The Role of the Credit Rating Agencies

   In the wake of the crisis of 2007-08 and the ensuing Great Recession, there was a torrent of criticism of the credit rating agencies, such as Standard & Poor’s and Moody’s, which had given their highest ratings—AAA—to mortgage-backed securities (MBS) and collateralized debt obligations (CDOs) that defaulted in massive numbers. One popular explanation for the faulty ratings has been that the agencies gave inflated rating to securities because they were paid by the issuers; if an agency declined to give a AAA rating, it wouldn’t be used and wouldn’t be paid. This explanation has been questioned by economists. I personally think (along with some economists) a better explanation is that the rating agencies were working from the same faulty assumption as everyone else: that a substantial decline in home prices was highly unlikely. Put slightly differently, the rating agencies (and almost everyone else) did not believe that there was a housing bubble.

   I’m not equipped to sort out those issues. Rather, I want to recount my personal experience with the role of the rating agencies in the MBS world immediately before the tsunami hit. I didn’t negotiate with any rating agencies and all my experience related to prime and alt-A residential mortgage securitizations (RMBS), so my observations may not hold for CDOs and subprime MBS.

   There are only four rating agencies: Standard & Poor’s (S&P), Moody’s, Fitch, and Duff & Phelps. S&P and Moody’s are the big fish, Fitch a smaller fish, Duff a minnow. An RMBS issuance needs ratings from two agencies, and no one would proceed with a RMBS rated only by Fitch and Duff. Which means that every securitization had to have a rating from either S&P or Moody’s. What’s more, investors occasionally insist on ratings from both the big guys, so that issuances rated by both S&P and Moody's were a bit more saleable. Under these circumstances, an issuer’s power to negotiate for higher ratings was limited.

   A RMBS issuance consists of a number of classes with different ratings from AAA to junk. As homeowners default on their mortgages, the lowest-rated class takes the losses until it is wiped out, then the next lowest-rated class, and so on up the ladder. In rating RMBS, the rating agencies principal decision is the level of protection that will justify rating a class AAA (or AA, or A, and so forth). That comes down to a decision as to how much padding there should be under each class. Should 5% of the issuance be rated below AAA, or should it be 5.5%?

   In my experience, issuers shopped for ratings, but that shopping was limited to how much of the issuance would be rated AAA. The decision was important to issuers because AAA classes fetched higher prices than lower-rated classes. But as a percentage of the offering, the differences between rating agencies was small, on the order of tenths of a percent. So in a $500 million offering, one rating agency might be satisfied with a $24 million cushion of below-AAA classes while another might want $26 million. A financial crisis is not made of such differences.

   As I said, my experience was with prime and alt-A RMBS. The ratings of CDOs, which were more central to the crisis, may have had a different dynamic.

   —Stan

The Earned Income Credit

In late June I posted a piece about the minimum wage. I was generally opposed to it, partially because of the possible effects on employment—I thought some people would lose their jobs—but also because I saw it as a reflexive, feels-good approach to an issue that requires serious thought. At the end of the piece I noted that some policy wonks had proposed an expansion of the earned-income credit (EIC) as an alternative to the minimum wage, and I said I would write something about the EIC one of these days. That day is here.

But first I want to talk about a personal interest in the minimum wage. My younger son refused to go to college, despite being probably the brightest one in the family. Instead, he spent a year at a technical school learning to be an auto mechanic. (He’d previously done a lot of mechanic’s work fixing beat-up vehicles at a Colorado wolf sanctuary where he spent a few years.) Course completed, he went back to Colorado and took a job with an auto dealership as a “lube tech”—the guy who changes your oil, rotates your tires, and does the safety check. All for $9 an hour, raised to $10 after his third month (though he does get to buy into the company health plan at a below-market price).

It’s not enough to live on, even for a single young man. The carrot is that eventually—though it could take several years—he would move up to a real mechanic's job, where the pay is better though still not great. In the meantime, he’s working Real Hard. It’s not like bagging groceries. For one thing, to do the work well you have to know quite a few things you don’t just learn on the job. And it’s hard physical labor, not to mention that you’re on your feet almost all day. On top of all that, it costs a bit of money just to be a mechanic; it’s the mechanic, not the employer, that supplies most of the tools, and lube techs may have to pay for damage done to cars, such as scratched paint, during oil or tire changes. Both the tools and the damage can cost thousands. Finally, doing it badly, especially the safety checks, could get someone killed.

