Why Forecasts Don't Work in Finance

Last Updated Sep 16, 2008 12:06 PM EDT

raincloud1.jpg
"Finance is an area that's dominated by rare events. The tools we have in quantitative finance do not work in what I call the 'Black Swan' domain," said author and former options trader Nassim Nicholas Taleb earlier this year.

Events in the financial markets have proven Taleb right -- at Lehman Brothers and Merrill Lynch, and from the looks of things, insurance giant AIG.

The problem is that predictive tools in the bankers' and insurers' arsenal are unreliable because they calculate the probability of future occurrences by referring to past events.

"Probability and forecasting can let you down when you most need it," observes Rafael Ramirez, fellow in strategic management at Said Business School in Oxford and co-editor of Business Planning in Turbulent Times. "People believe they can predict the future by calculating probabilities on what's gone before. But this assumes events fit a pattern. When turbulence sets in, as now, pattern is irrelevant."

Ramirez is an advocate of scenario planning, which considers a number of alternative future outcomes and is based solely on now and tomorrow. Scenario planning was most famously practised at oil giant Shell, where Arie de Geus and his colleagues applied what were originally military planning techniques to the business.

Nobody's saying it would've saved Lehman Brothers, but it's a good fit for today's ambiguous business environment. Ramirez explains:

Using plausibility, not probability

Scenario planning is what might be -- we make the implausible plausible. It asks: under what context could an activity become unviable? ood scenario work is only just plausible -- if you start by offering decision-makers probability, they will assume there's a solution and won't explore possible unexpected outcomes.
Outside perspectives
At Lehman, it was observed that "a boss less intimately wedded to his business -- Mr Fuld was the longest-serving chief executive on Wall Street -- may have seent he problems earlier and acted faster".
Fuld was so embedded in the Lehman world view, he couldn't see the alternative context in which the set of assets he'd built could become a set of liabilities.

This applies to other businesses -- your business strength may once have been your corporate culture, but it could've become a liability at some point. You need people outside of the business to point this sort of thing out.

Collaboration
The notion that markets will look after themselves doesn't hold up anymore -- taxpayers, regulators and businesses will need to work together.

There needs to be a minimum of collaboration among businesses within an industry. Look at football: if it were really just competition, Manchester United and Chelsea players would just shoot each other. They collaborate to establish a framework of rules and standards that make it possible to compete.

Scenarios help to re-establish the collaboration that makes competition viable. The (late) collaboration among bankers, regulators and central banks attests to this. Scenarios might have helped this to happen earlier.

Conversation enabler
Scenarios help to raise assumptions on which decision-making is based, then question them -- "what if?", "and then?", "what next". Keith Grint talks about how John F Kennedy did this to the military during the Cuban missile crisis. It's an enabler of strategic conversations.

When you create scenarios with others, they are not a final product, but invite further insights.

Succession planning
One bank recently wanted its next generation of leaders to present scenarios to current ones. This way you have an inter-generational conversation so the younger leaders own the scenarios they'll inherit, and they have more 'skin in the game'.

(Image: Dullhunk, CC2.0 )