Former Head ● HSBC Securities Services
In 1989, when the first issue of Global Custodian was published, I was in the custody division of a UK bank, wondering whether we would survive that decade let alone the next 20 years. Now, albeit partly retired, I see my former colleagues as the survivors of an environmental tsunami that decimated the industry and dramatically changed the landscape.
I was present for the latter days of the fatal gestation period of the London Stock Exchange’s abortive Taurus development. That precursor of the current CREST system fell to the destructive forces of scope creep, political power play and technological arrogance.
There was the million-pound development add-on to save a few thousand a year for the stock lenders. Then there was the blind belief in the delivery capability of the system vendor even amid the shifting sands of user requirements. And overshadowing all, there was the lofty arrogance of the LSE executives. They acted like the founding fathers of the Guild of Masters of the Universe.
Twenty years ago, on the international front, I was urging SWIFT executives to open the doors to fund managers, just two years after they had decided to cover the securities market. I can remember being ostracized for doing so by most of the world banks. Their custodian arms were often unaware that their payment colleagues were fighting to retain SWIFT as a bankers’ bastion. At one conference I was the only panelist who even realized it was an issue. Indeed, one clearing banker accused me of wanting to destroy SWIFT by opening its doors to every “Tom, Dick and Harry” in the City of London.
In 1989 SWIFT handled 296 million messages with a modest contribution from the custodian banks. Now it handles 3 billion to 4 billion, and well over half of them relate to either securities transactions or payments linked to them. Paradoxically, the board of SWIFT still seems more reflective of geography than functions. More worryingly, SWIFT volumes remain concentrated across a narrow range of message types. And several of these remain susceptible to migration to alternative channels, including the Internet.
In 1989 I was on the board of ISSA, and we contemplated the arcane world of corporate actions and the elusive panacea of same-day payments. We worked on the G-30 report-the first, and most influential, global report on the securities plumbing system.
Luckily, just before I joined the ISSA board, the memorable “Swiss Army Colonel” B. Imseng was replaced as chairman by Josef Landolt. Imseng was known for his precise timekeeping and reputably once reprimanded a young Ray Parodi of Citibank for arriving a minute late for the board. He would have made mincemeat of me and undoubtedly smashed my Blackberry if I had dared to even glance at it at one of his meetings.
“Clearance and Settlement in the World’s Securities Markets” gave the G-30’s nine action points to improve the efficiency of the post-trade space. CSDs, DVP, stock borrowing and lending and the use of ISO messages were at the core of that succinct report. I suspect most key players read it and concurred with its recommendations.
There were country analyses of compliance to G-30. National pride alone required almost total compliance and, so with an unbounded enthusiasm for opaqueness and dissimilation, countries proclaimed their markets met the standards.
Countries without same-day funds alleged compliance with the principles of DVP. Others with embedded local standards deemed them ISO compliant. One alleged they met the stock borrowing recommendation-at least for international stocks without noting that the practice was illegal for domestic ones.
ISSA, to its credit, adopted a monitoring role. But it was not immune to political pressure, and I learned a lot about the subtleties of drafting.
G-30 highlighted a problem that markets still face. Recommendations are produced. Unworkable global timetables for their adoption are suggested. A one-size-fits-all approach is usually adopted. Then investors call for action. The end result is progress, more in form than substance. And, as can be seen in the ongoing debate about security of assets or message standards, the gaps are either still there or someone has merely moved the goal posts.
In 2000 ISSA produced an updated series of recommendations. The new paper reflected new concerns, covering new products, better governance and clearer legal risk, as well as calls for a broader use of ISO standards.
It also drafted a “Clearing and Depository Risk Evaluation Guide.” The infrastructures lambasted us for our temerity in assuming the right to assess them. They succeeded in killing off the project. Ironically, its successor in spirit, the work undertaken by Thomas Murray, is at least equally unpalatable to many of them.
Thereafter, everyone started publishing recommendations, and the market was inundated with lengthy tomes and overlapping, but not always consistent, declarations.
In November 2001, we had CPSS/IOSCO and their 55 pages, complete with 19 recommendations. BIS also waded in with a series of papers, carefully crafted by apparently risk-adverse central bankers.