So, despite my principled objections to the minimum wage I might be in favor of Colorado enacting a $15/hour minimum as Seattle recently did. (Not that there’s much chance of that; only this past January 1, Colorado raised its minimum hourly wage to—drum roll—$8.) On the other hand, a raise from $10 to $15 might get my son unemployed; people might want fewer oil changes or tire rotations, or pay their neighbor’s kid to do it. I don’t think there would be a precipitous fall in lube tech employment, but it’s quite a risk when you consider that the job’s main attraction isn’t the pay but the opportunity for advancement. Think of lube techs (and perhaps many other low-wage workers) as extremely well-paid interns.

But why does a hard job that requires some technical expertise pay so little? The economics texts tell me it’s because there are lots of people who are able and willing to do the work for that pay. Supply and demand, if you will. Apparently there aren’t that many attractive jobs for intelligent high-school graduates. Still, it’s not clear why they should earn no more than sales clerks, who requires minimal training. Puzzling.

And now for the EIC. An EIC has the government pay low-wage workers a supplement for each hour they work. Right now the EIC is paid through tax code refunds, and participation and benefits are limited. Current proposals are to open it up to all low-wage workers and to make it more generous.

The attraction of the EIC is that there is no danger that it will increase unemployment. Indeed, it seems likely that it will decrease unemployment for the workers it covers. That’s because part of the EIC goes to employers rather than employees.

To see why employment might increase, suppose that the government starts to pay an EIC of $2 an hour to bakers, who previously earned $8 an hour. At $8 an hour, bakeries were just able to fill all their openings. (If they couldn’t, bakers’ wages would have risen.) But with a $2 EIC, the employer can now fill all the bakers’ positions while only paying $6 an hour, with the EIC raising the bakers’ total compensation to $8. But an employer who now pays a wage lower than $8 is likely to want to hire more bakers, since he can now sell more bread at a lower price, and this will raise the wage above $6.

Suppose that as a result of the EIC, the bakery winds up paying an hourly wage of $7—$1 lower than it paid without the EIC—while the bakers enjoys total compensation of $9 an hour($7 wage plus $2 EIC) , which is $1 more than they received before the EIC. The EIC raises bakers’ compensation while lowering bakers’ wages, with the difference being made up by the EIC. Both bakers and bakeries are better off, and more bakers are employed.

We don’t know a priori how much of the EIC will go to the bakery (in the form of the lower wage it has to pay), and how much to the bakers (in the form of a higher post-subsidy wage rate). That depends on market responses to the introduction of the subsidy. But statistical studies by economists indicate that employment has in fact responded to increases in the EIC, suggesting that the EIC has allowed employers to pay a lower wage.

Like the policy wonks, I prefer an EIC to a minimum wage. But most of the popular writing on both proposals refuses to come to terms with a critical question: Why do some workers earn less than others? (It’s the same question I wondered about for my son the lube tech.) Folk wisdom, too often reflected in media accounts, holds that employers are mean-spirited. The economists’ answer is that workers get the value of their marginal product; if that product is less valuable than the workers’ pay, their employer will fire them, lower their pay, or go out of business; if the workers’ product is more valuable than their pay, other employers will bid them away. Obviously, nothing in real life works that smoothly, but it’s hard to deny that the relation between the value of work product and work compensation exerts a force that over time will bring wages into line with what people are willing to pay for the workers’ product.

The direct cost of the EIC supplement will be borne by taxpayers, and there will also be indirect costs in economic efficiency: Workers will move between jobs to reflect the altered compensation landscape. Still, these efficiency losses should be less than with a minimum wage, and the benefits to low-wage workers higher.

—Stan (with help from an actual economist)

Corporations in the Dock: Addendum

Bank of America’s nearly $17 billion settlement with the Justice Department, announced yesterday, has all the evils I blogged about previously, plus a new ingredient. Once again, no individual is cited as having behaved criminally (though it has been reported that the government is getting set to charge Angelo Mozilo, Countrywide Mortgage’s former CEO). And once again, no thought seems to have been given to compensating investors who were victimized by BofA’s supposed criminal acts, though the government will be compensated—and then some—for money HFA and HUD lost insuring dud mortgages from BofA.

