FORTUNE -- A large part of the story of the financial crisis is that the big three rating agencies -- Standard & Poor's, Moody's, and Fitch -- gave top credit ratings to the complex securities that helped fuel the credit bubble and subsequent bust. Some former ratings agency employees told Congress that they inflated ratings to satisfy their paying customers, the bond issuers who were stuffing structured bonds full of high-risk, and sometimes fraudulent mortgages.
With amorphous financial regulatory reform legislation now casting a shadow over the rating agencies, McGraw-Hill, the parent company of S&P, has seen its stock fall 15% since the beginning of this year and nearly 60% from its pre-crisis peak. As politicians try to sort out the mess and regulate the industry, Deven Sharma, the president of S&P, agreed to speak with Fortune about what went wrong at his firm.
Sharma, who became the head of the McGraw-Hill (MHP, Fortune 500) division in 2007 as warnings about a sub-prime mortgage led meltdown were heating up, has seen his short tenure marked by lawsuits, fraud allegations, and internal turmoil. Here, he denies that issuers influenced S&P during the boom, says that regulators need to stop mandating the use of his company's ratings, and explains how making it easier to sue rating agencies will impact companies that need cash.
In a recent note from S&P, the firm says that all of its credit ratings performed inline during the crisis, except for structured securities made up of real estate assets. What happened in the division that oversaw structured finance?
There is no doubt that this was an analytical miss. [Since the meltdown] we've focused on what could have been done differently so we wouldn't have missed what was going on in housing. The fact of the matter is we were looking at home price declines. What we didn't foresee was the severity and the pace at which declines actually happened.
Had we thought that home price declines would be double what we assumed they would be, we wouldn't be where we are today. We'd have better performance in mortgage-backed securities. But even Warren Buffett in his response to the Financial Crisis Inquiry Commission said that no one saw it coming. And he's a smart guy, as are the many policy makers who also missed the problems. That doesn't make it right. We did miss the problems.
There were people who didn't miss the housing bubble you say the entire market overlooked, including large well-known mutual fund managers like Bob Rodriguez of First Pacific Advisors, who were vocal about a severe downturn exacerbated by mortgage fraud. Are you now more likely to engage in debate with someone like Rodriguez or a hedge fund manager like Steve Eisman or John Paulson over your assumptions?
I'll offer three observations. First, engagement with investors and others for different points of view is very important. We may not agree with them, but they do give us deep insights, and we've systematized the way that information becomes part of our process.
Ultimately we want to know what they think of the market and of us because if they believe we have something to offer them we'll thrive, despite any regulation.
Second, you want to talk to as many players as possible for a better sense for where risks are popping up because the market is very interconnected.
Third, as part of our checks and balances, we have a new risk management group that reports to me. Part of its role is to look at market changes and ask how it will influence credit risk, and how to bring those different points of view together without overreacting to one minority viewpoint.
But aren't those minority points of view very important? Why not model out even the worst-case scenarios, even if you think they're unlikely?
We have begun to publish stress scenarios alongside our assumptions and credit ratings. Investors can then debate with us over whether our final assumptions are right. And we've already begun to publish alternative scenarios in RMBS.
About a year ago we decided that our ratings had to be more forward looking. We formally recognized that the market expects stable ratings and that we had to be clear about how we came up with our ratings. We decided that our ratings must be comparable. Triple-A ratings must mean the same thing no matter what kind of security they rate, and every rating category must withstand a certain stress scenario. For example, triple-A securities must be able to withstand the Great Depression stress scenario and not default, whether it's an RMBS or a sovereign bond.
Do you think you're giving fewer triple-A ratings because of this?
We've been making a lot of rating changes in CMBS, RMBS, CDOs and CLOs. The answer is yes.
When we did this we got a lot of pushback from parts of the market that then accused us of going overboard. But we see economic stresses that could lead us to strong declines. Commercial property values in some places are declining by 25% to 30% levels, and those declines could greatly impact CMBS ratings.
About the idea of pushback from the market, aren't the agencies there to give the market that reality check and ignore pushback?
Did rating agencies feel pressure from issuers to give better ratings than they wanted to give to structured securities?
Let's look forward. As we have made the criteria changes, sell side analysts and others have raised questions about criteria, but we'll stick with it. We've put a stake in the ground about our ratings being forward-looking, comparable, and stable. We've chosen the stress scenarios we've chosen. As a consequence there have been deals we've not been asked to rate. We still published commentary on those deals and we'll continue to do that because we didn't think that risks were fully captured in other ratings that were done.
This allows all of us to make public our viewpoints and avoid rating shopping.
But did your firm feel pressure from issuers to give better ratings during the boom?
Both issuers and investors have different needs. Issuers may want higher ratings and investors will want lower ratings because lower for the higher yield. All things being equal, our job is to be independent.
We have to demonstrate that our firewalls between our analytics and commercial activities are intact.
Do you need to show that firewalls are strong because influence leaked in during the boom?
If you dissect why we missed some of the downturn, clearly some of the data we were getting from issuers about mortgages was a factor. Clearly the volume levels made people feel like risks had gone down. When you say pressure from issuers, issuers always want higher ratings. We've always pushed back and said we're going to say what we think is right.
Even during the same period, asset-backed securities made up of credit card receivables continued to perform as expected during the downturn. This happened amid the same culture, value systems, and types of analysts.
