A Hitchhiker’s Guide to AI – a different financial world
The reason why the auguring of Artificial Intelligence techniques makes so many people shiver with unease, at the national economy level, is not at all original - it is because the onslaught of mechanization of their core value proposition has so far eluded them. From the dawn of the first industrial revolution to the present day, much of the prevailing wisdom regarding economic advancement has been premised on the seemingly inevitable step-by-step transition from agriculture to industry to services. The world’s most economically advanced societies, from where this wisdom usually percolates, are all predominantly service-oriented economies. The greatest proportion of value added in these countries, and consequently, the greater share of employment of people is concentrated in the services sector.
Due to the advances in mechanization achieved in the period since the First Industrial Revolution, both agriculture and industry have been heavily mechanized. As a result, in these sectors, the utility of human resources continuously fell. This resulted in the shift in the employment of human resources away from agriculture and towards manufacturing industry first; and when the latter too started to get heavily mechanized, from heavy industry to services. But the learning element which transmitted information back from outcomes to influence further decision-making was still unmechanized. This is where the role of professionals and managers stood highlighted in importance. The prevailing theory became that “it is not the manufacture that is important; it is the decision to make that is”. The effects of the improvements to efficiency and quality, in the form of being continuously forced to adapt or perish, were to be faced not by the decision-makers and the learners but by those engaged in execution.
The real estate sector serves as a barometer for the health of the economy in general. For investors and managers engaged in the real estate sector, the various risks and uncertainties involved in the business can often be nerve-wracking. Entire fortunes can be wiped out overnight from being stuck in the wrong portfolio position at a time of significant market adjustment. Heavy repairs and maintenance costs can often wipe out significant chunks off the profitability of a venture.
In all these areas, the deployment of artificial intelligence technologies can derive significant efficiencies for the venture. From providing precise information regarding market price movements in the real estate sector on a real-time basis; to providing accurate information regarding repairs and maintenance costs and property utility expenses; to providing accurate information regarding customer payments and shortfalls enabling early intervention; to quickly analysing vast sets of data from a large set to identify what is “current” and “hot” in real estate management techniques from a customer perspective, artificial intelligence techniques can provide significant efficiencies for participants in the real estate sector.
For real estate managers, the deployment of artificial intelligence techniques can significantly reduce management overhead costs and enable them to be more proactive to market changes. For real estate investors, the deployment of artificial intelligence technologies can significantly reduce the “learning” time taken to know about the outcomes on existing investments as well as prospects for new opportunities to invest; and can help predict trends early with a higher degree of fidelity, thereby enabling more efficient real estate portfolio management.
For the “user” of real estate - the occupant - the deployment of artificial intelligence technologies can improve the efficiency of their stay. The use of AI in utilities management has already been mentioned. The deployment of artificial intelligence technologies on similar lines in other areas can help monitor quality-of-life indicators such as air quality, room temperature, particulate count, noise levels etc on an ongoing basis. This is already a reality in many dwellings in Dubai, Abu Dhabi and the other emirates.
Just like it has become possible with the deployment of artificial intelligence technologies to identify impending natural disasters early, isolate the area of risk and take precautionary action with much greater chances of success; the use case also appears to be quite valid in the case of impending disasters in the man-made financial markets and the capital allocation industry. Early warning of systemic customer events such as missed payments and check runs through enhanced data collection techniques can alert bankers as to the overheating of the credit markets.
Fraud detection can also be enhanced to levels of precision through the use of transaction monitoring and pattern recognition programs. With the dispersion of data collection and recording devices to larger portions of the population through ever more affordable devices, the practical establishment of credibility stands to get even more affordable and accessible. While the level of credit risk for the recently unbanked may still be high, it will still be lower than the credit risk for those who are entirely unbanked.
This has important ramifications in countries where large segments of the population are still not economically advanced and suffer from structural deficiencies, such as India. In only the last half-decade, the number of unique bank accounts opened, predominantly digitally, has been greater than the number of bank accounts opened in the entirety of the history of the country since 1950.
The opening of bank accounts and the digital recording of transactions enables a much greater range of potential lenders (who could deploy computer programs in place of human lending reviewers) to review financial records at a much faster pace and consider lending, faster and with greater precision. For a country where sluggishness is a byword for reality, the potential impact of the changes brought about by the deployment of artificial intelligence in capital allocation is seismic and revolutionary.
Much like in commercial banking, where lending decisions stand to derive a lot of benefit from the deployment of artificial intelligence techniques, investment banking and securities trading stand to benefit immensely through the use of these techniques as well. Already, there is a niche industry of whizz kids known as “quants” whose task primarily involves the use of advanced mathematical formulae and computer programming to sift through petabytes of commercial information to sniff out the best investments as per the mandate given to the manager.
As a Chartered Accountant, and an auditor of financial statements, this hits home on so many fronts. The formation of audit strategy and the outlining of audit procedures has always been a highly cerebral undertaking. It has involved deliberation over the key risk areas, the key matters over which audit comfort is required through testing and confirmation, and the most time- and cost-efficient ways of obtaining the said audit comfort.
As the identification of key risks and the subsequent formation of audit strategy are the outcomes of an iterative learning process containing multiple data points from a variety of sources (which is why experience is said to count for something!), this is a default case for the realization of efficiencies through the deployment of artificial intelligence techniques.
While experience will still serve to differentiate the wheat from the chaff, the scope for differentiation will narrow. Already, digital audit files are pre-populating risk assessments, lead schedules and even working papers based on the input of a few key preliminary metrics, vastly increasing the speed of the audit process.
The data collection element of the financial audit process has already been upended through the use of advanced data sensors and recorders (think drones for stock taking!). The quality and precision of data acquired in the course of forming an audit opinion stands to be enhanced to a great degree through the deployment of artificial intelligence techniques.
Where auditors can differentiate their value proposition (which is their desired outcome), is in informing better value judgements to the users of their output. While the Finance Function in larger organizations usually have the wherewithal to extract value from the data-crunching exercise themselves, for smaller organizations, the lines between being an independent, external auditor and an advisor to management start to get a bit fuzzy.
It is very important for external auditors to continue to avoid the threat of being seen to act in place of management (the “Management Threat”) either in appearance or in fact; however, within the strict bounds of professional values, it behoves auditors to take care to see that their work enables management to make better-quality decisions, ownership of which, it has to be made clear, rests with the latter only. The risk, otherwise, is that the financial audit profession will fast become irrelevant, and that too in the not-so-distant future, as compliance with regulations becomes increasingly automated.
That said, the utility of an external reviewer of an organization’s financial affairs and supporting systems is unlikely to subside anytime soon. The utility of an external auditor may change from strictly “auditor” to more of an “assessor”, a “sounding board” or even “agony uncle/aunt”, all of the latter of which are paying professions by themselves! The psychiatric skills of the auditor may need a big re-polishing in the years ahead. Given that the word “audit” is Latin for “to hear”; maybe it is indeed high time that he/she did what it says on the tin going forward!
That is what stands to be changed.
The need for members of the population engaged in these sectors to re-skill and re-tool themselves for the changed economic circumstances is only made more urgent by the rapidity with which these changes are taking place. The world will look very different in 2030.