the economy - update

[jaywalker]

"Historical data is always imperfect, but it is the only measure without bias that we have to measure risk and make hypotheses about the future".

ok, clearly at one level this is very sensible. Yet I am troubled in a number of ways.

Historical data is always-already selected: the principles of selection bias the notion of 'history' they then present. I think this is orthodoxy in the academy (at which point I get ridiculed, but there it is). You could have heard your words out of the mouth of Barber in 1971 - he had not read (or at least understood) Friedman's critique of the Phillips Curve.

My guess is that measuring risk is really only a projection of the status quo. If we have calm-ish weather and know the time of year then we can posit a distribution. But for the interesting risks we can't. There is, simply, no underlying distribution to posit. This is because complexity operates BOTH in the unknown of the initial conditions AND because the non-linear feedback loops characteristic of the mythologised 'the economy' (and I do mean that this is a myth) themselves are both poorly modelled AND CHANGE, at times chaotically. We have seen such phase shifts this year. The modelling effort really only homes in on precision when times are not a changing.

[jaywalker]

came across this tonight - very entertaining.

http://www.bbc.co.uk/news/business-38440946

[Lordship 516]

jaywalker Wrote:

-------------------------------------------------------

> "Historical data is always imperfect, but it is

> the only measure without bias that we have to

> measure risk and make hypotheses about the

> future".

>

> ok, clearly at one level this is very sensible.

> Yet I am troubled in a number of ways.

>

> Historical data is always-already selected: the

> principles of selection bias the notion of

> 'history' they then present. I think this is

> orthodoxy in the academy (at which point I get

> ridiculed, but there it is). You could have heard

> your words out of the mouth of Barber in 1971 - he

> had not read (or at least understood) Friedman's

> critique of the Phillips Curve.

Since 2011, both unemployment and inflation have fallen, with little sign of the inverse relationship associated with the Phillips Curve. The Philips Curve's usefulness has receded already as it became altogether incapable of producing the pattern of inflation and unemployment we see during the first five years of the recovery from the recent recession. What we see during the past few years is persistently high unemployment combined with low and stable inflation expectations and a falling observed rate of inflation. The Philip's Curve has to be put back in the tool box for now and not used until there is a significant change in circumstances.

>

> My guess is that measuring risk is really only a

> projection of the status quo. If we have calm-ish

> weather and know the time of year then we can

> posit a distribution. But for the interesting

> risks we can't. There is, simply, no underlying

> distribution to posit. This is because complexity

> operates BOTH in the unknown of the initial

> conditions AND because the non-linear feedback

> loops characteristic of the mythologised 'the

> economy' (and I do mean that this is a myth)

> themselves are both poorly modelled AND CHANGE, at

> times chaotically. We have seen such phase shifts

> this year. The modelling effort really only homes

> in on precision when times are not a changing.

You misunderstand the application of modelling uncertainties, particularly 'known' uncertainties. These can be given discrete interpolated values using custom techniques. By running these models dynamically over time we can get a pattern of outcomes that can help us further narrow their results. I mainly use these for micro scenarios - what if, what not to do, when to do, when to stop, when to go, how much, when to hedge, what to hedge - all for particular narrow scenarios and we can have quite a success rate using these techniques. I'm sure the BoE have far more sophisticated ABM/MAS models doing the same every day.

[jaywalker]

HiLordship, I think you misread my post:

1.

jaywalker Wrote:

-------------------------------------------------------

> "Historical data is always imperfect, but it is

> the only measure without bias that we have to

> measure risk and make hypotheses about the

> future".

>

> ok, clearly at one level this is very sensible.

> Yet I am troubled in a number of ways.

>

> Historical data is always-already selected: the

> principles of selection bias the notion of

> 'history' they then present. I think this is

> orthodoxy in the academy (at which point I get

> ridiculed, but there it is). You could have heard

> your words out of the mouth of Barber in 1971 - he

> had not read (or at least understood) Friedman's

> critique of the Phillips Curve.

