Disruption and the New Financial World Order (II)

How Advances in IT and AI will undermine--and eventually displace--the traditional investment management model

In the first part of this series, we highlighted the far-reaching impact of the investments industry's historical development on its normative features. There, we argued that a confluence of recent technological developments create a compelling alternative evolutionary path that threatens the prevailing institutional model in investment management. We extend that discussion here by delving more specifically into the mechanics of the impending changes. This review is not exhaustive, but it tries to cover some of the fundamental areas where emerging developments will, taken together, catalyse major deviations from the current industry architecture.

Redefining the tools of the trade

As discussed in part I, the emergence of the MPT as the dominant ontological framework in finance permeates all facets of its service delivery model, from the definition of risk to asset classification, to performance evaluation. It is inevitable, therefore, that any emerging paradigm will revisit these issues for fresh perspectives.

For instance, what exactly defines an asset class and how does it fit in to delivering solutions to clients? Formalised by the MPT, the industry's historical development encouraged asset classification centered on security-type/industry/geographical schemes. However, as demonstrated by its frequent failure to deliver the MPT-desired outcome precisely when most needed (i.e. as correlations converge on 1 during market crises), this approach may represent a mis-specification from the perspective of end-client goals.

One alternative view that emerges from our discussion is that, instead of acting as inputs to a generalised optimisation problem (as in MPT), asset classes can be interpreted as feature-vehicles to be configured to deliver customised outcomes for clients. Furthermore, under this interpretation, their categorisation may differ substantially from the current MPT-driven model, tending rather to a much broader variety of fluid, latent factors and pre-configured design options (e.g. using derivatives) that are driven more by end-goals in an individualised framework (e.g. target retirement income). We already see glimpses of this trend with the rapid proliferation and extension of the smart-beta and outcome-oriented (e.g. target date) schemes even under the traditional framework.

However, as focus shifts away from outperforming standard market indices--artifact of the industry's earlier trend towards institutionalisation--to securing life-cycle/lifestyle outcomes, questions arise about alternative performance evaluation paradigms. In some ways, this may not matter. In the private wealth space that has long catered to high net-worth individuals, portfolio performance has always taken a back seat to broad-based utility the service provides the client, and the growth of Ant Financial, which essentially leverages AI to cascade similar outcome-focused options to a much broader market, suggests that issues may not play out that much differently in the mass market.

As discussed in part I, Ant's experience and the regulatory response across the globe to the nascent era does suggests that while different industry architectures are possible, the platform approach--where technology companies become a de-facto client-facing interface--appears likely to dominate. This in turn suggest that as onus shifts to the client facing platforms to deliver on both client experience and ever more variegated and individualised outcomes under competitive forces, financial intermediaries at will increasingly focus on efficiently harvesting features from financial markets that more readily fit into this paradigm. Under this paradigm, lifecycle related issues like, longevity, inflation and downside protection for incomes--as opposed to index outperformance--then likely to emerge as the preponderant value propositions. This will likely reverberate upstream, potentially transforming the roles and organisation of operators up the investment management value chain and blurring the boundaries between financial services.

This consideration is also of significant strategic importance for financial service providers as they (re)position themselves along various points in the value chain in a changing market, with technologies like block-chain likely to drive the reconfiguration and consolidation of roles for market operators (witness the decimation of trading desks across the industry). Notably, with the various current and emerging portfolio optimization schemes becoming increasingly commoditised, it remains to be seen how the regulatory hurdles which currently provides some measure of protection for financial incumbents will hold-up against the onslaught of the more distinctive value provided by their client-facing technology rivals. It is clear that as the technology giants become better integrated with the financial system and the regulatory landscape, their threat to big banks and money managers will intensify.

A new economics

Like many economic concepts developed in the last century, MPT's enduring appeal partly stems from fulfilling a need for important constituencies. For academia, it served as an anchor for rigorous-looking postulations in the absence of solid empirical foundations. For practitioners, policy-makers and regulators it offered an operational framework; and for advisors it supplied a tidy sales pitch. This utility has been hard for the industry to forego and may have contributed to the stagnation of financial theory insofar as real-life applicability is concerned. It is also true that, with notable exceptions, much of the thinking in this area over the past half century has gone towards buttressing the extant framework rather than seeking fresh perspectives on what are undoubtedly complex phenomena. Thus, while its reliance on neoclassical economic concepts whose untenable simplifying assumptions (e.g. perfect information, rational agents and seamless competition) may have sufficed in the prior era, it is definitely ripe for an update in the information age.

Incidentally the social sciences (to which economics definitely belongs despite its periodic outbursts of "physics envy") is currently undergoing a subterranean revolution. Fueled by the explosion of data, algorithms and computational power, such multidisciplinary approaches as graph/network theory, are opening new and promising avenues for studying of socio-economic phenomena that will significantly enrich the field. These approaches rely, not on pre-defined or prescribed theoretical abstractions, but on empirical, data-driven methods for understanding agent interactions and their implications within a specific framework. In addition to being better adapted to the emerging computational paradigm, data driven methodologies have the decisive advantage of relying more on analytic approaches which reduces or obviates the need for assumptions that often turn the scientific method on its head. Moreover, whatever these new data-intensive approaches may lack in precision and elegance, they need not sacrifice rigour for efficacy--or vice versa.

This change definitely plays into hand of tech-industry and other challengers outside the MPT bubble as it coheres better with their operational model, giving them a head-start on the financial incumbents in melding into the emerging financial service landscape.

