Although economic and financial projections do not use the same level of abstruse mathematical techniques as, for example, thermodynamics or particle physics do, it does not make economic forecasting any simpler.
While the quantitative methods used in the physical sciences and engineering disciplines are complex, physical objects or sub-atomic particles in motion obey classical mechanics or relativistic physics.
In contrast, economic projections are more prone to error because human beings do not necessarily follow predictable patterns of individual or group behaviour.
It is evident from the on-going credit crunch and economic slowdown that CEOs of long-standing, globally significant companies and once highly regarded regulators in developed countries failed to address some obviously unsustainable risks.
For instance, they were lulled into "missing the wood for the trees" or they decided, for reasons of personal gain, to insist that returns in financial markets can far outstrip the rates of overall economic growth for indefinitely long periods of time.
Pension assets invested for the long term in stock markets have plunged in value, and in the developed West, hapless taxpayers are paying for financial sector and other bailouts.
In an 1820 book titled Elements of the Philosophy of Right, Hegel wrote: "Political economy is the science [emphasis added] which must go on to explain mass relationships and mass movements in their qualitative and quantitative determinacy and complexity."
In the 19th century thinkers such as Malthus and Ricardo felt they could establish political economy as a "science" which would be as exact as the physical sciences. Advances in the physical sciences and engineering innovations have had a profound impact by raising standards of living. It is likely, therefore, that there was a tendency to incorporate quantitative methodologies in economics to build it as a "science".
The mathematical techniques used in economics mostly involve calculus, matrix algebra and statistics. Financial engineering adds stochastic calculus and modelling derived from Brownian motion to this mix of techniques.
A major leap of faith in finance is that the movements of stock or bond prices can be modelled as Markov processes with an underlying assumption that markets are "efficient".
Stock and bond price distributions are invariably assumed to be Gaussian with constant mean and standard deviation, and historical price volatilities are used to price derivatives. This is done by arriving at closed-form solutions to differential equations as in the Black-Scholes option pricing model or by running Monte-Carlo simulations.
At the same time, scenario analysis can provide a sense of the dispersion of possible outcomes by using several sets of volatilities and assuming fat tail distributions. It follows that the use of mathematical methods in economics and finance has incalculably enriched our understanding and also improved decision-making. However, this has also perhaps given us a sense of false precision.
In the context of the alleged damage done by dependence on quantitative methods, Nassim Taleb and Pablo Triana sarcastically said in a Financial Times article that "when you see a quantitative expert shout for help, call for his disgrace." It was probably not Taleb and Triana's intention to only blame experts.
All major market and economic decisions in large private firms or governments are invariably taken by relatively generalist top managers. It is these managers who should have provisioned for the high levels of risk they were taking by setting aside adequate capital.
Regulators are usually well-qualified professionals and it was their responsibility to verify that firms had provided for adequate risk capital. All things considered, readers of this column should not be concerned if they are unfamiliar with stochastic calculus.
The simple point is that modelling is useful only if it is complemented with an understanding that in economics and finance there are unknowable unknowns, eg the underlying volatilities can change considerably over time and correlations across economic indicators move sharply and unpredictably.
In the last two weeks we have witnessed this incredible $50 billion "Ponzi" scheme fraud perpetrated by Bernard Madoff, a former head of Nasdaq.
Ponzi was an Italian who emigrated to the US in 1903 and had concocted this ingeniously simple scheme to make payments to earlier investors by using later investors' money. Obviously, "quants" were not to blame. It is yet another example of financial sector regulators failing to protect the interests of investors.
In India, it has been a story of remarkable ups and downs in the last few years. Economic growth was buoyant and capital inflows were burgeoning over 2007-2008. Some were convinced that stock markets would continue to boom, capital inflows would keep growing and the Indian rupee (INR) should be allowed to appreciate to INR 35 to a dollar, or higher.
A contrary view was that the INR exchange rate needed to be managed and faster utilisation of forex reserves was constrained by India's supply side limitations. With the benefit of hindsight it is clear that: (a) INR exchange rate policies were broadly correct; (b) the last few repo interest rate hikes were unnecessary; (c) the cautious movement towards capital account convertibility and opening up of the financial sector was justified; and (d) the short-term growth versus inflation trade-off has shifted decisively towards the need to push for growth.
As we have seen time and time again, it is difficult to make accurate long-term predictions about international capital flows, commodity prices (Goldman Sachs had projected oil prices at above $200 per barrel by December 2008), interest/exchange rates and equity market movements.
Therefore, it would be useful to be less doctrinaire about Indian monetary and exchange rate policies and stimulation packages in the face of unprecedented events in the global economy.
At the same time, for better-informed decision-making, we need to collate national statistics on a more systematically comprehensive and frequent basis. Of course, the National Sample Survey Organisation (NSSO) conducts periodic surveys and the Central Statistical Organisation is a source of detailed numbers.
As of now, however, these public sector institutions are not adequately funded or staffed. To summarise, we need to: (a) make pragmatic and step-by-step course corrections in our economic and regulatory policies in the near term as we adjust to domestic and international developments; and (b) on a longer-term basis, increase the frequency with which sample surveys are conducted and facilitate easier access to updated data.
The author is Ambassador of India to Belgium, Luxembourg and the European Union. Views expressed are personal.