Mpc Future Estimates Of Cpi

[Lepista]

OK, I've asked a few times for this information, but nobody has risen to the challenge.

I was bored tonight, so decided to have a play.

Each quarter, the MPC publishes an inflation report. In this, they produce a chart showing their future predictions for the CPI rate.

Bank of England Inflation Report, February 2010

Strictly, it is not a single estimate, but rather a spread of probability. The upper and lower bound limits of the graph are actually the bound of which 90% of their estimatees fall. i.e. there is a 5% chance that the actual CPI rate will be above the highest number, and a 5% chance that it will be below it (according to the MPC).

Each graph on it's own is quite interesting, however looking back through the series, and you can then compare how well the MPC have predicted the actual CPI rate. For ease, I have split their estimates down into specific future looking timescales, as follows:

6 month lookahead.pdf

The graph shows what the MPC were expecting the CPI to be in six months time, compared to what actually happened.

Actually, Im' not sure I've got this quite right - it looks far too spot on for it to be real. Might have something to do with when the MPC issue the report?

There is also a slight booboo on the charts that is excel's fault - the year label is shown for some reason at the end of the year, rather than at the beginnning.

Anyway, things get a little more interesting when we start to look at the longer range predictions. Here's the 1 year lookahead graph:

1 year lookahead.pdf

pre-2008, the actual CPI is quite close to what was estimated one year previously. post-2008, the actual CPI is much closer to the upper limit of the MPC estimates.

How about them looking 1.5 years into the future? Lets see:

1.5 year lookahead.pdf

Currently, we are bouncing along at the ceiling of their "worst case" upper limit. Oops. Also note that the "best estimate" is hovering around the 1 - 2 % mark. Shame it never avtually reached there, eh Merv?

The 2,3 and 4 year lookaheads are as follows:

2 year lookahead.pdf

3 year lookahead.pdf

4 year lookahead.pdf

Again, for all of these, we are consistently at the top end of their estimate.

My conclusion? That the MPC are actually doing a pretty good job of estimating the future CPI YOY percentage. As long as you look at the 2years+, UPPER estimate.

6 month lookahead.pdf

1 year lookahead.pdf

1.5 year lookahead.pdf

2 year lookahead.pdf

3 year lookahead.pdf

4 year lookahead.pdf

[Beggar Thy Children]

Producing the inflation report is easy.

At 2 years away, put a data point at 2%.

Then draw a squiggly line back to the current point.

[Bloo Loo]

Producing the inflation report is easy.

At 2 years away, put a data point at 2%.

Then draw a squiggly line back to the current point.

then you create supporting guesses to prove the guess.

Where is 150 dollar oil....

[Lepista]

Producing the inflation report is easy.

At 2 years away, put a data point at 2%.

Then draw a squiggly line back to the current point.

The graphs say otherwise.

It goes like this:

- carry out some detailed analysis of the long term inflation issues

- put the report in front of Merv

- Merv has a higher authority forcing him to keep inflation high.

- Merv alters the output so that the "best guess" line is actually now the upper estimate line (this way, they can always say that the actual inflation was within their confidence band)

- For the best estimate, make the 2+ year estimate at 2%, to show that you expect your actions to bring inflation back in line

- for the worst case estimate, put a scary deflation threat in there, in case anyone questions why inflation is so high. it's because the risk of deflation is very real, and we can't go there. silly.

[martingale]

Since you are working hard how about comparing their estimations with a naive filter which attempts to estimate the derivative of inflation and then projects this estimate forward n steps.

http://en.wikipedia.org/wiki/Alpha_beta_filter

If the Alpha-beta filter and BOE estimates look similar then there is a good reason for this!

It is because they use a similar technique (kalman filter this time) but then weigh it with a long term bull***t macroeconomic model which by some miracle always seems to return inflation to target over the long term.

http://www.bankofengland.co.uk/publications/workingpapers/wp179.pdf

[billybong]

The 2 year projection chart (which is the time zone they have claimed over and over again to be really targeting - their often stated 2% in 2 years policy) is interesting as the actual CPI broke the 3% extreme upper limit line a while back and then recently rebounded to it and is now gaining momentum upwards and away from it.

If the Mervyn King and the Manana Policy Committee (and the government) don't act very very quickly it wouldn't be a total surprise to see CPI soon through 5% and then quite shortly near 10% and up up and away.