The new wrinkle here is that most of the conduct cited by the government was committed by Countrywide and Merrill Lynch before those companies were acquired by BofA in 2008. It’s standard corporate law that a surviving company is liable for the acts and obligations of an acquired company, but that makes little sense in the criminal context.

One possible explanation advanced for penalizing stockholders for a company’s sins is that the prospect of criminal penalties will encourage stockholders to prevent illicit conduct. Given the limited abilities stockholder have to influence corporate conduct, it’s not much of an argument. But when the conduct occurs before an offending company is acquired (and the acquisition is for cash so that stockholders of the acquired company do not become stockholders of the acquiring company), the stockholders of the acquiring company could not possibly have prevented the criminal acts.

—Stan

Corporations in the Dock

In the last few months various federal and state agencies have gathered up enormous fines from the big banks. The Economist estimates the total at $35 billion, with Bank of America (rumored in the $15 billion range) and others still to be heard from. Among the recent contributors to the public fisc were JPMorgan Chase ($15 billion for dicey mortgage sales and foreclosures), BNP Paribas ($8.9 billion for money laundering), and Citigroup ($7 billion for misleading investors).

So the question for today is: Can the banks have done any of these things? Or more generally, in what sense can a corporation be said to do anything? After all, a corporation is simply an organizational form for doing business. It doesn’t have beliefs or desires, and it cannot act except through individuals. So when we say that a corporation “did” something—laundered money or lied to investors, for example—what we really mean is that individuals acting for the corporation did the laundering or lying. Saying that the corporation did something is merely a façon de parler: useful in many contexts, misleading in others.

There is a legal doctrine, respondeat superior, that makes a business owner (including a corporation) responsible for the tortuous acts of an employee. For example, if an employee driving a truck on the owner’s business negligently injures a pedestrian, the owner can be liable for the employee’s negligence. It’s not that the employee is excused; both he and the owner are liable. But it’s usually the owner that gets sued because that’s where the money is.

Early in the twentieth century the concept of respondeat superior was ported over from tort law to criminal law: A corporation could be criminally liable for the criminal acts of its employees. A corporation can’t go to jail of course, but it can pay a fine, lose a governmental license, or have certain business practices enjoined. Much later in the twentieth century, in the wake of Enron and other gaudy scandals, the penalties for corporate crimes were greatly increased. (An article by Joan McPhee, a partner at the Boston law firm of Ropes & Gray, explains the evolution.)

Of course, when we say the corporation pays a fine, loses a license, or has its conduct curtailed, we should understand that we’re really talking about consequences for individuals. For the most part, it’s the corporation’s stockholders, not “the corporation,” who bear the costs. (Similarly, if someone forges your name on a check, it’s you and not your checking account that will be the loser.) But why should the stockholders suffer for the illegal act of a corporate employee? There are a number of explanations, not all equally plausible.

First, if the employee’s conduct benefited the corporation, as it might where a corporate customer is defrauded, it seems only fair that the corporation/stockholders return the ill-gotten gains. But this explains only a small part of what’s been going on recently, where no one bothers to reckon how much the corporation may have profited and little thought is given to paying over the fines to the victims of the employees’ acts.

A second explanation is that the guilty employees usually don’t have the resources to fully compensate the victims. But this argument would never pass muster outside the corporate context: If your neighbor defrauds someone but can’t fully compensate his victim, should you have to pay the difference? In any case, little of the fines go to the ostensible victims of the fraud or other illicit conduct.

A better reason for making stockholders pay is that the possibility of punishment will encourage them to prevent illicit conduct. In a modern US corporation, the stockholders elect the company’s directors, who in turn elect and monitor the company’s officers. Unlike the directors, who only serve part-time, officers are full-time employees who run the day-to-day affairs of the corporation, including hiring and firing lower-level employees. If the stockholders can be punished with heavy fines, they will elect responsible directors, who will in turn appoint responsible officers, who will in their turn institute measures to prevent criminal conduct, including more closely monitoring the corporation’s employees.

