Analysts, strategists and fund managers polled by the South China Morning Post were almost spot on in their projections of where the Hang Seng Index would end last year. According to forecasts made more than 12 months ago, the index would rise an average 12 per cent during the year and clock out at 14,109 points. It did almost that, finishing at 14,230.14, or a mere 121 points above the average. The closest projections came from Celestial Asia Securities (for the second year), Daiwa Asset Management and HSBC, whose projections ranged between 14,000 and 14,300. The full range spanned 13,000 to 15,200. Their crystal balls proved somewhat hazier when it came to the H-share index. Most market observers agreed the index returns would be modest after a 152 per cent rally in 2003, which left the H-share index at a 61/2-year high on the final day of that year, but only three of 12 forecasters correctly foresaw that the index would fall. The H-share index hit its year high of 5,440.75 on January 5 - a level it never reclaimed after a sharp correction in April and May. It edged above 5,000 briefly in late November and early December and finished the year with a 5.55 per cent decline at 4,741.32 points, against an average expectation of a 3.9 per cent gain to 5,219 points. More and more banks and brokerages are moving away from setting index targets, especially ones that centre on a specific date such as the end of the year, choosing to focus instead on individual stocks. Perhaps that will be the way to go amid growing pressure for analysts to be held accountable for their calls. However, it is likely investors will still want to know what direction they can expect from their market. And let's face it - a projection is still a projection and should never be treated as science. They can be wrong.