The most dreaded words in the banking business must be too big to fail. For they imply, for the affected entity, that there will be less leverage allowed, more capital required and demand for a tough living will alongside a costly new organisation structure to enable legally sound ring-fencing of insured depositors from the effect of a bank insolvency.
There has been little public debate about the too big to fail suppliers in the securities services space. The reality is that, on a global or national scale, most of the top twenty providers would fall into this category. It is not so much that their clients are directly guaranteed, other than the retail population and some beneficiaries of pension funds, but that the political impact of a material default affecting people’s savings would be significant. And although the laws on beneficial versus legal ownership are quite robust in most jurisdictions, there remains a residual risk that needs to be quantified and managed, most likely through a specific capital allocation.
I see a series of key risks that need to be considered as part of the control tool box. The good news is that the regulators are doing a reasonable job in managing this process. The key risks, in order of their gravity that I would identify are fiduciary risk, transactional risk, ethical risk, credit and operational risk. The paradox is that regulation may have a good handle on credit risk, captured in general banking oversight, and operational risk, for which statistical data exists. But they have less of a handle on the other risks, mainly because they have little in the way of historic evidence or because the change in the game plan has been too recent for many to appreciate the scale of the challenge.
I see the main sources of credit risk being with transaction counterparties and also through leverage lending, mainly to alternatives. The main source of operational risk, excluding transaction risk which I deal with later, comes from corporate actions but, in reality losses in this area are usually of a scale that can be easily covered by revenue flows. These are generic banking risks, and as already noted, well covered by regulation and capital haircuts. It is the other risks that are more difficult to risk assess for there can be process rules and legal structures that reduce risk but it is unclear what capital or other buffers are needed to offer sensible protection. I will take each of the other risks I mentioned in turn and in reverse order to my perceived potential impact.
There have been many cases of failed ethical behaviour across the banking industry but securities’ services has been a relatively sound area. The exceptions have been in alleged foreign exchange mispricing, portfolio transformations and stock lending disputes. Information allows clients to measure their banks in these areas much more effectively now, but the reality is captive business is tempting and larger clients should ensure that meaningful transaction flows are put out to tender or executed on open platforms. The general focus on ethics by regulators has eliminated some of the behavioural risks, although it would be a greater optimistic than me to say this risk is behind us or that there are not further pitfalls to be discovered. However, internalisation of trade flows within universal banks is a risk as there is definitely a tempting conflict between the principal trader aim to maximise profits and the fiduciary duty to ensure that captive flows are subject to best execution principles.
Transactional risks are worrying. They are usually seen as part and parcel of operating risks. But the nature of the underlying business has so changed and the risks resulting have been so transformed that they need to be examined under a new light and apart from the historic performance of the market. It is for this reason, as an example, that regulators are credible when they look for an allocation of capital to support stock lending positions. Stock lending, including asset swaps and similar financial products, is more risky than in the past. Quite simply, markets are more volatile and therefore the risk of collateral shortfalls have increased whilst liquidity is scarcer with a resultant increase in realisation risk. And collateral management has meant a greater immobilisation of assets as markets move to secured transacting. The rules on re-hypothecation have been tightened, but the law is still either unproven or capable of challenge when it comes to establishing title on cross jurisdictional temporary allocation of assets. The risks in OTC markets have been amply discussed but the reality is the traded markets, now home to an ever greater portion of fund assets, are logically only a perfect indication of the balance between seller and buyer at the most recent point in time and a poor indicator of the eventual realisation price for an asset at a time of crisis. Regulators have struggled with pricing risk for some time; the danger is that they confuse market risk with technical pricing risk and make the agents liable for both. And I would also note that old bugbear of delivery versus payment, a simple concept first raised in the first G30 report in the 1980’s. It remains a loosely used, and much abused, term with intermediaries advising it operates even in markets where payments, as an example, can be subsequently unwound. Such a misleading definition of delivery versus payment has to be at their own risk!
The greatest risk though remains fiduciary risk. It has escalated over recent years, if only because of the tight asset safety rules adopted by the EU regulators and the stringent interpretation of administrator liability in cases such as Madoff. The fiduciary is now liable for losses across their administered funds, across their networks and within infrastructures, depending on the nature of the underlying client. And I have noted in earlier blogs that, irrespective of the low likelihood of such losses arising, the sheer scale of assets held makes this a potential wipe out risk, not just for the custodian or administrator but also for their entire group. And there are further fiduciary risks, in the general oversight package for managers being adopted by regulators around the world, that add to this risk albeit to a lesser financial extent but with significant reputational risk.
In 2013, I blogged in this journal that each trillion dollars of assets could need $400 million of capital (4 basis points) at a cost (using 15% returns) of $60 million per annum or 0.6bp of assets under custody (versus the 2-3 basis points gross returns being earned generally). I recall one of the top three global custodians suggesting my figures were too low. With hindsight, and given the risks discussed briefly here, I suspect they were right. The question is what precisely the capital allocation should be and the basis to be used for its computation. One has to query if this is not going to become a material issue for the future for our industry’s too big to fail brigade.