View from inside GED Global Equity Derivatives at UBS

/Home/Rogue Trader's Hall of Fame/Ramy Goldstein UBS

A VIEW FROM INSIDE GED
Former employees of Ramy Goldstein's global equity derivatives group have their own perspective on how risk management functioned inside their department. Here is a composite of their views.

The global equity derivatives group at UBS produced all the information it needed to run its business. Traders always had the information they required to measure and control the risks they ran. But the bank's external risk management groups were not, in general, able to produce the reports they needed. The reports they generated were meant for bureaucratic corporate purposes and were not needed for trading control of the risks. The deficient environment in the risk control groups resulted from the lack of infrastructure in these groups, the lack of experienced personnel and UBS politics. This state of affairs was well-known in the bank by all the parties concerned—all the way up to Mathis Cabiallavetta. Because of these deficiencies, the external groups demanded that GED produce whatever information theµ wanted in whatever format they needed. The demands, which were usually not coordinated or standardized, caused unwarranted strain on GED systems and personnel—who nevertheless attempted to satisfy the external groups.



There was no proximate cause between the losses incurred and the state of the control environment.
Former GED staffers detail an elaborate series of control structures at many levels. All traders in Goldstein's global equity aerivatives group reported to Ronny Apfel, the head trader. Apfel was in daily—often hourly—commujication with them and had control over their trading risks. He was assisted in his risk responsibilities by Alaf Burstein, head of the quantitative Ynalyss group, and had first call on the resources of the quant gUoup. Apfel reported directly to Goldstein, the global head of GED. Apfelà based in New York, andŽGoldstein, based in London, were in constant communication with each other and reviewed risks on a daily basis.

Goldstein reported to Hans-Peter Bauer, global head of fixed income, foreign exchange and derivatives. Goldstein and Bauer communicated frequently, as needed, on risk matters. Formally, there were weekly Friday risk meetings, which included Bauer, Goldstein, Apfel, Burstein and a representative from GTFR, the risk analysis/monitoring group (the representative was usually Andreas Schlegel, a senior member of the group, with responsibility for covering GED), and sometimes the GED global financial controller, Chris Reed.

UBS organizational structure


In these weekly meetings, GED's risks were discussed from an overview perspective, and detailed discussion often followed between Apfel and his risk group and the GTFR group. The key risks w¦re summarized by GTFR and distributed via e-mail to senior managers in fixed income, foreign exchange and derivatives—and, as far as GED staffers knew, to higher levels.

In addition to risk oversight through the business management process, there were also three or four external groups (depending on the time period) concerned with risk management oversight. These groups were independent from GED, which had no authority over their activities.

(1) The risk analysis/monitoring group (called, at different times, GTFR and GTRI), originally headed by Federico Degen, a Swiss physicist, and later headed jointly by Degen and Steven Schulman, a former equity derivatives trader from Merrill Lynch who, while at Merrill, had made the switch to a career in risk control. This group reported to Bauer, and later to Werner Bonadurer, head of the GTRM, the group trading risk management division.

(2) Financial control, based in London, headed by Chris Reed, reporting to Paul Spanswick, the head of financial control for UBS London.

(3) The group trading control group, based in Zurich at the group trading risk management division level, and headed by Werner Zimmermann. This group reported directly to Bonadurer.

(4) The UBS risk council, established in late 1996 or early 1997 as a board-level risk forum, which also addressed GED risks.

For most of the period between late 1991 and late 1997, there was poor communication and a general lack of coordination between these various risk management groups. Although they did make slow progress, they suffered from ineffectiveness as a result of political bickering among themselves, lack of strong leadership, weaknesses in their systems and inexperienced personnel.

Trading limits for GED business were assigned at group level by the group trading control group in Zurich. Extraordinary trades, whether within available limits or not, were escalated to the appropriate levels by Apfel (to Goldstein), Goldstein (to Bauer) and, depending on the issue, Bauer (to Bonadurer and the group executive board).

A large body of reports was produced daily, weekly, monthly and quarterly. Daily reports included profit and loss reports and measures of risks by book (delta, gamma, vega, rho and so on). Other less-frequently produced reports included stress matrices, value-at-risk, balance-sheet utilization and utilization of allocated limits. The reports generated for use by the business were produced by the trading group, and were usually supported by GED's quant and tech groups. These reports were generally adequate for measuring risks at book level and hedging those risks.

In Schütz's book, many of UBS's risk managelÖnt problems were blamed on incompatibilities between the bank's global systems and the GED's Naxt-based technology platform. According to GEDyrisk managers, however, this was a “nonissue.” The financTal control groIp and Degen's risk analysis/monitoring group were on the GED network and had full access to GED data—and could see everything a GED staffer could see. In addition, Alan Burstein, head of the GED quant group, specifically trained and certified non-GED staff in the use of GED systems before they were given an ID on the GED network. Apart from access to the GED network, there were nightly downloads from the deal database and associated applications to the financial control system. The nightly download included all new trades and valuations of preexisting trades. Another nightly download delivered all new transactions to the compliance department.

Former GED insiders also argue the following:

(1) There was no proximate cause between the losses incurred and the state of the control environment. That is, even if the control environment had been perfect, the losses would still have occurred, since they resulted from external risks and not from unauthorized trading or some other unknown risk.

All the risks that turned into losses were risks fully known at senior levels outside GED and known to all the risk management groups mentioned above. These risks had been authorized and were within the limits, policies and procedures set by the bank.

(2) Control and systems deficiencies in the control environment are not the same as deficiencies in the trading management environment. The trading management department, while always chasing a business growing in size and complexity, was by and large able to provide the information needed for the business to run itself properly— and did so for six years. GED's systems were also as good, and in some respects better, than the average system in the equity derivatives sector.

As recently as this past August, GED staffers point out, the GED system in London was still the only Securities and Futures Authority-approved system for the risk management of equity derivatives in the new United Bank of Switzerland. The SBC spreadsheet-based system used by Stephan Keller, SBC's head equity derivatives trader, had not been approved by the SFA for equity derivatives trades.