Pearlstein: Risk Management
Wednesday, November 28, 2007; 11:00 AM
Washington Post business columnist Steven Pearlstein was online Wednesday, Nov. 28 at 11 a.m. ET to discuss the limitations of risk management.
Read today's column: The Art of Managing Risk
The transcript follows.
About Pearlstein: Steven Pearlstein writes about business and the economy for The Washington Post. His journalism career includes editing roles at The Post and Inc. magazine. He was founding publisher and editor of The Boston Observer, a monthly journal of liberal opinion. He got his start in journalism reporting for two New Hampshire newspapers -- the Concord Monitor and the Foster's Daily Democrat. Pearlstein has also worked as a television news reporter and a congressional staffer.
His column archive is online here.
Boston: Regulators who rely solely on whether investment banks have a risk management system have clearly never sat in on performance review/bonus meetings at investment banks. How do you regulate greed?
Oh, my balance sheet is a little weak too right now--who do I call in the Emirates for a cash infusion and will the U.S. Government grease the wheels for me too on any regulatory hurdles?
Steven Pearlstein: Cynical, but probably pretty accurate.
Fairfax: The refrain of the past 18 months from the housing bubble blogs is appropriate here:
Privatize the gain.
Socialize the risk.
That's exactly what subprime mortgage brokers did. Once subprime mortgages are dismembered and integrated within umpteen tranches, traceability is totally lost, as is the ability of investment holders to negotiate a haircut on the sick mortgages.
Let's hear it for the free market. BARF!
Steven Pearlstein: Privatize the gain, socialize the risk. That's a good theme. Mind it I borrow it sometime?
Philadelphia, Pa.: I am aware of people with huge assets moving them steadily just under the radar to European, British and Swiss accounts to manage risk to their family assets from the falling dollar. Is there any way for people without huge assets to do the same?
Steven Pearlstein: If you have a brokerage firm that handles such things, you can buy soveign bonds of European countries. You can buy stock in European and japanese companies. You can buy foreign currencies or currency futures. It takes some doing if you are a small investor, so you probably have to go to a global bank or brokerage firm.
Annapolis, Md.: The problem of human denial is historically found in one form or other in major projects failures, e.g. IT, gov/commercial,etc. It was quite obviously a problem as you point out in the mess at Citi and Merrill. We have been working on software to "mine" the insight of people who do the actual work and present a characterization of the "ground truth". As Jack Welch wrote in a recent B-Week column, "bad news festers in the trenches until it's too late to avoid the risk".
My general question is what experiences have you heard about where the festering or as I call it, the "upward spin" has been successfully managed?
Steven Pearlstein: I'm not sure I have any experiences or anecdotes to relate, at least off the top of my head. But your general point is a good one: middle management is a terribly impervious layer when it comes to the flow of bad news or good ideas. It is a big big problem in many companies and it is very hard to counteract. You can take two approaches, I suppose. One is to develop credible back-channel lines of communication. The other is to do some serious cultural training to change the nature of middle management. The first is easier, the second more effective.
Boston: So in 2002 the price of oil was in the $30s, our big sacrifice for the war on terror was a tax break that accrued predominantly to the wealthiest 1 percent, we punted on higher auto fuel standards to protect a couple of domestic companies, and mortgage brokers were driving Jags and telling homeowners to "pick a number" for home appraisals.
Today, oil is around $100 and we've written trillion dollar petro checks to friends and enemies alike (that's consumption not investment), we've spent $500 billion on the Iraq War, private equity general partners whine that they will quit working if they have to pay the same taxes as everyone else, we still have no rational energy policy and our former Treasury Secretary has to go hat in hand to the Middle East to bail out our largest bank who had been glad to fund the now unemployed mortgage broker's Jag.
Did Rome feel like this at one time?
Steven Pearlstein: Maybe Britain.
McLean, Va.: Interesting column. It was somewhat similar to a piece in the Wall Street Journal a couple of weeks ago, that featured an interview with John Reed, the CEO who guided Citi through its crisis in the early 90s. One of his main points was that every organization (or, to keep it accurate, let's say every big bank) has its risk management procedures in place. But he felt that it becomes such a quotidian thing that it becomes too easy to ignore, especially in the face of the huge profits that can be made by ignoring it. It almost seems like this is just a human nature-based limitation. Aside from the regulatory solution that you advocate in your column, any thoughts on what an organization can do internally to overcome the problem?
