Skip Page Banner  
Skip Navigation

Dodd-Frank Section 953(b): Why It Is a Legislated Disaster

Wednesday, May 23, 2012

By John H. Lowell, P-Solve Cassidy

In my experience, what I would describe as reactionary laws are bad laws. In other words, when Congress runs across a particularly unforeseen problem, it has a habit of over-legislating in an effort to solve that problem. I'm not saying that every bill that is written this way is bad, or that even among the bad bills that the entirety of every bill is bad, but reactionary bills tend to have some horrible provisions. Perhaps this is why we are burdened with Section 953(b) of the Dodd-Frank Wall Street Reform and Consumer Protection Act.1

The requirements of this provision seem simple enough.2 Each issuer [of a proxy] is to provide three items:

• The median total annual compensation (as defined generally for proxy purposes) of all employees, not including the chief executive officer (CEO),

• The total annual compensation (TAC) of the CEO,

• The ratio of that median compensation to that of the CEO.

The third number is intended to be the key. It is supposed to give the reader of the proxy a hint as to whether the CEO is appropriately paid or not.

This subsection and most of the rest of Dodd-Frank were written in reaction to the financial meltdown of the last several years. Executive compensation was viewed as a big problem. Think of AIG, Bear Stearns, Fannie Mae, and Freddie Mac and you can see where the drafters were coming from. From where I sit, the problem is that they didn't understand what they were writing and Congress probably still doesn't understand what it enacted.

Reaction to Section 953(b)

I'm not the only one who thinks Section 953(b) is overly complicated. The bill was signed into law on July 21, 2010. As I write, roughly 20 months have passed. The Securities and Exchange Commission (SEC) has issued no guidance and has said that this provision will not be enforced until the SEC has written rules.

I'm also not the only one who thinks these rules are too cumbersome. Hundreds of commentators have expressed their opinions to the SEC about the problems inherent in the application of Section 953(b).3 While many commentators write in favor of the subsection, those who have commented on the actual provision have been, by my reading, unanimous in thinking that literal application of this provision will be an exercise of extremely large corporate expenditures compared to the value provided to investors.

As much as I don't like the provision, however, to my knowledge, none of the commentators have developed a potential solution that remains truly faithful to the stated intent of 953(b). Virtually all do a good job of simplifying, but the result will frequently not resemble the outcome that would be achieved with literal adherence to the law.

In this article, I will first address some of the technical issues with 953(b) as written. After that, I'll assume that the provision stays intact and discuss some things that companies might do to prepare for it. Here are the major technical issues as I see them:

• The pay ratio is backwards.

• Executive pay is more volatile than broad-based pay.

• Global companies have significant issues related to having employees in multiple countries.

• There are three common measures of average (mean, median, and mode). The median of a large group for a complicated statistic is neither easy to develop nor to approximate.

• Companies that sponsor defined benefit plans will have additional complications.

• Companies with defined benefit plans in multiple countries have additional complications that may provide a biased basis of comparison.

Pay Ratio Is Backwards

This sounds like a silly complaint, but really it's not. Here's why.

Most readers of proxies are not math people. That is, they don't have a tremendous facility with numbers. I'll show you with a few examples why it makes more sense for the ratio that is expressed in the proxy to be (B)/(A) rather than (A)/(B).

In the first example, let's make the math work nicely. Suppose the median total annual compensation (MTAC) for Company A is $50,000 and that the CEO total annual compensation (CTAC) for Company A is $5,000,000. Then, the ratio that I recommend produces a result of 100:1, or perhaps simply 100. People know what 100 is. They can get their arms around 100. They can count to 100.

How about the ratio prescribed in the bill? In this case, it is 0.0100 (I am using 4 decimal places for a reason). I know what that ratio means. You probably do as well. But, to the average American who stopped taking math classes the first chance that he or she had, that's a funny number. Is it one one-hundredth? Is it one one-thousandth? Is it something else? What does it mean?

Let's make example two a little bit less friendly, but perhaps closer to reality. Here, we'll keep CTAC at $5,000,000, but let's increase MTAC to $60,000. Now, the friendly ratio is approximately 83. Again, you know what 83 is and even the typical proxy reader has a pretty good handle on what 83 is. The other ratio is 0.0120. What is that? Is it better than 0.0100? Is it worse? Is it more? Is it less?