Since 2011, both unemployment and inflation have fallen, with little sign of the inverse relationship associated with the Phillips Curve. The Philips Curve's usefulness has receded already as it became altogether incapable of producing the pattern of inflation and unemployment we see during the first five years of the recovery from the recent recession. What we see during the past few years is persistently high unemployment combined with low and stable inflation expectations and a falling observed rate of inflation. The Philip's Curve has to be put back in the tool box for now and not used until there is a significant change in circumstances.

But that is what I said? The Phillips curve was destroyed by Friedman (1967 or 1968 I don't have it to hand) well before its empirical collapse.

2

> My guess is that measuring risk is really only a

> projection of the status quo. If we have calm-ish

> weather and know the time of year then we can

> posit a distribution. But for the interesting

> risks we can't. There is, simply, no underlying

> distribution to posit. This is because complexity

> operates BOTH in the unknown of the initial

> conditions AND because the non-linear feedback

> loops characteristic of the mythologised 'the

> economy' (and I do mean that this is a myth)

> themselves are both poorly modelled AND CHANGE, at

> times chaotically. We have seen such phase shifts

> this year. The modelling effort really only homes

> in on precision when times are not a changing.

You misunderstand the application of modelling uncertainties, particularly 'known' uncertainties. These can be given discrete interpolated values using custom techniques. By running these models dynamically over time we can get a pattern of outcomes that can help us further narrow their results. I mainly use these for micro scenarios - what if, what not to do, when to do, when to stop, when to go, how much, when to hedge, what to hedge - all for particular narrow scenarios and we can have quite a success rate using these techniques. I'm sure the BoE have far more sophisticated ABM/MAS models doing the same every day.

Of course. But I said that the interesting uncertainties are the unknown ones. The problem is (shifting into my own academic specialism) that humans define their situations contingently (Kitsuse). These are in non-linear feedback relation to the definitions of others. No amount of modelling of 'known' uncertainties can cope with this (for example, the failure of polling organisations in the general election, brexit and the USA election).

[JohnL]

jaywalker Wrote:

-------------------------------------------------------

> Of course. But I said that the interesting

> uncertainties are the unknown ones. The problem is

> (shifting into my own academic specialism) that

> humans define their situations contingently

> (Kitsuse). These are in non-linear feedback

> relation to the definitions of others. No amount

> of modelling of 'known' uncertainties can cope

> with this (for example, the failure of polling

> organisations in the general election, brexit and

> the USA election).

The moment everyone gives up and assumes they're

going to be relegated far to early and you need to

change the manager.

[Lordship 516]

Philip's Curve

Many criticize the Phillips curve because they believe it implies that growth causes inflation and repudiates the theory that excess growth of money is inflation?s true cause. Not so. One can believe in the Phillips curve and still understand that increased growth will reduce inflation. The misplaced criticism of the Phillips curve is ironic since Milton Friedman [one of the co-inventors - with Phelps - of its expectations-augmented version] was also the foremost defender of the view that ?inflation is always, and everywhere, a monetary phenomenon.?

I have a preference for the Lucas critique & the new classical school of economics. One important application of the critique is its implication that the historical negative correlation between inflation and unemployment, known as the Phillips curve, could break down if the monetary authorities attempted to exploit it [which has happened].

The Lucas islands model is an economic model of the link between money supply and price and output changes in a simplified economy using rational expectations. It delivered a new classical explanation of the Phillips curve relationship between unemployment and inflation.

The Phillips curve was lauded in the 1960s as providing an account of the inflation process hitherto missing from the conventional macroeconomic model. After 40 years or so the Phillips curve, as transformed by the natural-rate hypothesis into its expectations-augmented version [short-run Phillips curve], remains the key to relating unemployment (of capital as well as labor) to inflation in mainstream macroeconomic analysis. It is still a relevant touchstone - but just now it has to be accepted that conditions are volatile & the criteria for equilibrium are somewhat chaotic to regard the Philip's Curve as currently useful - but it will return to relevance, probably in a further development of the Lucas model after Kydland & Prescott.

