Ratings Agencies Still Coming Up Short, Years After Crisis | Hacker News Reader

Ratings Agencies Still Coming Up Short, Years After Crisis

nytimes.com

40 pointscinquemb30 comments

I am still flabbergasted that after rating junk swaps as high as possible and being paramount in both misleading investors worldwide and bringing the world to economic crisis that is still being felt, that Moody's and S&P not only continue to exist, but investors still look to them for ratings!!

"Flabbergasted" is the only word.

Also blah blah no one in prison.

By law, many institutions must buy rated paper.

So Moody's and S&P are like Ziploc bags at TSA lines - required by law.

Well there's some sweet regulatory capture.

Keep in mind that there are massive incentives for ... well, everyone, or at least every company ... to mislead these rating agencies, as lending to companies is also completely dependent on those ratings.

And next we ... insist on having competition by these rating agencies. They're being paid by companies selling debt, so effectively they're being paid to get a good rating (and for the attention in the press their ratings get).

This can't be fixed.

Because someone has to do the research, and rating agencies do it better than anyone else. Not saying that they do it perfectly...

"and rating agencies do it better than anyone else." Could you hint us to any evidence for your claim?

You're competing with people who have 6 figure incentives (at least) to reverse engineer how the ratings are derived and take advantage of it in any way possible to drive profits.

I guess the truth is closer to "Ratings agencies do the best they can with the people, technology and incentives they can afford"? How do you change this though?

There is no one else :)

There needs to be many more ratings agencies. The two big ones need competitors.

If there are enough of them then when synthetic CDOs do have bad batches that at least one will want to lead the way in downgrading. It would be much harder to pay off say 10-20 ratings agencies.

There has to be some self-interest among ratings agencies to one up another, by upgrades or downgrades because of better knowledge. I think the lack of competition for ratings is part of the problem that caused 2008. Maybe even ratings based on community as well as authoritative sources, right now they are all authoritative and mostly blackboxes in terms of processes to achieve ratings. The concentration in centralized power in ratings agencies is a bad thing and history says it will bite.

The number of ratings agencies isn't the problem, who pays the agencies is the problem: it's the issuer of the debt in most cases. This arrangement is more efficient than buyers paying for ratings, but misaligns incentives. Supposedly mutual funds do additional research on debt issues beyond looking at ratings, and that might help, but it isn't very public.

Maybe we need some rating agency rating agencies. And some more that will rate those too. If we can go up enough meta-levels, maybe it'll all become so confusing that it'll be impossible to effectively game.

(I get weird ideas when I'm tired.)

I don't think this is a weird idea. In fact, I really like this idea. It seems like a rating agency could be given a "score" based on the historical accuracy of their measurements, since although ongoing books are business secrets I would imagine you could get SOME proxy at least for examining the veracity of the ratings in hindsight. It could be done with very little overhead, and would be a first step in some small sort of accountability, at least, if only an accountability to the data.

Could anyone with an actual familiarity with this domain say if something like this exists, or if there's sufficient audit data for it to exist?

I keep hoping that'll be a role the SEC plays, but so far I've been pretty disappointed.

If we have a bunch of them then aggregate agencies will start popping up and we will have the same problem again.

This problem keeps popping up everywhere, when does a search engine become a content editor / censorship inc.?

We need to have a clearly defined line or figure out a different way of looking at companies like these (that effectively own a 'platform'), because right now we basically have a oligarchy...

If someone, including rating agencies, knew something useful about bonds that were not already priced into the market, then that someone could easily be a trillionaire just by buying and selling. Rating agencies are not populated by trillionaire analysts. Therefore, rating agencies and their ratings are not supplying any useful information.

So why do rating agencies exist? Three of them -- the profitable ones you hear about -- have licenses from the federal government. Those licenses entitle their ratings to a presumption of validity in front of bank regulators and under pension and investment rules. Yes, institutions that the government backs are entitled to invest or blocked from investing your money, whether you like it or not, based on what the rating agencies say.

So they're quasi government bodies with the power to defraud you through the banks and pensions you are mandated to use and to grant government bailout guarantees you will pay for in your taxes. In return, they make lots of guaranteed cash and empower the banking system to skim more and more pension funds and investment fees from you.

The sad bit is that this actually works better than lots of other systems for regulating banking and pensions that have been tried.

