Tag Archives: Economic Forecasting

ECONOMIC CHAOS ?

The piece below is over 2000 words long and I have just completed it for a client .

It is about the random nature of an economic system.

Have you ever wondered why ALL economic predictions are wrong? Have you noticed that in spite of a proven record of error, economists and politicians continue to bang their heads against the forecast-wall and refuse to do anything else but continue to predict outcomes which by now, they must realise will be incorrect?

They certainly use all the latest computer models which have been empirically derived and used for many years.

So, are there any incorrect assumptions about “fundamentals”?

Is the economic process Stochastic (a sequence of random variables)? Or is it Deterministic (when the output of a system is totally dependent on its initial state and  subsequent inputs – and therefore, predictable)?

(Mind you, to add to the confusion, deterministic systems may occasionally produce random  and therefore unpredictable results. )

Is economics a question of Stochasticity v Determinism?

Why do I ask the question? Because there appears to be a total absence the ‘stable equilibrium’ predicted by classical economists.

On the contrary, Market Economics behaves like a collection of dynamically unstable systems. The instability is attributed to external ‘shocks’ rather that any fault in the basic concept. There is what can only be described as ‘non lineality’.

One solution to this ‘non-lineality’ is CHAOS THEORY!

So far, no real evidence has been produced of ‘low – dimensional’ Chaos in economic processes but there are definitely discrepancies between the ‘expected’ according to classic economic models and the ‘observed’. Just look at any economic prediction within your memory. It was probably incorrect.

We still have a ‘mechanistic’ view of the world and economics as a ‘hangover’ from 18th century SCIENCE.

Scientific thinking is very simple: ‘Measure, predict and adjust until you no longer have any more surprises. Then keep measuring to confirm that what you measured in the first place can be replicated’.

Economics was conceived on that same principle . It was established as a ‘science’. That’s where the Determinism crept in.

It was at this time that man first considered the possibility of his own intellect being so unconstrained that he would eventually understand the ‘Universe and everything’ through the medium of scientific reasoning.

This principle was applied to all sorts of activities and thinking – including economics.

The so-called ‘Enlightenment Policy’ would help man in his pursuit of happiness. Especially in the sciences. Science was cool and now in the early 21st Century it is enjoying a bit of a revival.

Of all the subjects on offer, Physics became the admired Paragon for Enlightenment and so it continues.

The way Physics works is simple: Carefully describe an environment and you should be able to predict the outcomes of any experiment conducted within that environment.

Likewise in Economics:  Know the initial environment and you should be able to predict outcomes based on subsequent inputs.

The belief stemming from that philosophy is that EVERYTHING is governed by ‘NATURAL LAWS’ which are a set of ‘cause-effect’ regularities. That means that everything can be predicted.

These same principles have been applied to Economics.

A simple scientific rule is that ‘The state of any system is a consequence of what it was in the preceding moment…..and so on.’

In the beginning, random occurrences had no place in such linear thinking. Everything was governed by Mathematics and Laws.

However, there is one major flaw in the way that we ‘do’ science: That is our ignorance of the CAUSES which generate phenomena and events.

For instance, we know the effects of gravity – which we can measure but we don’t really know the CAUSE.

However, in spite of our ignorance of the exact causes of events added to the imperfection of our analyses, we still cannot have 100% certainty about the vast majority of phenomena.

Economists also appear to have forgotten both the imperfection of analysis and their ignorance or (at best) of the exact CAUSES of events.

What is the solution? What is to be done about our comparative blindness?

Our ‘crutch’ is the science of probability. Chance.

Current economic thinking is a throwback. In economics, the world is still viewed as totally deterministic.

‘STOCHASTIC’ is non-existent – as is uncertainty because uncertainty is treated as ignorance or a failure to understand the deterministic rules of a very complex system.

Yet, with ALL our processing power, no-one has yet been able to establish those rules which should  predict outcomes.

So, as Chaucer wondered in The Nonnes Priest Tale – Travelling from A to B:  Freewill or Predestination?

