Subscribe: RSS Feed | Twitter | Facebook | Email
Home | Contact Us

IMF Encourages Investors to “Rely On Their Own Due Diligence”

Posted by Larry Doyle on October 7, 2010 9:36 AM |

In the midst of the economic crisis, many business models have been exposed as broken. Other models have been exposed as downright useless. Somewhere in that realm lies the business model of our credit rating agencies. How will these entities, charged with providing meaningful credit ratings analysis, adapt to the changing economic and financial landscape? More importantly from my standpoint in trying to promote ‘sense on cents’, how should investors adapt? Well, my jaw dropped yesterday upon reading a report by none other than the International Monetary Authority on this topic. Let’s navigate.

The IMF recently produced a Global Financial Stability Report. Embedded in that report is an Executive Summary which highlights: 

The credit rating agencies have attempted to produce stable “through-the-cycle” ratings to satisfy clients who find it costly to frequently alter trading decisions that are based on ratings. The chapter shows that a typical smoothing technique used by at least one rating agency is deemed likely to contribute to procyclicality in ratings compared to a method that accurately reflects current information at a “point in time.” This is because a “through-the-cycle” approach waits to detect whether the degradation is more permanent than temporary and larger than one notch. However, this often means that the lagged timing of the downgrade accentuates the already negative movement in credit quality.

Overall the chapter suggests the following policies to lessen some of the adverse side effects that ratings and rating agencies may have on financial stability.

First, regulators should remove references to ratings in their regulation where they are likely to cause cliff effects, encouraging investors to rely more on their own due diligence (LD’s highlight!). Similarly, central banks should also establish their own credit analysis units if they take collateral with embedded credit risks.

Second, to the extent that ratings continue to be used in the standardized approach of Basel II, credit rating agencies should be overseen with the same rigor as banks that use the internal-ratings approach—credit metrics reported, ratings models backtested, and ex post accuracy tests performed.

Third, regulators should restrict “rating shopping” and conflicts of interest arising from the “issuer pay” business model by requiring the provision of more information to investors. A user-pay-based business model is difficult to maintain because of the inability to restrict access to ratings and their public good characteristic of aggregating difficult-to-obtain private information. Hence, mitigating conflicts of interest in the issuer-pay design through disclosure of any preliminary ratings obtained and how the ratings are paid for is preferred.

Not exactly a recommendation of the efficacy and integrity of our rating agencies. In fact, I would categorize this statement as a total vote of no confidence.

Navigate accordingly!!

Larry Doyle

  • Drew

    What did our hallowed Financial Regulatory Reform and SEC do to address this massively flawed business model of ratings agencies?

    De Nada!!

    Despite all promises to the contrary, business largely remains as usual on Wall Street and Washington …

  • Kiffmeister

    We didn’t mean this statement as a vote of no confidence in credit rating agencies. The point is that credit ratings should only be seen as one of several tools to measure credit risk, and not as the sole and dominant one.

    The other key point is that, regardless of how accurately or inaccurately rating agencies measure credit risk, regulations and rules that are hardwired to ratings make the rating agencies inadvertent potential financial destabilizers.

    (I say “we” because I was one of the authors of the report.)

    • fred

      Sounds alot like the warning on the side of a package of cigarettes.

      Warning, ratings issued by rating agencies may be hazardous to your financial health.

      A couple of comments:

      1) Is the problem the agency business model or is it the risk models being utilized by analyst that fail to measure fat tail risk accurately?

      2) As long as you have fiduciaries you will have the “cover your ass mentality” that requires blame for bad investments be placed elsewhere. “AAA” by one of the ratings agencies has always been the perfect scapegoat for poor judgement.

  • LD

    Thanks for your commenting on my post. I appreciate your taking the time to write. Although you may not have meant the statement as a vote of no confidence having read and reread it a number of times, I believe it very much comes across that way…and the fact is I think you are right.

  • Kiffmeister

    Fred, it may sound like a “cover your ass” mentality, fiduciaries and others need some sort of objective and third-party verifiable criteria, and ratings happen to be a convenient acceptable metric for that. That’s why we say in our Global Financial Stability Report that you can’t just go and pull the rug out overnight. A possible replacement would be an “internal model” but that might bring with it a whole host of other control problems.