In fact, I wanted to know what Americans think. I took a sample of 50 Americans, all with college degrees, all between the ages of 25 and 55, and all with white -collar jobs. I asked them the following question in writing so that they could think about it: If I calculate the ratio of median total annual compensation to CEO total annual compensation for two companies and the first one has a ratio of 0.0100 while the second has a ratio of 0.0120, which has a greater disparity between CEO compensation and rank-and-file compensation?

Only 23 of 50 answered correctly. Perhaps that is a reflection on our educational system. Perhaps it is a reflection of the way I phrased the question. But, while I didn't ask the question, I would bet that most of the same group would have answered correctly if I had used my preferred ratio.

Now, let's go to example three, a far more realistic one where CTAC and MTAC are not such round numbers. Let's make CTAC $7,654,321 and MTAC $58,472. In this case, my preferred ratio is approximately 131, while the statutory ratio is 0.0076. If only 23 of 50 people that I surveyed could answer my question correctly when the numbers were “nice,” how many of them do you think would have a clue when we start getting ratios like 0.0076?

I think this one is a simple fix, however. If a company can provide one ratio, it can provide the other. The SEC could simply ask issuers to provide both.

Executive Pay Is More Volatile

For almost all publicly traded companies, the TAC for the CEO and other top executives is much more volatile than it is for rank-and-file employees. It would not be unusual for the CEO to have a TAC composed of all of these elements (and more):

• base pay,

• bonus (annual incentive),

• long-term incentives, including stock grants and awards,

• increase in the present value of accumulated benefits in a broad-based pension plan, and

• increase in the present value of accumulated benefits in an executive pension plan.

In many cases, base pay will be dwarfed by those other components, each of which we would expect to have significant variability. For the rank-and-file, we would expect that base pay will be the primary component of TAC, especially as we get close to the median employee.

In 953(b) terms, what does this mean? We'll address the pension issues later, but with regard to the incentive payouts, hasn't there been a pronounced movement to tie executive compensation to corporate performance and shareholder return? So, in situations where shareholders have been rewarded, does it not make sense that the statutory ratio would be smaller (my preferred ratio would be larger)? How would shareholders view this? How would other interested parties such as labor unions view this?

Further, this TAC is generally measured from the end of the fiscal year to the end of the next fiscal year. Suppose the stock markets have a particularly good or bad day on Dec. 31. This could have a significant effect on CTAC, but will it then produce the type of ratio that the drafters sought?

Some companies make significant incentive payments to both executives and rank-and-file employees. If they use equity for these incentives, the calculations just got unnecessarily complex.

Consider how the median TAC of a group of 50,000 employees is determined.

(1) Determine the TAC of each of the 50,000 employees separately.

(2) Rank them in order.

(3) Determine the 25,000th employee from that group.

So, we can't just determine the value of the equity grant or award for one person. We have to do it for every one of the 50,000 that received a grant or award. Then, we add that number in with the other components of their TAC and rank them. That takes a lot of time, although a grant of one share on a set of terms on a specific date generally has the same Black-Scholes4 value for one employee as it does for another. It takes a lot of time, but does it provide a lot of value?

Global Companies

This is a legislated disaster of the most extreme level. For an American company with significant overseas employees, why does it matter what the ratio of the compensation of the median employee is to that of the CEO. This is not information that I would consider worthwhile. For those who disagree and consider the pay ratio worthwhile, let me re-frame this: how much would you be willing to pay for that information? Or, stated differently, how much should an SEC registrant have to spend to produce that single statistic? If it costs a company $100,000, is that too much? How about $1,000,000? You may think that these figures are ludicrous, but they may not be. It may cost this much to produce a single data point that, in my opinion, is of dubious value.

Let's consider a company in the tobacco business. Setting aside the issue of smoking, tobacco is a long-standing industry in the United States. Shareholders of tobacco companies are betting that those companies will have sustained profits.

I'm not going to pick a particular tobacco company. I would guess, however, that any of them would have operations in the top tobacco-growing countries in the world. According to a 2005 study by the Food and Agriculture Organization of the United Nations, the list of the top 10 tobacco-growing countries includes Zimbabwe, Indonesia, and Malawi. So, presumably, our hypothetical U.S. tobacco company employs a lot of tobacco pickers in those countries. Levels of pay in those three countries are not high. In fact, they are extremely low.

