For all the anxiety in the financial press about private credit — especially amid a spate of redemption headlines — much of the commentary overstates the systemic risk.
Without delving too deeply I would quibble on the basis for your size comparison. The epicenter of the problem was subprime mortgages not mortgages so comparing the size of private credit to the size of all mortgages seem wrong. Private credit is a sliver of the business lending environment just like subprime was a sliver of the mortgage market. I believe subprime and private credit were actually pretty close in nominal size. The part I haven't dug into deeply that does seem different is the notional value of CDS's on top of the underlying debt in the subprime episode. I'm not clear we have a massive amount of those derivative liabilities in this case like we did in 2008.
With quibbling, despite all the noise in private credit, the asset class with only one part exception is performing very well and it is only about 20% of the asset class, and that is software loans. The rest is performing quite well. And even in software despite the media hysteria, while perhaps the median software loan is trading at around 88-90, the stronger credits are trading at about 98-99, right where they are originated.
and most importantly the banking system’s exposure to PC overall is quite low, as is the insurance industry so even if there are losses it will be investors eating them, not the financial system in the main. So this is literally nothing like 08, even if we have a recession.
No need to reiterate your view. I understood it the first time. I’m undecided. I was simply pointing out a rather meaningful flaw in part of the reasoning that led to your conclusion, and hinting at something that might help you strengthen it.
This is a pretty weak article. 10-20% of most private credit loan portfolios paying in PIK. Even Fitch, who no one would accuse of being a tough credit analyst, says that 70% of PIK payers end up in default. No mention that the vast majority of businesses receiving these loans have no hard assets...unlike true secured loans of days long gone, when credit underwriting was a thing and "first lien" meant something. These loans are nothing more than securitizations on overstated, "adjusted" cash flow that are weakening rapidly and were over-levered to begin with. I look at your article and only see 1 hard data figure posted, that 1% of insurer portfolios are non-IG private credit. First, this is blatantly wrong. Second, it is just a figure you regurgitated from Torsten Slok's daily blog...who happens to be the chief economist at a major private credit provider (Apollo). Please try looking farther afield next and do some basic reading before you put content out there. There are a multitude of resources out there, even on Substack alone, that will challenge virtually everything you have written in far more detail than what you have scribbled out.
First, you seem to have missed the main points of the article. One of the most important parts is that PC at about 3% of total assets is 1/20 of the size of the mortgage assets that caused the GFC.
Second, banks provide the liquidity for the economy and their capital problems due to losses are what led to the GFC, and today bank capital is not only much stronger but their exposure to PC is very low, as is their exposure to the companies that borrow PC is very low as PC basically replaced bank lending to these companies which withdrew from that market in order to restore their balance sheets. So that systemic risk is very low.
Third, if there is significant increase in defaults, which I noted is possible, the losses will be absorb by INVESTORS, not the banking system, so the systemic risks, which led to the GFC are greatly reduced, as you don’t have the contagion risk.
Fourth, the amount of leverage applied to PC is a small fraction of the leverage used in investments that led to the GFC., like the CDOs which blew up Lehman.
Fifth, I don’t know if Tortsen Slot’s figure is exactly right or not, but here are a few things. He is a highly respected economist having worked at the OECD and the IMF and as a trained economist myself I find him to be one of the best and most informed with great data. You of course can have your own opinion. With that said, having worked at Prudential for 10 years and head of credit for their mortgage company and had to live with insurance company capital rules, I know that insurers as regulated entities must have the vast majority of their assets in investment grade assets because the capital they must keep against below that level makes it expensive. The fact is that for U.S. insurers overall, a good recent proxy is below-investment-grade bonds, which were 4.9% of total bond holdings at year-end 2024; the NAIC also reported that NAIC 3, 4, 5, and 6 bonds together made up 4.9% of total bond exposure. And not all of unrated assets are PC, so it is obviously a small percentage of their assets. Maybe it isn’t 1% but it surely is not a high figure that could threaten that industry.
Sixth, the CDLI index which goes back to 2004, so includes the GFC, has historical default rate of just 2% and with recoveries losses about 1%.
Seventh, lots of assets in PC are backed by assets including PC loans to infrastructure, to real estate, to ASSET BACKED lending overall. Now lots of PC is not backed by hard assets but also by companies with strong cash flows, and in businesses with in many cases low exposure to economic cycle risks, and leverage is relatively low and see the default rate.
Eighth, while defaults have gone up from the historical 2%, the last quarter they were all the way up to 3% (down from 5% in the fourth quarter).
Ninth, re PIK loans, there are actually two kinds of PIK, the “good kind” where it is in the covenants to grant a LIMITED use for companies with very strong relative credit with fast growing businesses and the default rate on the good PIK is similar to the 2% overall rate. The bad kind is where defaults are likely to be high. So you would have to know what % of PIK is good vs bad.
Sorry I forgot to address your false statement that PIK is 10-20%. Actually the fourth quarter CDLI index reported PIK of just over 7%, and again not all is the bad kind of PIK.