    On you’re first question, in the case of structured finance products, the problems are many, including models, inputs, and maybe some conflicts of interest, but I wouldn’t overplay the latter. There certainly are perceived conflicts of interests, and the regulators and rating agencies have to work on that. However, I must admit that I think that the way that the rating agencies model structured finance product risk is fundamentally wrong, and I have to chuckle when they introduce “methodology” changes, which are really only input or parameter changes.

    On the other hand, I’m encouraged when I hear the rating agencies talking about using more common sense in their structured finance rating processes, and often over-riding their model ratings when they don’t make sense.

  • Kiffmeister

    Larry, as I said to Fred, their are some fundamental problems with structured finance ratings, which is why for years we’ve been calling the the rating agencies to stop using the conventional fixed income rating scales (AAA, AA, etc.) or at least add a structured finance identifier (e.g., an “SF” superscript). That’s happening in Europe, but last time I checked, it’s not going to happen in the U.S.

    However, our GFSR chapter is all about sovereign ratings where their record is actually pretty good. I think we’d find the same thing with their corporate ratings. When I say “pretty good” I mean that they do a good job of doing what they say they are aiming to do – rank order credit risk. We think they should be aiming for cardinal accuracy (i.e., where each rating is associated with a specific default probability) but that doesn’t seem to be what customers are asking for.

    We think they could do a better job of incorporating sovereign debt composition (i.e., levels of short-term debt) and contingent liabilities. However, it’s pretty hard to blame them for looking a bit flat-footed on Greece, for example, when the information that would have resulted in a smoother downgrade trajectory was not made available to them. In other recent spectacular failures, such as Enron and Parmalat, there was a fraud angle, and I don’t know if any third-party rater could have caught those ones.

    Anyways, I still wouldn’t call it a “no confidence” vote, but I will say that the rating agencies could improve in some of the ways we mention in our chapter. However, I still view them as the best place to start when I want to understand the credit risk of a corporate or sovereign, and even a structured finance product. But where we say things have gone awry, is that’s where many stop their analysis. Ratings should be only a part of the assessment process.

    • fred


      If you are in a position to influence policy, please keep these points in mind:

      1) Independant rating agencies are “demanded” by investors, fiduciaries and BOD’s.

      2) If the “Abbaccus GS” or “AIG” situations taught us anything.

      a. a rating should not be contingent upon portfolio insurance.

      b. In stressful times, (high volitility, low liquidity) modern portfolio theory fails. Portfolio risk takes on the characteristics of the best or worst securities within the portfolio. It should be assumed that over the life of a portfolio, stressful times will be experienced, therefore prudence demands ratings should reflect a worst case scenario. To put it another way, rating agencies should assume a “fiduciary” as opposed to “caveat emptor” role within the process.

      Thanks for your input, it was both inciteful and thought provoking.

  • coe

    I can recall the early days of the non-agency securitization market when public pension funds and many other institutional investors automatically defaulted their “diligence” to the work of the ratings agencies. My sense is that as the markets and structured products got more complex, these same institutional investors learned how to lift the hood and examine the engine parts for themselves. That clearly didn’t work out as well as expected. And, the commentators are right in acknowledging the statutory, regulatory,legal, and policy importance of the role of the rating agencies in the investment process. Whether that is dangerous or even appropriate is part of the debate. In my opinion, somewhat like the regulators themselves, the good folks at the rating agencies mean well, but their models and analyses were/are some combination of flawed, dated, manipulated, and clearly inadequate as a valuation tool and as a legitimate predictor of risk (i.e. WRONG!)- yet so much was riding on these models and analytics nonetheless. I’m inclined to agree with the IMF – do your own diligence!! Last I looked, nobody buys a house without hiring an engineer to do a detailed inspection – they would never rely on the representations of the seller or that the realtor suggests the house is in good order and take it on faith “as is”. The rating agencies pretend to be dispassionate engineers, but really aren’t good enough – it’s that simple. An institutional investor should not automatically buy the ratings-based imprimatur – they are not the best place to start at all, and they are certainly not the only place to rely on…caveat emptor

    • fred


      Your response is appropriate for large institutional investors who have the “in house” expertise to effectively evaluate a rating. Along these lines, the report was written as a policy guide for central banks suggesting that they establish risk control units.

