Syndicate content

FPD Forum 2010: Inconvenient truths on the future of finance

Editor's Note: John Nellis was a Senior Manager in the World Bank's Private Sector Development Department. He is now Principal of the consulting/research firm, International Analytics.

On March 2 and 3, I attended the annual World Bank Forum on Financial and Private Sector Development in Washington. Last year, as we teetered on the rim of the abyss, the keynote speaker was Nassim Taleb, author of The Black Swan, who (as I wrote) left no one in the room uninsulted. This year, as recovery stumbles its way forward, or sideways, both speakers and audience evidenced lower blood pressure, and, alas, less bombast and theatricality. Some reflections:

The first speaker was the very shrewd and able Moises Naim, presently Editor of Foreign Policy journal.  He has been an Executive Director (i.e., member of the Board) of the World Bank, a professor, and, in the 1990s, Minister of Industry and Trade in Venezuela.

With apologies to Al Gore, Naim said he had some "inconvenient truths" for us. These were:

  1. The decline in investment and growth stemming from the crisis will last a decade; maybe more.
  2. Despite all the hoop-la, the architecture and rules of the game governing the financial sector will not be radically transformed; there will not be any new, comprehensive, global financial regulatory system. Reforms will emerge partially and piecemeal, in different countries and regions, and they will differ substantially from one another. This will present much opportunity for policy arbitrage; i.e., gaming the differences in the systems for profit. As another speaker put it, "regulation breeds circumvention."
  3. As patchwork measures surface to limit some activities of formal banks, the "shadow banks" -- the informal systems that arose in the period 1998-2008 and contributed greatly to the crisis -- will flourish.  As for banks proper, they will be "de-risked" and come to resemble utilities.  The wrong target will be struck.  (Not stated at the conference, but an adage from elsewhere that seems appropriate: "All initiatives end by punishing the innocent.")
  4. The crisis has formalized the split in economic thinking; there is now a western and an eastern approach to ownership, property rights, finance and regulation. Much of the thinking in the west on private sector development uses the formal, publicly-traded firm as the building block; but the more common and quite different firm in the comparatively successful parts of the world (for now) is the family-owned, closely held, non-traded firm, the behavior of which is insufficiently analyzed in formal economics. And finally,
  5. The crisis has significantly lowered the barrier to the adoption of bad ideas (an hypothesis confirmed daily in legislatures around the world, spectacularly so here in Washington).

Not on a pessimistic par with Mr. Taleb's Black Swan address of a year ago, and delivered in a far less flamboyant manner, but unsettling stuff nonetheless.

I attended the second plenary on March 3, this time with a keynote delivered by Niall Ferguson, Professor of History at Harvard (and holding appointments at the Harvard Business School, the Gunzburg European Center, Jesus College, Oxford, the Hoover Institute at Stanford, several private consulting firms and the BBC -- don't you dare say you're busy). He spoke on the question: "Can developing countries learn from the history of money or are they doomed to repeat it?"

Ferguson describes himself as "a reasonably engaging speaker," and he is certainly that. He also says he "is proud not to be an economist." He is an historian. His last book was The Ascent of Money; he is now working on a history of the west's ascendancy. This book will be published as a companion piece to a 6 part BBC series on the same topic (in which, presumably, Mr. Ferguson will do for development what Richard Attenborough did for iguanas).

Niall Ferguson

He begins by asking: is the contribution of the financial sector to long-term growth negative? The opposite view is held by most scholars of the topic. Indeed, Ferguson cheerily notes that the premise of a thoroughly positive association between density and efficiency of financial intermediation and long-term growth underpinned his latest and quite successful Ascent of Money. But the crisis, and some recent research of his own and others, leads him now to wonder if this is correct. 

He notes that the current crisis is one basically of western advanced industrial economies, and that countries with smaller and more constrained banking/financial systems, mainly but not exclusively in Asia, weathered the storm far better. Smaller financial sectors, he says, are associated with less debt, fewer bubbles, and current account surpluses. Perhaps we've had it all wrong?

He expands the notion to query what he considers as the conventional explanation of the crisis: that de-regulation of banks was the fundamental factor precipitating the boom and then the bust. Mr. Ferguson insists that problems in the banks was only one of 6 factors, and not the most important one, that contributed to our present mess.

