YOU MAY RECALL seeing the headline on the lead story of our front page on Saturday - 'Flats data demolishes housing supply fears'. The reason for this sudden confident claim by our government was a new release of statistics, supposedly making the future transparent to us at last - 15,000 units unsold in completed projects, plus 54,000 units under construction, less 6,000 units pre-sold, yields 63,000 units available for sale. Now let me tell you something that I learned from the 18 years that I spent as an investment analyst in this town. It is that a new property analyst takes much longer getting up to speed on his sector than an analyst covering any other sector. The reason for it is simple. There is far more published raw data on property than there is on any other sector, and I mean mountains more. It is my guess that the readily available figures are more exhaustive than anything you could find in New York or London. Leave alone the detailed statistical series put out by government - broken down for almost every conceivable category of supply, demand and price, and which covers every sort of building and plot of land - there are further figures in detailed academic studies, from estate agents and, of course, from the big property companies themselves in their annual reports. Thus, when officialdom tells us that all was previously murky but now stands revealed because A plus B minus C equals D, I imagine that every property analyst in town greeted the news with a loud horse laugh. Leave alone that the equation could long be worked out from existing data (a property analyst who does not do it on his first spreadsheet ought to be sacked), these wonderful new figures are actually less than we used to get. At one time the Lands Department made a regular three-year forecast of demand and supply, then a two-year one and now effectively nothing at all that is really news. What is more, I think the government's much ballyhooed conclusion of there being an abundance of residential supply is demonstrably wrong. In fact, it is the polar opposite of the real state of affairs. The red line in the first chart represents a 12-month average of the number of flats completed over the past 20 years in both the public and private sectors. You will notice that it is hardly distinguishable from the trend of the blue line. This represents a 12-month average of the number of flats authorised for construction. The twist in this chart is that I have moved the data for the blue line 30 months to the right of the data for the red one because, as should be perfectly obvious from the chart, completions of new flats follow authorisations by about 30 months. The blue line thus gives you an accurate forecast of completions up to 2006 and it tells you that the coming trend is straight down. It suggests that new supply in 2006 will be only 27,000 units as against a present 20-year average of 69,000 flats a year. Despite a booming economy, supply of new flats will be at a more than 20-year low by the middle of next year and then worsen. There is more that needs to be noted here, however. The shortfall is not only in the number of flats, but in the amount of space they offer. As the bars in the second chart show, the usable area of the average private-sector flat in Hong Kong is tiny by any standard and has changed very little over the past 30 years. In fact that average flat was larger in 1975 than it is now. I shall grant you a range of possible error in these figures but not much. Yet over this period Hong Kong has grown wealthy. Real gross domestic product per capita has almost quadrupled. We have outgrown the shoebox days but our homes have not and the pent-up demand for better accommodation has become so great that, on a per square foot basis, large flats now cost three times as much as small flats. Let me stress this. I am not talking of price per unit but of price per square foot. Please join me in that horse laugh. Our government has scored another triumph of fantasy over fact.