Comments 31

A Financial Early-Warning System

Aug 27, 2015 Nouriel Roubini

Recent market volatility – in emerging and developed economies alike – is showing once again how badly ratings agencies and investors can err in assessing countries’ economic and financial vulnerabilities. More than ever, the world needs an early-warning system for financial tsunamis.

NEW YORK – Recent market volatility – in emerging and developed economies alike – is showing once again how badly ratings agencies and investors can err in assessing countries’ economic and financial vulnerabilities. Ratings agencies wait too long to spot risks and downgrade countries, while investors behave like herds, often ignoring the build-up of risk for too long, before shifting gears abruptly and causing exaggerated market swings.

Given the nature of market turmoil, an early-warning system for financial tsunamis may be difficult to create; but the world needs one today more than ever. Few people foresaw the subprime crisis of 2008, the risk of default in the eurozone, or the current turbulence in financial markets worldwide. Fingers have been pointed at politicians, banks, and supranational institutions. But ratings agencies and analysts who misjudged the repayment ability of debtors – including governments – have gotten off too lightly.

In principle, credit ratings are based on statistical models of past defaults; in practice, however, with few national defaults having actually occurred, sovereign ratings are often a subjective affair. Analysts at ratings agencies follow developments in the country for which they are responsible and, when necessary, travel there to review the situation.

To continue reading, please log in or enter your email address.

To access our archive, please log in or register now and read two articles from our archive every month for free. For unlimited access to our archive, as well as to the unrivaled analysis of PS On Point, subscribe now.

Featured

Our Infant Information Revolution

Jun 15, 2018 Joseph S. Nye
The Anatomy of Global Debt

Jun 15, 2018 Howard Davies
Can the Euro Be Saved?

Jun 13, 2018 Joseph E. Stiglitz
The Divine Right of Donald

Jun 11, 2018 Elizabeth Drew
The Republicans’ Protectionist Pedigree

Jun 13, 2018 Jeffrey Frankel

Nouriel Roubini

Writing for PS since 2007
119 Commentaries

Nouriel Roubini, a professor at NYU’s Stern School of Business and CEO of Roubini Macro Associates, was Senior Economist for International Affairs in the White House's Council of Economic Advisers during the Clinton Administration. He has worked for the International Monetary Fund, the US Federal Reserve, and the World Bank.

0 Comments on this paragraph, 35 in all 35 Comments on this article

Carlos del Pedregal Oct 6, 2015

ojo
Suhan Gurer Sep 20, 2015

We need financial secularity. The rating agencies have to be totally immune from political pressures of any kind.

This doesn't mean that rating agencies are good to begin with. How qualified are they to rate is another question. My view for the last 8 years is, they are not even halfway to being qualified. They need to create much more diversified and complicated methods to individually assess every country's risk. Totalitarian models do not work.
Chris Cowsley Sep 7, 2015

Here's a thought experiment. Suppose there were an accurate, accepted forecasting mechanism that people trusted and acted upon.
Everybody in the zero sum game would make a profit.

So by a reductio ad absurdum argument there never will be.
jatin meswani Sep 2, 2015

The key is how to identify between noise and actual risk. There are many economist who keep on raising red flags for not right reason.
John Lock Sep 2, 2015

This all sounds clever, and many of the comments on it sound even cleverer. The idea of using big data and predictive analytics in one form or another is of course attractive. The trick is, that if you ever get such a system working, you mustn't tell anyone! Because then the new information you add to the system, will change the system. It's a classic Heisenberg uncertainty scenario. Step 1) You perform a meta analysis of the system, capturing a sampling of relevant information (no, it is quite possible you don't need to have all information, or know internal relationships (causality)). Step 2) You build and validate (on historical data and even resampling of historical events) a model that has excellent predictive power! Hooray!. Step 3) You tell people about the "warnings/opportunities", and maybe the model too (well, a warning system isn't much use if on-one is warned, and opportunities not taken are sooo annoying!). Step 4) People start routinely using your meta analysis of the system to interact with the system - uh oh, the meta analysis is now a feature of the system (kind of like black Scholes). Step 5) The system starts to drift away from the predictions of your not-so meta analysis, due to the pre-emptive impact of those predictions. Step 6) The model fails. Step 7) You build a new model and the cycle repeats.