[jaywalker]

Gosh, this is taking me back (over 35 years to when I decided to give up economics so forgive me if I get the details wrong :-) ) - we were talking at cross purposes because I read the Friedman critique as a refutation of the Phillips Curve (although that still haunts the political imagination as in the current idea that a return to fiscal expansion in the USA will cause real effects even though capital is highly mobile and exchange rates are floating).

Moving on, in terms of expectations, you have got to be joking :-). Friedman's model tacitly incorporates rational expectations as a limit (the vertical curve once you get positive feedback between expectations and policy) but of course relied on adaptive expectations in the mechanism of the model itself. This seems to me to be very sensible.

The Lucas (or as I remember it, also Barro) approach is also very interesting - and I get the as it were bottom up and distributed 'island' idea you said you were working on with more recent AI components (one that could be exploited in many other arenas of social research).

However, sticking with more 'macro' theory: the experience of the 1980s was that RELATIVELY sticky prices in a floating exchange rate world with perfect capital mobility INCLUDING an assumption of rational expectations undoes Monetarism (Dornbusch). We saw that writ large with Thatcher. Since then the orthodoxy (not politically very palatable) has been that only tiny interest rate movements are safe (and fiscal policy is impotent).

The great intellectual merit of the (Muth)-rational expectations assumption was that it sutured the model to include that model (where correct and taken to be correct) in the modelling of the actions of agents in the model. This must be right (although, again, this is not instant and might be quite poorly adapted to). But if the model is not correct it might still be taken to be correct. Also, correct models (from, say, the point of view of logic or rational interest) cannot be correct if they are not taken to be correct - in a strong sense rational expectations then fails because the social is bootstrapped by definitional processes that are contingent on all sorts of things, rarely calculated or rational ones, and these cannot be anticipated. Economics is therefore a doomed enterprise.

[Lordship 516]

jaywalker Wrote:

-------------------------------------------------------

> Economics is therefore a doomed enterprise.

Economics versus the economy - these are not always compatible but we must keep doing our best in an ever changing & diversified world where the main players are at odds with each other. We can only find a wormhole through the whole chaotic morass of political & commercial shenanigans. A plan is better than no plan - laissez-faire is not a real option - Smith, Keynes & Hayek all acknowledged that.

[????]

Fuck me guys - it is self-idnulegeunt pseuds corner on here with you two. You really are confusing speaking an over complicated flowery academic language with knowledge and insight. Ironically there's a whole sociology on coded language (read it) .....a lot of hot air. You are both not suggesting much confidence in your ideas/hypothesis if you can't or are not willing to express them in a more accessible language - it's translatable but tedious as is. Have you no confidence in the merits of your arguments to express them clearly rather than in academic bullshit?

[Seabag]

It's all got a bit boring up there. I'm sure there's merit to all the wordiness, only it's near impossible not to glaze over whilst reading. And I like business/models/analysis very much.

But this Independent article is manageable

This is the year that Britain will allow working families to slide into poverty - The Independent

https://apple.news/AAWMtDbGdRTiNrMv83GADOw

[Lordship 516]

???? Wrote:

-------------------------------------------------------

> @#$%& me guys - it is self-idnulegeunt pseuds

> corner on here with you two. You really are

> confusing speaking an over complicated flowery

> academic language with knowledge and insight.

> Ironically there's a whole sociology on coded

> language (read it) .....a lot of hot air. You are

> both not suggesting much confidence in your

> ideas/hypothesis if you can't or are not willing

> to express them in a more accessible language -

> it's translatable but tedious as is. Have you no

> confidence in the merits of your arguments to

> express them clearly rather than in academic

> bullshit?

