There is a school of thought that wholesale and investment banking is immune to true disruption by financial technology – that the protective complexity of these businesses will give the incumbents time to assimilate new technologies. In a recent report, for example, the consulting firm McKinsey found the majority of fintechs aspire to have financial institutions as their customer, rather than to muscle in and transact with the end-client.
It’s an appealing vision, because it strips away some of the hyperbole surrounding fintech and presents it as nothing more than an expansion of the vendor landscape. It’s as if, following a geological catastrophe, evolution took a time-out to give the smarter dinosaurs a chance to re-group.
It is also complacent and wrong. Investment banks’ complexity does not immunise them against disruption (see box, How banking’s protective walls are crumbling).
Admittedly, from the fintech perspective, it can be forbiddingly difficult to understand what investment banks do and how they work, but the many barriers preventing banks from embracing and integrating millennial technology – described in the first article in this series – leave them prey to less-encumbered start-ups and competitors. And it is already possible to identify core investment banking businesses that are permeated by the very same factors through which the banks’ grip on funding and lending is being loosened today.
This article looks at three of these at-risk businesses, starting with an area in which the outflow of revenues from investment banks to challengers already runs into the billions of dollars.
Market-making offers a perfect illustration of the combined effects of regulation and technology. In the many discussions I had as a practitioner regarding what read-across we could make from equities e-trading to foreign exchange and fixed income, my equities colleagues would stress the much tighter regime of regulation around equities and the greater requirements for transparency.
Whether this regulation benefited clients – or simply created something akin to a designed environment for flash crashes – lies beyond the scope of this article. But it certainly created opportunities for trading firms that could amass and deploy quant and tech resources more quickly than banks.
Trading firms’ comparative advantage was agility. Market microstructure was analysed with an arsenal of quant techniques. Specialised hardware such as FPGA and microwave lines was deployed to confer decisive latency advantage. All of this was directed by people who understood their markets, their quant methods and their technology. It represents a level of knowledge fusion that a large bank would take years to match. This combination of market insight, maths, software and hardware to implement a business in the form of algorithms on machines is precisely what characterises millennial technology.
In my own experience, I found that when my team first started researching statistical techniques for the full automation of forex pricing, our core quant group wasn’t interested; it was “the wrong maths” for quants schooled in traditional derivatives pricing. Similarly, getting a trading network set up with appropriate characteristics for automated trading took years, since there were so many other demands on the various infrastructure teams. Progress only came with reorganisations.
During this time, ‘non-traditional participants’, who already dominated equities trading, took top positions in EBS, once the primary interdealer forex network, before rolling on to fixed-income products – for example, firms such as Citadel now feature in the top rankings for credit default swap index trading on Bloomberg’s ALLQ.
This is not just about high-frequency traders. The growth of all-to-all networks is also set to continue. It remains to be seen exactly where and to what extent platforms such as Algomi, which can be regarded as socialisation technologies, come to displace the role of dealers willing to warehouse client risk.
More recently, XTX Markets, a 2015 start-up that aims to de-weaponise electronic markets for the benefit of natural clients, shot from nowhere to fourth place for spot foreign exchange in last year’s Euromoney rankings – traditionally the home ground of the large investment banks. This year, XTX was named Risk magazine’s flow market-maker of 2017.
One of the defining functions of an investment bank is to advise its clients where to invest. The value of research to clients has come under close scrutiny amid growing regulatory pressure to charge for it explicitly. Fund managers routinely complain about being spammed with bank research, and the use of research aggregators has become ubiquitous. When fewer than 7% of analyst recommendations are to sell, there is a natural suspicion of the objectivity of the analysis and only a select few analysts are routinely followed. Over the past four years, the number of analysts at investment banks has fallen by 10%.
Now, millennial technologies promise new ways to analyse markets. Their key capabilities are (1) new data sources; (2) new techniques for analysing this data; (3) large numbers of students trained in these techniques.