Without delving too deeply I would quibble on the basis for your size comparison. The epicenter of the problem was subprime mortgages not mortgages so comparing the size of private credit to the size of all mortgages seem wrong. Private credit is a sliver of the business lending environment just like subprime was a sliver of the mortgage market. I believe subprime and private credit were actually pretty close in nominal size. The part I haven't dug into deeply that does seem different is the notional value of CDS's on top of the underlying debt in the subprime episode. I'm not clear we have a massive amount of those derivative liabilities in this case like we did in 2008.
With quibbling, despite all the noise in private credit, the asset class with only one part exception is performing very well and it is only about 20% of the asset class, and that is software loans. The rest is performing quite well. And even in software despite the media hysteria, while perhaps the median software loan is trading at around 88-90, the stronger credits are trading at about 98-99, right where they are originated.
and most importantly the banking system’s exposure to PC overall is quite low, as is the insurance industry so even if there are losses it will be investors eating them, not the financial system in the main. So this is literally nothing like 08, even if we have a recession.
No need to reiterate your view. I understood it the first time. I’m undecided. I was simply pointing out a rather meaningful flaw in part of the reasoning that led to your conclusion, and hinting at something that might help you strengthen it.
This is a pretty weak article. 10-20% of most private credit loan portfolios paying in PIK. Even Fitch, who no one would accuse of being a tough credit analyst, says that 70% of PIK payers end up in default. No mention that the vast majority of businesses receiving these loans have no hard assets...unlike true secured loans of days long gone, when credit underwriting was a thing and "first lien" meant something. These loans are nothing more than securitizations on overstated, "adjusted" cash flow that are weakening rapidly and were over-levered to begin with. I look at your article and only see 1 hard data figure posted, that 1% of insurer portfolios are non-IG private credit. First, this is blatantly wrong. Second, it is just a figure you regurgitated from Torsten Slok's daily blog...who happens to be the chief economist at a major private credit provider (Apollo). Please try looking farther afield next and do some basic reading before you put content out there. There are a multitude of resources out there, even on Substack alone, that will challenge virtually everything you have written in far more detail than what you have scribbled out.
Let’s see if I can address your issues.
First, you seem to have missed the main points of the article. One of the most important parts is that PC at about 3% of total assets is 1/20 of the size of the mortgage assets that caused the GFC.
Second, banks provide the liquidity for the economy and their capital problems due to losses are what led to the GFC, and today bank capital is not only much stronger but their exposure to PC is very low, as is their exposure to the companies that borrow PC is very low as PC basically replaced bank lending to these companies which withdrew from that market in order to restore their balance sheets. So that systemic risk is very low.
Third, if there is significant increase in defaults, which I noted is possible, the losses will be absorb by INVESTORS, not the banking system, so the systemic risks, which led to the GFC are greatly reduced, as you don’t have the contagion risk.
Fourth, the amount of leverage applied to PC is a small fraction of the leverage used in investments that led to the GFC., like the CDOs which blew up Lehman.
Fifth, I don’t know if Tortsen Slot’s figure is exactly right or not, but here are a few things. He is a highly respected economist having worked at the OECD and the IMF and as a trained economist myself I find him to be one of the best and most informed with great data. You of course can have your own opinion. With that said, having worked at Prudential for 10 years and head of credit for their mortgage company and had to live with insurance company capital rules, I know that insurers as regulated entities must have the vast majority of their assets in investment grade assets because the capital they must keep against below that level makes it expensive. The fact is that for U.S. insurers overall, a good recent proxy is below-investment-grade bonds, which were 4.9% of total bond holdings at year-end 2024; the NAIC also reported that NAIC 3, 4, 5, and 6 bonds together made up 4.9% of total bond exposure. And not all of unrated assets are PC, so it is obviously a small percentage of their assets. Maybe it isn’t 1% but it surely is not a high figure that could threaten that industry.
Sixth, the CDLI index which goes back to 2004, so includes the GFC, has historical default rate of just 2% and with recoveries losses about 1%.
Seventh, lots of assets in PC are backed by assets including PC loans to infrastructure, to real estate, to ASSET BACKED lending overall. Now lots of PC is not backed by hard assets but also by companies with strong cash flows, and in businesses with in many cases low exposure to economic cycle risks, and leverage is relatively low and see the default rate.
Eighth, while defaults have gone up from the historical 2%, the last quarter they were all the way up to 3% (down from 5% in the fourth quarter).
Ninth, re PIK loans, there are actually two kinds of PIK, the “good kind” where it is in the covenants to grant a LIMITED use for companies with very strong relative credit with fast growing businesses and the default rate on the good PIK is similar to the 2% overall rate. The bad kind is where defaults are likely to be high. So you would have to know what % of PIK is good vs bad.
I hope that is helpful.
Sorry I forgot to address your false statement that PIK is 10-20%. Actually the fourth quarter CDLI index reported PIK of just over 7%, and again not all is the bad kind of PIK.