Sorry about your bother. I avoided using more technical language & simplified my contribution as much as possible. There are some issues that require a vocabulary - easy enough to look up the meanings. Not all academic/theoretical expression is bullshit except to those who want to distill everything into meaningless soundbites. Just because you cannot understand something doesn't make it bullshit !

A lot of what I read here I find boring also but I just move on. One person's meat is another person's poison. Tolerance is the name of the game !

[JohnL]

All I remember with Thatcher was every time a money supply figure failed

She invented another - quite brilliant

M1, M2, M3 ...

[JohnL]

jaywalker Wrote:

-------------------------------------------------------

>

> Fat tails: # it might be that the Supreme Court

> just allows brexit on May's timetable. But there

> is also a possibility that they will - in

> practical terms - block it entirely. # Russia. #

> French elections (these still look deeply

> troubling) # pensions crisis (really this is a

> crisis!) # stagflation. And of course unknown

> unknowns.

Rumour is she's after the Supreme Court now

Human rights- transfer to Supreme Court as now then

make sure she can appoint members of the Supreme Court.

Why should she even need law experts there ;)

[malumbu]

Gosh. I've been away for a few days and missed lots of excitement.

Whilst I don't always understand the discussion I do feel a little more enlightened.

My economics is pretty minimal. I had unfettered free markets, freedom of movement and global trade beaten into me a few years ago when working sort of in this area. Then in 2008 I found that this was all wrong and we needed some controls and interventions on financial markets.

I then looked into the case study of Harley Davison who should have been allowed to go the wall. But was more successful in being propped up by the US government than our shoddy motor industry (Harleys at the time were unwanted apart from the die hards and a generation behind the Japs). So we want to prop up uneconomic sectors again?

And then I see the masses want improvements in quality of life, the latest X or Y, with lifestyle slickly sold to them, but there aren't enough people around to pick the fruit, make the coffee or clean the toilets. And whilst my modest savings stagnate others can get a loan with a guarantor at a rate just under 50% APR.

Am I now oversimplifying things?

[jaywalker]

Seabag Wrote:

-------------------------------------------------------

> It's all got a bit boring up there. I'm sure

> there's merit to all the wordiness, only it's near

> impossible not to glaze over whilst reading. And I

> like business/models/analysis very much.

>

> But this Independent article is manageable

>

> This is the year that Britain will allow working

> families to slide into poverty - The Independent

> https://apple.news/AAWMtDbGdRTiNrMv83GADOw

There are probably only two possibilities then.

One gets this all the time. You expect ordinary language to be able to cope with any issue. But then you go to the doctor to understand your complaint and look at the file and it is written in gibberish (to you).

Opposed to this, priests (in secular form) constantly up the ante on esoteric language to further their prestige (bien sur).

So it is undecidable, isn't it? Unless you form a view about the specialism.

When I wrote that "models (from, say, the point of view of logic or rational interest) cannot be correct if they are not taken to be correct - in a strong sense rational expectations then fails because the social is bootstrapped by definitional processes that are contingent on all sorts of things, rarely calculated or rational ones, and these cannot be anticipated" does that sound like theology or reason?

[jaywalker]

JohnL Wrote:

-------------------------------------------------------

> Rumour is she's after the Supreme Court now

>

> Human rights- transfer to Supreme Court as now

> then

> make sure she can appoint members of the Supreme

> Court.

> Why should she even need law experts there ;)

did you see the brilliant Short Cuts section in the London Review of Books this week? Andrew O'Hagan on the absolute poverty of the government case in the Supreme Court: and its sinister claims.

[jaywalker]

???? Wrote:

-------------------------------------------------------

> @#$%& me guys - it is self-idnulegeunt pseuds

> corner on here with you two. You really are

> confusing speaking an over complicated flowery

> academic language with knowledge and insight.