Looking at the unpredictability of economic outcome, we move from linear to non-linear dynamics, from certainty to probability, from Economic Theory to Chaos Theory.

Theories of economics have been shaped by the assumption of ‘Rational Man’ who behaves in accordance with a known set of rules.

The evolution of economics into a science was ‘booted’ into becoming a science when it was ‘mathematicised’. Formulae arrived and suddenly, it became a bona fide branch of Applied Mathematics.

Many of the original people who translated economics into a mathematical form were physicists, engineers and mathematicians…… and it still shows. At that time, their view of the world was ‘linear’.

Does that work in economics? The short answer is ‘no’. That is why economists are struggling, interpreting and making excuses.

Marshall in his ‘PRICIPLES’ compared the study of economics to the study of tides. The number of variables affecting tides means it is impossible to create a consistent dynamic picture.

Even nowadays, there isn’t enough processing power to generate an accurate picture of such a dynamic system, especially as the number of variables affecting such a system is, for all intents and purposes – infinite.

Imagine random stones being thrown into the sea or small outcrops of rock or variations in the seabed. They all have an effect on the ‘shape’ and speed of the tide.

And so it is with an economic system: lots of rocks, stones and other variables.

It is not possible to formulate or predict a picture of such an infinitely dynamic system.

Currently, economic theory appears to predict that any shock to such a dynamic system will (obviously) have an effect on the system but that it will ultimately converge-to or seek either a new equilibrium or ‘tend’ towards its original equilibrium because, after all – that’s what ‘systems’ are supposed to do!

Economic Theory assumes a tendency towards stability and equilibrium with certain ‘oscillatory happenings’ on the way.

So we have a situation where economic thought was (and still is, in most cases) linear, deterministic and quasi-dynamic. That is to say, the ‘set-in-concrete’ notions of certainty, invariant economic laws and sameness……………..rather than approximation, probability and infinite variety.

For instance, the Bank of England  predicts an inflation rate one year ahead, based more on hope than fact or perceived fact. But when such predictions are (always!) wrong, there is no revisiting of the thought process, merely another prediction with little or no basis in anything-in-particular.

Often, both ‘inputs’ and predicted outcomes are decided by committee and vote!

All predictions appear to be based on an assumption of an ultimate convergence of economic process to stability, via those periodic cycles which, although not understood are treated with a certain sense of fatalism.

Chancellors are so locked into predictions based on erroneous facts that they will even massage their outcomes in order to land somewhere near the expected landing point – purely in order to retain credibility not only for themselves but also for ‘the system’.

What cannot possibly be countenanced are the random fluctuations of what is most likely a permanently unstable economic system. We don’t do that sort of thing because it may suggest a lack of control!

Let’s have a look at non-linear Economic Dynamics.

Actual (REAL) economic results indicate little resonance with the symmetry and regularity suggested by a linear mechanistic dynamic system. (Something that moves predictably along a pre-determined path).

On the contrary, fluctuations and movements are totally unpredictable. That means that regular Deterministic Laws cannot apply.

If we look at an economic situation in say, the Eurozone at a particular point in time, we may try to predict an outcome in say, 10 years’ time.

However, a small variant or an incorrect assumption in our analysis of the initial economic situation will have an effect on the ultimate outcome. The earlier that variation occurs, the more devastating will the effect be.

For instance Greece’s hidden debt at the time of its accession to the Eurozone, undetected at the time, is having a huge effect on the Eurozone’s economic outcome.

Meanwhile,  the economists, bankers and politicians crave and need the comfort of ‘stability’.  They know that the further the Eurozone travels from the initial conditions at Greece’s entry into the Euro, the more anomalies“The Greek Effect” will generate. It’s a self-amplifying issue.

Consequently, the bulk of the  work of Eurozone politicians is  now concentrated on creating a series of ‘faux’ stabilities.

It is the fallout from Stochasticity which is causing  fear with Determinism being their comfort and shelter.

It was only 60 years-or-so ago that stochastic considerations were appended to classical economic theory.

But the so-called New Classical Macroeconomics was no more than a compromise. “Let’s introduce a Factor X because we can no longer ignore it.”