Steven Pearlstein: First, it requires a top management that won't let itself get pushed around by Wall Street and short-term investors. and that requires compensation packages that are less tied to medium and short-term movements in the stock price.
Second, you have to change the culture and incentives within the firm (monetary and career incentives). You can celebrate the guys who bring in the big money, but you also have to celebrate the people who save you from disaster.
I think the directors are somewhat to blame as well. They are supposed to be the wise outsiders. And when they see a particular line of business is suddenly delivering outsized profit margins for what looks to be modest risk and involves no great skill or originality, a light should go off. And they should engage someone from the outside to take a look at the business and decide if there are inbedded and unacknowledged risks in the line of business that need to be addressed.
Maybe I'm naive in thinking directors can do this. But they are the best candidates I can think of who might have enough distance that, when a company is raking in the money, doesn't think that the only reason is because of the brillance of management.
Phila., Pa.: My experience with risk management came from my six years in the military. We had plans A, B, C, D.... You trained your replacement early on. A reporter asked a consultant working on the Iraqi Constitution about the possibility of missing a milestone and was there a plan B. The consultant said no, once you develop a plan B, you have already abandon plan A.
After the invasion of Iraq, a criticism has been expressed that there intentionally was no plan B as a fall back, after it became evident that the welcome was not as warm as expected. Civilians with business backgrounds seemed to over-rule the more conservative military voices.
Risk management is being treated as a recent new concept in the business world and not universally accepted.
We seem to have two schools of thought, the business school and the military school. As a project manager, I have received much pushback from peers who get annoyed "playing" what if, what if, what if. Business seems to be allergic to redundancy and is willing to risk a 100 million dollar project when one component fails or one process becomes a fatal bottle neck, both identified as risky prior to the project starting. What is taught in business school? Is risk management considered "negative thinking"?
Steven Pearlstein: It is a good question about the business schools and I don't know the answer. But I doubt that's where you need to deal with the problem. After all, the military planners don't have MBAs, for the most part, and they think about Plan B. The problem, to repeat myself, is cultural. You have to have a culture that encourages risk-taking but also encourages people to remain very clear about the costs and benefits of those risks and celebrates people who pull the plus or go to Plan B when that balance gets out of whack.
East Coast: Not a questions, but a comment.
I became a banking regulatory attorney right after law school, during the S&L crisis. It was appalling to me that things had gotten as bad as they did, and we were cleaning up that mess for years afterwards. I eventually left that practice because the role of the regulator kept decreasing until there was little left for us to do. The final straw was when the Glass-Steagall Act -- separating banking from more risky insurance, investment banking and trading activities -- was effectively repealed in the late 90s.
The regulators do too little to rein in the banks, but they also know too little about what actually goes on in banks. The amount of unchecked risk-taking I observed in-house and in private practice was astonishing. It was all an exercise to confuse and avoid the regulators.
Steven Pearlstein: You know, this is such an important comment I want to send a copy to the Comptroller, Fed Chairman Bernanke and Sheila Bair at the FDIC. They really need to shake up their operations and reconceptualize what bank regulation is about. Unfortunately, they seem to be moving in just the opposite direction with Basel II, which is even more reliant on risk management systems than before, as I understand it.
Arlington, Va.: In your article today, you had this little gem
"Many times I have been sitting across the table from an energy trader and I would say, 'Your portfolio will implode if this specific situation happens.' And the trader would start yelling at me and telling me I'm an idiot, that such a situation would never happen," he says. "The problem is that, on one side, you have a rainmaker who is making lots of money for the company and is treated like a superstar, and on the other side you have an introverted nerd. So who do you think wins?"
That is almost exactly how the Geologists and other technical professionals who are now warning of a fundamental discontinuity in the oil supply paradigm. The military and a few local governments have started performing risk mitigation for peak oil, but the psychology favors pretending that everything will continue along as it has in the recent past. Right?
Steven Pearlstein: Maybe. But I'm not sure the analogy is perfect. The peak oil enthusiasts overstate their case, in my opinion. And I have no particular stake in focusing on additional drilling.
Danvers, Mass.: Did Kaminski talk about how easy it is to do the statistical correlations that pop out of the data that they have, and how that differs from having an understanding of "how things work" which might not be revealed in the data at hand?
Steven Pearlstein: He said a lot of things that, frankly, went over my head, because, of course, I don't have the skills and training to do risk analysis. But I think he said something like that.
