Procyclicality

[chih22]

I am a little confused about what contributes to procyclicality (seems like its everything)...

Based on my studies this far, I thought I read that Rating Agencies (which use a though-the-cycle approach) contributes to procyclicality (e.g. they are too late to issuing downgrades when the economy slows and contributes to tightening of credit in a credit crunch).

However, I also thought I read that an "at-the-point" in time approach also contributes to procyclicality because it will result in looser lending during economic booms and tighter credit during economic downturns, thus prolonging the crisis and leads to procyclicality.

Also, I think I also remember reading about how the IRB method in BASEL II can also lead to procylicality (but IRB on the graph is in-between a pure point in time and pure through the cycle methods). I think it said that it can lead to procyclicality because the credit cycle lags the economic cycle and ratings are not forward looking, but isn't this more applicable to the standardized ratings based approach and not the IRB approach since it depends less on the ratings from rating agencies?

Basically, I'm a little confused from what I remember reading and don't remember if I read them correctly because it is mentioned in 3 separate areas. It would be helpful, if someone can summarize and link the 3 different areas where I remember procyclicality was mentioned as it just seems that everything leads to procyclicality because bank lending itself is procyclical?

 

[David Harper CFA FRM]

chi

These are truly excellent observations! I agree procyclicality appears three times ("procyclical" will be a key theme in the cram session, btw). Briefly, I think you are mostly correct above. The FRM historically emphasized this: point-in-time-approach is procyclical.

However, with the addition of Aschraft's subprime securitization case study, IMO the FRM now reflects a more progressive concept: any rating-based approach, external or internal, is procyclical. (I'd go further: any capital adequacy is likely procyclical). But let's add, keeping in mind testable concepts, if we must choose between point-in-time or through-the-cycle as to which is "more procyclical," the safe answer remains "point in time".

In regard to the three procylical mentions:

* de Servigny (Ch 2, p 46, paraphrase). Generally says any approach (internal or external) that links PD to capital requirements could be procyclical; i.e., recession implies less lending implies exacerbate recession. But argues that point-in-time approaches (e.g., Merton model) are "more procyclical" because they tend to overreact with higher volatility. (And, also argues that a focus on EL rather than PD may be more procyclical; e.g., Moodys = EL, Standard and Poors = PD). This is the basic procyclical idea that we just want to know. What is it?

* Aschraft's Subprime Securitization Case Study argues that through-the-cycle for structured credits are more procyclical. Yes, this is different than de Servigy's point and, IMO, this argument is more subtle: they are saying that the credit agencies' motive to maintain stable through-the-cycle ratings is itself procyclical. (see thread here with their Fig 11). As in, a senior tranche in an MBS is rated AAA so it has a low (coupon) cost. Economy enters recession. Because agencies want to keep rating stable, they increase the credit enhancement: they de-leverage the structure (less senior, more junior tranches) which increases the average cost of subprime MBS structure. Specifically, to "de-leverage the structure has knock-on effect on economic activity" (Ashcraft p 50-51)

* In Basel II, there is a long story here. IMO, it is safe to say both (standard & IRB in Basel II), consistent with above, are procyclical. There are arguments as to which is worst. Re: standard, you have a good point about ratings, but on the other hand, the design is not extremely leveraged; e.g., 100% only goes to 150%. It is not a punishing system on the downside. Re IRB, many banks will employ point-in-time to IRB and this itself will tend to make it more procyclical, but on the other hand, the design of the IRB formula deliberately "mutes" the penalty for higher PD. Bottom line on Basel: either could be procyclical (this is why some public figure, I forget where i've read, has recommended that the Base II capital ratio should be lower right now so banks can use the buffer rather than shrinking their balance sheet). I don't have a handy link on this. GARP dropped the relevant reading and we don't have AIMs on B2.

Hope this helps. Thanks for surfacing a THEMATIC IDEA that involves three assigned readings!

David

 

[chih22]

Thanks David for your quick response and for clearing things up! I love your website and screencasts!