Employ AI-powered predictive models to address ratings variance – Prof. Agbloyor

Employ AI-powered predictive models to address ratings variance
Professor Elikplimi Agbloyor, Senior Lecturer in Finance at the University of Ghana Business School (UGBS)

Amid the ongoing tango between government and Moody’s rating agency over the latter’s recent downgrade of the country’s bonds, a Senior Lecturer in Finance at the University of Ghana Business School (UGBS), Professor Elikplimi Agbloyor, is recommending the application of artificial intelligence and machine learning-powered predictive models by the state to better gauge its drivers of sovereign security ratings.

This, according to him, would better prepare the state for impending rating scores, provide data-backed arguments as well as boost public confidence in the event of counterclaims being made.

We need to let research drive policy. Do we really understand all the drivers that move our ratings? If we were to build machine learning models to produce our ratings when a rating agency comes to tell us this is our rating, we would be better informed and have more clarity,” he told the B&FT.

The state, through the ministry of finance, strongly disagreed with the last downgrade by the New York-based agency, and in a terse statement accused it of an “African bias”.

Prof. Agbloyor however noted that widespread adoption of the aforementioned models would easily prove or dispel such a notion, based on data after accounting for system political risk.

“With the talk of an African bias, is the claim supported by research?” he quizzed. “We could build a machine learning model and include all the African countries in it, and we would then be able to clearly tell. And in the event that it is true, then we would have a strong case because the statistical data would be there to prove it. But if we merely say it, then it is one word against another.”

He added that such an initiative would help propel Ghana toward being the technology hub of the continent.

Rating through-the-cycle 

While current tensions are over bind rating grades, which could have implications for access to additional credit on the international market as well as direct foreign investments, Prof. Agbloyor believes the free market’s pricing of the nation’s bonds should be of greater concern to managers of the economy.

This, he explained, is because rating agencies rate ‘through-the-cycle’; they are unlikely to make sharp changes from one period to the next unless they detect permanent changes in credit quality in the medium-term. The market, on the other hand, gives its verdict on a daily basis.

Following a recently held Non-Deal Roadshow organised by the Ministry of Finance, the market responded favourably as Ghana’s sovereign US$ yield to maturity (YTM) index – as tracked by Cbonds – showed a glimpse of an uptick.

This, however, was short-lived – as the index began to decline scarcely a week after, largely on the back of the country’s recent downgrading by Moody’s credit rating agency; a reflection of an increasingly difficult task that government faces in addressing its entangled liquidity and debt challenges.

Domestic interventions

The Senior Lecturer of Finance at the UGBS said that in addition to measures such as addressing revenue shortfalls and rationalising expenditure, serious steps must be taken to develop capital markets.

This, he said, would minimise the problem of original sin – the situation where emerging market (EM) economies are unable to borrow abroad in their own currencies, thereby adversely affecting their exchange rates and foreign reserves. He also called for the institution of strong domestic rating agencies.

Leave a Reply