Reinterpreting risk

One of the manifestations of the foregoing issue with economic theory in finance is its conceptualisaton of (market) risk. Its identification of risk with volatility or statistical variance has been a cornerstone of the discipline, at least since MPT, though this specification--even by the admission of its originator--was mostly for convenience. The mean-variance formulation, along with the broader parametric approach it implies, has many widely acknowledged problems but the most important may be that it is a proxy which by its very nature relies on somewhat simplistic assumptions and is typically assessed unconditionally (i.e. without regard to underlying context), besides being unobservable in real time. Despite many important refinements to its specification and measurement, the current state of the practice falls well short of a satisfactory resolution of the conceptual problems underlying risk metrics.

One of the definitions of risk I consider most useful is attributed to Warren Buffet, who reportedly describes it as coming from not knowing what one is doing, mainly because it (correctly, in my view) identifies risk as an epistemological issue which should be recast in informational terms. This opens the door to potentially richer, more fruitful, data-driven formulations where, to the extent we can resolve unknowns in any phenomena by a process of reductive conditioning (which is a staple of the rapidly developing fields of statistical data analysis, machine learning and graph theory), its risk component can be apprehended in a more subjective and contextual setting, and thus more directly and comprehensively addressed. With wider adoption, this approach provides a window to reorienting financial practice towards customisation and individualisation, and away from generalised macro-frameworks like the MPT, again playing to the advantage of the tech-industry challengers.

 Geopolitical undercurrents

The feature-oriented approach to financial services we have been describing clearly presents an immense opportunity, which has not escaped the attention of the US tech giants. However, reflecting the very real impact of structural and socio-political factors in the evolution of any major industry, these have had to contend with both political and regulatory obstacles. The case of Facebook is particularly interesting because not only is its product portfolio akin to the more successful Chinese tech leaders, the company was angling to add a particularly potent weapon to an already fearsome arsenal.

In proposing a crypto-currency launch as it recently did, it brought technology's disruptive potential to the existing financial order into full relief. It is hard to fully comprehend the impact of a universally transmissible, instantly available means of exchange on global finance and commerce, but for good measure, it shrewdly targeted a so-called stablecoin, i.e. one tied to tangible assets (in this case a basket of global currencies), preempting any potentially deleterious problems with trust and acceptance by sidestepping some of the biggest challenges facing crypto-currencies. The implications of this move deserves its own essay, but suffice it to note here that its capacity to fully globalise financial service platforms, which till date have essentially remained geographically bound despite massive increases in cross-border financial flows in recent decades, elevates our reflections on this issue to a whole new level.

If politicians and regulators, globally, had been dawdling on a coordinated response to the emerging techno-financial order, they were certainly awoken by Facebook's Libra plans and reacted with such near-unified indignation that Facebook has apparently put the idea back in the cooler while its putative partners, possibly enlisted to make the idea less threatening to the existing order, have quietly slunk away. However, vis-a-vis how the similar issue has played out in China, it bears evaluating if this was really the appropriate response from western financial overseers, even after taking into consideration the different cultural, socio-political and institutional context of these events, and the valid concerns raised about the company in particular, and the idea in general.

Would more sober considerations, shorn of the surrounding socio-political hysteria, perhaps have called for a different approach in the light of the broader dynamics which, after all, involve competing players and jurisdictions with like ambitions in what may be a make-or-break moment in global financial history? News emerging recently that the People's Bank of China is launching its own cryptocurrency which closely resembles Facebook's Libra certainly suggests so, nor are the PBoC's intentions obscure. Incidentally, this geopolitical dimension imperils the regulatory protection currently enjoyed by financial services incumbents as hinted at earlier, because it dramatically alters incentives for western financial regulators, likely forcing their hand in any dilemma around protecting the status quo.

Indeed it would be a mistake to consider the Faceook Libra matter closed. The stakes are simply too high. With players like China increasingly acknowledging their nascent tech prowess as a viable path to challenging the existing US-led financial hegemony, matters will likely to continue to evolve in this arena, coming to a head sooner than later. At any rate, for emerging and frontier markets (like Nigeria) and their investment management industries, these developments represent an existential threat as we shall explore in future essays.

Let us conclude with a few takeaways:

• Technology is on the verge of catalysing a major re-invention and re-organisation of the global financial system, and the investment management industry, which has largely been shaped by the MPT over the last half-century, is one of the key segments staring fundamental changes in the face.
• There is growing evidence that such factors as the compelling cost-benefit dynamic that drove the passive investing revolution is also at play in the emerging automated investing paradigm.
• The technology/data behemoths will probably be at the bow and stern of these changes, being best placed to capitalise on them.
• Being a conservative industry, there is substantial inbuilt inertia confronting would-be revolutionaries in finance, arising from its historical evolution and the manner this intertwines with vested institutional, governmental and regulatory interests.
• However, the build-up of the trend towards regime change bears all the hallmarks of the "gradually, then suddenly" archetype, and with ready availability of data and emerging AI to grapple with it, we just may have crossed the suddenly threshold.
• As with all major shifts there will be winners, losers and very big losers, and this calls for strategic awareness, astuteness and readiness at all levels, institutional, regulatory and governmental.

The issues raised in this essay are fundamental and far-reaching and there is a lot to unpack. We will revisit many in subsequent essays. Please share your thoughts in the comments to help guide the areas we will cover in the series. Thank you!