The Citi Sometimes Sleeps

Citigroup's board of directors may well dismiss its CEO Charles Prince today. They will thank him for his four years of service--during which Citi earned profits of a magnitude too big for me to wrap my brain around--give him a golden parachute, and send him on his way. They will do this on a special meeting called on a Sunday.

Why? Why fire him now? Why not ease him out gracefully at the next annual meeting, and have him say that it is time for him to spend more time with his family and give his subordinates much-deserved promotions? What mistakes did he make that were so dire?

1. He trusted his subordinates when they accepted the rating agences' valuations of CDOs, CMOs, and thus of SIVs.
2. He had an opportunity to come clean in August and September and say that there is a huge problem--that the rating agencies' largely
   backward-looking procedures relied on an implicit assumption that America's different regions' real-estate markets were not thathighly correlated, and this assumption was false.
3. He did not come clean in August and September--meaning either that he tried to hide the seriousness of the losses in the hope that things would break in a favorable direction, or that he did not understand the seriousness of the losses.
4. In either case, he should not be CEO of Citi.

So, at least, the board thought today.

Trusting the rating agencies before August is understandable. They have a pretty good track record over the past century, after all. It takes a big burst of inflation like the 1970s or a Great Depression to drive their judgments massively awry, and we have none of those--no big bursts of inflation, no mass unemployment, not even an asset price crash to produce steep sudden falls in any tranche of the underlying. To state that the rating agencies had gotten this one massively wrong would have required that one be (a) paranoid, (b) far-sighted, and (c) enough of a lover of risk to stake your reputation on long odds indeed, for the rating agencies had not only to be wrong about what would happen in a storm but the storm had to actually blow in on a schedule convenient to your prediction.

After August, trying to keep things quiet for a couple months in the hope that things will break your way is natural but not understandable or forgivable. It is natural--humans do it. It is not understandable--financial professionals are trained not to do it, and know enough to understand why. And it is not forgivable.

After August, not recognizing the seriousness off the losses is natural and understandable but not forgivable. Information is rarely transmitted undistorted. By the time a problem had been passed five links up the chain of hierarchy, each link has shaded it a little more in the direction of being a not so serious problem that we can handle. A CEO may understand that his VPs are shading things to work better. But only a rare CEO grasps in his or her bones the consequences of this happening at every stage up the line. Some do: CEOs who regularly show up on trading floors and watch; the Kangxi Emperor's scattering the hereditary slaves of his family line throughout the bureaucracy, at all levels, with instructions to report to him directly and personally about what was really going on; Lloyd Bentsen scattering his long-time senate staff theoughout the Treasury bureaucracy. Most do not figure out the difference between a small bureaucracy--where people who know what is going on tell you small lies to make themselves look good--and a large bureaucracy with its mendacity-multiplication effects.

Not to figure this out is understandable, but not forgivable.

A decade ago I asked a moderately senior Citigroup executive whether the most senior people understood what risks the organization was running. I think I now have my answer.

How serious are the losses? Ah, Grasshopper, if I knew that would I be spending my time writing this weblog?

Comments

[Bruce Wilder]

This may seem somewhat tangential to the general narrative of not properly assessing and rating risk, but I think it might be closer to some larger economic truths.

Last year, as the yield curve lay flat or actually inverted, there was a lot of idle speculation from economists about whether the yield curve was still a good "leading indicator" for recession. As always, there were those who were ready to volunteer that things had changed in the financial markets, and the yield curve no longer meant what is used to mean.

In economics, there is always a hidden tension between narrative meaning and analytic function. And, there were a few of us -- well, me, actually -- crying out, as in the wildnerness, that the yield curve is not simply an indicator, but is representative of an actual mechanism in the economy -- a critically important mechanism: the yield curve represents the state of things in the carry trade: the vital practice of intermediation, where an intermediary borrows at low short-term rates and lends at higher long-term rates, and, not incidentally, invests in information and expertise so as to be able to lend long intelligently, minimizing the higher risks of lending long, which keeps most short-term money out of the long-term debt markets.

A flat or inverted yield curve represents a squeeze on financial intermediation, which, ceteris paribus, will bring long-term lending down, slowing the economy or inducing a recession. When the Fed raises short-term rates, to slow the economy, the mechanism they are using is the yield curve, squeezing the intermediators, which used to be banks, to curtail long-term lending, and, again not incidentally, to tighten the credit standards which intermediators use to prevent themselves from making bad long-term loans.

But, thoroughly modern economists have abandoned narrative accounts that emphasize mechanism, for narrative accounts that emphasize "expectations", reflecting the fashion in functional, mathematical analyses imposed by enthusiasm for rational expectations theory a generation ago. (Never mind that the "expectations" of the narrative accounts bear only a tenuous relationship to the "expectations" of the formal functional mathematical theory.)