  1. Yes, the banks (and he notes most western banks were culpable, not just those in the US) were excessively leveraged, but this by itself would not normally have caused a collapse.
  2. AAA abuse:  the failure of the rating agencies and system was breathtaking.
  3. Massive error in monetary policy by the US Fed, which he says dates to 1999 when a group of high level economists (Larry Summers strikes again?) persuaded the Fed that rapidly rising asset prices -- houses -- could safely be ignored as long as the consumer price index was not tracking housing prices.  The Fed did nothing to prick the housing bubble. (Note that Eugene Fama and other Chicago economists say there is no such thing as a bubble. Maybe Ayn Rand wrote this someplace too, thus persuading Greenspan?) The brakes were never applied; the throttle was left open. (A Toyota metaphor lurks here.)
  4. Financial insurance mechanisms went wacky. In his best line, Mr. Ferguson stated that insurance and swap devices were built not to see that "risk was borne by those best prepared to take it, but rather by those least prepared to understand it." Well done!
  5. Politicians contributed to the problem as much as CDOs and model-driven economists by giving incentives to home ownership to more and more "marginal borrowers." Ferguson asks what percentage of citizenry home ownership is needed to precipitate a financial crisis? The answer appears to be 69.
  6. And grand geo-political issues intrude as well; China's fueling the US debt furnace is another contributing factor.

What should emerging countries do to avoid such a mess? Follow Canadian rather than US regulatory systems. Realize that asset prices count. Don't push home ownership for its own sake. Realize that insurance should be based on calculable and understandable risk.

All sound and sensible, but a bit of a let down following the breathless build-up. Still, it was presented with verve and a flourish, with asides and jabs and interesting slides. But we have heard most of this before, and the tone of the presentation --- I've put all the pieces of the puzzle together before anyone else -- is not justified.

Ferguson is, as he states, an historian; he looks for the whole story. The economist's search for the essential, for regressing to the key element that explains the largest "percentage of the variance" is antithetical to his approach. I enjoyed Ferguson's talk, learned a good amount, and was provoked to reflect at a number of points. But the rich fabric of his story on the crisis does not add up to a tight explanation.  Mr. Ferguson should not be quite so proud of not being an economist.


Great post John, thanks for sharing. Just a comment on your closing remarks: one brand of economists (think Arthur and the Santa Fe complexity crowd) would mostly sympathize with Mr. Ferguson's approach. Systemic knowledge is knowledge too, and it just fits some problems better than regression. I got to the end of your post under the impression that you thought the crisis is one such problem, and Mr Ferguson's treatment of it is quite convincing. If his approach works well, why should he not be proud of it?