This is much like the inter-dependence of opinion in Keynes' beauty contest, although with a more explicit time component. Still. hopefully Nouriel gets some investors!
Manuel G Samuel Aug 31, 2015

Nice article. Mr. Roubini is upholding the use of scenario planning for financial early warning purposes. Nothing wrong about that. Big Data mining, crunch and predictive analytics newest developments make scenario generation a lot more interesting. A global financial early warning system based upon predictive analytics can be very helpful. Even if the future is impossible to predict, a financial scenario generation system (notably based on predictive analytics, digital modeling and even social analytics), may possibly warn with some degree of accuracy to the likelihood of one or more alternate financial negative scenarios. This may be crucial to avoid inflating underlying financial conditions and even to possibly avert financial crashes. It is not entirely safe, but it is a good start at this point in time as predictive analytics are rapidly developing. Some rating agencies are already trying to develop skills in this regard. The descriptive and proactive stages are already old fashion. Rating agencies need now to be predictive, very much in the sense of Nouriel Roubini's early warning proposal.
Enrique Woll Battistini Aug 31, 2015

There will always be a need for a financial early-warning system; but how relevant, important or critical that system would be to peace and prosperity, is inversely proportional to the realism and efficiency, and thus the value and stability, of the existing resource allocation systems in countries and the globe. The better the articulation and coordination of the private and public entities involved in socioeconomic development, and the better they are able to promote comprehensiveness and balance in investment and production outlook, freedom of choice in consumption, and fairness in the distribution of income and the accumulation of wealth, the more inclusive the world will become, the less room and opportunity for corruption there will be, and the more stable it and all its socioeconomic systems will be. Thus, the improvements in technological innovation, and production efficiency, and in all areas of human endeavor, needed to attain world peace in this century, would require an for-profit articulation of all entities charged with addressing comprehensive socioeconomic development, both private and public, governmental and multilateral, in order that investment be directed in optimal ways. The need for this articulation would be best served along three north-south axis, Asia-Oceania, The Americas, and Europe-Africa. Moreover, to create and maintain fairness in the distribution of production among the labor forces at both ends of the scale, and generate the concomitant incentives and savings to maximize individual productivity and prosperity, widespread tax reform would be needed. Obviously, rampant capitalism will not meet these requirements, as Thomas Piketty has already shown, and it must be curbed. The following documents posit some practical possibilities addressing the above identified requirements, and provide and informal proof of Piketty's Thesis:

1. On Articulation of Development Entities:

https://www.academia.edu/12823841/Mathematical_Model_and_Simulation_for_A_Partnership_for_Development_with_the_United_States_of_America_-_December_1999

2. On Tax Reform:

https://www.academia.edu/13062837/La_Reforma_Tributaria_del_Siglo_XXI_The_XXI_Century_Tax_Reform_-_2011100411

3. On Rampant Capitalism:

https://www.academia.edu/13062623/Informal_Proof_of_Thomas_Pikettys_Thesis-2014060802
Farid Novin Aug 31, 2015

Paul Daley's comment worth repeating:
"Roubini is basically claiming that the world needs a theory to predict economic events and that he has what the world needs. The market reaction - "Prove it!"

To have a reliable proof, it would be nice to have the models predictions for the GDP growth, balance of trade, ten-year bond interest rate , budget deficit for the third quarter of 2016 for Japan, France, Italy, south Korea and Australia.

