(FORTUNE Magazine) – Now that private pension plans in the U.S. have amassed assets totaling roughly $1 trillion, the quality of pension fund management is a major economic issue. Unfortunately, the record looks disappointing. Examples abound of fully funded pension plans that are run very well, both by corporate sponsors such as U.S. Steel and General Electric, and by professional money managers such as Batterymarch, Capital Guardian, and Alliance Capital. But the pension system as a whole is earning subpar returns on investment. Some fundamental changes in pension fund management might give employees more assurance of a secure retirement. A paucity of data long made it difficult to arrive at an informed judgment about investment performance. Almost all of the publicly available information focused on performance by individual money managers, who typically invest only part of a corporation's pension fund. The broader data that did exist indicated that pension funds as a group barely kept pace with inflation over the past two decades. Pension funds invested in stocks apparently did worse than equity mutual funds. In other words, pension plans seriously underperformed other large investors and failed to provide a real return over 20 years, a period that has to qualify as the long run. Better data that have recently become available confirm that dreary impression. Wilshire Associates, a California consulting firm, has assembled information from some of the major banks that administer master trusts for pension plan sponsors. From these data Wilshire created its Trust Universe Comparison Service to enable corporations to evaluate their entire pension plans as well as the performance of their individual money managers. Wilshire's statistics indicate that the median manager of stock market investments achieved an annual rate of return of 14.5% over the five years from 1980 through 1984. That was a tenth of a percentage point lower than the annual total return (capital gains plus dividends) of the Standard & Poor's 500 index for the same years. The median manager of fixed-income investments achieved a 12.4% rate of return over the five years, a tenth of a percentage point lower than the Shearson Lehman index of government and corporate bonds. Those figures overstate the returns that employee benefit plans actually received. First, the data exclude the results of managers whose performance was so bad that fund sponsors fired them before the end of the period. Omitting their results boosts the median performance about four-tenths of a percentage point a year. Second, the results are before management fees, which average around a quarter of a percent of the assets. Third, the median manager's performance does not include the expenses of shifting a portfolio from one manager to another; the new manager usually liquidates most of the portfolio and invests the proceeds in different stocks, generating substantial commissions. Dennis Tito, the president of Wilshire, calculates that those three factors reduced the net returns received by pension plan sponsors by nearly one percentage point a year. Thus median managers appear to achieve stock market and bond returns that are approximately one percentage point lower than the returns on broad market averages. In other words, the active management of pension assets has generated lower returns than the corporate sponsors could get by passively investing in ''index funds'' that replicate the market averages. Tito estimates that a pension fund has only a one-in-three chance of beating the market averages over a five-year period and only one chance in four over ten years. Why have managers performed so unimpressively? The principal reason appears to be the sheer size of the assets invested in pension plans. Pension funds can't beat the market because they are the market. They own about one-third of all the stock in U.S. corporations and half the long-term corporate debt. Fifteen years from now pension plans will own half the stock. I have two suggestions for improving this sorry record. First, the U.S. Treasury should issue bonds that are indexed for inflation. Such securities are available in many other countries, including Britain. The British government finances half of its new debt with indexed bonds that pay a real rate of return of about 3%. Indexed bonds would enable pension funds to lock in a positive real rate of return. Other investments might have higher expected returns over a long period, but indexed bonds would guarantee a solid performance in every period. Entire cohorts of workers whose pensions are based on pension fund performance would no longer run the risk of reaching retirement age when the stock market was in a deep slump. Furthermore, indexed bonds would provide a perfect benchmark for judging the performance of active managers. No longer would ''good'' performance by a money manager be determined solely by comparison with worse performance by others. My second solution involves rethinking the basic philosophy of pension fund management. Pension assets have grown so large that when one fund sells its stock in General Motors, the buyer is apt to be another fund. Viewed as a whole, the pension system is becoming a permanent stockholder. Pension funds as a group simply cannot get out of a stock that suddenly looks unattractive to them because there aren't enough other investors around to quickly absorb their holdings. To thrive in this environment, pension plan sponsors should replace the active management of securities with the active management of investments. Instead of trying to beat the market, they should get involved in the long- term operations of the companies that their pension plans own. Instead of trying to add value by juggling stocks, pension fund sponsors should add value through active, informed ownership of the companies they invest in. The lines between management and shareholders clearly will have to be redefined to accommodate more energetic owners. And such a change obviously will require new skills and competencies. Present money managers have neither the inclination nor the training to act as proprietors of the companies whose stocks they buy. Either they will have to acquire those characteristics or new institutions will develop to fill the need. RADICAL as this proposal may seem, it amounts to little more than a return to the traditional capitalism of the late 19th and early 20th centuries. The pension funds obviously can't take over day-to-day management of the companies they control, but they can speak up when shareholder interests clearly aren't being served. Some of them already do. Money managers such as Dean LeBaron of Batterymarch and Jesse Unruh, the California state treasurer, have been outspoken opponents of attempts by corporations to adopt anti-takeover amendments. Their efforts exemplify the type of involvement that all money managers should practice. A voluntary pension system that encourages excellent investment results must remain the primary source of America's retirement income. But performance that, on average, is consistently mediocre will not long escape the attention of Washington policymakers. The pension system itself should act to improve performance before the legislators are tempted to intervene.