Model Predictions

On the heels of the October 2008 subprime mortgage credit collapse has emerged not only the worst financial contraction since the Depression, but “collaterally,” discreet burials of patients bearing their surgeons’ mistakes. That is, country risk analyses (CRA) gone off track. When the patient dies, who suffers? The client, of course. ¹ But since most commercially available country risk profiles are proprietary, client confidentiality also seals survivors grieving CRA errors.

What of CRA’s embedded assumptions? While there are many, let us consider a case study based on public information for a partial review. To review a few:

Venezuela’s mishandled nationalizations of foreign direct investments (FDIs) and why this happened.

This same analysis can be applied to assessing the wrongs of our contemporary financial crisis. The bottom line is that risk modelling, in failing to properly account for unpredictable dynamics of complex human behaviour, can only go so far in calculating risk. Obviously, predictive capabilities of abstract, math-based, quantitative mod- els is suspect. Should they be supplemented with — or replaced by — a more dynamic, qualitative, evidence based historical approach?

HUGO CHAVEZ AND NATIONALIZATION

According to Forbes, in 1999, in Hugo Chvez’s first year of government, Venezuela’s FDI dropped by 42%. In 2000, its’ net FDI moved higher than its’ 1998 figure. But come 2002, FDI plunged to US$800 million from its 2001 figure of US$3.7 billion; net FDI was negative for the first time under Chvez. Since then, FDI has fluctuated. In 2006, a presidential election year, gross FDI was negative for the first time under Chvez.²

As of May 1, 2007, Chavez nationalized the last oil production sites under foreign control in the Orinoco Oil Belt . Until then, companies could negotiate terms of their nationalization.

Oil production in “the Belt” covers an area of roughly 55,000 square kilmotres generating an estimated 1.3-trillion barrels of “oil in place,” ranking the area among the world’s largest extra-heavy oil reserves. After 2007, it has been co-ventured between Venezuela’s state-owned PDVSA (ownership stake, 60% plus) and foreign subsidiaries (making up the difference). Previous arrangements gave PDVSA a minority stake. ³

On Feb. 28, 2007, CBS reported Chavez had previously announced intentions for a majority stake by May 1, 2007 in four heavy oil-upgrad- ing projects run by British Petroleum PLC, Exxon Mobil Corp., Chevron Corp., ConocoPhillips Co., Total SA and Statoil ASA. ⁴

According to the U. S. State Department, Chavez decreed in February 2007 that the strategic associations convert to joint ventures, giving PDVSA a 60% stake by Apr. 30, 2007. ConocoPhillips and ExxonMobil refused to migrate their investment stakes in three of the four associations. Result; Venezuela took over their investments. Both companies claimed expropriation, filing claims against Caracas. ⁵

Whether or not this does in fact constitute “legislated theft” is not the point. The point is that such sentiments, as well as (principally American) FDIs’ perceptions of Chavez as a dictator, raise critical CRA issues. For example:

What was the earliest awareness of Hugo Chavez as an emergent player in Venezuelan politics?

Was there any legitimacy to the way the government treated FDI players in that country?

Did anyone’s analysis include a review of the foreign investors resenting Chavez’ policies?

These questions highlight CRA’s transaction orientation. Any enterprise making significant transactions and/or partnerships within another country needs to know the history and stability of its foreign partner. Why? Because if it commits resources to an area, it expects a return on its investment (ROI). Hence, due diligence includes knowing relevant history about a foreign jurisdictional partner.

Hugo Chavez led an unsuccessful army coup in 1992 against then-president Carlos Andres Perez. Prior to that, the BBC reported, Chavez and fellow officers founded a secret movement inspired by Simon Bolivar. So, if Chavez was not on radar before 1992, exactly when was he factored into CRA Venezuelan analyses? In other words, why would ConocoPhillips and ExxonMobil cry foul as of Apr. 30, 2007?

CRA’s transaction orientation sets itself up to fail for two reasons:

It doesn’t properly monitor real events capable of impacting future business affairs.

It victimizes itself on the basis of its evidence/data bias selection. ⁶

NUMBERS AND HISTORY

What, then, is the bridge between CRA modelling and better incorporating “long-tail” historical analysis? A diversion may help.

