If you have been following the Malaysian House Price Index published by the National Property Information Center (NAPIC) in the past few years, you would probably have seen some sharp drops in house price growth. You would think, all the government and the central bank’s initiatives were working well.
But if you were really paying attention to it, you would notice that the series has been suffering from drastic revisions since 2010. You would realize the revised drops were shallower than before, or it was not really a drop at all. Which makes me unsure about the effectiveness of all those tightening by the authority.
I have noticed this for quite some time now but it was only recently that I decided to look into it. And true enough, the latest data point from the index usually underestimates the rise in prices. Or more accurately, the preliminary reading underestimates the final data.
You can see the difference between the revised data and the preliminary data below (red is the revised/actual index and blue is the series if there is no revision at all):
The revised series is just the latest series available from NAPIC, which is reproduced by Bank Negara Malaysia in its Monthly Statistics Bulletin. For the unrevised data series, I had to go through old databases maintained NAPIC and the central bank and collate all preliminary readings for each quarter.
You can see how drastic the revisions can be from the chart above. If you like numbers, the root mean square deviation for 2003-2009 is 0.9% while the average RMS for 2010-2013 is 3.3%. I chose 2003-2009 and not earlier because earlier unrevised figures are somewhat unreliable because they were not published every quarter.
The drastic revisions give a very wrong impression of reality (assuming the revised series gives the right version). That may sound obvious but if you know nothing of the revisions, the series may scare you. In 1Q13 for instance, you would see a sharp collapse price growth, which would probably push homeowners into panic mode (see the blue series below):
If pre-revised data is to be believed, house price growth slowed to 6.0% YoY in 1Q13 vs 12.2% YoY in the previous quarter. The end is near, one would say. Finally, BNM tightening were working, you would say. But after revision, growth was at 10.7% YoY (the short red line in the second chart), much higher than the preliminary release.
From the look of the revised index, house prices growth are decelerating, but only slowly.
More importantly, since the preliminary data is subject to large revisions, there is essentially a 6- or 7-month recognition lag. It is only after 6-7 months that you can say something about house prices with great confidence from reading this index. If you do not have the patience, then the only thing reliable from the preliminary data is the direction of change.