> Ironically there's a whole sociology on coded

> language (read it) .....a lot of hot air. You are

> both not suggesting much confidence in your

> ideas/hypothesis if you can't or are not willing

> to express them in a more accessible language -

> it's translatable but tedious as is. Have you no

> confidence in the merits of your arguments to

> express them clearly rather than in academic

> bullshit?

ok, so you want me to read and respond to something - let me know what it is. Otherwise, you are authorising your views in a way that cannot be gainsaid.

[red devil]

[????]

So. This morning.

Service Sector PMI smashes expectations and out performs all 16 economists polled by Reuters (any of our resident economist on that per chance?) - this joins Manufacturing & Construction Indexes published earlier this week. Guess what? Above forecasts.

FTSE at highest levels ever this week.

"Ah but inflation" i hear the sages call - currently 1.2% compared to Eurozone 1.1% - the reality is that rather than 'brexit' as the pessimists keep banging on about it's MAINLY the appreciation in the $ as a result of interest rates rises teir that has pushed the ? lower. The ? is on a parity of where it was vs the Euro in 2013.

Even the Guardian published this morning an article on why has the economy performed above post-referendum (note: not Brexit) expectations - hint because most economist were (and are) subject to Cognitive Bias on this...

[JohnL]

???? Wrote:

-------------------------------------------------------

> So. This morning.

>

> Service Sector PMI smashes expectations and out

> performs all 16 economists polled by Reuters (any

> of our resident economist on that per chance?) -

> this joins Manufacturing & Construction Indexes

> published earlier this week. Guess what? Above

> forecasts.

>

> FTSE at highest levels ever this week.

>

> "Ah but inflation" i hear the sages call -

> currently 1.2% compared to Eurozone 1.1% - the

> reality is that rather than 'brexit' as the

> pessimists keep banging on about it's MAINLY the

> appreciation in the $ as a result of interest

> rates rises teir that has pushed the ? lower. The

> ? is on a parity of where it was vs the Euro in

> 2013.

>

> Even the Guardian published this morning an

> article on why has the economy performed above

> post-referendum (note: not Brexit) expectations -

> hint because most economist were (and are) subject

> to Cognitive Bias on this...

Not an economist but things often accelerate most

before a crash.

[????]

That may be true John - but it has bugger all to do with any predictions by economist about the post-referendum economy! None said "There will be an economic boom followed by a crash". Personally I think their useless predictions have potentially fired an unsustainable boom (see increase in lending and falling Savings Ratio) as Carney cut interest rates and did more QE post referendum as things were going to fall of a cliff.....apparently. Not covered themselves in glory the dismal scientists.

[Earl Aelfheah]

This sums it up pretty well: https://www.ft.com/content/15837254-d272-11e6-b06b-680c49b4b4c0

[Seabag]

rahrahrah Wrote:

-------------------------------------------------------

> This sums it up pretty well:

> https://www.ft.com/content/15837254-d272-11e6-b06b

> -680c49b4b4c0

I'm guessing you're a subscribed reader rah

Many are not, so the link is pretty useless

[Earl Aelfheah]

To summarise (apologies, forgot about the paywall): Economists got it wrong about the immediate impact of a vote to leave the EU. The error stemmed from a failure to predict the strength of household consumption following the vote. The assumptions made about household spending were reasonable, but never the less mistaken.

Rather than rising, household savings fell throughout 2016. The savings ratio dropped to an extremely low level in the third quarter, as consumers went on a borrowing and spending binge not seen since before the financial crisis. The reasons why this might be should become clearer later in the year.

It's worth remembering that any forecast, necessarily involves a degree of guesswork. You have to make assumptions when you're predicting what might happen in the future and sometimes you'll get it wrong.

None of this is a reason to think we should stop taking experts seriously, or that an expert forecast has no more to recommend it than Dave the cabby's.

[red devil]

Didn't the economists when making their post-referendum predictions base them on A50 being triggered straight away, as that was what Cameron said he would do if Leave won. I think all predictions whether gloomy or rosy are on hold until A50 is triggered...won't be long now!