"A sad chapter in the history of American business," intoned William Donaldson, chairman of the Securities and Exchange Commission, last week as he announced that 10 of the biggest names in US investment banking were coughing up $1.4bn (£871m) for publishing research they knew was wrong.

A sad chapter it may have been, but it is not the end of the story. The final chapter in the extraordinary narrative of the 1990s internet boom is still being written. It will not end until the prison gates slam on at least some of the participants. But we can already begin drawing the morals of the tale.

Analysts at the miscreant banks wrote enthusiastic notes, particularly about internet and technology companies, that they knew, privately, were shaky. They did it because, in a ferociously competitive market, it helped their investment banking colleagues win those companies' business.

Thoroughly reprehensible, as Mr Donaldson said. Yet reading the SEC's case against the firms, it is clear that the analysts' dilemma was not whether to produce honest or misleading research. It was whether to produce misleading research or find jobs as zoo keepers, couriers or celebrity party organisers. Writing bogus research notes was not a fall from grace; it was the essence of the business.

Companies looking for investment banking services made clear what they wanted: research notes from the banks that said their prospects were great. The investment banks were happy to oblige. Analysts were expected to attend "pitches" to potential investment banking clients. The analysts' (enormous) bonuses were tied to the amount of investment banking business they generated. Their investment banking colleagues' opinions played a crucial role in the appraisal of the analysts' perform-ance, remuneration and prospects.

Some analysts felt tugs of conscience, but they soon learnt the rules of the game. "Unwritten rule number one: that if you can't say something positive, don't say anything at all," a CSFB investment banker told one of the firm's analysts who downgraded a client company.

Except that not saying anything at all was not an option, either. After CSFB underwrote the initial public offering for a company called Digital Impact, the firm's analyst regularly rated it a "buy" or a "strong buy", even as its share price sank from $50 to less than $2. When a new analyst took over, he attempted to duck out of covering the stock. No way, the investment bankers told him. He continued to describe the company as a "buy" until he screwed up the courage to downgrade it to a "hold".

Some analysts, like dissidents in a repressive dictatorship, published numbers showing the companies doing well, but used the commentaries to hint obliquely that they should not be believed.

Who were the victims? Small shareholders who relied on the research. "I feel bad for those naive investors," an institutional investor e-mailed a Lehman Brothers analyst. "Yes, the little guy who isn't smart about the nuances may get misled - such is the nature of my business," the analyst replied.

Under last week's settlement, the 10 firms will separate their investment banking and research activities. For five years, they will have to commission outside, independent research.

So what are the lessons? First, companies that want to be trusted need to decide which customers they can serve. The banks did not treat all their customers badly. They gave their investment banking clients exactly what they wanted. It was their small retail investors they short-changed. Few companies treat all customers equally: the highest spenders get better treatment. Companies need to decide what level of service to offer the others. They need to set a standard and stick to it. Whatever it is, it should indicate some respect for less profitable customers. They may be paying less, but they are parting with their money all the same. Less luxurious service may be in order. Lying is not.

Companies also need to decide what to be trustworthy about. Glen Urban, professor at the MIT Sloan school of management, says there are some areas where companies, however well-intentioned, may not be able to deliver. For example, airlines cannot promise punctuality; too much is outside their control, such as air traffic control and the weather. They need to develop their relations with customers based on something other than timekeeping, such as swift check-in or on-board service.

In a recent paper*, Prof Urban says companies that do not have the right product for customers should suggest they go elsewhere. Customers will discover in the end that what they have bought is unsuitable, and the rapid exchange of information over the internet will mean others will find out too. Companies that tell customers when they do not have the right product for them are more likely to be believed when they say that they do.

Honest investment banks might have decided they could serve corporate customers or retail investors but not both.

Can they serve both now? The Chinese walls that, under last week's settlement, will separate the investment bankers from research may ease the pressure on analysts, but it will not disappear. Small investors may, in any event, want to ask themselves why companies that deceived them yesterday deserve their business today.

*The Trust Imperative. http:// papers.ssrn.com/sol3/ papers.cfm?abstract_id=400421 .............................................................................michael.skapinker@ft.com