Modelling, myth vs reality, map vs territory, a 2003 Hewlett Packard study, ⁷ says important things about modelling. First of all, it points out that modelling, which is critical to country risk analyses, is based upon abstraction. It likens the use of models to the use of maps. But “what must be understood is that the map is not the territory,” the HP study points out. “The only model that will behave as the original system is [the original] system itself.” And the ‘original system’ upon which the model is based is dynamic, the study further notes.

“Chaos theory has its implications, suggesting that forecasting is a black art and should be left well alone,” the study says. “This is wrong. Chaotic systems are not necessarily impredictive, they just have limits on how long trajectories can be trusted. Weather forecasting can be considered reliable for five days. To say that a system is chaotic without the timeframe is pointless when considering whether a model will produce useful predictions.”

Transferring HP’s analysis of complex issues affecting business modelling to CRA, where do models calculate humanity’s passions? Where do CRA analyses address issues of justice, discrimination, exploitation, hope, despair and skewed male/female demographics creating sexual competition? Leaders arise from nowhere, correcting injustices “their people” collectively seek to eliminate. Others simply capture power, exploiting the mantra of “justice to the people,” thus re-triggering the same grievances later.

Against this bottom up backdrop of human drama, commercial corporations are top-down, command-obedience structures designed to respond to market challenges. Political jurisdictions (assuming no ideologies of justice and social mobility) can be seen in the same way:much like businesses, they establish laws to create social cohesion, protect sovereignty and predict human behaviour. CRA modelling, as practiced between corporate and political entities, may have validity, but “the map is not the territory.”The CRA model is not the nation where citizens live.

MODELLING THE FINANCIAL CRISIS

CRA modellers have the reality game turned inside out. Instead of looking into their models, software and tools, they might do better to look at how the world “got this way.” Is how North America “got this way” just a case of modelling selectively adopted strategies and abstracted worldviews for the purpose of maximizing return at someone else’s expense?

The nuances of human behaviour can help explain (and may have even predicted) the current credit crisis in a way that mathematical models did not. Subprime mortgage lending — i. e. low-interest, adjustable-rate mortgage lending by financial institutions to NINJAs (people with no income, no jobs and/or assets) — followed a subprime ethical standard, the potential consequences of which lenders ignored while the going was good. Simply put, when the American economy was performing, 40-year mortgages performed. These marginally collateralized instruments were then spun off into derivatives sold globally through third-party financial institutions to Swedish farmers, Australian fishermen and who knows to whom else.

This happened while North American manufacturers and service industries outsourced to foreign, low-cost labour markets. When petroleum prices were hiked in 2008, and employment decreased in concert with high debt (a result of the mortgages, the trade deficit and Iraqi and Afghan military expenditures), did it take a genius to figure out that credit would collapse?

Historians are rarely surprised by these seemingly inexplicable, sudden turns of events. According to Niall Ferguson, “self-satisfied bankers” living in the moment dismissed as alarmist any warnings of “a drastic decline in…liquidity through the global financial system.” Ferguson himself warned prior to the crisis:”We should be cautious about expecting the good times to last indefinitely.” ⁸

Early warnings about higher subprime mortgage defaults were also ignored before October 2008. One warning, for example, could be found in The Evolution of the Subprime Mortgage Market, ⁹ published in January 2006 and written by Souphala Chomsisengphet and Anothony Pennington-Cross. Also, in his May 2007 report to the Berkshire Hathaway AGM, board chairman Warren Buffet warned that: “rising defaults in the subprime mortgage business have been exacerbated by securization of these loans…Subprime mortgages are offered to poorer people with blemished credit records. The loans are often packaged up and sold…to investors as mortgage-backed securities. ‘Once you package those things and sell them through major investment banks, discipline leaves the ^{system.'”10}

Why didn’t anyone listen? Did models account for Niall Ferguson and Warren Buffett as Cassandras when cash flowed like champagne from the spigot?

