Investors should reduce their holdings in bank stocks to a market weighting, said Lawrence Cohn, an analyst with Paine Webber Inc.

Bank stocks have ridden the three-year-old bull market for about all its worth, said Mr. Cohn.

In his view, a company s earnings improvement compared with other companies is the single most important issue in relative stock performance. He forecasts that bank earnings as a whole will rise 20% next year - the same as for corporations in general

"On an industry basis, there's no reason for bank stocks to do any better than the market," said Mr. Cohn.

When he changed his view of the industry from aggressive overweighting to a market weight in October, he put "sell" ratings on Bankers Trust New York Co. and J.P. Morgan & Co. His top buy recommendations: Bank of New York Co., BankAmerica Corp., and Bank of Boston Corp.

Mr. Cohn had considerable success in his stock picks. According to a survey by The Wall Street Journal, Mr. Cohn ranked No. 1 among bank analysts in stock picks for 1992.

Q.: Which stocks did you pick or not pick that were so successful?

COHN: The winners we had in 1992 all shared a common characteristic: They were the bottom of the barrel, they had substantial problems in the prior cycle, and they were on the rebound. You had to find the lousy banks.

Whether it was Midlantic, Shawmut, Chase, Bank of Boston, Bank of New York, even Citicorp at one point, they were stocks that had had some problems.

Q.: Are there investment themes this year?

COHN: I'm not sure this is a theme-driven market. It is tougher to make money in these stocks, and we've recommended that people reduce their exposure in the group. The absolute levels of valuations are such that the upsides are much less than two years ago.

Two years, when I started to recommend the group, the only stocks I would recommend were those where I could see at least a doubling in value. A year ago, I was only recommending stocks where I could see a 50% gain. There are none of those guys out there anymore.

Q.: So you've gone to a recommendation of market weighting?

COHN: Around six weeks ago, when stocks were higher, we looked out toward 1994 and 1995. It was clear that relative earnings momentum was starting to run out of steam.

Relative earnings improvement is the single most important issue in relative stock performance.

Our forecast for the banking industry is a 20% earnings gain, the same as nonbanks. So on an industry-aggregate basis, there's no reason for banks to do any better than the market.

The industry is running out of earnings leverage. Margins are not expanding anymore. Loan-loss provisions are reaching their lows.

But at some point next year, if the economy stays reasonably strong and lending volume is O.K., lots of banks will have to start raising their loan-loss provisions just to maintain their coverage ratios, because the portfolio will be growing.

Every cycle I've gone through, we've reached a point in the cycle where earnings growth looks fine but the stocks don't do anything.

It is a very frustrating period. What you tend to get in the mid-to-end period of the cycle is stocks that lag behind the market and then get the violent catch-up moves before lagging again.

You can't make the argument that this is a growth industry. It is very cyclical. Portfolio managers have learned the hard way that analysts are terrible at predicting the end of the cycle.

As you get into the cycle and the stocks get close to fair value, they say, "the hell with it, we just won't play," rather than take the risk of holding the bag when the game ends. So the group gets very volatile. We are in the early stages of that in my opinion.

Portfolio manager are getting a lot more selective. They aren't putting new cash into the group. They are looking for banks where there is real clear differentiation, a story that will draw money into a group that is otherwise a dull group.

Q.: How do investors view the lower loan-loss provisions?

COHN: Two years ago, when we started recommending the group, we looked at normalized levels of earnings, what banks could earn once their credit costs came down to average levels over a cycle. In 1991 and again in 1992, reported earnings were dramatically lower than normal earnings.

Now we are at a point where we are starting to see lots of banks whose reported earnings are above their normalized earnings. Their loss provisions are dramatically below their long-term sustainable levels.

And it is at that point in the cycle where you start running into real problems making money in the group, because investors don't know how much to value the earnings stream.

Q.: How much above are some banks?

COHN: Bank of Boston's loss provision in the third quarter was $10 million. We estimate that the normal provision is $47 million. Their normalized per-share earnings on a quarterly basis are 59 cents. They reported 82 cents.

We normalize Bankers Trust on a different basis, but the result is the same. That's why we put a sell on them and Morgan. We looked at average trading results over a period of time.

They were $73 million higher in the second quarter than we estimated the average to be. As a result, we came up with normalized earnings in the second quarter of $2.30 per share, they reported $2.97.

Q.: Would they say in response that derivative activity is picking up?

COHN: They do say that. So does J.P. Morgan. But I don't believe that argument holds water. Look at the price of a 30-year constant maturity bond. Over the last five years, a constant investor in a 30-year bond would get a rate of return in price appreciation of about 10% a year.

That's exactly the same rate of growth in the trading profits. What that says to me is that the growth in bank trading revenue is related to the fact that we've been in a bull market for bonds for the past five years.

What has yet to be tested is what happens when we go through an extended period of time when prices go down. We hypothesize that trading profits will drop a lot.

Q.: So in a themeless market, what do you pick?

COHN: There's a continuation of the old turnaround theme in BankAmerica. The West Coast is still the last area where we can get some real kick out of an improving economic outlook, and BankAmerica is the most leveraged of the big California banks to that improvement.

There is a California and a non-California part to BankAmerica. The revenues are about 50-50. The non-California part is showing terrific growth, and the California part is deteriorating. The net of that is no growth.

You don't have to get California growing for the bank to show really great earnings. All you have to do is stop California from bleeding. The growth in the rest of the company is so substantial, it will carry the revenue base.

Bank of Boston is a company in transition and a stock idea in transition. The original basis for the recommendation was an improvement in asset quality. Now we are betting that they can get a lot more efficient than they have been and that they can make acquisitions efficiently and effectively, unlike in the past.

Bank of New York still has some improvement coming from declines in the loss provision. They have a great credit card operation and a terrific processing business. Those two businesses are a little over half of the bottom line, and they are showing really good earnings. The rest of the bank gets carried along with it.

Both Bank of Boston and Bank of New York are very cheap stocks. The market doesn't accord them the kinds of valuations they deserve based on their growth characteristics.

Q.: Don't investors have plenty of reservations about those stocks?

COHN: Bank of New York has been my single most frustrating stock this year. There is a huge chunk of the world that lumps the New York banks together. If they own a Citicorp or a Chase, there's no room in their portfolio for a Bank of New York.

Second, there's this incredible fear out about dilution. The company doesn't have a history of diluting, and in fact, the last couple of acquisitions have been antidilutive.

Bank of Boston's history is not one that gives you a whole lot of confidence. We are arguing that this is company that in the future will do a good job of making acquisitions. They will price them right and integrate them well.

Second, this is a management that hasn't done a great job of controlling its expenses and is now beginning to demonstrate that it is.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.