The European Commission, through Alberto Giovannini, tired of recommendations and created lists of barriers. The first Giovannini tome in 2001 was a mere 95 pages, although his second report in 2003 added a further 62.
The deadlines all expired long ago and yet little progress has been made. There are several reasons. The recommendations wrongly saw Europe as a self-governing entity rather than part of a global marketplace. The barriers were agreed but there was no clear framework (other than a SWIFT-led initiative on standards) for their adoption. Practitioners were then drowned by the further maelstrom of added directives and recommendations.
Eventually, we had the father of them all when, in 2003, G-30 produced their “Global Clearing and Settlement” report with 20 recommendations and a mere 137 pages of closely debated text. How many people have read it? And how many recall its contents? But it actually is worth reading with an eye on the trauma of the last two years or so.
After all, one recommendation was to expand the use of central counterparties. Another was to address the failure of a systemically important institution. There was the call for improved valuation methodologies and close-out netting arrangements. How useful they would have all been in the world of Bear Stearns and Lehman Brothers.
The occasional report was influential. But most have been filed away, even by those who carefully crafted them to justify their own business models-but there have been winners and losers.
GSTPA, happily, has now been almost forgotten. It cost the industry-if one includes the jettisoned internal developments across its planned user community-hundreds of millions of dollars. It was an example of ambition and aspiration exceeding budget and reality.
It failed for similar reasons to the Taurus development. It was over-engineered. It was over-dependent on fundamental and costly market change, especially at investment managers, who were the smallest financial beneficiaries in the process. It was not easily adaptable to different business models. The Anglo-Saxon distributed model drove its design, yet many powerful voices, especially in Europe, operated a global bank model where the bulk of the GSTPA functions were internalized. Omgeo’s CTM and SWIFT’s Accord seek to fill the gap, but they still do not succeed in the GSTPA dream of the seamless interchange of real-time data across all parties to that complex post-trade confirmation and information chain. The original idea was a good one in a perfect world of altruistic service providers; alas, that is not the market I have lived in over the last 30 years.
Elsewhere, CLS proved itself. But is its future in more of the same or will it be tempted by diversification? It survived, unlike GSTPA, partly because there was no alternative. The central banks insisted the market had to put in money to create a system to eliminate timing and settlement risks in foreign exchange. The alternative would have been even more costly capital penalties. The systemic risk in the “plain vanilla” foreign exchange market at the time was real and potentially life threatening for some major institutions. There was ever-greater concentration of high multi-billion dollar exposures (and before CLS everything was settled as dealt) across a small number of banks at both the global and country level. It is, with hindsight, unfortunate that the same rigor was not applied to more complex instruments.
Other infrastructure change was mooted over the period. Euroclear and Clearstream never merged despite periodic senior stakeholder enthusiasm. LCH did merge, but with Clearnet, and subsequently proved that politics and national cultures can get in the way of any clear business logic for rationalization. And, if markets really followed user need, LCH.Clearnet would have merged first with Eurex before even talking to DTCC.
But users did not force such issues, yet they had the buyer power as they controlled their trade flows to the markets (which in turn fed the CCPs) or decided where to hold their assets. Perhaps, secretly, many liked the in-built market inefficiencies and opaqueness of the status quo. And infrastructure management had an easier life when they avoided the big issues that mergers of equals or like businesses would create.
Infrastructures also discovered that monopolies and duopolies can be profitable. The ICSDs hit the jackpot and were, for a long time, the thought leaders in the global market. They may be losing this role now to the DTCC and also some of the mega-custodians.
Much is made of the hybrid nature of these infrastructures. Are they infrastructure or are they global custodians in disguise? There is an unhealthy preoccupation with the concept of profit or not for profit. These are far less relevant than the fact that European custody and settlement, even allowing for scale impacts, is expensive relative to the US. And, in time, there is a risk that the emerging mega-markets of China, India or Brazil will reinforce this unfavorable cost comparison.
That is when the inevitable will happen, especially if DTCC and others adopt a more global approach to business.
But ICSD performance over the last two decades has been spectacular both in terms of the robustness of their infrastructures, richness of their product and improved client service ethic. Euroclear Bank and Clearstream Banking have seen assets grow in that period from $1.1 billion to $12.1 trillion. Asset servicing is obviously now cheaper in absolute terms although by no means commensurate with such volume increases. And, as for others in the sector, that reflects growing complexity, risk and service offerings as well as a fattening of those profit margins.