I looked at some real disclosure information for one such U.S.-based tobacco company. It does, in fact, employ lots of people in each of those three countries. More than half of the employees worldwide of this company are tobacco pickers in predominantly third world nations. Standards of living are not high there. Pay is not high there. Since the median employee of this company will likely be one in a third world country, the ratio of MTAC to CTAC is going to be a very small number. Is this bad? That's a matter of opinion.

Now let's consider the effort that it will take for this company to determine the MTAC. They need to find the compensation of every employee, including those in all the Third World countries. That's not easy to do. Then, for each of these people in each of these countries, they need to normalize this by converting the compensation to U.S. dollars. In some of these countries, the company offers pension plans, albeit neither broad-based nor generous. We'll get into the complexities of dealing with pension plans later, but suffice it to say that they add unbelievable complexity to the TAC calculation for each covered employee.

That's a lot of effort for this company. And, to what end? What will you, as an interested observer, do when you see that this company's pay ratio is 0.00004? Or, suppose the SEC decides that the ratio should only be expressed to four decimal places with traditional rounding. Then, how will you react when the company discloses its pay ratio as 0.0000? That's a scary thought, isn't it?

Defined Benefit Plans

This is where the calculation gets really bad. You see, for proxy purposes, and therefore for pay ratio purposes, compensation for an individual includes the increase in the present value of accumulated plan benefits for that individual. Present value is to be calculated using the actuarial assumptions used for the company's financial disclosures under Generally Accepted Accounting Principles (GAAP). Among those assumptions is a discount rate. U.S. GAAP, specifically ASC 715, indicates that for a defined benefit plan, this discount rate should be the rate at which those obligations could effectively be settled. Further, this guidance asks the issuer to look to the rates on high-quality corporate bonds of similar duration to the liabilities of the plan.

So, here is what we have. A company, in consultation with its actuary and its auditors, chooses a discount rate for each plan. The actuary calculates the present value of accumulated plan benefits for each employee as of the last day of both the most recently ended fiscal year and the one before that. This involves calculating the accrued benefit for each employee, developing an anticipated stream of payments from the plan and then discounting that stream of payments back to the measurement date at the discount rate that was appropriate at that measurement date.

Discount rates virtually always change from year to year. Discount rates often differ from plan to plan. In general, higher discount rates produce lower present values, and conversely. So, for an employee, the change in the present value of accumulated plan benefits includes both the change due to new accruals and the change due to the change in discount rates (as well as other assumption changes, such as the mortality table assumption).

For employees in traditional defined benefit plans, present values of plan benefits have a significant age bias. That is, the present value of accumulated plan benefits generally increases much more rapidly for employees close to retirement age than it does for younger employees. That change will go into the calculation of TAC for each employee, which will affect both the determination of the median employee and the determination of the compensation for that median employee.

Does that sound like a lot of work to generate a single number? It is. I am one of those actuaries who might have to do those calculations. For a single broad-based plan in the United States, it's a fair amount of work, perhaps more work than the value of the result would justify. Suppose instead that this company has multiple pension plans. In that case, if there are two plans, the amount of work is doubled. If there are three plans, it is tripled, and so forth.

These are just U.S. plans. We live in a global economy. What of the multinational companies with defined benefit plans in multiple countries? Perhaps, during 2013, discount rates in the United States will rise, but in most of Africa, they will fall. Then, in our calculations, the influence of U.S. pensions will be relatively less in that year than in other years while the influence of African pensions will be relatively more than in other years.

What of the currency conversion? That will influence the process as well. If the U.S. dollar suddenly strengthens, should the CEO of the company be paid more or less? Or, should it not matter?

Is this what Congress intended? Will the users of these data understand this? Does it have value to those users?

Who Really Uses This Information?

On a website maintained by the AFL-CIO,5 you can find information about pay ratios, inverse to those specified in the law (CTAC divided by MTAC). On its Executive PayWatch website, the AFL-CIO explains why this ratio matters. It discusses what the appropriate ratio, in the union's opinion, should be. It lauds companies with lower ratios and disparages those with higher ratios. It neither tries to explain differences nor worries about the reasons.