One example of this is the analysis of ‘news’. This can be structured news, from for example a Reuters feed, or unstructured data, such as that available from scraping websites and Twitter. Start-ups such as InfoTrie and Cytora claim to mine digital sources for knowledge relevant to finance that is not available over traditional news channels. Others, such as Almax, are working on hybrid services that combine machine-based data analysis of structured news feeds with mathematical models of the way traded instruments respond to changes in the analysed data.
Aside from grazing the data already digitally available, other firms are taking matters a step further by generating the data they want to analyse. For example, Orbital Insight is putting up satellites to track US parking lots as an indicator of retail activity, and storage tanks as an indicator of true crude oil volumes. Along similar lines, at least one promising new start-up is about to launch with the intention of forecasting commodity prices by applying algorithms to proprietary satellite imagery.
It is possible firms such as these will be absorbed one way or another by investment banks and disrupt nothing; the eventual adoption of such technologies by research departments can certainly be expected. However, the more significant truth is that the satellites, the machine learning and the people creating these capabilities come from new quarters. The end-client – be it a fund manager or an individual investor – does not require the credit intermediation nor intellectual validation of a bank to deal directly with these new suppliers.
There is no XTX yet that has stolen solid research revenue from banks. However, the application of better technology and data science is well placed to prove what is so far a matter of surmise: that research can be profitably traded. This will transform its value to clients. Like lending, the rationale for doing this on modern technology over digital channels without a massive legacy cost base is compelling. It offers high disruption potential to an outfit with greater aspirations than to sell its wares to an incumbent.
In looking for the one area of banking that has the air of living on borrowed time, the eye is drawn to the large private banks. The sector has no protective wall of mystery around its purpose and practices: it’s about telling people how best to invest their wealth. The simplicity of mission and the mundanity of the available tools mitigate against entrenched advantage and make it ripe for competition.
Competence in wealth management comes down to optimisation of returns, customer engagement and minimisation of wealth loss through tax and costs. In the first of these factors, the large private banks may enjoy some advantage from organisational proximity to an investment bank. This is not due to the ability to leverage the investment bank’s research, since that is typically firewalled away from the private bank. And we have seen the advantage of proprietary research is itself subject to fintech challenge. Rather, the advantage of having an in-house investment bank lies in its ability to construct a wide range of investment vehicles, such as capital-protected structured notes. However, the universe of investable assets does not lack either depth or variety, and it is hard to see how products from the likes of BlackRock or Fidelity fall short of most people’s needs.
Investors who want a more active and engaging investment experience can already see their options improving. The poster child for disruption in wealth management may be Nutmeg, which is less than six years old, already has around 100,000 clients and has inspired the creation of the new robo-advisory sector. The more significant transformation may come from the wash-through to established investment firms, such as Rathbones and Investec, which have orders of magnitude more assets under management, if they can leverage IT transformation to deliver an effective service to a broader market. Sub-mega banks are in the same hunt: Sweden’s Handelsbanken, for example, like the UK’s Nationwide Building Society, is essaying a ‘bricks-and-clicks’ strategy, combining a digital platform with a traditional branch network.
Further tech-driven incursions into private banking territory come from firms such as Saxo Bank, which has long offered platforms for clients to trade a wide variety of assets. To round out such a platform for someone who is managing a big chunk of their overall wealth rather than simply day trading, Saxo, and no doubt others, have already begun to add the ability to trade investment portfolios too. It would be wrong to dismiss this as merely gamifying the serious business of providing for retirement. The benefits of giving the customer a more immersive, digital interaction with their own wealth management can be considerable. While today’s retail platforms do tend to thrive amid the voodoo of technical trading, the more far-sighted of the suppliers are repositioning themselves towards a more balanced and investor-orientated demographic.
For a Saxo Bank, for example, to channel development effort away from charting and towards portfolio diversification tools, and to target, say, 30-to-50-year-old female investors rather than 25-year-old male day traders is essentially just a board decision, and one the bank has probably already taken. On the other hand, for a large private bank to invent for itself a compelling digital presence is an overwhelming proposition, and probably one at which it has already failed.
These are matters of technological reality rather than strategic choice. Once you have a coherent digital platform you can choose, say, your robo-advisory tooling to suit your purpose. A mega-bank that has been actively building proprietary systems for decades is unlikely to find itself in this position and will consequently struggle to promote even the best of visions beyond white-paper format.
