediary is granted irrespective of whether asset segregation is put in place or not. This does, of course, not exclude that asset segregation may be of use when proving proprietary interests in securities held with intermediaries (proof of claim).

Further and in line with general legal principles applicable in most Model 1 and Model 2 jurisdictions, cash balances qualify as mere claim against the respective intermediary under Swiss civil law. This is mandatory and cannot be amended by agreement. Accordingly and in contrast to some other jurisdictions, where cash balances can be segregated from the custodian’s estate (client money), the segregation of cash balances would not lead to an ownership interest in such balances under Swiss civil law. In other words, like for non-cash assets, such as securities, client asset segregation does not have a direct influence on ownership interests: In case of cash, segregation has no direct impact because it would not lead to the establishment of a valid ownership interest over a cash balance, while, in case of securities, segregation has no direct impact because it is not required to establish a valid ownership interest over a securities account.

In accordance with Swiss civil law, Swiss regulation does not impose asset segregation of whatsoever kind. However, as in other jurisdictions, regulatory provisions may indirectly lead to asset segregation. For instance, under the Too Big to Fail legislation (TBTF) which became effective in March 2012, systemically important financial institutions are expected to prepare their organizational and operational set-up in order to assure that their specific recovery and resolution plan can be executed rapidly and effectively. While not explicitly required, asset segregation may well be part of such a plan in order to facilitate the portability of the respective assets to another financial institution.

5) Conclusion

When assessing asset protection arrangements, investors should be aware that, despite a general trend away from the traditional co-mingled omnibus accounts, asset segregation is not an obvious call. Rather, it is a complex decision that requires rests on multiple factors, such as laws and regulations (as applicable along the whole custodychain), asset classes involved (cash vs. securities), quality of segregation (“on behalf of all clients” vs. “on behalf of the client”) and level of segregation (“with custodian” vs. “with upper-tier intermediaries”). In this context, it is particularly worth noting that not all jurisdictions confer additional rights when asset segregation is implemented and that, wherever put in place, asset segregation will inevitably have an impact on the model of operating global custody and, thus, likely have costs and subsequent pricing implications. Today, it is therefore yet unclear whether asset segregation will effectively be established as an SOP all over the holding systems or whether it will be limited to specific patterns and situations. The fact that global custodians appear willing to invest in these changes and that clients appear willing to pay an increased price for custody services is, however, evidence of how recent events have changed, perhaps forever, our view of modern asset holding.