Edited January 18, 2011 by billybong

[Little Professor]

Why make these pdf files rather than plain images?

[Lepista]

Why make these pdf files rather than plain images?

How do you do that?

[tinker]

A bit off track, but I thought it was funny.

NASA do charts predicting 'Sun Spot' activity... they are just like the BoE and their inflation/growth fan charts.

Sun Spot Chart animation (Too big to post here.)

BTW current predictions are at a level comparable with The Maunder Minimum of 1675-1715.

"It’s tough to make predictions, especially about the future."

Philosopher Y. Berra

Edited January 19, 2011 by tinker

[Lepista]

Here's a couple more.

This is the entire prediction for each quarter. If the MPC were being consistent, then the future prediction value would track through as we get closer to the time. Lets see, shall we?

Best Estimate Lookahead.pdf

Lower Limit Lookahead.pdf

Upper Limit Lookahead.pdf

Well, their best estimate is fairly consistent, in that they are predicting the CPI generally hoverring around the 2% level. it's also consistently way too low.

Lets not even talk about the lower bound.

The upper bound is a pretty good fit to the actual CPI, and there is a clear trend showing expecting the CPI to relatively slowly creep upwards.

Taking the upper bound graph as the best predictor of the future CPI, this is showing that CPI will continue to rise in the future, getting towards 4.5% fairly soon, and 4.8%+ longer term.

I reallly do think that the upper bound limit is what the MPC is actually expecting to happen.

Best Estimate Lookahead.pdf

Lower Limit Lookahead.pdf

Upper Limit Lookahead.pdf

[Bloo Loo]

The graphs say otherwise.

It goes like this:

- carry out some detailed analysis of the long term inflation issues

- put the report in front of Merv

- Merv has a higher authority forcing him to keep inflation high.

- Merv alters the output so that the "best guess" line is actually now the upper estimate line (this way, they can always say that the actual inflation was within their confidence band)

- For the best estimate, make the 2+ year estimate at 2%, to show that you expect your actions to bring inflation back in line

- for the worst case estimate, put a scary deflation threat in there, in case anyone questions why inflation is so high. it's because the risk of deflation is very real, and we can't go there. silly.

I do beleive this is not far from the truth of it.

I mean, if you beleive your policy is correct, then your forecast MUST show the correct result...otherwise, you have some explaining to do.

trouble is, they are NEVER brought publicly to account....and when was the last major MPC reshuffle...they have got it very very wrong in the last 5 years.

But then again, they are not there for anything other than to be a lifeboat for insolvent bankers. Lets not forget the whole reason for the Bank itself. Its not just to have a quality building in the most expensive real estate area in the known Universe.

[Guest_FaFa!_*]

How do you do that?

Tinypic, imageshack are your friends. Whack them up there, then use the insert image button and link to them from your post.

[babesagainstmachines]

The Bank of Emgland's policy on controlling inflation is easily summed up :

Oooh, we didn't expect that! Nevermind, it'll be alright in the long run.

[Lepista]

The Bank of Emgland's policy on controlling inflation is easily summed up :

Oooh, we didn't expect that! Nevermind, it'll be alright in the long run.

ACtually, they are expecting it - this shows they are very good at predicting the future CPI rate. At least, that is if you look at the upper limit line that they predict...

[Sledgehead]

The graphs say otherwise.

It goes like this:

- carry out some detailed analysis of the long term inflation issues

- put the report in front of Merv

- Merv has a higher authority forcing him to keep inflation high.

- Merv alters the output so that the "best guess" line is actually now the upper estimate line (this way, they can always say that the actual inflation was within their confidence band)

- For the best estimate, make the 2+ year estimate at 2%, to show that you expect your actions to bring inflation back in line

- for the worst case estimate, put a scary deflation threat in there, in case anyone questions why inflation is so high. it's because the risk of deflation is very real, and we can't go there. silly.

That's interesting.

Edited February 16, 2011 by Sledgehead

[yellerkat]

FT Alphaville with charts back to Feb 2009 - note the sudden change in the Y-axis this month.

LINK.

[Sour Mash]

ACtually, they are expecting it - this shows they are very good at predicting the future CPI rate. At least, that is if you look at the upper limit line that they predict...

