CLOs are riskier than current ratings suggest

Is the COVID-19 economy going to repeat history?

The economic impact of COVID-19 is eerily similar to the recession in 2007-2008. Back then, credit ratings agencies were slow to downgrade structured finance products like asset- and mortgage-backed securities, and collateralized loan obligations (CLOs), failing to accurately convey risk to investors. Today, history appears to be repeating itself.

Despite significant financial stress across a wide-swath of companies – and market forecasts calling for bankruptcy levels higher than that in 2009 – there is a significant disconnect between credit agency ratings and CLO risk. Between March and June 2020, S&P and Moody’s downgraded approximately 25% of the collateral feeding into CLOs, yet the value of CLO tranche downgrades was approximately 2%, increasing to 5.5% when also considering credit watches.

Another way to look at CLOs is through the lens of real estate. If you build an apartment complex on a flood plain, the basement and first floor will have a much higher risk of flooding than the upper floors. But if there is a huge flood, everyone’s risk should increase, even those on the upper floors. During the COVID-19 pandemic, it is like we are having a huge flood, yet the risk of the upper floor apartments has not changed. How is it possible that the water is rising, yet that risk is not reflected in the ratings?

If investors rely on incorrect risk ratings, they can lose a lot of money. And policymakers should be genuinely concerned about repeating the economic crisis of 2007-2008. That is why the ratings need to be unbiased, timely, and accurate.

We examined possible explanations for this mismatched relationship between asset and downgrading behavior and came up with two likely explanations.

First, the ratings agencies could be slow to act because they are basing decisions on subjective factors. When our tools and models cannot explain their ratings, one of the only things left to consider is something beyond quantitative factors that is not visible to the public.

If subjectivity is impacting the ratings, then we should be concerned about conflicts of interest. We saw this in the last recession, when disconnected ratings led to government investigations. The current mismatch of ratings and risk could trigger similar investigations.

However, it is up to the government to decide how long is too long before it investigates. A few months? A few years? So far, an investigation has not happened, but policymakers would be wise to keep an eye on rating agency actions.

Second, portfolio managers could be strategically managing their investments to make them appear less risky from the perspective of a rating agency. Financial models will deem a loan due in one year less risky than a loan due in five years. To use this to their advantage, managers could be window-dressing collateral pools by purchasing loans with shorter maturities. The risk has not changed since managers will need to find replacement investments as loans mature, but the pool of loans is window-dressed to appear safer.

If ratings agencies’ approaches can be window-dressed, investors should be worried because it could lead to inaccurate ratings. Hopefully, ratings agencies will respond and update their models – or they risk losing investor trust in the ratings.

In either case, investors need to do more due diligence and look beyond the ratings before they invest. And policymakers need to take notice, especially because this disconnect impacts many investors through structured investments and pension plans.

MIT Sloan School of Management Visiting Finance Prof. Jordan Nickerson and University of Texas-Austin Prof. John M. Griffin are co-authors of “Are CLO collateral and tranche ratings disconnected?”