20th Street and Pennsylvania Ave., NW: Mr. Pearlstein,
About the only thing I typically agree with you on is our shared opinion of Doug Jemal but I have to give you a bravo on this morning's column.
I was especially taken by the paragraph where Dr. Kaminski describes sitting at the table questioning his trader about a deal and the trader calling him an idiot. Since I do credit risk management at a non-bank real estate lender I've been in similar situations and the rain makers always win because management always sides with the revenue producers when times are good.
I do not have a solution to offer other than somehow changing the existing employment agreements for rainmakers and management to allow for clawbacks in the event debacles like we're seeing come true. Seeing the executives that went along with the deals walk away with millions is galling while I sit here sweating through another reorganization by yet again another management team because I'm an at-will employee not protected by an employment contract.
Bulls can get rich, Bears can get rich but Hogs can, and should, get slaughtered.
Steven Pearlstein: I agree with you about the clawbacks. I'd actually make them holdbacks -- that is, don't give them the money until the end of the next downturn.
Princeton, N.J.: Steve, this is somewhat off topic, but the actuaries at SSA and CBO have been using similar techniques to guess at the 75 predictions for Social Security. Specifically, they use Monte Carlo models to guess the unknowable statistics like economic growth which depend on future events, i.e. they take 500 random guess and take the average. This is no better than reading the entrails of a goat.
Steven Pearlstein: Yes, although as I understand it, they could also apply a "stress" test to their central conclusion, and do a "what if" based on a extremely unfavorable set of circumstances. At that point, the question ought to be what steps might be taken to reduce the probability of that happening.
Prescott, Ariz.: In your column, you wrote:
"Risk management is an art, not a science.
And its latest failure is turning out to be an expensive learning experience, not only for Wall Street, but for the rest of us as well.
Certainly it is a failure on the part of executives whose jobs, reputations and fortunes are on the line."
I want to know, are any executive's fortunes really on the line? They got cash bonuses for the short term gains they could eke out, even if it was going to fall apart long term.
That is the key problem with risk management you don't address. Someone who gets rewarded by a series of 3, 6, or 12 month carrot and sticks is going to have different priorities than a company or stockholder that wants to be around in 10 years. Until Wall Street changes their pay system to reflect what they want out of the market, nothing will change.
Steven Pearlstein: Well, I agree with that. In fact, I wrote that two weeks ago myself. But it is not true that they have nothing on the line. Even those who took away big bonuses and cashed in on stock options still have lots of stock and options in their name, so they still have skin in the game. But it is a lot less painful, I agree, when you've already taken home enough money to live a very, very lavish life for the rest of your days.
Washington, D.C.: As a recent college graduate I obviously do not have the experience of real world implementation of risk management, however, I believe DR. Kaminski was over-zealous when in his statement that risk-management is a failure.
Such quantitative analysis systems should always be used only as a supplement to the entire risk analysis process and not undermine the effects of qualitative factors.
Frankly the writing was on the wall, three years ago when housing prices were peaking in some major metropolitan areas, creating a fanatical drive in the housing market.
The problem is that managers in many businesses have become to engrossed in short-term effects rather than planning for long-term outcomes.
Impatience is what undermines the economy and the sin of making a quick buck over the prudent long term diversified investor.
Steven Pearlstein: I don't think Vince would say that risk management is doomed to failure. He is saying it has failed because (1) imperfect models are relied on too extensively and (2) top executives don't listen enough to the risk managers. In other words, he thinks it is worth doing, but needs to be done and used better.
Columbia, S.C.: Your piece reminds me of a story within the LBJ administration. Apparently he had one upper staff member be his "Devil's Advocate" so he could bounce policy ideas off him. The only problem was that it became known what the guy was doing and he became ignored. Kind of like setting your clock ahead 5 minutes, sooner or later you will just subtract the minutes.
Risk Management seems to have the same, well, risk. If you put your risk management department in a stove-piped department, doesn't it just become the guy in the room that no one wants to hear? The crank?
Shouldn't a company desire to have it's employees all think about risk? Then it would be harder to avoid. But then, I guess they couldn't point to a "risk management department" and say they understand risk.
Steven Pearlstein: A very good point. I suppose you need both -- a separate department, because there are some specific skills involved and you do need the risk managers not to report to the line managers, plus a cultural change that welcomes and celebrates caution within the operating divisions.
