Reply to: 1 in 7 chance of house price falls by 2020

[BlasterBates]

Yes, and they can't predict sh1t - this is not because they use wrong numbers however, but because the market is not driven by random stuff, it is driven by expectations that follow cycles - right now the market is on the up, so they all blow the trump implying nothing can go wrong, so quickly they forget what happened just 5 years ago: the awakening will be rude.

Exactly....

...and P/E rations are lower, much lower than they were 5 years ago.

5 years ago the market was at the same level as it was now, but that was 5 years ago, all the same companies make more money thatn they did then so they are worth more.

[AtW]

as far as I know all the Phd rocket scientists in the finance industry are using psuedo-random numbers.

Yes, and they can't predict sh1t - this is not because they use wrong numbers however, but because the market is not driven by random stuff, it is driven by expectations that follow cycles - right now the market is on the up, so they all blow the trump implying nothing can go wrong, so quickly they forget what happened just 5 years ago: the awakening will be rude.

[IR35 Avoider]

This improves uniqueness of numbers

The purpose of adding the numbers together is to get numbers for a Normal distribution from a generator for the Uniform distribution, as a consequence of the Central Limit Theorem.

[IR35 Avoider]

With pleasure: pseudo-random generators actually just offer series of numbers following certain formulae which may appear random to an untrained sight. However these numbers are not random, and they also loop in cycles - this undermines whole premise of "Monte Karlo" method because it relies on randomness, not some pre-determined sequence which is not random.

A "psuedo-random" sequence must obviously have sufficient resemblance to the real think to deserve the name. There's nothing in what you've said that explains why you differ with all the rocket scientists in thinking the resemblance is insufficient.

To know exactly why you need to talk to a PhD in numeric analysis - Threaded.

I don't know how you generate a real random sequence, as far as I know all the Phd rocket scientists in the finance industry are using psuedo-random numbers.

[BlasterBates]

But just think for a moment, if there is a 6% chance of a loss, doesn't that mean the return is garbage over 15 years.

[sasguru]

Atw - Even you must be tired of the bollocks you spout. You are increasingly becoming like Threaded. This is what a lack of female company does to one.

[DimPrawn]

It won't work cos Atw says so and that's good enough for me.

[AtW]

rand()+rand()+rand().. (12 times)

This improves uniqueness of numbers, I actually used this method just few days ago.

However the premise is flawed in the first place - preudo-random number generators are just that - PSEUDO, they generate sequences of numbers that loop (often with very short cycles), and thus flawed.

To know exactly why you need to talk to a PhD in numeric analysis - Threaded.

[IR35 Avoider]

Can you explain the inference, if any, used in reaching your conclusion "hence their conclusions are wrong" fron the given premises?

I also want to know. The lecturer in Quantitive Finance who wrote my textbook recommends rand()+rand()+rand().. (12 times) - 6 as a quick and dirty method for generating normally distributed random numbers in Excel, and I use this for my Monte Carlo simulations. I would like to know why this won't work, so I can tell him to correct the next edition.

[AtW]

Can you explain the inference

With pleasure: pseudo-random generators actually just offer series of numbers following certain formulae which may appear random to an untrained sight. However these numbers are not random, and they also loop in cycles - this undermines whole premise of "Monte Karlo" method because it relies on randomness, not some pre-determined sequence which is not random.

But in any case their research is Bull - markets are driven by expectations and not even Pinocchio's Karlo will change that.

HTH

[DimPrawn]

Looking further afield using the PWC model, we can see house prices drop to zero after the global nuclear war in 2057 that results in the complete destruction of mankind.

HTH

[Mailman]

Isnt that better odds than global warming actually being real?

Mailman

[expat]

Monte Karlo is based on a random number generator. I am certain that PWC losers actually use pseudo-random number generator, hence their conclusions are wrong.

Can you explain the inference, if any, used in reaching your conclusion "hence their conclusions are wrong" fron the given premises?

[Buffoon]

PriceWaterhouseCoopers based its prediction on the "Monte Carlo Simulation" which marries future prediction of house price growth with analysis of how the market has behaved in the past.

Spot the journo with no knowledge of economics, maths, science, ……

[AtW]

Monte Karlo is based on a random number generator. I am certain that PWC losers actually use pseudo-random number generator, hence their conclusions are wrong.