Submitted by Xavier L. Simon on
John, hi! Excellent comments. I enjoyed very much your most recent entry and that of a year ago, including about Nassim Taleb’s talk. For a long time I have subscribed to Taleb’s underlying model, and I have been following Niall Ferguson’s work very closely. Although neither Taleb nor Ferguson are aware of it, their theses are intimately connected. The connection is the behavior of a particular class of complex interconnected systems called “multiple-loop non linear feedback” systems. Such systems learn by experiment, which means trial and error, continuously overshoot and correct, and therefore exhibit cyclical behavior. For more years than I care to admit I have been trying to understand development through the prism of such systems. Recently I began to settle on a methodology that allows me to understand and explain better, and then propose concrete solutions to societal and economic development issues that had baffled me for many years. In my approach economics and sociology are intimately connected and need to be addressed jointly. Inherent to learning by trial and error is that things go wrong. That is the basic underlying premise of Taleb’s Black Swan. You can build very fancy math models but inevitably everything eventually boils down to experiment or trial and error. All that my fancy MIT and Harvard degrees gave me was a capacity to make better educated guesses, but nonetheless these all remain guesses that can only be validated through experiment. In studying the behaviors of many societies, past and present, I came across a development mechanism that appears to be common to all societies, ranging from early hunter gatherers and persisting through the most advanced societies today. And within that mechanism there is a particular characteristic that explains why societies that do develop do so at different long-term rates and have different life spans. Generally, societies that are better able to generate innovations, adapt these to their needs, and then more homogeneously absorb the changes, including the new rules each change requires, throughout the society will grow more and last longer. There are other steps relevant to the change process but for the purposes here these should suffice. This mechanism works quite well as long as all of the players remain relatively small and the group or society also remains sufficiently homogeneous. Things go astray when the system develops lumpiness or unusual growths and becomes more heterogeneous. Since the learning, including key elements of the change adoption mechanism, operates by trial and error and is cyclical, lumpiness tends to result in oscillations that the system is less able to handle. When the lumpiness becomes too large, or in the lingo of my methodology there is excessive bigness, the effect of trials that result in error can be destabilizing. When two or more of these cyclical events coincide the result can result in system failure. The latter is exactly what happened that led to the financial crisis. The more obvious bignesses, if you allow me that word, that led to the financial crisis were (1) Fannie and Freddie; (2) the larger private financial institutions that resulted from the repeal of Glass Steagall; (3) new financial instruments, the use of which grew too large before they inevitably failed and we were thus able to learn more about their properties; and (4) the massive reversal of international capital flows that acted as an accelerant. A fifth more elusive bigness could be that of the central government itself, and perhaps even an excess of regulation. Notice that I didn’t include regulatory failure. The reason is that I would argue that the regulatory failure that did occur was an effect rather than a cause. In a system of learning by trial and error our practical knowledge and understanding of the behavior of any system will always lag change. With change always comes a need for new rules. But we can regulate only that which we can foresee, and change always has unforeseen effects. We only learn about them after the fact. I thus argue that the regulatory failures were inevitable and that their effect on the system was large only because the other four or five changes were each very large and when they came together and amplified they totally overwhelmed the system. I have tried to understand individual regulatory failures like, for instance, the loosening of mortgage lending or the failure to adopt the 1998 CFTC proposal to regulate derivatives. I believe that each would have dampened the overall effects but that we would have nevertheless ultimately had some large effects. There were just too many changes happening simultaneously and some were too large. To me the real big one was the reversal of capital flows. If that hadn’t happened the other smaller failures would have eventually reared their ugly heads but without the large consequences that we saw. Unfortunately, I think that without the crisis also the fact that Fannie and Freddie had gotten too large and flawed might have remained hidden and would have led to some other later crisis. Regarding a possible excess of regulations, historically the systems or societies that performed better are those with the larger number of internal checks-and-balances that operate down to the level of the individual. I am beginning to believe that in advanced societies regulations grow excessively to where they form a large all encompassing strait-jacket that pretty much determines how the society as a whole functions. If so, what happened that led to the crisis is that the real culprits were the very large inflows of foreign capital and that the system gave where the strait-jacket was weakest, namely housing. If the overall system had been looser then maybe the large inflows would have been spread more widely throughout the economy and a crisis could have been averted or at least dampened. Moreover, if I am correct that the US has become over-regulated, then that could in part explain why we are having a harder time coming out of this crisis, and it doesn’t bode well for the future. My bottom line is that the real key to solving future problems is in the “too-big-to-fail” formulation. If I were king of the world I would run around with scissors cutting up anything and everything that looked too big! And this, by the way, is one area where the behavior of social systems connects with the behavior of economic systems. The connection is economies of scale that lead to bigness, and much more often than we give credit to bigness of the too-big-to-fail type. I will not go down that road at this point except to say that while economies of scale results in important economic savings, these come at a social cost that we seldom, if ever, consider. I have traced that effect in some situations that affect developing countries, particularly my own, Mexico. The result are periodic large social swings that impact adversely on the economy and retard overall long term economic growth. This is what I think Neil Ferguson is observing and writing about without fully understanding the dynamic mechanisms that are at work. It also explains Black Swans. In Ferguson I first observed it in his 2006 “The War of the World: Twentieth-Century Conflict and the Descent of the West,” where he comes ever so close to identifying the mechanism and effect of bigness. The effect is also quite visible in his “The Ascent of Money.” Since I said my methodology or model can help understand problems and suggest concrete solutions, let me mention two others I would pursue if I were still king after stepping on so many toes. Regarding economies of scale, I would form a team to explore further how these affect social systems with the concrete objective of developing a methodology to quantify or at least take into account formally the social cost of economies of scale. With respect to Mexico and countries affected with the same problem, I would make my priority, far ahead of any other, the elimination of petty bureaucratic corruption. Earlier I mentioned that to function well a society has to be fairly homogeneous. Petty corruption in many developing countries separates the people into two large social classes, those that participate fully in the more modern and developed economy and society, and those that don’t. It is the tension and occasional flare ups between the two that slow down the longer term development of those societies. Solving the problem of petty corruption is probably going to be much harder than it seems, and may require working together with the sources of the morality and ethics systems of those societies. The reason is that the problem is very deeply ingrained in the norms and traditions of the people at least of Latin America. The way petty corruption works in Mexico can be traced to the mechanism utilized by the Spanish Monarchy for collecting taxes and before that to those of the Catholic Church. And that is only with respect to the more modern Mexico. Before that you also have the huge number of indigenous people with their own norms and traditions. An interesting parenthetical note is that the Church might be much more adaptive than we normally give them credit for. They could even prove a good partner if social scientists and political players only stopped demonizing them. Recently I watched a documentary about life in a village in Oaxaca, Mexico. On Sunday the cameras were unable to attend the Catholic service. The doors were closed to cameras because the particular village church still practiced some form of sacrifice during the service!

Submitted by Xavier L. Simon on
Please note that I continued my note above as a comment to a new report by John Nellis titled "Eight centuries of financial folly and counting," that is at:

Add new comment