Sunday, May 02, 2010

Reputational Capital and Incentives in Organizations

The following passage, jarring in light of recent revelations, appears in the opening pages of Akerlof and Kranton's recently published book on Identity Economics:
On Wall Street, reputedly, the name of the game is making money. Charles Ellis' history of Goldman Sachs shows that, paradoxically, the partnership's success comes from subordinating that goal, at least in the short run. Rather, the company's financial success has stemmed from an ideal remarkably like that of the U.S. Air Force: "Service before Self." Employees believe, above all, that they are to serve the firm. As a managing director recently told us: "At Goldman we run to the fire." Goldman Sachs' Business Principles, fourteen of them, were composed in the 1970s by the firm's co-chairman, John Whitehead, who feared that the firm might lose its core values as it grew. The first Principle is "Our clients' interests always come first. Our experience shows that if we serve our clients well, our own success will follow." The principles also mandate dedication to teamwork, innovation, and strict adherence to rules and standards. The final principle is "Integrity and honesty are at the heart of our business. We expect our people to maintain high ethical standards in everything they do, both in their work for the firm and in their personal lives."
If the preservation of its reputation for serving the interests of its clients was a major organizational goal for Goldman, then something clearly went terribly wrong. Consider, for example, Chris Nicholson's report on the manner in which the bank managed to shed its holdings of mortgage backed securities shortly before they collapsed in value, allegedly serving itself "at the expense of its clients." Nicholson reproduces the following email from an employee at the European sales desk to the head of mortgage trading: 
Real bad feeling across European sales about some of the trades we did with clients. The damage this has done to our franchise is very significant. Aggregate loss of our clients on just these 5 trades along is 1bln+. In addition team feels that recognition (sales credits and otherwise) they received for getting this business done was not consistent at all with money it ended making/saving the firm.
Felix Salmon considers this email to be "particularly damning" for the following reasons:
Illiquid things like CDOs are sold as much as they’re bought, and Goldman’s highly-paid sales team was aggressively going out and selling instruments which were at one point on Goldman’s balance sheet and which wound up cratering in value.
The effects were twofold: firstly, the Goldman clients who got stuck with this nuclear waste when the music stopped were understandably none too impressed with Goldman. And secondly, Goldman managed to stick the losses on those instruments to its clients, rather than taking those losses itself, and as a result its profits were billions of dollars higher than they would otherwise have been.
Was the hit to Goldman’s franchise value a hit worth taking, given the billions of dollars it saved? Probably yes, until the SEC and Carl Levin came along.
But the possibility that the SEC and Mr. Levin would eventually come along was always there. This is a form of tail risk that is not unlike that taken by the folks at the AIG financial products division when they sold vast amounts of credit protection in the mistaken belief that they would never be faced with significant collateral calls. Raghuram Rajan, in a remarkably prescient 2005 paper, described this process as follows:
Consider the incentive to take on risk that is not in the [compensation] benchmark and is not observable to investors. A number of insurance companies and pension funds have entered the credit derivatives market to sell guarantees against a company defaulting. Essentially, these investment managers collect premia in ordinary times from people buying the guarantees. With very small probability, however, the company will default, forcing the guarantor to pay out a large amount. The investment managers are, thus, selling disaster insurance or, equivalently, taking on “peso” or tail risks, which produce a positive return most of the time as compensation for a rare very negative return. These strategies have the appearance of producing very high alphas (high returns for low risk), so managers have an incentive to load up on them. Every once in a while, however, they will blow up. Since true performance can be estimated only over a long period, far exceeding the horizon set by the average manager’s incentives, managers will take these risks if they can.
As in the case of tail risks arising from the sale of credit protection, damage to the firm's franchise value does not appear in standard compensation benchmarks. The problem in Goldman's case was not that such damage was "a hit worth taking" but rather that the incentives faced by its employees did not adequately reflect the value of the firm's reputation in the first place. To the extent that employee behavior is responsive to such incentives, the sacrifice of reputation for immediate profit will be made regardless of whether or not, in the broader scheme of things, the damage to franchise value exceeds the short term gains.
How, then, might a firm accomplish the subordination of short term goals to long term objectives in practice? There are two possibilities: one could hire individuals who are predisposed to behave in a principled manner even in the face of incentives not to do so, or one could design compensation schemes that adequately reward actions that preserve or enhance reputation. Economists, being fervent believers in the power of incentives, usually tend to favor the latter approach. But in this particular context, there are two possible problems with this. First, the contribution of any given transaction to the reputation of the firm is generally much more difficult to ascertain and quantify than any contribution to the firm's balance sheet. This makes it difficult to assign reward appropriately. Second, in order to serve as credible commitments to clients and customers, compensation schemes must be easily observable and not subject to renegotiation after the fact. This is seldom the case.
The alternative is to hire individuals who are predisposed to behave in a manner that meets organizational objectives: to place a premium not only on ability but also on character. But would this not create incentives for potential employees to simply misrepresent their values? As Groucho Marx famously said: "The secret of life is honesty and fair dealing... if you can fake that, you've got it made." 
Fortunately, the consistent misrepresentation of personality traits is often infeasible or prohibitively costly. There is an interesting line of research in economics, dating back to Schelling and continuing through Hirshleifer and Frank, that explores the commitment value of traits that are costly to fake. Hirshleifer went so far as to argue that the "absence of self-interest can pay off even measured in terms of material selfish gain, and... the loss of control that makes calculated behavior impossible can be more profitable than calculated optimization... we ought not to prejudge the question as to whether the observed limitations upon the human ability to pursue self-interested rationality are really no more than imperfections -- might not these seeming disabilities actually be functional?" 
One could take this a step further: not only might limitations on the unbridled pursuit of material self-interest be functional for individuals, they may also be functional for the organizations to which they belong. And in the long run, firms that manage to identify and promote such individuals will prosper at the expense of those who are unable or unwilling to do so.


