The cost of the innovation revolution

Innovation should be focused on (expensive) cutting-edge technology rather than products, according to John Gubert.

I recently suggested, in an interview, that innovation in the future must be more technology than product based.  The reality is that the environment militates for a major reengineering of the industry’s technology. The drivers are simple. A regulatory focus on real-time. Buy-side increasing demand for digitisation. A realisation in an age of robotics, distributed ledger technology and ever more powerful processing, that legacy and even recent applications are fast becoming redundant. And the imperative for cyber security in a far more dangerous world than even a decade ago.

The cost of this is going to exceed the wallets of all bar the mega players. And so, alongside this technology change, I believe there will be need for a reform of market infrastructures and a spate of corporate restructurings. At the infrastructure level, it would be logical for a greater amalgamation of trading platforms as liquidity declines in the face of the capital cost of proprietary trading, causing buyers and regulators to question the effectiveness of the price formation process. Downstream, we need to see true CCP interoperability with hopefully same day settlement eventually moving their risk management role clearly into the futures space. CSDs are an anachronism and, although local point of contact is logical, national markets, for cost reasons alone, need to move onto a common processing utility, perhaps using one of the ICSDs. And the regulators need to impose fundamental market changes including, as examples, digitisation of new issue information, mandatory same day trade matching and an extension of market settlement rules to the collective investment sector. Corporate players need to look at collaborative solutions, perhaps with technology providers, or full mergers to gain the needed scale, if this proves too problematical.

Regulators imposed CLS on the foreign exchange markets because the open position risked market stability. The same issue, on a similar scale, arises today in the securities markets as we consider the open position and the operational risks remaining despite decades of G30 and other mandates. Regulatory reporting is in close to real time and major clients have similar demands. The move to T+2 was painful for some but it also led to a reduction in the faxed contract note syndrome between fund managers and custodians, thus weakening substantially one of the great barriers to real time transaction processing. The bank cash markets, even at the retail end, have moved to same day or even same minute cross enterprise settlement. It is inevitable that securities markets will do likewise. Most custodian platforms operate in real time but there remain multiple friction points primarily due to two flaws in the market IT architecture. The first is the lack of real digitisation and the second is cloned and conflicting ledgers. Real digitisation operates when markets use uniform message and business standards and users, either buy side, sell side or intermediary, can switch off applications in house and depend on third parties for data. Where the data has proprietary as distinct from market generic characteristics, it exists when users can capture the external information they need and import it into their platforms without adaptation or restructure.

Robotics are used to a limited extent across the industry. In time they will eliminate the vast bulk of back office roles. As anyone who has worked on outsourcing must be aware, well over 90% of all transactions can be executed using a quite simple rule book. The market remains inefficient because it focuses on the 10 per cent. Paradoxically, that 10 per cent is a huge component of the cost base. And it does not follow a normal rule based process for multiple reasons, the most important of which is the inherent inefficiency of a party during a transaction life cycle. The processing breaks resulting are all avoidable. They may cover the erroneous client instruction which should be tackled with hefty repair fees as the source of material cost by comparison to an STP transaction.  One can eliminate the failed settlement due to stock out on loan or securities tied up as collateral by real time monitoring. Counterparty fails are being attacked through the settlement discipline regimes although penalties need to be toughened. UK due date settlement pre-CREST was around 80% and it is now above 99%. Any market can get to equivalent performance as long as the cost of delinquency is tackled resolutely.

We also need to upgrade our IT platforms for the markets are changing. Often people talk of millennials and their thirst for automated and real time, tablet or iPhone compatible solutions. They should recognise that tech savvy can also be applied to previous generations. Personally I have been aware of mainframes for all my working life, although they were quite bulky in the early days. And I have been a PC user for over 30 years.  Communication is digital and instant and that demand from the individual will feed back into the entire securities industry value chain, especially as wealth is likely in future to be managed like a multi-product bank account whilst pensions cease to be an employer benefit but an individual’s personal tax efficient savings plan.

And then we have the big daddy of them all in terms of cost and risk.  This is the burgeoning cost of cyber security in an age of state and criminal inspired hacking for malevolent or mischievous purposes. The cost of building the right firewalls in an age of multiple gateways to platforms, both in house and external, is enormous. The risk of a cyber breach can be costly in terms of potential direct financial loss and catastrophic in terms of loss of client confidence. Legacy platforms are often the most vulnerable to cyber-attack. The effort needed to police ones IT platforms and the technical skillset required are often prohibitive mainly through a skills deficit although finance is always also an issue.

So what will the true innovator do? First they need to consider the accuracy of my comments on cost in respect of their organisation as that will dictate their wallet using standard cost recovery and return on investment criteria. Then, and I assume the majority of firms, they will find they need either talk of a ten to fifteen year programme that will be out of date before final completion or think outside of the box. And the main new solution I can see would be a consortium of buyers creating a joint processing venture with a group of technology vendors for it is unlikely that any single vendor will have the competence and reach required to offer a solution on a stand-alone basis. And, at the same time, the industry should use its buying power to force through the infrastructure change that will take out complexity before they build it, yet again, into their IT and operating models!