Deregulation is good for economic growth but it can do crazy things to the cash flows of sleepy companies that are suddenly exposed to competition. Case in point: electric utilities. A new analysis by NERA, a New York-based economic consulting firm that's part of insurance brokerage Marsh & McLennan Cos., shows that the volatility of cash flows of U.S. electric utilities nearly doubled between the early and late 1990s. The increased choppiness was a foreshadowing of the even wilder swings that have marked 2000 and 2001.
NERA analyzed the utilities using a method it recently developed called cash flow at risk, or C-FaR. For each company, it assembles a group of similarly situated companies. If one member of the group has recently suffered a big cash-flow shortfall, it's taken as a sign that the same thing could happen to the others. The four key measures that determine how a company is grouped are market capitalization, profitability, stock-price volatility, and the riskiness of the business it's in. Take the stock-price component: The volatility of the Standard & Poor's 500 electric utility stocks shot up to about 22% recently after averaging around 13% during the 1990s (chart). (Roughly speaking, that means there's a one-third chance of the typical utility stock rising or falling 22% or more in the coming year.)
Although NERA lumps all the big electric utilities into the same riskiness group, companies don't have to be in the same industry to be lumped together. For instance, the model puts Dell Computer Corp. in the same category as Bed, Bath & Beyond Inc. because they're similar in all four criteria.
According to an article by Harvard University economist Jeremy C. Stein and other NERA associates in the December issue of Electricity Journal, in the early 1990s, there was a 5% chance that the cash flow of a typical utility would fall short of expectations by an amount equal to 1.8% or more of its assets. By the end of the '90s, that potential shortfall had jumped to at least 3.3% of assets.
The cash-flow-at-risk method isn't perfect. Since it's based on historical data, it can't forecast the impact of a change in conditions. So although the method would have picked up some increase in electric-industry volatility in the 1990s, it wouldn't have foreseen anything like the California mess. Then again, almost nobody else saw that coming either.