Mr. Roubini criticises the rating agencies for their ratings because they " are often backward-looking", and " downgrades occur too late. Yet the model suggested in this article "using 200 quantitative variables and factors to score 174 countries on a quarterly basis" apparently can only provide rather old information. In other words, all the information that is provided by the model with regard to China, Brazil, Ireland, Spain and Greece are already quite well known.

Using 200 variables, even for about 30 of most advanced countries that have relatively reliable statistics, requires either a well specified structural model which at present would be impractical or some dimension reducing techniques. Using nonparametric techniques , and assuming perhaps some conditional quantile function to satisfy smoothness, is challenged by the rapid loss of estimation precision as the dimension or the number indicators increases. If this is true then the assessment of risk would be highly unreliable.
Using principal component analysis, principal factor analysis, maximum likelihood factor analysis, and so on are challenged by the reliability of the eigenvalues and eigenvectors. Given that resampling of data is not an option, presumably the model uses analytical expressions to provide useful insight with respect to the major sources of error. The validity of these assessments would be based on subjective assumptions that are employed in the analytical expressions, which would depend on the predetermined assessment of risks.
Richard Stewart Aug 31, 2015

This reads like a paid commercial.
Todd Clay Aug 30, 2015

What I am really hearing is that you need a better way to get out at the top!
Todd Clay Aug 30, 2015

The early warning system is there...people refuse to heed it and Wall Street uses it to their advantage by manipulating data (aka Greece, Louisiana, etc.) which makes them able to play both sides and why they're always on the other side when the crash comes...because they are the crash.
Andrew Benington Aug 30, 2015

Plenty of people spotted 2007/8/9, eurozone sovereign debt and emerging economy vulnerabilities years before they happened including Mr Roubini. But he lacks a grasp of driver, concentrating on symptoms.
1/ Rating agencies have limited liability. Only a fool paying for advice from an accountant or lawyer would trust that advice when it's not backed by family home, pension and future income.
2/ The last decade's problems are the price driven dismantling of vertical structures and their replacement by the horizontal structures of participation, control and ownership inherent in tech.
gene sperling Aug 30, 2015

???Realistically, the banks and financial institutions (now including insurance companies and brokerages) have successfully shifted all social risks to the people.
I don not think this scenario will not work without a complete overhaul of all three of the aforementioned entities.
steve from virginia Aug 30, 2015

Right now there are all sorts of warnings and nobody pays attention climate warnings, cost warnings, resource depletion warnings, corruption and cartelization warnings, over-dependence upon central banks, etc. Economies are too fragile, they cannot bear the weight of the truth.

Even though unraveling is well underway, the various media components of big business are busy informing the proletariat (and each other) that everything is fine and that 'sustainable growth' is right around the corner.

Warnings -- early and otherwise -- are self-propelled so there is no place for them. The central banker can never say there is an asset bubble or that the bank is helpless to affect outcomes. A money-panic would occur in either case.

This in turn reveals the extent to which market capitalism is dependent upon fraud and participants ability to embrace it.
eusebio manuel Aug 30, 2015

When most economic experts and analize China economy in the last 35 years they tell us that this time sounds like overdue market correction
Chris Menon Aug 30, 2015

Steve Keen has done some valuable work here on the importance of private (as opposed to public debt) and is one of 3 wise men worth heeding on this matter. If you'd like to learn more, read this: www.safestocks.co.uk
Mila Kasalica Aug 29, 2015

Since when stock markets pay attention to risks when the only relevant scenario deals with causing panics. Last one did not work so well as expected because to default China is much more serious than Greece and EU together. Somehow those matrices for EWS will never focus on essence because developing world learnt decades ago the lesson- volatility is only permanent definition of markets and that economists predictions (positive or negative) regardless which model use are not reliable and will never be reliable.
Stephen Stanley Aug 28, 2015

I've yet to see Economics as a science, at least not in the sense I'm used to. If a law of physics failed to predict something at least as big as objects in a gravitational field accelerating toward each other, roughly equivalent to economics missing the entire Great Recession, there would be a flurry of scientific research into why the model was wrong, not how to justify the failure of the model. Come on, guys, you missed forecasting Hurricane Katrina and you want to be taken seriously? What about the notion that your elegant mathematical models are just that, models, with little grounding in reality.
- steve from virginia Aug 30, 2015
  
  Hurricane Katrina was accurately forecast the National Hurricane Center. The response to the forecasts was a crime.
  