Update (5/2). At the end of the post I linked to above, Felix Salmon asks:
Let’s say you work at an investment bank and you’re in charge of a book which includes a $1 billion barrel of toxic nuclear waste. You know that barrel is going to zero sooner or later, and you manage to sell it to some European dupes just in time, for full face value, saving your bank from $1 billion in losses. How much of a bonus, if any, should you get on that deal, and where should the money come from? And should you feel bad about avoiding the losses and sticking them to your clients instead?
In a comment on that post, The Epicurean Dealmaker responds:
The answer depends. If you are a proprietary trading shop your job is to make money (and avoid losses) at all costs. In fact, you have a fiduciary duty to somebody (eg, limited partners or shareholders) to do so. You sell the crap and never look back. Caveat emptor rules.

If you are a traditional investment bank, you find out which morons allowed $1 billion of concentrated toxic risk to accumulate on your balance sheet and you fire their incompetent asses for cause (ie, no bonuses, no golden parachutes). Then you convene the Executive Committee to decide whether it is worth permanently damaging your franchise as a supposedly neutral market maker by offloading the waste before it blows up onto your clients, or whether you should eat the loss as punishment for failure...

This is why trying to run a large proprietary trading operation inside a traditional market-making investment bank introduces a fundamental, highly dangerous conflict of interest. At a small scale, this kind of stuff happens all the time, and should, in a traditional investment bank. However, it should never reach the scale that threatens the short- or long-term future of the bank.
This gets it about right in my opinion. See also the many excellent comments along these lines on Mark Thoma's page. Mark links to a related post by Richard Green, who in turn mentions pieces by James Surowiecki and Yves Smith that are both worth reading. Here is Yves' bottom line:
Legal issues aside, it isn’t merely the great unwashed public that is taking an increasingly dim view of Goldman. What is striking is the change in sentiment among professionals.

