Friday, August 19, 2011

CONTEST: design an effective ratings agency

The primary conflict of interest at credit rating agencies is well known: The company is paid by the same issuers (banks and companies) whose securities it is supposed to objectively rate.

At Moody's, this created incentives to give Moody's clients favorable ratings, lest the clients fire Moody's and take their business to one of the other ratings agencies.

The obvious organizational design question is how to design an organization or an industry that would give more objective ratings.  Succesful entries will (i) address the conflict of interest (ii) describe how the company would get paid; and (iii) how it would get access to confidential issuer information?

The best answer (in comments below), will win a signed copy of the book (worth about $100) and a coffee mug.


  1. The people who are supposed to truly benefit from the ratings that credit agencies provide are the investors. However, currently it seems that the burden is on the banks and companies who have investments to sell to pay for the ratings. In return the banks have the expectation to be “advertised (rated) in a favorable fashion. My solution is as follows, and one I suspect will make Professor Froeb cringe.
    Why not mandate through regulation and/or legislation that the ratings be paid for by the investors who use them (i.e through a fee ,tax on all investments). This would put the burden on credit raters to provide objective ratings without worrying about the implications of upsetting the banks, which is a concern under the current system. Under this proposed system, the rating agency would have to be worried about upsetting the investors who are now paying the agency if their ratings turn out to be inaccurate, and it return one would suspect that more objective and detailed ratings result as the conflict of interest which exists in the current system is eliminated. Finally, again through regulation, banks would be mandated to release appropriate information to the rating companies in order to allow them to have all of the information needed to make the ratings. -

  2. The same problem exists with businesses & their accountants.

    I suggest payments be made to an association and they randomly assign the rating agency to evaluate the securities ( or business). Keep the govt out of any implementation function.

    In this suggested way, the payor does not have direct influence over the rating agency.

  3. 1. The rating decision should never include anyone from the product division.
    2. The analysts should just have the data to make the decision and not the name of the issuer.
    3. Data they use (such as the housing market forecast) should come from an outside agency.
    4. The analysts division should report directly to the CEO. The analysts and the CEO compensation should not be tied to the number of ratings they do.
    5. Need government oversight/audit to make sure the analysts are doing their job right
    6. There should be heavy penalties on rating agencies for incorrect ratings
    7. Government should make sure the issuers are not threatening to leave the rating agencies for poor (albeit correct) ratings.
    8. The rating agency should be paid the same amount irrespective of the rating of the product.

  4. The rating system is a business. The customers should be entirely made up of those seeking reliable information on the risk involved in investing in a company or product. My company would set itself apart from the others by guaranteeing that I would not take any payment from the company being rated. I would issue reports based solely on objective data without the bias of subjective opinion (data in data out). I would get the confidential information by listing the company as either one that allowed me to evaluate non-public data or one that did not, which would diminish the credibility of those that did not. This should be an incentive for the companies to give the non-public information to my company, as they would appear to be hiding something if they did not give all the information.

    I would be paid by the entity requesting the information, and my success/failure rate (based on default rates of the companies I rated) would be made public to evaluate how accurate my ratings were.

    This would be a private business, and the entities requesting the information have a choice as to where to get the information they need. Government oversight is unreliable and often corrupt, so they should stay out of my business and let the market decide if I am a reliable business, and my success should be tied to that.

    Jodi Johnson

  5. Given the low consumer confidence in CRA's currently, there is urgent need to obviate the conflict of interest.

    This can be achieved by having the current CRA's supervised by a monitoring agency, such as an international SEC like agency. Given that the global economy is so closely tied, it would seem reasonable to have an international monitor, who is not paid in any way by the banks/companies being rated.

    Issuer data should be blinded. All of the issuers should be mandated to provide meaningful public and non-public data to be able to provide accurate ratings.

    The rating agencies should continue to be paid as they are currently. However, monitoring agencies should be able to heavily penalize CRA's for inaccurate performance and should even be given the capacity of "disqualifying" underperforming CRA's. On the other hand, consistent and accurate CRA performance should be rewarded, thereby better aligning CRA goals with investor goals.

