The New and the Old in Risk Management

I have been writing about the new and the old in risk management over the past year. This starts with the slow pace of adoption of FinTech by incumbents in financial services. I have suggested that an important component of the change needed includes incumbents amending and enhancing risk management frameworks to reflect new FinTech innovations. (See my last post on the subject.)

Recently, I came across an article from McKinsey that makes a similar point in the context of model risk and the adoption of artificial intelligence (AI) and machine learning. It turns out I am in good company! 

McKinsey’s article notes that banks have developed and implemented frameworks to manage model risk, including model validation reflecting specific regulatory frameworks, in this case from the US Federal Reserve (here). They recognise that the implementation of these frameworks is not appropriate to deal with the model risk associated with AI and machine learning. Banks are therefore proceeding cautiously and slowly introducing new modelling approaches even when these are available.

The article then shows how a standard framework for model risk management is used to identify extra considerations required for this framework to cover appropriately AI and machine learning models.  The key message is that the challenge of adopting AI and machine learning can be addressed through a careful consideration of existing approaches. 

Two further thoughts from McKinsey’s article. Firstly, the article rightly refers to model management rather than validation. It is always useful to reiterate that model validation undertaken by the risk function is just a component of how models are managed in the business. Secondly, model management should not apply only to internal models used to calculate regulatory capital, but should apply more widely to models used in the business such as those used for pricing, valuation of assets and liabilities.

The article ends with a cautionary tale of an unnamed bank where the model risk management function took initial steps to ready itself for machine learning models on the assumption that there were none in the bank. It then discovered that an innovation function had been established and was developing models for fraud detection and cybersecurity.

If you found this post of interest, you can subscribe and receive further posts by email. See the box on the right-hand side of the blog's screen or click here.

Freakonomics and Insurance Regulation

A post from the Freakonomics blog (here) highlights an interesting paper (here) about insurance regulation by Ralph Koijen and Motohiro Yogo.  The paper looks into the growth in the use of ‘shadow reinsurance’.  Freakonomics is concerned that there may be a bubble building up. 

I believe the underlying issue represents a good example of the law of unintended consequences that applies sometimes to financial regulation. This post will explain that and offer some thoughts from a Solvency 2 perspective.  

Some context – shadow reinsurance

There isn’t a formal definition of shadow reinsurance. Koijen and Yogo refer to shadow reinsurers as “affiliated and unauthorised reinsurers without an A.M. Best rating” in the US or offshore. (The paper used A.M. Best data, hence the reference to them in the definition.) 

The use of shadow reinsurance has also been covered in a recent report (here) from the NY State Department of Financial Services (DFS).  The DFS has a similar definition in mind, other than the ratings consideration.

Unintended consequences

The likely chain of events leading to an increased use of shadow reinsurance is set out below.   

Firstly, supervisors introduce new reserving requirements for certain term and life insurance products, so called XXX and AXXX reserving requirements respectively. 

These new reserving requirements require the use of stringent assumptions, which have been described as resulting in reserves “over eight to 10 times that of an ‘economic’ type reserve, such as a FAS 60 reserve under U.S. GAAP” (here).  I don’t really know if this estimate has passed the test of time – the paper is from 2005 – and the quote is offered as an indication of the conservativeness of the underlying assumptions.  As a footnote, it is worth noting that industry papers tend to refer to the delta between economic and AXXX / XXX reserves as ‘redundant reserves’.

Secondly, firms take action to reduce the impact of non-economic regulation. 

This requires an understanding of the prevailing regulatory environment.  Insurance companies would then realize that they can reduce the impact of these stringent reserve requirements by transferring some of these liabilities to special purpose vehicles (SPVs) in such a way that an insurance licence is not required. There are then some additional issues which give rise to the need of letter of credit to the originator. However, the key point is that if the recipient of the liability is not an insurance, the stringent valuation requirements would not apply.  

So far so good from a company's perspective.  However, what appears surprising is that state supervisors have a role in approving these transactions. There is even a sub-group of the NAIC Executive Committee called “Captive & SPV use sub-group” (see last page of this report).  It is therefore not entirely surprising that one of the DFS report recommendations is that state insurance commissioners stop approving these transactions:

“state insurance commissioners should consider an immediate national moratorium on approving additional shadow insurance transactions until […] investigations [by the Federal Insurance Office, Office of Financial Research, the NAIC and other state insurance commissioners] are complete and a fuller picture emerges.”

It would appear that the same supervisor that introduced stringent reserving requirements has a role in approving transactions aimed at reducing the impact.  This begs the question of what is the value of these onerous reserving requirements. 

Back to economic rationality, it is not entirely surprisingly, that the data in the paper also show that the (mean) reinsurance shifted to shadow reinsurers increased during the financial crisis when the uncertainty about valuations probably increased (Table 2 of the paper).

What about Solvency 2?

The essence of the unintended consequences means that it is difficult to foresee if similar transactions would happen under Solvency 2. 

The key point is that there would less incentive for these transactions to take place because it would not appear that the Solvency 2 reserving requirements would be uneconomical as in the case of AXXX and XXX requirements.  Although depending on the on-going discussion about the treatment of long-term guaranteed products, it may be argued that for certain insurances the Solvency 2 reserves are uneconomical. 

However, there are two additional considerations that would make the outcome in the paper less likely under Solvency 2. 

Firstly, there are specific articles in the directive about SPVs used for reinsurance purposes which require authorisation and the application of solvency requirements.  This would mean that shifting liabilities to SPVs would not necessarily mean a reduction in reserving and capital requirements.  (If you want to see the detail of the likely implementing measures, please check EIOPA’s advice to the Commission - here).  

Secondly, within the Solvency 2 directive a group is widely defined.  Personally, I would be surprised to hear that this could be read to exclude such SPVs.  More generally, the default approach to calculate the group solvency requirements is based on the consolidation of all exposures, which would mean that the effect of the transfer to an SPV cancels out.

The Solvency 2 implementing measures will  provide additional details about these issues but, as ever, regulation can have unintended consequences in particular when there are uneconomic requirements.