Princeton, N.J.: I do not understand how one can make these predictions about statistics which are basically event driven. For example, imagine an economist in 1830 trying to make 75 year predictions. He might have predicted the civil war, but he certainly would not have predicted the boom in the last third of the 19th century due to the railroads. Similarly someone in 1920 would have to have predicted the Great Depression and the rise of computers. How much weight should we give to the worldwide retrovirus epidemic in 2027? Or the boom due to asteroid mining in the late 2030's? Or the cheap cure of all diseases in 2040 due to nanotechnology? And so on. Remember the predictions in 2000 of "surpluses as far as the eye can see."
Steven Pearlstein: There is a difference between making accurate predictions and pointing out a big risk out there with very bad consequences that has a non-trivial probability. That might be a 5 percent probability, for example, but if the cost of it happening were very big, then maybe you need to change your strategy in a way that reduces the probability, or reduces the damage from that 5 percent event, at a small cost to current profitability. I think that's what we're talking about here.
Former Banking Regulatory Attorney: It's more than increasing and changing the focus of the regulatory activity, though. The financial institutions have to agree that if their risk is to be socialized (great slogan, BTW!) then they have to accept a greater level of informed oversight. They were masterful at coming up with new loopholes and regulation-avoidance schemes, knowing that they would not be allowed to fail.
Steven Pearlstein: Thanks once again.
Alexandria, Va.: You've previously written that you considered the people at Fannie Mae and Freddie Mac to be smarter and more worthy of public trust than the people at other institutions. Have the events of recent weeks (exposing the failures of risk management at Fannie and Freddie) caused you to revisit that opinion?
Steven Pearlstein: I don't think I ever said that -- that's just the way you took it. They do have a good group of smart people, because they are able to pay good money to attract them. And to the degree that their institution has public-oriented goals it has to meet, they may operate with a more public interest mindset than, say, Citigroup. But I wouldn't want to push that too far.
Nor am I in any way surprised that their balance sheet and income statement have taken a hit from the meltdown int he mortgage finance market. How could it have been otherwise? They insure a ton of mortgages which are now suffering default in much higher numbers. They hold lots of mortgages and mortgage-backed securities which are being discounted because of the fear of non-payment, so they have to market these to the current market price. They are in this situation because their sole job is to provide financing to this one segment of the housing industry, the conforming mortgage segment, and make good returns doing it. Unlike some the banks that ganged up to try to kill them, they are not withdrawing from that market now that it is unfashionable and unprofitable.
Princeton, N.J.: But Steve, you miss my point. I guess I didn't say it clearly. Because we have NO idea about future events, we can't even talk about 5 percent probability. In 1830, how could you assign a probability to railroads in 1870?
Steven Pearlstein: I don't think anyone in risk management focuses much on 40 year planning horizons, except maybe in nuclear power or something like that. Ten years is a bit easier.
Darien, Ill.: Managing risk is an important topic but I think you may be missing the bigger picture that bank's absolutely ignore the advice of their own risk managers. Too many decisions are supported with sophisticated computer modeling without regard for common sense and historical events which almost always repeats itself. Take Continental Bank, LTCM and now Citi. In all of these historical failures their BOD members were mostly NOT bankers. How could a non-banker understand the complexity of GAP asset/liability reports or even understand the consequences let alone the meaning of SIV's?
Steven Pearlstein: Oh, I bet they could figure out something....
Columbia, S.C.: Posting again: I guess my larger point about the pointlessness of a risk management department is that it makes risk an external consideration. You would be more likely to try to "get around" or "ignore" the risk managers. Risk needs to be internalized. So, I could see people with the risk skill set being embedded in teams instead of in a formal department.
As a side note, this reminds me of my experience in accountability/auditing. Too many people rely on external accountability - test scores and the like. Accountability is most effective when it is internalized such that there are more seemless checks and balances (lesson plans with objectives and internal tests, internal measurable goals).
I don't think for effective organizations that such vital functions can be segmented into different departments.
Steven Pearlstein: You make some good points.
Chattanooga, Tenn.: People behave irrationally. Analytical models never seem to take this into account.
Steven Pearlstein: Behavior economists have, in fact, found patterns to supposedly irrational behavior, so it can be modeled up to a point. But I think Vince's largest point is the better one: someone with experience and smarts can see big problems developing without having to precisely model them.
Steven Pearlstein: That's it for today folks. "See" you next week.
Editor's Note: washingtonpost.com moderators retain editorial control over Discussions and choose the most relevant questions for guests and hosts; guests and hosts can decline to answer questions. washingtonpost.com is not responsible for any content posted by third parties.