Paolo Sironi is the global Fintech Thought Leader of IBM Watson Financial Services and he just published his fourth and latest book “Financial Market Transparency”, launched in Davos during World Economic Forum. After his acclaimed “Fintech Innovation”, this latest essay crowns ten years of professional research at the crossroads with finance, economics, technology, science and philosophy. Paolo explores the biological underpinnings of financial markets and re-writes economic theory to explain investment and economic decision-making in the presence of fundamental uncertainty. This is a breakthrough approach to bolster the global economy’s immune system in today’s volatile times. According to Paolo, transparency is a foundational regulatory requirement which must be reflected in business model design to implement successful digital banking architectures. Technology and transparency must go hand in hand to transform “digital change” into “social progress” (environmental impact, governance, social responsibility). We are glad to host Paolo today with some questions:


In your recent book titled “Financial Market Transparency “ you navigate the reader throughout the abyss of markets transparency. Do you think that the use of technology via Fintech or Regtech applications will eventually lead us to a better monitored and self regulated/operated market?

Since the Global Financial Crisis, it has become evident that something was broken in the main mechanisms that regulate or attempt to self-regulate financial services. Alan Greenspan, former Chairman of the FED, declared in a public hearing in front of US congress after the default of Lehman Brothers that “he found a flaw in the model, because investors and institutions are not rational and markets not efficient to avoid collapse”. 2009 was also the year of Satoshi Nakamoto’s blockchain paper, in the bold not necessarily complete attempt to re-write incentives and interaction of the main economic agents which regulate modern economies.

I truly believe that a solution to today’s banking conundrum requires to think anew not only about technology but also economic theory, which reflects the way we believe the world works. Transparency about incentives, costs and consequences of economic action is a regulatory requirement underpinning the European MiFID II. Most of all, it is a powerful mechanism to create a consequentialist ethic that generates value for banking clients and global ecosystems. Transparency helps to overcome current limitations of economic theory and market functioning, in particular it will help to mutualise the “systemic put” which is the central bank’s system guarantee that got largely amortized since the GFC.

However, making banks more transparent and clients more responsible would not be sufficient without helping the broader public to deal with investment related issues and cunning financial decision. Here comes Fintech role, to simplify access to banking and support people’s understanding of financial services with artificial intelligence. I would say with “augmented intelligence”.

How close do you think we are in seeing the machines, smart applications and systems, replace human traders?

A large part of trading has already been automated well before the appearance of trusted artificial intelligence. However, I believe that this trend will not take over the whole investment management industry. In particular, the crucial activities based on investment relationships in retail and private banking, as well as many insurance businesses, cannot be easily replaced by automation or artificial intelligence.

This is due to biological needs of individuals and cognitive biases which make relationships very sticky against digital convenience. Insurance is sold more than bought, so investing, because households have cognitive biases when it comes to money and don’t behave like in the consuming world. A fancy user experience is not enough to change this mechanism. Digital brings many benefits to streamline the processes in financial services, but front office disintermediation could easily create financial exclusion in the western world because many households operate in a “push” modality; only a few self-directed ones would know how to “pull” financial products.

This is the reason why the growth of first mover Robo-Advisors was initially very promising but then faltered, while firms like Vanguard and Charles Schwab can still grow fast on digital due to their capability to optimise marketing costs on their existing client base. Being “pull” means going on digital with a purpose, like looking for a specific product on Amazon. None of my friends has ever invited me to take a look at what is happening on Amazon, but clearly many readers have been searching for my literature on line. However, very few households would google for the next UCITS compliant investment fund: the majority would ask a friend, a bank or an advisor about their recommendations.

Clearly, this is a problem of financial knowledge and education which is the core of the asymmetry of information that has benefitted banks’ balance sheets and created an industry which is largely configured as a distribution channel of investment and insurance products, not necessarily advice. Clients are “sold” financial products, in essence, but the meta-truth is that many of them “buy” a fiduciary conversation, a comfort zone to make financial decisions.

This is why FinTech hybrid models of cooperation (B2B and B2B2C), which merge humans and technology, seem to be a more solid FinTech trend than many initially thought. However, this doesn’t mean that ultimately banks cannot go fully digital because things are starting to shift due to artificial intelligence. When A.I. will become truly conversational, digital might ultimately go from PULL to PUSH and the digital touchpoint could start dominating the human relationship. Voice will be new marketing.


And that happens; we have seen that some of the trading algorithms or machine learning applications have an inability to predict so far irregularities or the extremes when it comes to the trading activity. Do you think sooner or later technology will be able to give us a solution that could be adapted to behavioural finance as well, predicting and addressing major fluctuations not only on the figures but also on the phycology of the traders, institutions, markets?

The core of my fourth and latest book “Financial Market Transparency” relates to how humans make financial decisions in the presence of fundamental uncertainty over the irreversible time. It is a biological approach because it explains why human brain tends to underestimate long tail probabilities which is not simply psychological but deeply biological. It therefore explains why behavioural finance “nudges” might be incomplete to gain back agents’ rationality which is usually a fundamental assumption of most economic and trading theories.

Focusing on portfolio optimisation is not appropriate because it is based on the assumption that price dynamics have physical properties. “Physics envy can be hazardous to your wealth”, as Prof. Andrew Lo admirably noted. No single minute of financial markets can be truly replicated because markets are biological: if humanity fails there will be no prices any longer. Moreover, uncertainty is typically seen as exogenous to the decision-making framework and we tend to forget the Black Swan to facilitate a more “reassuring” risk-taking appraisal.

Same happens to many mathematical models which are largely based on the assumption that available data is sufficient to calibrate the algorithms. Deep learning might fall into a similar trap because uncertainty is not part of the big-data dataset needed for training the algorithms. “Financial Market Transparency” provides some needed reasoning to keep modeling and financial thinking open instead of closing the reference framework and face collapse, as the GFC demonstrated.

Therefore, I believe it is not just a matter of writing better algorithms but we need to revise our mindset about how markets work, thus the way we believe the world works. Responding to Greenspan comes first, building better trading algorithms comes consequentially afterwards.