With all the muscle and billions of dollars of annual budgets among custodians, it is surprising that a substantial part of the plumbing remains in such poor shape.
Some areas have grasped the need for automation. The trading end has recently been the best example, much to the detriment of the bottom lines of the traditional exchanges.
MTFs now have up to 30% of total market trading activity, and that really shakes up the establishment. The CCP market has seen material price reduction with the advent of new entrants. Could the same happen on the depository front? It should not be excluded, especially if legislation and regulation drives activity away from its traditional locations.
The performance of the post-trade world outside of core functions for core instruments remains sub-standard.
Twenty years ago the target was to eliminate paper in the share ownership and trade settlement process for listed securities. That has worked pretty well. The market has also improved automation of corporate events, with a lot of help from SWIFT and many CSDs around the world.
But it has to be noted that there is still further progress to be made. Global Custodian surveys show corporate actions as the greatest single area of friction in the post-trade market. Beyond that, the sector remains a poor example of the 21st century and the power of technology.
The 20% or so of activity that falls outside the standardized world must be the source of around 50% of industry cost and 80% of industry errors.
We have seen, many times in the past two decades, a development cycle of manual process without volume evolving to automated process post-volume and post-crisis. Markets, even allowing for the current credit crunch, now tolerate too many costly and inefficient manual processes.
Investors and issuers have little interest in automation as long as intermediaries do the repairs and contain the problems. I have never understood why each new issue cannot have a page containing the data required for processing. Perhaps it is, as one lawyer explained to me, because of the need to “ensure we retain the important nuances” contained in prospectuses. Or is it that they feel protected by ambiguity?
Sadly, today, knowledge bases are being eroded and investment is being curtailed as costs are cut in the face of declining revenues. Developments are being directed to the main “sacred cows” of the regulators, albeit with regional jealousies potentially impairing sensible global solutions.
The regulatory burden is ratcheted up every year. The market, trapped in the “bad bankers” syndrome, appears scared to challenge some of the sweeping powers being accorded to the regulators. Yet there is little proof that more regulation is needed or that absolute regulatory power is the barrier to further crises.
Basically, there is just proof that current regulation has not been applied effectively. After all, if it had, would Madoff have really happened? And was RBS’ unsound business model a revelation or just confirmation of a basic law of risk management? And was the risk of leverage a 2007 discovery, even if it was the root cause of LTCM and many other illustrious collapses?
The problem is that the end-to-end securities production line looks ever more fragile with both capacity and liquidity unlikely to be adequate for any upturn in markets. In other words, there might have to be further consolidation. I suspect the survivors will be those whose organizations have managed to avoid the pitfalls of the last few decades. Custodians have already almost become an endangered species.
Future consolidation will risk creating an oligopoly that the survivors can exploit. I always remember the biggest global custodians voicing their concern to me where, as a sub-custodian, HSBC did not have at least one strong competitor in any single market. It is likely that investors will find real choice to be minimal in many sectors of the market in the future. And that does not bode well for progress, although it may attract the bankers’ profit and loss managers.
I recently looked at some of my old strategy papers. How dated many appear. In 1987 we talked about the need “to transform our securities processing cost centers into commercial businesses.” The focus appears now strangely inward looking and organization centric. The 1997 strategy identified the need to create an integrated custody, administration, cash and treasury operation with major investment in people and technology. This time the strategy appears very product orientated.
It was only in a further assessment, some three years later, that we put enough focus on the customer and the environment. We moved away from targeting pure organic growth to calling for acquisition where it added to client franchise, product and geography. The Bank of Bermuda was the best of those acquisitions. Adding HSBC’s balance sheet and capital market strengths to a service-oriented administrator in the alternative funds market provided real value added.
In some ways my experience typifies much of the industry. In the 1980s custody was emerging from the operational bowels of the major banks and slowly becoming a business. In the 1990s, consolidation started and the focus was on economies of scale through market share and major relationships.
The current decade was, until recently, a period of growth on the back of the biggest asset bubble the world has ever known. Everyone made superb profits until it all stopped about 12 to 18 months ago. For that reason, all those fault lines, which did not matter in the age of plenty, will have to be mended. And fast.