Yet, the economists still needed their ‘models’ – because deep down they were still the mathematicians and physicists of old.

A formula was devised (SLUTSKY) which took the linear dynamic business cycle model and added random (not necessarily economic) terms which attempted to explain the real ‘actualités’!

At last, an attempt had been made to explain ‘exogenous shocks’ to an economic system by the introduction of nothing more than random error terms.

But  what was REALLY missing in classical economic reasoning was the concept of  NON-LINEARITY.

So, the battle was between a Linear Model with a Stochastic Term (a fiddle factor) versus a pure Non-Linear Model.

Obviously by now – 200 years from the beginning, we have to assume that the evidence for linearity in economics has been overestimated!

So, if we agree that we do need a new non-linear model of econonomics, what are we searching for? What are the other ingredients and how do we ‘work them in’?

Do we want a synthesis of economics, psychology, politics and sociology? Or do we simply stick to the notion of determinism?

Human evolution is viewed as a random process (although the way it is often expressed makes it seem as if scientists view it an ‘inevitable linear’).

The evolution of an economic system is also pretty random, except that, applying psychology, politics and sociology, it can never be a system that can develop naturally. (For example, Survival of the Economically Fittest).

Mind you, economists have already had several attempts at introducing the concept of non-linear economics.

Followers of Keynes developed theories which generated Real Business Cycle Theory but any exogenous shocks to the new non-linear system were considered as merely ad hoc disturbances.

Economists could NOT break away from LINEAR THINKING. Linear thinking was being applied in an attempt to imprison a loose and free system, which tended to CHAOS.

The result? More economic models that you can shake a stick at!

It is only fair to say that our understanding of economic phenomena has been greatly enhanced by all these models and formulae…… but still no cigar. No General Theory of Economics. No equivalent of E =mc2….+εe

So Chianella, Pun, Goodwin, Kaldor, Baldrin, Woodford, Barmal, Benhabib etc have all done their bit but we’re still NOT QUITE there.

Unfortunately, for all intents and purposes, many of the models did no more than introduce the concept of economic ‘white-noise’.

Chaotic systems generate their own randomness without need for external input. Therefore in a chaotic system, predictions can ONLY be very short term and even if there were deterministic rules within such a chaotic system, an inability or failure to 100% ‘book’ the initial conditions of the system will always yield forecasting errors.

This all suggests that economic forecasting (except that on a very short time-scale) is a nonsense. PLUS – the bigger the system, the bigger the CHAOS.

That would suggest that a proposal such as a EUROPEAN ECONOMY is a flawed concept because there is very likely to be an exponential amplification of Chaos.

The dynamic of a mega-economy is very different to a housewife balancing the books at home – although economists are still applying the same principles to both.

Unfortunately so far, classical economists continue to resist economic chaotic concepts.

The reason for this apparent intransigence is simple: it is very difficult to extract evidence of chaotic dynamics from economic data – especially on a meaningful scale. Especially if another dose of chaos is injected into the ‘mix’ by erroneous or spurious data.

In order to predict in a chaotic system a VAST (infinite) amount of data is required – far more than is normally available and so far, the search for Chaos in economics has not been successful.

Meanwhile it is Chaos which is making long-term economic forecasting totally impossible and increasingly sophisticated and precise measurement of ‘initial conditions’ incredibly difficult and potentially prohibitively costly.

If we imagine an economy to be like a cloud – subject to all those forces that clouds are subject to, we can  see the impossibility of a mathematical model which can predict the size, shape and exact direction of the cloud or even its shape and volume as it travels.

Its ultimate shape will always remain a mystery.

Politicians, bankers and economists ought to be able to say ‘I don’t know’ without us constantly expecting magic answers which do not exist.

For example: ‘Mr Chancellor or Mr Banker – what will be the effect on the economy of billions in Quantitative Easing?’ Correct answer? ‘We don’t know.’

“The initial conditions of a system are always uncertain, while Chaos guarantees that these uncertainties make prediction impossible.”  (Heisenberg)

THAT is the essence of Chaos within an Economics System.