If you make a huge spread on a fanchart then of course you stand a pretty good chance of the actual figure lying somewhere within that spread. That's not 'predicting' anything, it's covering all your bases.

[Lepista]

If you make a huge spread on a fanchart then of course you stand a pretty good chance of the actual figure lying somewhere within that spread. That's not 'predicting' anything, it's covering all your bases.

That's not what I was saying though - My thesis is that the upper limit IS their prediction, that they have consistently achieved.

The fan chart is just the cover for allowing them to say that they were trgetting the 2% value, whilst staying within the bounds of the spread.

[Sledgehead]

That's not what I was saying though - My thesis is that the upper limit IS their prediction, that they have consistently achieved.

The fan chart is just the cover for allowing them to say that they were trgetting the 2% value, whilst staying within the bounds of the spread.

So Merv meets Jezzer Clarkson at a city bash and asks why he's never been asked to take a spin around the Top Gear track. "Is it cos you're worried that my record on inflation targetting suggest I'll be over-steering and under-steering all the way 'round?"

"That," says Clarrkson, "and the fact that by the time you're on the show they'll be no such thing as a 'reasonably priced car'."

[ccc]

That's not what I was saying though - My thesis is that the upper limit IS their prediction, that they have consistently achieved.

The fan chart is just the cover for allowing them to say that they were trgetting the 2% value, whilst staying within the bounds of the spread.

Do they just do the opposite with their GDP predictions ?

Tonight's homework for you !!

[billybong]

FT Alphaville with charts back to Feb 2009 - note the sudden change in the Y-axis this month.

LINK.

From the comments in that link

He also asked the journalists what they would have done differently in terms of policy, but noone seemed willing to say that they should never have accomodated the asset price/house price boom from 2002 onwards.

Or willing to say anything for that matter.

One of the journalists could have politely said something like, "Well after 5 years of mostly failure in meeting the target and now inflation going beyond the pale I would have insisted on trying to meet the written remit from the government and as my only lever is interest rates I would have had to put interest rates up. That would be my job and it is defined in the remit. Of course then the government would be free to increase the inflation target in the remit and then I would be able to keep interest rates low for as long as inflation met the new target - but that would be a government decision."

Of course that's where the "we're all in it together" comes in with BoE/government connivance clearly the method of operation along with having poodle journalists at these meetings.

It's time to jettison the target altogether.

Edited February 17, 2011 by billybong

[billybong]

http://www.marketoracle.co.uk/Article26314.html

The more widely recognised measure of Inflation RPI stood at 5.1% and real inflation at 6.6%, as the official inflation indices have been systematically doctored to under report real inflation by successive governments for several decades resulting in serious and compounding under reporting of the real rate of inflation as experienced by the British population.

The Bank of England MPC members continue with their mantra of temporarily high inflation due to short term factors. One could cut and paste from any inflation statement from MPC members of the past 12 months to hear the same propaganda out of the Bank of England. The question everyone should be asking the BoE is when does temporary high inflation stop being temporary? Originally it was for a couple of months, now it is over a year, will high inflation still be temporary a decade from now? For that is how long I expect the Inflation Mega-trend to run.

The gold fish memory broadcast and mainstream media fed by ivory tower academic economists continues to tow the line of temporarily high inflation by focusing on core inflation that excludes, food and energy costs because off course everyone in the UK has stopped feeding or heating themselves. Despite that fact that food and energy are far more relevant to the British population than for instance the price of a 50 inch Plasma Screen.

....

Bank of England Targeting Nominal GDP Not Inflation.

My ongoing UK Inflation analysis concluded in the fact that the Bank of England has clearly not been targeting 2% inflation for some time but more likely been targeting a sustained trend to above 2% GDP, which given the Q4 2010 dip ensures that the Bank of England's MPC will continue to sleep walk the country into a wage price spiral.

.....

The fuel duty hike due in April 2011 of 5p per litre ensures that Britain will experience another so called 'temporary' inflation shock as the fools at the Bank of England looks set to blindly continue to brush aside every single inflationary factor that crops up each month as just being temporary whilst the pressure builds to beyond boiling point on the wage price spiral.

Edited February 17, 2011 by billybong

[Sour Mash]

It's notable that on the singular occasion when it looked like inflation might actually undershoot the target significantly, they were extremely quick to move with massive base rate cuts.