Recall Goldman’s reputation: that of being the best managed firm on the Street, and its boasting about its risk management as key to its superior profits... its once-vaunted risk management, which led observers to believe that Goldman was doing a better job of managing exposures, now increasingly looks like the firm was simply more systematic and aggressive than its peers in not just shifting risks onto customers but engaging in further profit-maximizing strategies that look downright predatory...

Goldman is increasingly beleagured. Its lobbyists are now pariahs. More private lawsuits are coming to the fore. There are rumors it is in settlement talks with the SEC. But the once-storied firm apparently turned its well oiled machine to ruthless profit-seeking. It is an open question how much damage the firm will sustain from the well-deserved backlash, and whether it can change its conduct. 


Update (5/4). As usual, there are a number of excellent comments on Mark Thoma's page. Here's Roger Chittum:
Goldman and the other IBs are in a very different business now than they were in the 1970s. Formerly, their business was representing real-economy clients for generous fees in facilitating mergers, acquisitions, spin-offs, IPOs, bridge loans, restructurings, and other balance sheet transactions, in which they sometimes took participations. Their reputations were quite important to relationship building with the few firms that could afford their services and with the investors to whom they returned again and again. In fact, they used to regard proprietary trading as a business that was beneath them. Thus, Solomon Brothers, for example, which early on was very active in trading, especially in government securities, was not in the top tier reputationally.

In recent years, the businesses and cultures of the IBs have been transformed. Something like 75% of their profits now come from proprietary trading, and the top executives have come up through trading instead of the white shoe service businesses. They are essentially now hedge funds with advisory groups appendages. And they are dangerous and unapologetic predators.

It may be possible to have commercial banks and investment banks as formerly constituted have common ethical cultures that are concerned with the longer term and institutional reputations, but I don't see how that can happen that can happen in hedge funds and proprietary trading units. Each of the big IBs has a story of internal culture struggles as their businesses changed, and notice the recent reports of culture clash arising out of the dysfunctional forced marriage of BofA and Merrill.