Rating agencies are more or less out-sourced research houses - performing work that would be too costly for others to do by themselves. It an "economies of scale" type thing and someone would have do this research. So, to say that "their ratings are not supplying any useful information" is absurd. There is an extremely high correlation between default rates and ratings.

Admittedly some of the fees for ratings are absolutely ridiculous. When you're doing a structured finance deal for $X billion, then 25 basis points is neither here nor there, but in $ value terms is a crap load of money. Rating agencies are cash cows, and that may be part of the problem.

I worked for a rating agency for a couple of years.

A whole lot of comments here assume that the rating agencies are doing this because they're totally driven by greed. I think what people neglect to mention is understand is that rating agencies are overwhelmed by complex securities partly because they're made up of B-players.

In b-school, it's very common to see all the finance A-players gravitate towards PE, banks, HF and prop shops. Then the student sub-layer below them go to boutique versions of the aforementioned entities and it's usually Group B.

Believe it or not, this has a harsh impact on the culture there. Ask anyone you know who works at a lower level bank about how pointlessly harsh the culture is. My hypothesis is that: being made up of B and sub-B professionals, rating agencies just don't have the in-house intellectual capacity to do the levels of diligence they'd be expected to do to prevent the A-players in banks from cheating the market

Read "The Big Short" to see exactly how true this is.

If we were holding journalists, economists, meteorologists (or climate change specialists) accountable for their forecasts, we would need to build many new prisons. People need to be realistic about the capacity of rating agencies to predict credit defaults. All they can do is express a judgement on the credit worthiness of something from the outside.

What I found working with them is that the new regulations are rather making the process more error prone. Rating agencies are now paranoid about any communication with financial institutions and you just can't pick up the phone and call them. It leads to them going away, making their own assumptions and interpretations and making it very easy for them to make a mistake that cannot be spotted by the people who understand the numbers.

Forecasts? Why you are talking about forecasts?

When Moody's says a particular investment has a "AA" rating, it means it has a certain probability to make a loss over its lifetime. It is effectively a forecast of expected losses.

This sounds like a problem where transparency is the solution.

Post the formula for grading and let the public draw their own conclusions.

The ratings agencies get access to the books under nda. It's tough to tell if they're doing their jobs. Probably safest to assume they're not.

So if the S.E.C. can't punish these agencies forcefully enough to seriously damage their margins, this may never change as they'll have little incentive to shape up.

Stocks are actually way too complex things for the average person to understand, otherwise there would have been no need for Moody's or S&P ratings, just the Amazon/Ebay ratings would have done the job!

However, the "average intellectual" person is evolving and getting tech savvy at an unprecedented rate globally. I am willing to bet that a stock rating system similar to ratings on Amazon/Ebay is not that far in future.

This has nothing to do with stocks. The role the ratings agencies played in 2008 was about bonds.

I also think it's unlikely that bonds see an Amazon-style ratings system when the market for a lot of these specialized products is three pension funds in Dusseldorf. We're talking almost entirely about institutional investors.

For any bond with reasonable trading volumes the ratings are irrelevant, because you can just look at the spread over the relevant benchmark - UST, Bund, JBG, Gilt etc. The market tells you how risky it is; think back to 2010 and the way Greek-German govie spread ballooned out. The problem is with illiquid bonds. Some credits don't trade frequently, so a last trade price may be days or weeks old.

This article and similar debates seem to be taking place under the assumption that investors actually trust the credit rating agency ratings, a notion which I see very little evidence of. Even before the financial crisis, most investors were creating funds and investment strategies without regard to the ratings given by the government-backed ratings agencies (GBRA); the GBRAs have basically only been a rubber stamp required to determine reserve requirements and other regulatory issues. For example, if investment banks had really cared about GBRA ratings, they would get ratings before packaging together loans, not after.

If I am right, then GBRAs are not to blame for the financial crisis, except for the fact that they may have given the public and the regulators false confidence in the credit-worthiness of borrowers. This view really says that the GBRA are worthless overhead, which only serve to help the government pretend that regulators know what is going on and are protecting the citizens.

Uninformed question for the more informed on here: What would a "startup rating agency" look like?

The problem is the incentive, not the agency itself.

Unless you find a solution for that you'll end up just as useless as the existing agencies.

Reply on news.ycombinator.com