It is accepted wisdom that the condition of the nation's banks reflects the condition of the nation's economy. In reality, bank stability leads economic activity, becoming worse one or two years before the economy deteriorates and improving one or two years before the economy improves.
The key to understanding this is to recognize the difference between bank stability and bank performance.
Rear-view mirror. Common performance indicators like the charge-off ratio (charge-offs to recoveries) or return on assets are actually lagging indicators of bank stability.
Newer indicators like interest-rate risk or volatile liability coverage give advance warning of weakness in banks by gauging the amount of risk against the bank's capacity for risk-taking. When risk exceeds capacity, the only unknown is time. (For a sampling of the factors used to determine if a bank is stable, see, "Indicators of Instability.")
Since December 1987, my firm has been applying a wide variety of advanced stability measures to U.S. banks with less than $1 billion in assets, using data from the regulatory agencies.
In our 1987 analysis, the banks in Massachusetts, which were performing very well, showed as much instability as the banks in Texas, which were being devastated by the economic downturn in the Southwest. And by mid-1988, the New England economy was seen to be in free-fall.
More recently, beginning in late 1988, banks in California exhibited a dramatic, sustained drop in stability. In June of 1989 the stability of California banks was at 67%, but by June 1990 the level fell to 55%; thus only a little more than half the banks in California were stable at that time. Articles about banking and general economic problems in California began appearing in mid-1991, a year or so later.
The stability cycle. Experience shows that stability in banking is much harder to rebuild than it is to lose.
The last five years have provided three examples of state banking systems going through this cycle. Each currently is in a different phase, so from all of them the cycle itself can be understood.
Texas has come the furthest. The bottom for Texas was in 1985-86. Exhibit 1 shows the last seven quarters' stability analyses of banks in Texas, the most recent data available.
Real improvement in the number of stable banks was not apparent until late 1991, fully five years after the "bottom."
Texas, which was until recently a unit banking state, lost one-third of its banks before recovering.
Massachusetts' problems on the other hand became apparent in late 1988 and were most extreme in 1990-91. Hope can be seen in the reduction in "troubled banks" and the increase in "stable banks" in the latter half of 1992. This could be the bottom of the cycle in that state, with true recovery coming in 1994-95.
In California, the cycle is in its earliest phases: as of the end of the fourth quarter, "troubled banks" represented a large percentage of the total, and represent 37% of all distressed (unstable and troubled) banks in the state (see Exhibit 3).
Based on the Texas and Massachusetts examples, the next phase in California will be a further drop in the number of banks, followed by three to four years of rebuilding, with recovery in about 1997.
National outlook. Applying what we know about the bank stability cycle to the U.S. as a whole, the outlook is good.
There are clear trends of increased stability, as demonstrated in Exhibit 4. In fact, median stability by state increased from 75% to 84% in the first half of 1992 alone--the largest change since we've been tracking stability. To keep up with this recovery--merely to hold position individual states had to show dramatic improvement in bank stability. And, in fact, there has been improvement in every state except California.
Using the historical "lag" relationship between bank stability and economic activity, this indicates a dramatic economic recovery in the second half of 1993 nationwide. …