  6. The conflict of interest is that the ratings agencies are paid by the banks, thereby incentivizing them to give the banks favorable ratings. If the credit rating agency was set up as a subscription, the consumers could subscribe to the CRA in order to get the information they need to make decisions regarding investments, loans, etc. As long as the incentive comes from the consumer and not the banks, the analysts will be incentivized to give accurate ratings that are meaningful to the public consumer. The credit agency would have to go through some type of accreditation process which would allow them to gain access to the confidential issuer information. The accreditation would be issued by a government agency that mandates the type of access they can have based on the accreditation the agency earns. Unless the issuer's confidential information is made public there would have to be some type of government mandate that specified what information had to be made available to which agencies. The agencies with the highest accreditation would have access to the most information, thereby incentivizing the agency to meet the highest level of accreditation, thereby making them more desirable to the consumer. If the consumers are allowed to subscribe to the agency they will pick the one with the most reliable, accurate information. This model would be similar to the consumer reports available for other products.

  7. The proposed rating entity would be entirely independent of input or influence by the businesses to be rated. Strict exclusions of any form of compensation or any other type of inducement would be mandatory for all analysts and managers of the rating company. Internal decision making regarding final rating would be by consensus decisions, with analyst and manger votes of equal value in the final grading.
    The proposed rating entity would be allowed to invest as a corporation in the businesses being rated based upon the score of that corporation of the corporation being rated. The rating entity would distribute profit/loss across all departments dependent on percentage amount of investment compared to other divisions.
    The proposed rating entity would be given highest level of recognition (ie preferred status) by the SEC and other governmental entities. Those companies wishing to do business with Federal governmental branches would therefore want to be rated by the proposed entity to insure access. Additionally, corporations would want to be included for purposes of preferred recognition (similar to JD Power rating of consumer products.)

  8. I believe the most feasible system to implement (ie avoiding more regulations) is to keep it as simple as possible. Have major funds/investors rate the reliatbility of the rating agencies. (similar to Amazon and Ebay customers rating sellers) This makes the risk shared. Companies want a good debt rating. The agencies want a good rating from investors. And now there is some onus put on funds to do more than say "Not our fault - it was AAA rated." Ultimate winner.... Joe Sixpack who put his savings into the fund. Everyone has some skin in the game.
    Justin K

  9. Competing private ratings agency sell subscriptions to their reports. Agencies promote themselves with backward-looking scorecards that benchmark prior reports with performance by rated issues. Agencies with the best performance would have the most subscribers, issuers would provide access or be absent from the ratings of the most popular issuers.

    i - No conflict of interest, agency strives to deliver most accurate information to best serve subscribers.
    ii - Paid by subscribers.
    iii - Access provided as investors demanded a rating from the most accurate agencies.

  10. (posted on behalf of Vaughn):

    There are legitimate roles for the federal government taking complete control of a crucial activity (national defense, enforcement of contracts, getting us to a 16 team college football playoff). Transparent financial markets fall in this category. I would propose a federal agency tasked with objectively setting ratings on companies. Ratings would be voluntary for companies but a government rating would be a legal prerequisite for most financial transactions. The independent agency would have legal authority to verify access to confidential information if a company wanted a rating. This organization would not be politically appointed and would have strict privacy/confidentiality constraints. This would be funded by a user fee/tax paid by the companies desiring rating (or by Warren Buffett if he prefers to donate). Any private organization is at higher risk of bias and/or corruption. Arguments against this solution would include government incompetence and the risk of political influence, however there are legitimate governmental agencies that function well with the right leadership and structure (our military).

  11. Conflict of Interest
    Have the analyst that is evaluating a company incentive base rating provide by the market place. For example, create a Yelp type system where investors would rate the outcome of the rating provided publically. The more realistic rating an analyst provides, the more credibility he or she would have overtime. The individual would start with no incentive until after 1 to 2 years to build the creditability. They will be offered 1st or 2nd incentive only after results would be recorded officially.

    How the company would get paid?
    Moody’s would get paid by Government through taxes that companies would have to pay in order get their ratings issued. This will enable Moody’s to get paid regardless if the outcome of the rating is good or bad.

    How would it access to confidential issuer information.
    The government should empower Moody’s and others the ability to review confidential information. If information is leaked out, Moody’s would be held liable to the company.

  12. the whole concept of a credit rating agency is flawed. people should do their own work and invest accordingly.

  13. When you think of your home or business, you think of a safe haven. We offer florida security systems and comprehensive security.