"Risk is the hottest topic in the compensation committee meetings in the boardroom in the current environment,'' says Irv Becker, national practice leader of Hay Group's U.S. Executive Compensation Practice. "Everyone's talking about risk," not just financial services firms.
Not too surprising, when pay packages on Wall Street were weighted 95 percent toward annual incentives. And caused a near meltdown of credit and financial markets, when the seemingly simple risk that housing prices might decline and foreclosures surge was not anticipated.
But there is no easy way to create a benchmark that will be easy to implement and at same time provide a valid basis for rewarding chief risk officers, operations managers, fund managers, traders and their supervisors for managing and mitigating the risks of buying, selling and investing in increasingly complicated securities.
"It's a challenge. No one's figured it out yet. And everyone's working on it,'' said Becker.
Such a benchmark is critical, if there is to be meaningful compensation at year's end to good risk management, just as there are year-end rewards for trading profitability. Beat the number and you get a bonus. Miss it, you don't.
"'That's exactly what this is all about,'' says Allan D. Grody, president of Financial InterGroup, a New York consulting firm who is working on a framework to account for risk and along the way make it possible to give managers financial rewards which are based on quantified ways of managing risk well.
The measure that is recommended as a starting point by Grody, who is a retired professor of risk management systems at New York University, Robert M. Mark, the CEO of Black Diamond Risk and former chief risk officer of the Canadian Imperial Bank of Commerce, and Peter J. Hughes, managing director of financial modeling firm ARC Best Practices in the United Kingdom, is a return on capital that is adjusted for risk.
The chart shows the overall calculation. Its pieces are risk-adjusted revenue and an amount of capital that is set aside to cover unexpected losses, known as "economic capital."
The risk-adjusted revenue is divided by the economic capital to produce the overall benchmark of risk management performance called Risk-Adjusted Return on Capital.
Two of the pieces of economic capital are fairly well-measured. They are market risk and credit risk. For these set-asides, securities firms, wealth managers and fund managers have extensive historical information that can guide them.
In fact, in their operating budgets already should be adjustments for, say, a 5 percent drop in overall market prices. In that case, a company managing $10 billion in assets should already have set aside $50 million, in reserve.
But a good risk manager should be allowing for an unexpected level of market risk. This would be resulting from an unforeseen event or sequence of events, such as 9/11 or the disappearance of two or more large investment banks from global markets.
In that light, a good risk manager might set aside enough capital to cover another 10% drop in market prices. That manager would have put aside another $100 million, or $150 million all told.
The case in credit risk would be essentially the same.
A company might historically have found that it fails to get paid back on 2% of the loans or lines of credit it extends to customers. In that case, a firm extending $5 billion of credit to customers might be normally setting aside $10 million to cover a normal rate of nonpayment.
But then there are those times when, say, housing prices collapse. The ripple effect, even on securities trading, might lead to unexpected losses three times as great. And a good risk management team would be rewarded if it set aside another $30 million to protect the firm from the unforeseen risk. Most often overlooked, though, says Grody, is operational risk.
"The operational risk number is the tricky one. All the stuff we know intuitively that managers do, managing the risk of their operating environment," Grody said. "Now the requirement is to quantify that in some way.''
What needs to be recorded effectively are the ever-present and usually uncaptured accumulation of drains on profits that occur from faulty actions by humans, their interaction with data gathered by the firm and the systems they use to accomplish tasks.
Most firms, Grody notes, do not have significant or lengthy histories that trace the frequency or severity of losses from operational risks. The Basel accords on capital management have defined such activities to include losses resulting from human errors, implementing new technology that doesn't work, existing technology that temporarily craters, burps from department reorganizations, instances of fraud or even the risk of an improperly designed product or financial model.
So how much money to set aside for unexpected losses from such operational risks?
Grody and colleagues estimated five years ago that about an additional $4 billion of the $100.9 billion of regulatory capital that Citigroup maintained would be needed to be put aside to cover operational risks.
Here's another way to develop a rule of thumb. Grody says that banks, typically, adjust the value of their assets for risk, using risk weighting factors. Then, they set aside about 8 percent of the value of these "risk-adjusted" assets as capital to cover unexpected risks and additional amounts to meet mandated minimum requirements for reserves designed to meet lending risks. Of the capital number, operational risk amounts to approximately 12 percent.
That's the equivalent of 1 percent of overall capital, very roughly. So, in the absence of a good set of industry-collected data on what level of losses can be expected from operational risks, then a company managing $10 billion in assets could set aside $10 million for expected losses. And then put another amount, say another $10 million, into its economic capital account to combat unexpected losses.