Back to history’s long tail. North America’s post-1945 prosperity was an unnatural economic bubble, a historical warp, created by America’s emergence as the world’s most powerful economy. With the recovery of Europe and Asia, North America faced greater foreign competition, leading ultimately not to “the decline of America, but rather about the rise of everyone ^else.”11 Ferguson’s bankers may have lived in the moment up to 2008, but Ferguson contends the West has been in decline since 1945, despite its military victory in the Second World War. ¹²

Do CRA practitioners take into account this kind of long-tail, historical research? According to one study, 274 military coups occurred between 1946 and 1970 in 59 countries. 13 Taken out of any historical context, do these numbers have any predictive power for a risk manager?

In The Legacy of Secrecy, ¹⁴ published in 2008, the authors argue that the same members of the American mafia — under Carlos Marcello — assassinated President Jack Kennedy, his brother Bobby, and “brokered” the murder of Rev. Martin Luther King Jr. The authors used evidence-based research for their devastating analysis. Given that it took 46 years for evidence of a triumvirate murder conspiracy to come to light, how, one might ask, could a contemporary model possibly assess the potential for a future coup elsewhere? On what basis? Some predictive algorithm? Or evidence?

Coups are like Christmas presents: they are planned in secret and delivered overnight as a surprise! What third party, other than a co-conspirator, would know of one in advance? Evidence based research is the only thing that cuts it.

Of course, if risk managers are really bent on forecasting the future, they could always ask an astrologer. But be careful: as the BBC reports, astrologers can be arrested for making political predictions. ¹⁵

1 Pablo Triana, Lecturing Birds on Flying; Can Mathematical theories destroy financial markets, John Wiley and sons, Hoboken, NJ., 2009. Pg 246.

2 As Venezuela Nationalizes, Investors Worry, Oxford Analytica. 09.10.08,6:00 AM ET http://www.forbes.com/2008/09/09/venezuela-chavez national izationcx_091 0oxford.html?feed=rss_business

3 Venezuela Decrees Nationalization of Last Foreign Controlled Oil Fields. February 27th 2007, by Gregory Wilpert –Venezuelanalysis. com http://www.venezuelanalysis. com/news/2245

4 Venezuela To Seize Foreign Oil Projects State Oil Companies Will Take Majority Stake In Foreign Oil Projects; Chavez Calls Bush “King Of Liars” CARACAS, Venezuela, Feb. 26, 2007, http://www.csnews.com/stries/2007/02/26/world/main 2519183.shtml

5 Venezuela 2008 Investment Climate Statement-Venezuela Openness to Foreign Investment, http://www.state.gov/e/eeb/ifd/200

8/101026.htm”

http://www.state.gov/e/eeb/ifd/2008/1010

6 Wikipedia.

7 Modelling, myth vs reality, map vs territory, Hewlett Packard, 2003. http://www.hpl.hp.com/techreports/2003/HPL-2003- 246.pdf

8 Niall Ferguson, The Ascent of Money, Penguin Books, London, 2008, Pg 7.

9 Souphala Chomsisengphet and Anthony Pennington-Cross, The Evolution Of the Subprime Mortgage Market, http://research.stlouisfed.org/publications/review/

06/01/ChomPennCross. pdf

10

http://www.marketwatch.com/News/Story/subprime-defaults-made-worse-securitization/ story. aspx?guid={817 DADC5-3315-4D00-97BA-6D6F91543 C58}&dist=MorePulse

11 Fareed Zakaria, The Post American World, W W Norton & Co Inc., New York. 2008.

12 Niall Ferguson, The War of the World, Penguin, London, 2006.

13 Explaining Military Coups d’Etat: Toward the Development of a causal model, Ekkehart Zimmermann, April, 1978. http://www.springerlink.com/content/p81063625k8475m5/

14 Lamar Waldron, Thom Hart mann, The Legacy of Secrecy, November, Counterpoint Publishers, Berkeley CA.,2008.

15

http://news.bbc.co.uk/2/hi/south_asia/8120859.stm

———

CRA modellers have the reality game turned inside out. Instead of looking into their models… they might do better to look at how the world “got this way.”

———

We should be cautious about expecting the good times to last indefinitely.