  "Great job, Brownie!"
- Robbie Jena Aug 29, 2015
  
  Thank you...
Henry Rech Aug 28, 2015

Why is this so problematic?

Busts follow booms - booms in asset prices, booms in goods prices, booms in investment, booms in private and government debt.

Surely it is not beyond the ken of appropriate authorities to monitor abnormal trends in the above areas and measure them against yardsticks?

When macroprudential yardsticks are breached, action is needed. The problem is that authorities and governments seem to be loath to intervene in markets for fear of losing electoral favour with some misbehaving group.

An example is the booming Australian property market heading for a trainwreck. Not only have authorities been slow to intervene and when intervening have not been determined enough to alter behaviour, but the Federal Treasurer was yelling from the rooftops, buy!, buy!, buy!. Utterly irresponsible.
- Diana McCoy Aug 30, 2015
  
  "When macroprudential yardsticks are breached, action is needed. The problem is that authorities and governments seem to be loath to intervene in markets for fear of losing electoral favor with some misbehaving group".
  And therein lies the critical realization. How many times have we seen "crisis" evolve from precisely this lack of moral / ethical responsibility?
  
  Read more at http://www.project-syndicate.org/commentary/financial-early-warning-system-by-nouriel-roubini-2015-08#J1DEGPAksMCWESAP.99
Zsolt Hermann Aug 27, 2015

In order to create an early warning system one needs to know, understand the cause and effect processes of the system otherwise one cannot make calculations, predictions since the foundation of such calculations is missing.

Based on my very limited knowledge of chaos theory trying to explain weather patterns for example is very difficult because we do not know the precise initial conditions the system. Still since the weather is based on the laws of the natural system we exist in we can create mathematical equations to try to at least have a shot at those cause and effect processes and come up with more or less accurate results.

The problem with our economic, financial system is deeper. They are based on an illusorical system dreamed up by human beings. We created a coordinate system that has no natural foundations, it does not follow the laws and principles of the natural system. Thus our human system including economy and finances are simply up in the air.

It seemed to be working for a while until we existed in a bubble we could seemingly stretch infinitely.
But today we reached the boundaries of our bubble where we started bumping into the natural system. As humanity has saturated its living space, as we are reaching the limits of human and natural resources. Thus has become obligatory to adapt to the natural laws and principles that are binding to everybody in the system, and as a result our artificial human illusion is falling apart.

There is nothing to predict except that our paradigm is unsustainable. It cannot be repaired, tweaked, "solved", or "bailed out".
We have to rebuild our human coordinate system but this time based on natural foundations. Then we will know where we are, what our goal is and where we are found on the path of evolution.
slightly optimistic Aug 27, 2015

"an early-warning system for financial tsunamis may be difficult to create; but the world needs one today more than ever"
In the UK, for example, legislation has long required the regulators and external audit to bite the hand that fed them [comment below]. Moreover they were awarded independence to carry out the vital role.
But following their official warnings about the coming 'financial tsunami' in 2008, and the huge losses if banks were allowed to continue, the independence was withdrawn.
The Chilcot report?
Erikwim During Aug 27, 2015

See also the work of Didier Sornette: http://www.er.ethz.ch/. In the light of the insights of Minsky that stability breeds instability and vice versa, it would however not fundamenttally change the boom-bust dynamics of markets and human behavior (common analogy: safer cars stimulate more reckles driving behavior)
Curtis Carpenter Aug 27, 2015

If only Mr. Roubini's model had accounted for that 201st vital statistic , instead of just stopping at 200...