Over time, the thumb-rule approach would be replaced by experience and collected knowledge.
The Operational Riskdata Exchange Association in Zurich, for instance, has been collecting reports on losses from operational problems since Jan. 1, 2002. Its database on incidents from 51 member banks in Europe now contains 102,500 losses, each exceeding 20,000 euros in value. The reports include date of occurrence, date of discovery, different categories to classify the event by and a Basel-designated business line to assign it to.
In Toronto, a software company, Algorithmics, has pulled together nearly 12,000 publicly reported losses that can be traced to operational risks.
But, that still amounts to precious little guiding data on losses from all forms of operational risk, on either a granular or aggregate basis, for companies to precisely predict losses in the future, Grody notes. Credit card companies use billions of transactions for their calculations. That said, for the example here, the denominator of the risk performance measure is complete: $100 million into economic capital for unexpected market risk, $30 million for unexpected credit risk and $10 million for unexpected operational risk. All told, $140 million.
Calculating the numerator-the return-on the measure is more straightforward. A firm can start by looking at its income statement.
Add in operating revenue. Subtract operating expense. Subtract any losses already allowed for, in normal course of business, from risk. And then add back an interesting quotient: The actual monetary return the firm can expect to earn from setting aside all that economic capital, to guard against unexpected losses.
So, if the company is earning 6 percent on investing the capital it holds, then the firm should add in $8.4 million to that numerator.
Shortcut to the final result: Let's say operating revenue minus expenses and expected losses is $11 million and the return on the economic capital is $8.4 million.
Then the risk-adjusted return on economic capital is $19.4 million divided by $140 million, or 13.9 percent.
Beat that, and a company's risk management team should be entitled to a slice of that as incentive compensation. Come in under that and it's wait until next year.
Of course, if you come in under the target, you might also be rewarded by the target being lowered for the next year.
"If your losses are less than your unexpected losses, then you adjust your model because you're putting away money that could be leveraged into your business, rather than simply putting it aside," Grody said.
There are other approaches. UBS, the Swiss investment bank trying to restore its reputation as a careful, responsible operator, has adopted an approach of long-term incentives that includes a "bonus bank." If a top performer earns a $3 million bonus, the person only gets a payout of $1 million. The remaining $2 million goes into the "bank." If operating objectives are met, the remainder can be drawn out over three years. If the objectives aren't met, some or all of the "banked" bonus is lost.
The ultimate answer, Grody contends, is to produce an entire system of "risk accounting,' to manage centralized and dispersed risks in a company, just like standard methods of "cost accounting" were developed, to manage centralized and dispersed costs in an organization.
"When a CEO, like Richard Fuld at Lehman Brothers or James Cayne at Bear Stearns, loses a billion dollars, they clearly have downside risk exposure. So you can't say they were motivated to undertake excessive risks. They made serious mistakes,'' says Ira T. Kay, director of Watson Wyatt Worldwide's compensation practice. "They didn't find the true risk. But it was more likely to be the risk model that was the problem, than the executive pay model."
If risks weights were put on each transaction as it was executed, then risk managers could watch the risks in their securities portfolios ebb and flow, accurately, Grody said.
"You could acTtually see risk exposures building up in the organization, not relying upon some models from the past to predict your losses, "Grody said. "It's almost like the difference between sitting having breakfast underneath a volcano and the only thing you know about is that the biggest eruption caused 33,000 deaths. Wouldn't you like to have some seismic monitor telling you there's a little bit of rumble and that it's getting bigger and bigger?"
RISK & REWARD
One formula proposed as a benchmark for awarding incentive compensation that rewards traders, operations managers and top executives at securities firms for managing all the risks a firm faces is called is return on capital, adjusted for risk.
The base calculation:
Risk-Adjusted Revenue ÷ Economic Capital
Risk-Adjusted Revenue =
Operating Revenue -
Operating Expenses -Expected Losses
+ Return on Economic Capital
Economic Capital =
Capital Set Aside for Unexpected Losses from Market Risk + Credit Risk + Operational Risk
One way the benchmark could be applied:
Trader earns $1 million bonus.
$500,000 is paid out immediately.
$500,000 is deferred.
During the deferral period, the individual is entitled to a payment equal to the risk-adjusted return on capital on the deferred amount.
So if the company's overall risk-adjusted return on capital is 9%, the person gets an additional payment of $45,000 in that first year.
This article can also be found at SecuritiesIndustry.com.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access