Private credit risks highlighted amid AI developments

It seems that central banks and financial regulators have been warning about the risks inherent in private credit for some time now; much of it due to its opaque nature. The sector has grown rapidly and its exposure to the AI boom has increased sharply, even if it does not appear to be acute. Blue Owl Capital has stirred the debate by preventing investor access to one of its funds, exacerbating a fall in its stock price, which has been going on for the past year. And it is easy to see why some make comparisons between what might be about to unfold now and the early throes of the GFC in 2008.

The boom today is in AI-related lending; back in 2008 it was activity in housing-related finance, specifically US subprime mortgages, that was the focus. As we've just mentioned, there's not a massive amount of transparency in private credit and many would see this as similar to the opacity of the myriad of exotic derivative products used to leverage the subprime boom nearly twenty years ago.

Back in 2008 the collapse in house prices strained lenders substantially, causing some financial firms to fail. This time around, it is the weakness in AI-related equities, such as software, that is seen by some as the likely instigator of new financial stress. Again, like 2008, those financial firms most at risk are thought to be those that cannot rely on any financial backstops, such as private credit. In 2008, it was investment banks, some of whom folded while others became banks in order to get access to Fed backstops.

We could continue making more comparisons. Perhaps JP Morgan boss Jamie Dimon made another one when he said recently that he sees signs that some people are doing “dumb things” just as many were doing ahead of the GFC. But Steven Barrow, Head of Standard Bank G10 Strategy said, before we panic, there would be a number of issues to consider.

First up, the proportion of AI-related loans made by private credit firms is still pretty low. The BIS puts it at around 8% of lending, although it has risen pretty rapidly, and within the average figure there are undoubtedly some that have far more exposure to AI, presumably including Blue Owl.

Another distinction we would make is that the 2008 crisis was based around the use of exotic derivative products linked mostly to the US subprime mortgage market. Investors lapped up these derivative products as they ‘reached for yield' given that the Fed had pushed rates down as far as 1%.

In short, financial firms had created exotic higher-yielding assets out of what was a very staid market; the market for mortgages for those that could not get traditional mortgages. The AI fervor today seems to be the opposite. For here, private credit (and other lenders) is utilizing pretty staid products, like straight lending or debt issuance, to finance something that is not at all boring given that many see AI as truly transformational in many respects.

Now clearly, the equity market may have gone over the top a bit and priced the AI-related sector too highly relative to the longer-term benefits that are likely to accrue. But surely it is far too early to make a definitive conclusion about this and, even if there is a repricing of AI-related risk and subsequent difficulties for private credit as a result, Steven Barrow finds it hard to believe that the financial infrastructure is as vulnerable today as it was back in 2008.

“Just to repeat, this does not necessarily imply that any private credit concerns will blow over without some incidents, of which Blue Owl might be the first. But we see this as more like the US regional bank episode that we saw with SVB's demise in March 2023, which blew over in a short period of time, rather than the first signs that a new GFC is brewing”, said Steven Barrow.