MM in his comment below has it right, I think.

I can't recall any time when, for every "Happy Days are Here Again" prediction, there wasn't a contemporary "End Times are Upon Us" prediction to balance it out. And in both cases, the predictors were relying on pretty much identical data sets. I'm skeptical that BIGGER data sets will change that. The Butterfly Effect is real.
Paul Daley Aug 27, 2015

Roubini is basically claiming that the world needs a theory to predict economic events and that he has what the world needs. The market reaction - "Prove it."
jagjeet sinha Aug 27, 2015

Big Data indeed has all the signals embedded in the enormous data that is captured and available for analytics.
Greece was easy to predict - perhaps many similar predicaments too.
The remedial actions that will prevent the meltdown is perhaps also possible to produce.
But despite the diagnosis and the remedies, the hospital fails to carry out the operations.
Endless possibilities of differences in course of action brings a stalemate.
Always because those who must, are not accountable - leading to Democracy itself being questioned.
Big Data Analytics may still not prevent the recurrence of Greece.
Instead of targets like Inflation alone, perhaps additional Metrics democratically implemented.
Together then, actionable possibilities are likely.
Osman Gani Aug 27, 2015

Chians economics is complex. I can means that the global faced economics problem. It is one reason that we have forgotten our Lord.
http://couponappa.com/supplies-outlet/
- Robbie Jena Aug 29, 2015
  
  "Chians economics is complex"
  
  Perhaps it is.... It is based on Systems Design and not Economics...as it was done by Engineers and not Economists...that is how it is...
M M Aug 27, 2015

The current global financial system is not fit for purpose, everyone knows it but no one is willing to touch it, let alone changing it, of fear of creating another “2008 Tsunami”.
Per Kurowski Aug 27, 2015

We should of course welcome any alternative early warning system. That could at least put some pressure on credit rating agencies to do a good job so as to keep on being relevant.

But that said the faulty credit ratings are not the real problem… these have always been and will frequently be wrong.
The problem was that even though banks already clear for credit risk information, like with size of exposures and risk-premiums, the regulators also made their capital requirements for banks to be based on credit ratings.

And we all know, that even if the information is perfect, you will get it all wrong, if you consider that information excessively.

http://subprimeregulations.blogspot.com/2011/04/basels-monstrous-regulatory-mistake.html
Val Samonis Aug 27, 2015

Independently financed Colleges of Wise Men are needed to use big data in early warning efforts and regularly publish conclusions for the public benefit.

Val Samonis
Robbie Jena Aug 27, 2015

So, the World is collapsing? What is the solution Mr. Roubini? Time to tell the fix for better growth of the Planet...Thank you.
Michael Public Aug 27, 2015

It is normally the target of the rating that must pay the fees of the rating agency. No sane person will ever bite the hand that feeds them. Of course, you are not immune to the same concerns. You say you have recalculated present positions of sovereign risk, but you you don't show your workings and you don't disclose your source of income to do this work. Did Brazil not pay your invoice? Or are you acting with academic integrity and warning investors with an altruistic intention? I would like to believe it is the latter but give the reader something to work with here.

http://prosyn.org/UaDBLAa;

Log in/Register

A Financial Early-Warning System

To continue reading, please log in or enter your email address.

Featured

Our Infant Information Revolution

The Anatomy of Global Debt

Can the Euro Be Saved?

The Divine Right of Donald

The Republicans’ Protectionist Pedigree

New Comment

Email this piece to a friend

Contact us

The Singapore Summit’s Uncertain Legacy

Financial Equality for Africa’s Women Farmers

Greener, Greater Cities

The Anatomy of Global Debt

Our Infant Information Revolution

Putin’s Economic Dilemma

The Power of Empowering Refugees

Agriculture’s Bad Medicine

Cookies and Privacy