Page 299

cent to fixed-income arbitrage, 15 percent to distressed securities, 3 percent to convertible arbitrage, and the remaining percentage allocated to various other strategies such as global tactical asset allocation, emerging markets, and currencies. Casscells refers to these as absolute return strategies. These strategies should make money regardless of whether stocks and bonds or rising or falling, excluding periods of liquidity crisis.

In these areas, the skillful manager will add value to the upside and reduce risk. She makes the analogy between these managers and underwriting insurance. Good underwriting of risks helps trim losses in difficult markets.

Selection Criteria

The selection, monitoring, and rebalancing of managers is done internally. There are eight people in the investment area. More than one person meets with a manager. And the allocations are approved by the senior staff, which meets weekly.

What are the selection criteria? Casscells says that a five-year track record is attractive. "It is important for them to have experience as an independent firm in difficult markets such as during 1994 and 1998." She observes that many of the managers who worked at investment banks during these difficult times were not independent. They personally did not receive the extreme margin calls. It is for this reason that she prefers managers with longer track records.

To date, all the managers they've allocated to have been U.S.-based since there are few non-U.S. managers in these categories. Nevertheless, most of the managers they allocate do trade on a global basis. For example, in market neutral, a number of the managers cover the United Kingdom and Europe. In the fixed income arbitrage category, G-7 and G-11 countries are included. Merger arbitrage managers trade in Europe. Distressed, however, has not yet successfully expanded outside the United States. This may be due to the more challenging bankruptcy systems abroad.

On the issue of asset size, Stanford prefers that the managers not be too large because it is often difficult to deploy lots of assets. Stanford also prefers that managers do not go substantially outside their specialty. Casscells observes that those managers that got hurt in 1998 in