Let me ask in a different way. Managing directors at rating agencies have testified that they were pressured to give good ratings to make issuers happy. Did that happen?
The answer is no. I have seen people say they were pressured to give good ratings, but I never said to do that. And Kathleen Corbet [who was the president of S&P until 2007] testified that she never gave that order.
What would you say to Frank Raiter, a former S&P employee who has many times said that issuers pressured your firm's management to give higher-than-deserved ratings on structured securities? You've said you can understand how people could think that this was the situation.
In our testimony before Congress we've said he's wrong. The real point that he makes is that we weren't investing in our models. I wasn't here at that time, but I've looked at the facts and seen that we've modified and adjusted the model many times, and we continue to do so. To your point about the issuers pressuring us, when I engage with people internally, I don't get the impression that people were pressured to give higher ratings.
Another question raised by the hearings is whether S&P had enough staff to keep up with demand for ratings on structured securities during the boom. Were analysts stretched too thin to scrutinize data from issuers?
At that point in time we were adding a lot of people and bringing in resources as we needed. Looking forward, we've instituted an analytical certification program because we think analytical capabilities need to be refreshed. We also did this to address people who kept asking whether our analysts were capable of rating structured securities. The program was created with the NYU Stern School of Business.
As you mentioned, there have been cases where issuers were giving you faulty information about the security they wanted you to rate. Will S&P start relying on third party sources to vet issuer information?
In the RMBS area we've already strengthened our criteria for due diligence. We are focused on the quality of due diligence done by our issuers on the underlying collateral of securities. We've also expanded the data sets we look at and we're using some of the new data providers that look at information from issuers. Where there is a need for third party services, we will use them. The point is, we will get access to more data and underlying information.
The reputation of our analytics is critical for us. We've made a lot of internal changes - analytical changes, addition of new checks and balance, new risk management procedures. We've redefined some employee roles and have made new hires. I'm proud of how people have stayed focused and delivered.
Of all the new rules for rating agencies proposed by politicians, are there any that you think get it right?
I spoke at [a securities]conference last week and was encouraged by a discussion about regulation. At the end of the day, regulation influences capital flow, which influences economic growth. There is a strong link between regulation and economic growth. Regulators there talked about educating legislatures around the world about that link. Legislation that may make sense a single country could work at cross purposes or undermine rules in other parts of the world. Regulation needs to make sense globally, even that passed in the US.
Is all regulation of the industry good? Most of the regulations that calls for greater accountability, greater oversight, and greater transparency will do a lot of good. It will create confidence in the marketplace and support recovery.
What is disconcerting is the feeling in Washington that increased accountability is the same as increased liability. But these aren't the same things. In terms of regulation, we question having a discriminatory liability standard for rating agencies compared with the rest of the market. Of course as a business we'll cope with it if it's passed, but it will impact the marketplace.
What will the impact be?
There are some securities at the lower end of the rating scale that are more volatile that we'll have to take a hard look at when we decide whether to rate them. These are typically new companies or private equity sponsored companies that are at the lower end of the rating scale.
Are you trying to say that you wouldn't rate them because they're at the lower end of the rating scale, more likely default, and thus more likely to get you sued by investors who get burned?
I don't know if we wouldn't rate them, but at the end of the day there would be an impact on access to capital and at what cost a company can access that capital. Anytime liability goes up, you take a look at operations internally.
Should ratings be taken out of regulations and things like swap agreements for capital requirements for banks?
We have a clear public position on this. We believe ratings play an important role as a trade risk benchmark. However, there should be no mandate that requires the use of ratings. When ratings are in regulations, the marketplace is required to use them. Rather, we should have to convince investors that our ratings have an analytical value add, just like we do when with our indexes. No one forces an investor to use the S&P 500 index over the MSCI or FTSE or Russell.
If regulators aren't going to take ratings out of regulations and contracts, then they should make it clear that investors should be looking at other benchmarks as well so that the rating is not used as an investment recommendation. That's not what ratings were intended for.
Inclusion in regulations means that there is a captive market for your ratings. Would customers have abandoned S&P during the height of the crisis had they not had a mandate to use your ratings?
It's hard to speculate. Now we intend to focus on convincing investors that our risk benchmark is useful in their decision making. If we can provide that product, we'll do fine.
The mortgage market was a huge miss. Other than that, all of our ratings have performed as expected in this environment. Our other businesses, like indexes, continue to be well recognized and received. In equity research, about a dozen of our researchers received a best of the street award last year, the same number as Goldman Sachs (GS, Fortune 500). Goldman does research for institutions, and we do it for the private wealth segment, where supposedly research is less rigorous.
And of course Capital IQ continues to do extraordinarily well. While we get a lot of criticism, people still want to listen to the point of view we have to offer. I tell our people that we still have opportunities to serve the role of bringing transparency to the marketplace. People still want to engage with us.
Moody's (MCO) and S&P have been called a duopoly, most famously perhaps by Warren Buffett, who used this as a rationale for investing in Moody's. Will new companies jumping into the credit ratings fray be able to break up this duopoly given the reputational damage that bad ratings have done?
With our criteria and analytical changes, we want to demonstrate the value of our ratings and show why it's different and better. If we can demonstrate this, we will create more demand even if there is more competition. There will be different business models and we want the market to be able to fairly choose which model it believes in.