Corporate officers don’t like to explain to more senior officers or to directors why their operations incurred governmental wrath, not to mention heavy fines. But the problematic conduct usually would have occurred up and down the line: Employees and officers engaged in illegal conduct, and more senior officers, directors, and stockholders were not sufficiently attentive. However, as one moves from employees to officers to directors to stockholders, the ability to control the conduct of underlings, and the degree of culpability, rapidly erode. So why does the least culpable group, the stockholders, take nearly all the hit?

One reason is that the process is controlled by prosecutors and corporate officers, and it is in their mutual interest that things work this way. One manifestation of this mutuality is that while corporate crimes must be committed by the corporation’s employees, the guilty individuals usually suffer no punishment more grievous than losing their jobs. (Their legal costs are almost invariably paid by the corporation—that is, by the stockholders.)

For a prosecutor, pursuing corporate employees for criminal activities raises all sorts of problems. For one thing, it requires amassing detailed evidence as to the employees’ activities. Moreover, juries are much more likely to punish an abstract entity like a corporation than a flesh-and-blood employee, so indicting individuals can prove embarrassing. (In the Tap Pharmaceuticals case, for example, the company settled for $875 million and an onerous corporate integrity agreement, after which a jury acquitted the 13 corporate employees whose supposed crimes were the basis for the charges against the company.) And massive fines, of a size that no employee could pay, are profitable for the government and produce headlines that can make a prosecutor’s career (the Rudi Giuliani effect).

On the corporate side, no one wants to go to jail, least of all well-paid corporate directors and officers. Far better to take the question of individual guilt off the table and lumber the stockholders with the fine. (Directors and officers are usually stockholders too, but they bear only a miniscule percentage of the economic damage.) Most important, a criminal conviction—even an indictment—can be a death knell for a company. (The accounting firm Arthur Andersen collapsed after being indicted. Its subsequent conviction was posthumously overturned by a unanimous Supreme Court.) Better to settle for a sum that, while impressively large, does not threaten the company’s existence.

The result is that corporate criminal prosecutions have turned the Justice Department into a kind of protection racket. Since no individuals are charged and the corporation cannot risk an indictment, let alone a conviction, prosecutors do not have to put any facts before an independent judge and jury. Instead, the prosecutors and the corporation simply agree on a dollar settlement that will be large enough to burnish the prosecutor’s resume while permitting the corporation to survive without having any executives go to jail.

Viewed this way, corporate criminal prosecutions resemble certain types of civil actions. In most states there is “strict liability” for death or injury caused by defective products. That is, the victim needn’t prove negligence in the design of the product, only that it was defective. Even though no one at the manufacturer was derelict in their duty, the corporation is still required to compensate the victim. The usual explanation for such a rule is that where someone is victimized by a faulty product, it’s the corporation that was best placed to avoid the injury by preventing the product defect.

Whatever you may think of strict product liability, the similarities with corporate criminal prosecutions are dwarfed by the differences. Most obviously, a plaintiff in a product liability case will have to prove to a court that the product was truly defective, that the defect caused the injury, and the amount that will adequately compensate for the injury. In a corporate criminal prosecution, in contrast, prosecutors will not have to prove that a crime was committed or the amount of the damages because the corporation cannot risk going to court. Moreover, no part of the settlement need go to the ostensible victims. (As Floyd Norris noted in The New York Times, in Citigroup’s recent $7 billion settlement of charges that it misled investors in mortgage-backed securities “there seems to have been no effort to quantify just how much Citigroup’s improper behavior cost investors,” and “there is nothing in the settlement to benefit most of those affected.”)

It’s understandable that prosecutors will want to raise money for the government and to make their careers, and that corporations will want to avoid oblivion. But what makes this situation possible is public anger over particular corporate scandals such as Enron and, more generally, over the recent recession. But anger is a poor guide to action. Anger over 9/11 leads us to blame Muslims generally, with results  both silly—stopping a mosque near the World Trade Center site—and tragic—invading Iraq. In each case there is what we might call a metaphysical confusion, which leads us to punish innocent individuals who are connected in our minds with an abstract entity, Islam or the corporation. (Strangely, many people who attack the Supreme Court decisions in Citizens United and Hobby Lobby by denying that a corporation can have speech or religious rights seem willing to accept that a corporation can have the mental state—mens rea—requisite for the commission of a crime.) The result in the corporate context is a corporate version of lynch law where, without any determination by an independent fact-finder, innocent stockholders are hung, drawn, and quartered.

—Stan