By James Kwak
Searching through my RSS feed*, I observed that not many people have commented on Goldman Sachs’s stunning compensation announcement (except for Felix Salmon), perhaps because it came out on the same day as the “Volcker Rule,” perhaps because bloggers are not wired to say nice things about Goldman (GS). But I’m going to make the sure-to-be-unpopular statement that Goldman did the right thing here.
We all know that Goldman made a lot of money last year: $35.0 billion before compensation and taxes, on my reading of the income statement (that’s pre-tax earnings plus compensation and benefits). Many people think that it made that money because of government support, but that’s beside the point here; right now, this is purely a question of dividing the spoils between employees and shareholders.
Historically, investment banks have given a large proportion of the profits (here, meaning before compensation and taxes) to the employees. For example, in 2007 Goldman gave $20.2 billion out of $37.8 billion to its employees, or 53%. There are undoubtedly many reasons for this. One reason is the idea that investment banking is a business that depends on individual “talent,” and therefore employees have to be paid their marginal product or they will leave for other firms. More insidiously, investment banking executives tend to see their employees as younger versions of themselves, which creates a sense of solidarity; at traditional investment banks, the management committee was composed of partners who worked their way up through the ranks. Contrast this to, say, Wal-Mart, where top management has very little in common (socially, educationally, economically, politically, etc.) with the vast majority of their employees.
As a result, investment bankers are overpaid. Now, before all the bankers get all indignant on me, let me say that bankers should make more money than average people, at least according to the normal rules of our society; for one thing, they are, on average, better educated than most people. (As I’ve written before, I don’t think there’s any moral reason why people should make more money simply because they are better educated, or have unique skills, or are more intelligent, or work harder; but that’s the way the world works, and most people are OK with that in principle.)
How much more overpaid? I’ve previously written about the paper by Thomas Phillipon and Ariell Reshef on wages in the financial industry. According to Figure 10, you would expect wages in finance to be about 30% higher than in the economy at large because of higher education and lower job security; yet, also according to Figure 10, wages in finance were over 70% higher than average earlier this decade. 40 percentage points divided by 1.7 implies that wages should come down by about 25%. This an industry-wide figure, however, and recent wage growth has been much higher in investment banking than in the rest of the industry (largely banking and insurance), so 25% is probably a very low figure for investment banking. But without more data, that’s the best I can do.
Now, Philippon and Reshef’s data only go through 2006. In 2006, Goldman paid $16.5 billion out of $31.1 billion of its profits to employees (53%, the same as in 2007), which worked out to about $620,000 per employee. (In 2007, that figure was about $660,000.) Subtracting 25%, we get that Goldman employees should have earned abut $470,000 on average. This is probably still much too high, because that 25% figure is undoubtedly far too low for investment banking. But it’s a starting point.
Now, what did Goldman do for 2009? Through September, they had already set aside $16.7 billion for compensation, a 57% payout ratio; annualized, that comes to a stunning $22.3 billion, or $700,000 per employee. In Q4, however, they did the unthinkable; it reduced its compensation expenses by $500 million.** This lowered the annual compensation pool to $16.2 billion, or $500,000 per employee, and lowered the payout ratio to 46%.
In their press release, Goldman trumpeted the fact that compensation was down 20% from 2007 and the payout ratio was the lowest ever.*** They neglected to mention that total compensation was the same as in 2006 and over 30% higher than 2005, when per-employee compensation was $500,000. (So on a per-employee basis, we’ve just rolled the clock back to 2005.) But still, $500,000 is better than $660,000 (2007) and a lot better than $700,000 (2009 through September).
I’m sure most people wrote this move off as a public relations stunt, and maybe it was. Maybe management leaked to employees that 2010 bonuses will be extra-good to make up for 2009. But there’s another possibility, as Salmon pointed out, which is that Goldman realized it simply doesn’t have to pay its employees as much. Goldman is the premier investment bank in the world, and the gap between it and its rivals has gotten much bigger; if someone is unhappy with his bonus, where is he going to go? Citigroup (C)? Bank of America Merrill Lynch (BAC)? If Goldman’s management team really wants to maximize shareholder value, then this is exactly what they should be doing. (The big problem, as the New York Times points out, is other banks that are paying big bonuses despite having bad years–like Citigroup, whose payout ratio is over 100%.)
The test, of course, will be next year. Goldman should reduce its per-employee compensation expenses even further, and should try to push the industry to a new equilibrium where the payout ratio is in the 30-40% range and average compensation for investment bankers is in the $300-400,000 range. And Goldman’s shareholders should apply pressure to make this happen; basically, they should try to squeeze labor.
Will they? Shareholder governance is something we usually celebrate about our economy. But I wonder if it works for investment banks. I wonder because the institutional investors that control most of the shares are the same kind of people as the bankers who work at those banks. Yes, it’s the buy side (people who buy securities) versus the sell side (people who sell them), but they went to the same colleges, they go to the Hamptons together, their kids go to the same schools, and so on. It’s probably easier for an institutional investor to swallow $600,000 per year compensation at Goldman than it is to swallow $10 per hour at Wal-Mart. Maybe class bonds outweigh economic interests.
My prediction is that this is just a PR stunt and next year (assuming it is a good one for the banks), per-employee compensation at Goldman returns to at least $600,000 and maybe $700,000. But I would love to be wrong.
* One of the best things about reading news and blogs in an RSS reader like Google Reader is that it works like a perfectly-targeted meta-search: You can search your preferred information sources all at once in one place, as opposed to using Google web search where you get the entire universe in one haphazardly-ordered list.
** This means that it reduced its bonus pool by even more, since during the quarter it had to accrue its employees’ ordinary salaries and benefits; assuming $40,000 per employee, that comes to $1.2 billion, meaning that it reduced the bonus pool by $1.7 billion.
*** Goldman uses compensation divided by net revenues, which gives them different percentages from mine, but the substance is the same.