So, here, we are. We have had a yield curve inversion, and a lot of insanely stupid speculation about what the inversion "meant" -- including bizarre mumblings from the now Fed Chair about savings gluts. And, we have the new fangled and evidently under-regulated intermediators respond, not by raising credit standards, but, rather, launching themselves into the brave new world of predatory lending.

And, now we have had what we used to get, before effective regulation of banking intermediators, from a sustained yield curve inversion: a bank run. Only this was, in Krugman's phrase: a non-bank bank run.

I have no doubt that the rating agencies have a role in this breakdown. But, let's not lose sight of how this episode has represented a breakdown in the functioning of the institutional mechanisms of financial intermediation -- of borrowing short and lending long, with expertise -- that are vital to the functioning of the economy.

The rating agencies did not do this alone. Large parts of the banking sector, having shed their institutional responsibility in long-term lending, embarked on various schemes of predatory lending, which exacerbated the situation. The statistics on the explosion in Alt-A and subprime lending are unmistakeable.

We have had a lot of innovation in the financial sector that has changed the institutional mechanisms for intermediation. We need some well-informed, well-intentioned scrutiny of just how well those mechanisms are working.

And, economics, as a profession contributing to the public discourse and public policy, might want to examine the applicability and utility of "expectations" narratives, that completely leave the mechanisms of intermediation out of account.

[Bernard Yomtov]

Bruce Wilder,

Very interesting comment.

[howard]

let me join bernard yomtov in noting how well put bruce wilder's post is.

and then let me return to my original reason for posting, which is to say that prof, i think you let "trusting the rating agencies" off a little easy here. it wasn't hard for me to know (and i've posted it often enough on this very site) that come the crunch (and given that we still live under market capitalism, a crunch was going to come) we would fine out if everyone had hedged correctly.

which is another way of your asking whether senior management grasped the risks correctly.

an appropriate ceo behavior in this environment is to remind the troops, over and over, that the four most dangerous words ever uttered on wall street are "this time it's different."

[Ponzi Q. Globalization]

The assumption that the CEO is working for the long-term benefit of the corporation and not for him or herself is a big problem in all this type of analysis. The executive that makes tens of millions in the short-term and then walks away with $169 mil or so is a big winner in the game that matters most to them. Such an executive wins even if he or she leaves a corporate corpse behind.

Loyalty to corporate institutions isn't all that strong these days. Almost any employee would gladly sacrifice their company on the altar of their self-interest. Why do we assume top executives are the exception to this rule? Do they really have better morals than the rest of us?

[Yves]

To your point about having spies throughout the organization, Prince should have known well before August. Gillian Tett of the Financial Times had been writing since January about the riskiness of CDOs, their high embedded leverage, the fact that they were going into the hands of weak buyers who didn't understand them. In May, Joseph Mason and Joshua Rosner wrote an extensively researched paper, "Where Did the Risk Go? How Misapplied Bond Ratings Cause Mortgage Backed Securities and Collateralized Debt Obligation Market Disruptions," which paints a particularly damning picture of rating agency procedures.

Someone like Tett is only as good as her sources, and similarly, while Mason and Rosner did very solid work, they too had to be clued in on certain matters by industry participants. So if Prince were properly informed, he should have had an idea, no later than the meltdown of the Bear hedge funds in June, that the value of these assets was tanking and Citi too might be exposed. He should have started asking questions then.

[Roger Bigod]

I've never seen an estimate for the amount that Enron vaporized, but this looks to be much larger, with wider systemic effects. And Citi's share may be on the same order of magnitude as Enron. Several of the themes recur here: off-balance sheet entities, blessings by impeccable outside entities (Arthur Anderson = ratings agencies), conscious fraud at lower levels (Fastow) with self-deception at the top (Skilling, Lay), use of derivatives to abstract risk and pass it on, deodorized, to unsophisticated buyers, self-congratulation at the wonderful scope for creativity provided by the free market.

For better or worse, the events here are boring and tedious to unravel, so the book and movie possibilities are limited. This diaster has no colorful narrative of gunslinger traders, sex bomb executrixes, an oleagenous CEO, gladiatorial corporate infighting, the giddy possibilities of the internet. Indeed, one of Enron's major problems was that it produced way more narrative color than it could consume locally.

[ferd]

Did most top finance people fail, for years, to foresee the mounting risk? Or, was there inside scuttlebutt for years that a stench was rising?

I'd like to hear what a guy like Warren Buffett would say about that.