Glass Steagall put the wall between commercial banking and investment banking. Perhaps commercial banking and fee-based investment banking in the 1970s style could thrive with a common culture, but proprietary trading needs to be hived off. There are irreconcilable differences there.
To which mrrunangun adds:
Another difference is that in Whitehead's time, GS was a partnership and the great bulk of the partners' wealth was their interest in the partnership. Any failure in the partnership had the potential of creating losses for all of the partners. The long-term success of the partnership was crucial to the long-term security of the partners.
In the public company that GS became, none of that discipline remained to restrain the people running the company from exploiting their position in order to maximize the short-term profits on which their claims to enormous annual pay packages were supposedly based. In contemporary practice, boards of directors are generally much more sympathetic to management than to shareholders. Boards put shareholders first only in small companies where the board is made up of the owners, who have a significant stake in the success of the enterprise, and perhaps their attorney and/or auditor. In medium to large company practice, the CEO generally recruits the board, probably always if the CEO serves as board chairman as well. Board members who are not inside directors rarely have a critical portion of their own wealth in the companies on whose boards they serve.
Then there's Bruce Wilder:
On the issue of "aligning incentives" for individuals, I can tell you what the right scheme looks like: it looks like a fixed salary, with modest "options" in the form of (mostly honorary in magnitude) raises and bonuses.
At the very top of a business enterprise, it makes sense to make top executives buy a substantial interest in the firm -- to essentially become "partners". The top executives don't have to own a large part of the firm, but their ownership interest has to be a large part of their personal wealth. And, it should not be a gift -- no free options; make them buy it, and make it hard to sell: a large part of their ownership interest should be tied up in trusts.
The top executives of large corporations really shouldn't be paid in "performance" contingent options. Shocking I know, but executive leadership has little to do with the piece-work of a sweat-shop or a cucumber field.
And, they shouldn't be paid in magnitudes that could make them independently wealthy in a single calendar year. 
Magnitude matters, and it can easily overwhelm any contingent "alignment" of incentives... If you promise to pay someone a vast amount, realizable in the short-term, contingent on some abstract score-keeping scheme, you are incentivizing (horrible word) them to corrupt the score-keeping. You are asking them to lie.
Rajiv Sethi comes around to this issue, via personal character and the difficulty most of the non-psychopaths among us have, in lying.
He might also consider that Goldman is a vast, hierarchical organization, embedded in a market-exchange network, all of which -- hierarchy and network -- consists entirely of generating and reporting numbers, as part of a complex scheme of compound control.
This system cannot function, if the participants have too much of an incentive to corrupt the reported numbers. If people at the top of the hierarchy, or the clients, or the counter-parties, or on-lookers in the same and related markets, get the "wrong" numbers, things go terribly wrong.
 Bruce also points to the following from Mike Konzal of Rortybomb:
To keep this in economic terms, this crisis has shown a wave of agency problems embedded inside financial institutions. The best phrase for why people were motivated to do the things they do is simple: IBGYBG–I’ll be gone, you’ll be gone.
The other obvious market failure is that in a ruthlessly competitive arena that is judged primarily on quantitative measures, any ability to juke your statistics forces others to participate in that as well. We see this in a variety of ways I can describe if people are interested, but if your competition is repo 105ing their balance sheet to make it look like they are getting better returns with less leverage, they are going to get better deals on customers and capital than you are going to get and put you out of business. So you better do that as well. The notion of Milton Friedman-ite self-regulation through reputation effects has been a complete failure.
This echoes Richard Serlin's comments below. And then there's this from an anonymous source:
My thought reading Sethi's post and also TED's latest was this:
Decades ago the investment banks stopped hiring the scions of wealthy families expecting a sinecure, and started looking for people who were smart and "hungry" (as Grisham has put it). They asked applicants "What would you do if you won the lottery?" And if you made it clear that you didn't really care about the money, you weren't the right person for them.
And as TED makes clear the people who make it to the top at the investment banks are the hungriest sharks. The people who aren't really hungry (and I think there are plenty of them), make enough and leave, or walk when they feel they're asked to do something that compromises their integrity. So there's a huge endogeneity problem here.
Finally, paine is skeptical of the idea that "we can create eisenhowerish org men and then blend em with fresh new corporate norms" on the grounds that competitive pressure transforms character. It's a fair point.


  1. A big thing is that the life of a CEO is so much shorter than the life of a company. How many CEOs are in their 50s, 60s, or 70s? There is so much harm that they can do for great short run gain that won't really become apparent until close to, or long after, they retire.

    With good corporate governance CEOs and other top management could be watched very closely and re-evaluated every year in a thorough, high level intelligence, high dimensional, flexible, non-simple formulaic way way, like basketball team coaches are. But, 1. corporate governance is extremely bad, with diffuse ownership and boards that have little time or resources and are often captured (see the book Pay without Performance, by Harvard's Lucian Bebchuk and Berkeley's Jesse Fried). 2. What's in vogue instead of high level, high dimensional thinking, is simple minded "objective" formulas that are extremely inaccurate in complicated situations like this (for more on this see:

  2. Here's James Galbraith on this, from an October PBS interview:

    They let all of this run, because they were getting a superficially stronger economy out of it. The ownership society, all that was a scam, basically, designed to lure people who could never afford these mortgages into accepting them. And yes, I think they, any rational person, certainly people in the industry, knew that this was not going to last. There was a little industry code, I've learned, IBGYBG. "I'll be gone. You'll be gone."


  3. "His first is that the business would capture the regulator"

    We have a long history of regulation still doing far more good than harm. Look at banking regulation post depression until the recent Republican dismantling. Look at anti-pollution regulation. Look at product safety regulation.

    Plus, this problem could easily be obliterated with an amendment banning business political spending, or public campaign financing so generous that there's little incentive to cater to businesses for their money.

  4. Oh sorry, the above comment was supposed to go to Richard Green's blog.

  5. Richard, thanks for your comments... certainly failures in corporate governance and incentive structures are an important part of the story. There are several comments on Mark Thoma's page along similar lines (see my 5/4 update to this post).