That's not necessarily bad. There are many CEOs who, in my opinion, are paid far more than their performance merits. That an organization with as broad a membership as the AFL-CIO points this out is good for public policy. It does inform the public of egregious levels of pay. A key question in my mind, however, is whether the 953(b) statistic is necessary for the AFL-CIO (and others) to inform the public of egregious executive compensation. Is it even the best possible statistic to use to inform the public? Does the value of having this one statistic justify the cost of producing it? If you were to ask Richard Trumka, president of the AFL-CIO, he would undoubtedly say that it does. He has in the past.6 If you were to ask me, I would say that it does not.

Preparing for SEC Rules

Section 953(b) of Dodd-Frank is law. Barring its repeal,7 at some point, the SEC will provide rules on its application. The commission has received thousands of comments,8 many suggesting simplifications that commentators say would not materially change the statistic or its value. Thus far, the SEC appears to have concluded that none of these simplifications is faithful enough to the statute or its intent to adopt. It appears that sometime soon (within the next year or two), the SEC will promulgate rules to explain to us how to apply Section 953(b). Given its complexity, issuers need to prepare.

There is no perfect way to do this. In fact, each issuer has different circumstances. Some have only U.S. employees and all of them are on a single payroll system. Others are in more than 100 countries with defined benefit plans in many of them and equity compensation in many others. Payroll may be decentralized. Different actuarial firms may be used in each country. I'm going to focus here more on situations like this complex one than on the relatively simple situation of a U.S.-only company with no equity compensation and no defined benefit plans.

Here are some steps I would take (generally the order in which the steps are taken is irrelevant).

• Centralize as many functions as possible to the extent that those functions inform the process. Putting everyone on a single payroll feed and consolidating all pension calculations and equity valuations will save considerable time, effort and cost.

• Make sure that all of your actuaries are prepared for this calculation. Ideally, a single actuary (or firm) would handle all of the plans and be intimately familiar with Section 953(b).

• Have all equity grants and awards for the year in a single file. Have the file set up to automatically perform Black-Scholes calculations (or similar ones) based on end-of-fiscal-year data.

• Have files from foreign countries set up to come to the United States with currency converted.

• The SEC may allow issuers to exclude certain employees from the calculation. Have identifiers set up to handle such exclusions. Possible exclusions could be employees who work on some particular part-time basis, certain seasonal employees, and those who work less than some specified number of hours during the year.

• Have one person internally in charge of the process. That person should work with outside consultants to develop a budget and a work plan for the project. Ideally, that person should have some knowledge of every one of the components of compensation that will go into the process including cash, equity, and qualified and nonqualified retirement plans.


There is not going to be a perfect solution. It would probably take a supermajority Republican Congress and Republican president to repeal Dodd-Frank. Since this section of Dodd-Frank is such a hot button with organized labor, the Democrats, who as a group are more heavily supported by organized labor, are extremely unlikely to support repeal of 953(b).

Nearly two years after passage of Dodd-Frank, the SEC still finds itself unable to promulgate rules on this topic. Among the flood of comment letters, it has not seen a simplification technique that it feels does the statute and its intent justice. The very good news from the SEC is reiteration that issuers will not need to comply with 953(b) until rules have been proposed.

On the other hand, the bad news is that someday such rules will be proposed. Be ready.

John H. Lowell ( is a consulting actuary with P-Solve Cassidy in Woodstock, Ga. He serves on the board of directors and executive committee of the Conference of Consulting Actuaries as vice president for practice areas and communities. His practice includes consulting on the design, funding, investment, compliance, and administration of both broad-based and executive benefits and executive compensation matters.


This document and any discussions set forth herein are for informational purposes only, and should not be construed as legal advice, which has to be addressed to particular facts and circumstances involved in any given situation. Review or use of the document and any discussions does not create an attorney-client relationship with the author or publisher. To the extent that this document may contain suggested provisions, they will require modification to suit a particular transaction, jurisdiction or situation. Please consult with an attorney with the appropriate level of experience if you have any questions. Any tax information contained in the document or discussions is not intended to be used, and cannot be used, for purposes of avoiding penalties imposed under the United States Internal Revenue Code. Any opinions expressed are those of the author. The Bureau of National Affairs, Inc. and its affiliated entities do not take responsibility for the content in this document or discussions and do not make any representation or warranty as to their completeness or accuracy.

©2014 The Bureau of National Affairs, Inc. All rights reserved. Bloomberg Law Reports ® is a registered trademark and service mark of The Bureau of National Affairs, Inc.

To view additional stories from Bloomberg Law® request a demo now