The bollocks still flows

It's been a while since I debunked some mainstream media "reporting".  That's not because the media have got more accurate, but because I got bored of it.  It was the same old tripe, over and over again, but still no-one picked up on it.

To begin with, let's start with some real news: Mortgage approvals rose in April.  This is true, there's no bollocks here.  However, the bollocks starts immediately afterwards; the BBC has done quite a good job, for a change, of being neutral in that article quoting both the "conditions are still difficult" and the "recovery is here, whee!" camps; it's other less reputable journals of opinion that are full of it. 

There are certain patterns, which is no accident, it's deliberate brain washing.  The most common one is that these press releases "confirm the stabilisation of the housing market".  Why is this bollocks?  Because the very notion that there is every such a thing as a "stable" housing market is bollocks.  Just take a look at the famous Nationwide house price index, plotted as a graph.  (And, in case you think I'm taking that graph from a biased source - and it is a biased source - you can find the same thing in the monthly Nationwide press releases, but they're PDF's so I can't link direct to the graph.)  You see that the market goes up, and the market goes down, the market is never "stable", except for brief periods when it's way below the trend line.

It is true that, at some point in the future, there will be another house price boom.  There always has been in the past, and there will, undoubtedly, be one again.  On that score the perpetual media optimists are right; but that's where it stops.  For there will, after that point, be another collosal unpredicted crash; the media will deride the very possibility, they'll point out that "the problems of the late 2000's credit markets have been solved", and other such platitudes, but events will conspire, they always do.

But the real bollocks is the amount of gun jumping that's going on.  While mortgage approvals did rise, they rose to the 43,000 level.  This is only half what is generally regarded as the neutral point; a good price predictor over time is that levels less than 90,000 result in price falls, and higher than 90,000 price rises.  Obviously this is a rule-of-thumb, and circumstances always vary, but it's been historically highly accurate.  So mortgage levels of 43,000, in traditionally prime house buying months, isn't very strong at all.  It's slightly stronger than before, but that's not saying much.

And even though 8% sounds like a lot, at that rate it would still take a year to get to the much fabled 90,000 level.  In fact, considering where mortgage approvals appears to have bottomed out, then the recovery is only 80% below a stable level.

That last statistic, the 80% one, is of course bollocks too.  I was just using it as an example, what I'd done was measure the current mortgage approvals on a line from the low point to 90,000.  It means nothing; but that, of course, is what most other statistics are too.

On the day (yesterday) that the Land Registry announced their most recent house price statistics, which, incidentally, are still negative; the Telegraph published this fluff-piece.  Despite the Land Registry being the most accurate measure of house prices, the author chose to instead report other statistics.  Again, it's no coincidence that both were published on the same day, it's a deliberate attempt to control public conceptions; to drown out the "unhelpful" but accurate statistics with cherry-picking from the mass of noise.  The author's business is selling mortgages, so it's no surprise that he's bullish, but still, that doesn't make of the rest of that article more accurate.

His claim is that prices are actually rising, on a three-monthly basis, at 4.2%.  If true, this would be not only "stable"; it would not only be "growth", it would be an annual rate of nearly 17%, a rate of growth that (despite all the property programmes on TV telling you otherwise) is actually quite rare.  And such rates are always unsustainable, but they don't like to dwell on that.

(If you apply the same analysis to the Rightmove house price index, you'll see they're claiming price rises of 46% annualised; but not even the most cheer-leading press bothers to quote them anymore, reality and the Rightmove index parted company a long, long time ago.)

Fortunately, that statistic was reached through faulty means, house prices have not risen 4.2%.  So what were the faults?  There were many:

Firstly, he chose the Nationwide index.  Whilst this isn't the worst one available, it's not the best either, it ranks somewhere in the middle; it's from a trustworthy source, but it has it's biases.

Secondly, he chose to ignore all other indexes.  Even the Halifax, which is produced by similar means, and therefore would corroborate the Nationwide data if a genuine trend was happening (but it isn't, so it hasn't).  (Although the next set of Halifax data is due out next week, I think, so that might corroborate Nationwide; but that's in the future.)

Thirdly, he chose an arbitary three month window.  Why three months?  Presumably because monthly numbers are useless, they're mostly noise (always have been, always will be), and annual numbers are still (strongly) negative.  The fact that the three months in question (March, April, May) are traditionally the strongest house buying months didn't hurt either.  The two end-points were also untypical: the month before the start of the three was an untypical drop, and the final was an untypical rise, this makes any direct comparison look greater it really was; the Nationwide's own "3 month on 3 month" statistic is saying -0.5%, by comparison.

Just swap Nationwide with Land Registry and see what difference it makes.

The Nationwide numbers are also seasonally adjusted, which should theoretically make them more accurate; but the Telegraph article quotes non-seasonaly adjusted numbers.  The same statistic, using seasonally adjusted figures gives a three month rise of 1.9%, which could be explained just by the line noise of the boundary cases.

What we're seeing is what we always see at this stage in the cycle, a slow down in the rate of falls.  In order for this to happen there does have to be a rise in "activity", but "activity" is a proxy to the rate of change, not the change itself.

There is also an element of the long predicted "dead cat bounce" at work too, by the looks of it, but it's not bouncing very high.

The final note on stability: the numbers of people remortgaging is quite strongly down, so not very many are taking up the offers of low fixed rates.  Hmm...