Can we take the Nationwide seriously?

julieq

Reports of the death of the BTL business were exaggerated

https://forums.moneysavingexpert.com/discussion/comment/16052273#Comment_16052273

geneer

Sibley wrote: »

How would you feel if you found out Geneer had a low paid job, lived in a bedsit and had about 30 quid a week disposable income?
Would any of you bother replying to his rubbish?

I see Simpley is continuing his campaign to elevant me to his "arch nemesis". Which part of "shrugs" does he not get.


Though thats nothing compared to JulieQs amazing stalkerish behaviour in the last half dozen or so posts. !!!!!!!:rotfl:

geneer

julieq wrote: »

Macaque, your desperation is touching, but it remains the case that the NW figures correlate very closely with the LR. Where they vary they've been much lower in the recent past.

Really it's pathetic to argue that this is misreporting of data to talk the market up. There's no reason for them to do that - it's exactly the wrong thing for a lender to do anyway because they'd be overvaluing their collateral and INCREASING their risk by lending more against false valuations.

Here we have data which basically confirms the LR numbers within a small error. You don't like it because it shows a YoY rise through random noise, but that doesn't mean it's invented. In 12 months the situation could well be the other way round.

And you're so dreadfully parochial anyway. Don't you think that the same techniques could have been tried in the US? Or in Eire? The difference is that in those countries there isn't a supply/demand imbalance to shore up the market.

Why would *I* blush, anyway? I'm not the president of the 70% club, and I haven't been wrong time after time with increasingly desperate predictions of falls, recalibration of the definition of a crash, and so on. If anyone should be glowing like a beetroot, it's you.

Lordy. You have to laugh.
NW is as reliable as Land Reg, if you accept the fact that the NW figures (being based on a fraction of 10% of the market) are subject to both huge volatility and a skew towards london. :rotfl:

Don't give up the day job.

julieq

But Geneer, they're completely correlated. On graph after graph they match the land registry to within a small margin.

I have no idea what you're trying to prove here, but you obviously don't have a clue about interpreting data. I mean what really gets me is that you even posted an article which explained you need to combine all the indices to get a view and you're still hanging on to the LR being the only truth on the basis that on a two sample point straight line it happens to show down rather than up!

Completely bonkers.

robmatic

julieq wrote: »

But Geneer, they're completely correlated. On graph after graph they match the land registry to within a small margin.

I have no idea what you're trying to prove here, but you obviously don't have a clue about interpreting data. I mean what really gets me is that you even posted an article which explained you need to combine all the indices to get a view and you're still hanging on to the LR being the only truth on the basis that on a two sample point straight line it happens to show down rather than up!

Completely bonkers.

You're probably going to have to explain the concept of correlation. Geneer didn't do stats at school.

geneer

julieq wrote: »

But Geneer, they're completely correlated. On graph after graph they match the land registry to within a small margin.

I have no idea what you're trying to prove here, but you obviously don't have a clue about interpreting data. I mean what really gets me is that you even posted an article which explained you need to combine all the indices to get a view and you're still hanging on to the LR being the only truth on the basis that on a two sample point straight line it happens to show down rather than up!

Completely bonkers.

:rotfl:+1.6% or -3.2%. A tiny margin.

julieq

One last time, then someone else can have a go.

The data series are correlated closely. You can see that by looking at them. In absolute terms they match within a small margin of error.

If you take the difference between two points then you amplify the error because you're including random variations over two points rather than one. Even within a closely correlated series you can get quite wide differences in the gradient of a line between any two arbitrary points, and that's what you're seeing here. We are not correlating the YoY or MoM numbers, we are correlating the base data and seeing them essentially as the same. In effect the NW numbers confirm the LR numbers given the lag in LR, and the picture that emerges is stagnation with slight oscillation.

Imagine you have two straight lines X1 = 15 and X2 = 15. They correlate exactly and are the same for all Y values. Now imagine you can have a + or - uncertainty of 1 when you measure at different Y values. At Y1 you measure X1 with a -1 offset, at Y2 you measure X1 with a +1 offset, then at both Y1 and Y2 you measure X2 at +1. You have a difference between your X1 points of 2, and a difference between your X2 points of 0, but the data set is still correlated.

So you have two correlated data sets - reflecting the same value exactly of 15 - with the same error probability, but showing a wide variance in the difference between two arbitrary points.

That's exactly what's happening when you're comparing the NW YoY with the LR YoY. They do have different volatility - LR is more heavily damped (slower reacting) and NW less so, but they represent to a close degree the same underlying data. You can't reject one over the other just because it seems to show a direction you don't like, to get the whole picture you look at as much data as you can.

Which is what your article explains. That's the funniest thing about your masterclass in ignorance.

wotsthat

geneer wrote: »

:rotfl:+1.6% or -3.2%. A tiny margin.

No he's still not got it.

John_Pierpoint

geneer wrote: »

Lordy. You have to laugh.
NW is as reliable as Land Reg, if you accept the fact that the NW figures (being based on a fraction of 10% of the market) are subject to both huge volatility and a skew towards London. :rotfl:

Don't give up the day job.

The Land Registry is far from 100% accurate.
There is a built in incentive to "bend" the prices, it is called Stamp Duty Land Tax (In France it is called "The attorney's bladder" - at the appropriate moment the attorney gets a call of nature and the buyer and seller are left alone in the room to sign the paperwork..........}

I know for a fact that there is a house near me, where the sale price has been copied with two figures transcribed, making the best part of a £60K difference in the price.