TSHILIDZI MARWALA: Can AI help to bring morality to exchange rates?

The evolution of the international monetary system has resulted in the erosion of the terms of engagement for developing economies, which has deepened marginalisation and underdevelopment — the variables that make domestic monetary policy management an uphill undertaking for the Global South.

The latter-day infusion of artificial intelligence (AI) infrastructure in the foreign exchange market buoyed sector analysts, but this did not tackle the host of impediments that abound and is unlikely to ever aid in realising equity in the market, resulting in the augmentation of poverty and the problems that ensue in developing economies.

Interestingly, in the beginning of it all global monetary policy was equitable and transparent. Gold was a medium of exchange and its simplicity encouraged the emergence of trade and formative globalisation of the time. This approach to international finance was passed down until it received universal recognition in the 1870s, with the joining of Western Europe and the latecomer, the US. Notably, this system demonstrated the philosophical claims of Ricardo’s theory of comparative advantage: those with surplus gold reserves and an export product advantage would succeed in global economic competition, regardless of the size of the economy and subjective perceptions and insinuations.

As with anything in the finance realm, instances of malfeasance came to define this era, especially during the interwar years. In theory, the use of the gold standard would provide ample opportunity for supply and demand of a country’s exports/imports to cause objective movement in the exchange rate in respect of an equilibrium value. In reality, however, speculators manipulated the markets by selling weaker currencies short to such an extent that the currencies of developing economies were no longer reflecting true market fundamentals.

As an observer from a country whose currency is continuously targeted by rent seekers and rogue traders, I believe the Bretton Woods conference missed an opportunity to right the wrongs of poor governance in the international monetary system. Granted, the leading interlocutors, especially John Maynard Keynes, were preoccupied with promoting demand-side stimulus to engineer post-war economic reconstruction but developing economies were still left to the devices of speculators and rogue market players, denting prospects for any semblance of post-war economic prosperity.

The devastating consequences of this great omission have been that post-independence, African economies have largely been constrained by weakening currencies, leading to heightened surges in prices and increasing poor sentiment. Though internal structural issues play a role in these outcomes, more often than not exchange rates do not reflect domestic economic fundamentals. There have been some exceptions, notably Ghana (Cedi) until recently and SA early on after the democratic elections (1994-2000).

In SA at the time Reserve Bank governor Chris Stals managed the exchange rate through a peg. In the late 1990s Stals spent countless billions of US dollars defending the rand from speculators, but this was not an effective firepower, nor was it sustainable, a result of decades of market manipulation of the rand by banks and hedge funds. 

These instances of immorality and poor governance within the sector has led to University of Chicago finance professor Luigi Zingales declaring that finance is a rent-seeking service sector. In Dark Towers, New York Times finance editor David Enrich wrote compellingly about market manipulations by Deutsche Bank, detailing how self-interest — in the form of bonuses and lakeside holiday homes — by market players is the root cause of these evils.

Evidently, the age of AI is firmly with us, and with the surge in technologies and digital innovations foreign exchange trades are now increasingly being facilitated by high frequency trading. In this environment the human hand is said to be completely eliminated; instead technologies and algorithms trade automatically as invisible hands.

However, theory is one thing, while the reality on the ground is often something else. Indeed, the problem of poor governance in the market remains, much as it did throughout the era of the gold standard. Exchange rates do not entirely reflect economic fundamentals even with the advent of AI.

Simply stated, demand and supply still does not exclusively determine asset prices, realising ineffective market performance, systematic risks, distorted market competition and untold suffering to the ordinary masses. The shameful alleged manipulation of the currencies by a cartel of both international and domestic banks reiterates the need to improve governance structures with the current AI world.

The major requirement of the current age then becomes the need for morality within AI. Thus, we should encourage theorists of AI hue, including market participants, to grapple and debate a litany of profound questions in respect of the applications of AI in service areas. This is a fundamental herald in the march towards a more just society.  

• Prof Marwala is a vice-chancellor the University of Johannesburg. This is an edited version of his foreword on Rabelani Dagada’s book ‘Digital Commerce Governance in the Era of 4IR in SA’.