      Let’s not forget that many fiduciaries lack the knowledge, skill, or budget of large institutions or central banks, and do, in fact, rely on the ratings of rating agencies when making investment decisions.

      With the little guy in mind, I strongly support the position that the quality of ratings by rating agencies must be improved to reflect maximum potential risk: adding a disclaimer just doesn’t cut it.

      This discussion is all part of “THE ISSUE”, (as well as underfunded liabilities), the “2,000 lb. gorilla in the room”. Afterall, if the recession is over and everything is moving foreward, why all the mtg default delays, recent under performance of financial stocks, Fed purchases of mtg securities, currency debasements and chatter for QE2?

  • Kiffmeister

    Fred, as you say, “independent rating agencies are “demanded” by investors, fiduciaries and BOD’s” which is why we also call for heavier official supervision. This is also needed because, as you point out to Coe there are some investors who don’t have the scale of operations or sophistication to do their own analysis. However, I still think that the authorities should do everything they can to discourage the kind of mechanistic rating hardwiring behind the “demand” you speak of. And that has to start by leading by example!

    • coe

      As I was reading these posts, Fred and Kiffmeister, it occurred to me that in the same way we can and perhaps should “go back to basics” with the GSE reforms re new world order rules for the conforming secondary mortgage markets, it would make equal sense to sanitize some of the complexity away in the new world order of structured products. You guys are both very direct and accurate in your thoughts, and there are major differences in dealing with well resourced sophisticated institutional investors vs those not so fortunate nor the retail investors. Should there be a better effort by the ratings agencies – yes, for sure…Is it a little bit of a stretch to compare the potential damage that clearly can occur if there is another episode of “Capital Markets Gone Wild” to the gun or tobacco debates – Is it the product itself.. Is it the misinformed/misguided innocent “user”.. Is it the charlatans and criminals in the product channel..Is it the manufacturer.. or..Is it the regulators/third party firms – eg in capital markets, the accounting standard setters and rating agencies – that should be on the hook for governance/accountability? Perhaps everyone has to carry some weight..

      If we rolled back the complexities, perhaps very little would be lost, and much more comfort would be gained…probably a bit of wistful naivety, but one can think fondly back to “those were the days”..just a thought

  • Kiffmeister

    Coe, I agree with you completely, and we said as much in our October 2009 Global Financial Stability Report:

    “Most products could usefully be standardized at least to some extent. This should increase transparency as well as market participants’ understanding of the risks, thus facilitating the development of liquid secondary markets. Although there will always likely be investors that demand bespoke complex products, securitization trade associations and securities regulators should encourage standardized building blocks for securitized products. It would also be
    useful if some standardization could be imposed on the underlying assets to maintain higher quality pools or at least verifiable pools.”

    “Valuation difficulties could also be alleviated if securitization products were simplified. Some of the product complexity was well intentioned, such as excess spread traps and triggers designed to bolster the creditworthiness of the senior tranches.20 Others, such as micro-tranching, were designed to game rating agency models. In any case, this product complexity has made some securities extremely difficult to value and risk-manage, and to the extent that regulation or market practices encourage such complexity, these components should be eliminated.”

    At the time we wrote that, we were very discouraged by the interest in pursuing this. However, we are seeing some activity on this front in Europe, but the U.S. securitization community is still sitting on its hands on this front, at least as far as I can see.

    BTW, you can download all of the IMF GFSRs here:

    • coe

      Thanks for the site..I’m not sure I would say that US securitization is sitting on its hands, however – with all the regulatory uncertainty – skin in the game/FDIC/accounting developments/foreclosure would be hard to imagine how the factory will grind the rust off the gears – especially in non-conforming space

      • Kiffmeister

        What I meant was that I’ve seen no U.S. efforts to encourage securitization product simplification and standardization. The authorities are certainly very active on all other fronts – maybe a bit too active!

  • phil trupp

    The question that constantly nags: How accurate are corporate balance sheets? We are uncomfortably familiar with the quarterly adjustments made by publicly traded companies, and we know these adjustments are designed as window dressing. There also remains the problem of the simplified line-item report, or summary, in which items such as “loss,” “carried interest,” etc., require specialized, detailed inspection before any rating can be assigned. How much truly bullet proof data will public corporations expose regarding their financial health? If history is any judge, the answer to the latter is, “not much.”

Recent Posts