The various beneficial services a private bank can offer a client – market insight, range of investments, portfolio analysis – can be imagined as the outer layers of a Russian doll that are in turn being opened by technology innovation. Insofar as tax evasion was ever the secret inner doll, it has gone too. At the centre of what may look an increasingly empty brand, what stubbornly remains is the cost base. For the type of private bank that is part of a larger, universal bank, this is an irremovable, dense core generated by a franchise that is at once manual yet still bears the costs of decades of legacy technology. For a digital competitor, those core costs are dramatically lighter, and over the years in which the client’s wealth is invested they will impose a far lower drag on performance.
Challengers and incumbents
Regulatory disaggregation effects and millennial technologies combine to offer firms such as Saxo Bank and XTX and an emerging generation of Research innovators new opportunities to take business from the investment banks. It doesn’t stop there. The context in which investment banks transact is itself changing. A dozen funds are already actively trading using machine-learning algorithms. The decline of the human trader – responding to inputs from a bank of screens on a trading floor – will continue. Threats to private banks will also come from asset managers that offer direct-to-client products. Legal & General, for example, has started doing this already; so far it accounts for under 5% of its investment management business, but it can be expected to grow, and for more firms to offer it.
Although retail banking is not our subject, it is notable that companies such as Virgin and Tesco have been able to offer banking services when it has suited their purposes to do so. In ClearBank, we have recently seen the arrival of the first new clearing bank in the UK for about 200 years; naturally it’s cloud-based. What impediment is there for Amazon, say, to move into foreign exchange or Paypal to offer settlement services?
The investment banks that find the narrow path through this will not necessarily be those that invest in the most fintechs, re-label their automation projects as “robotics” or have the coolest labs. The surviving firms will be those quickest to recognise the disruptions already underway, to develop the best partnership networks in the new economy and to refactor their operations so the greatest possible proportion of expenditure is directed coherently towards the future rather than the past.
How banking’s protective walls are crumbling
The defining premise of investment banking has been an efficiency of intellectual scale: by accumulating large pools of the smartest analysts, traders, lawyers and technologists, investment banks have been able to devise esoteric services in any market that offers a sufficient marginal return. The enmeshed nature of the resource pool historically made it almost impossible for a challenger to break in and locate a niche in which they could steal territory.
In recent years that premise has been doubly undermined by generational changes in both regulation and technology. Post-crisis banking reform has sought to correct flaws in the valuation of risky businesses and to deliberately challenge economies of scale in wholesale banking when they present in the form of ‘bundling’. Many investment bank trading businesses are now either outlawed (prop trading), unprofitable (most exotics lines), or fundamentally re-structured (much of flow fixed income). Credit and lending businesses require far more risk discipline and are thus curtailed in scope.
In many markets, the burden of remaining a principal market-maker has significantly increased. Research, which was once regarded as a primary driver of sales, must now be paid for by consumers at a point when belief in its value is at a low, with many investors moving funds away from active management. Far more business now has to be centrally cleared.
More than ever, suspicion hangs over any form of implied cross-charging, in which one service to a client is explicitly or implicitly subsidised by a different revenue source.
These changes are eroding economies of intellectual scale for the large banks, and it is becoming easier for organisations at the periphery to bite away at banking services in niches of their choosing.
The coincident arrival of millennial technologies lands the second punch: these enable new entrants to create breakthrough platforms. Unlike regulatory changes that were framed with the express aim of disrupting certain banking practices, the so-called disruptive technologies are neutral in intent. However, we saw in the last article that, while everyone sees the potential of millennial technologies, the path to their actual adoption is exceedingly narrow and there has been little progress along it. For example, one European bank recently issued a 25-page white-paper on the uses of artificial intelligence (AI) in banking; the only instances cited of actual use of AI at the bank are a support tool for low-level tasks such as password resets and a beta implementation of Amazon’s Alexa to answer client questions. It’s hard to be agile when you have 60,000 staff and 5,000 systems that enshrine pre-millennial operational practices.
A version of this article was published by Risk in May 2017.