Financial markets, like other
markets, are where supply and demand meet and price discovery takes place. And
when financial asset prices are discovered, the market fulfills another
important function - it prices risk. A price, in turn, can be interpreted as an
implied probability-weighted average of all possible outcomes.
It is intuitive, therefore, to
approach fundamental valuation by running this logic in reverse; identify
possible outcomes, determine their impact and estimate their probabilities. The
sum product of all of them represents a 'fair' valuation.
Risk and uncertainty
A problem arises when probabilities
cannot be estimated and instead have to be guessed.
When you can make estimates, you are
dealing with risk. When you are best-guessing, you are dealing with
uncertainty.
The concepts of risk and uncertainty
are quite obviously not exclusive to financial markets. Cambridge University
has profiled a series of papers of risk and uncertainty more broadly. Their
Professor David Spiegelhalter said "Making important decisions in the face
of uncertainty is unsettling and difficult."
Needless to say,
"unsettling" is not a pleasant feeling in financial markets. There
are a number of examples.
Elections are part risk and part
uncertainty. We can use opinion polls and other inputs to try to assign
probabilities to various outcomes, but uncertainty remains, well, a certainty.
How likely is a gaffe, revelation or external event? Will the politicians
actually do what they've promised, and if not, then what instead?
As the number of moving parts
increases, so does the difficulty in estimating with confidence and so
therefore does uncertainty. Political outcomes are normally harder to predict
when different sides are forced to negotiate; between different parties or differently
controlled different chambers (as in the US) or to form a government at all (as
in most European countries). How will negotiations shape up and what if they
don't produce a budget (as in the US) or a fragile coalition government (as in
Italy) or no government at all (as in Belgium not so long ago)?
A sting in the tail
Tail 'risks' are a special case.
Though they are called 'risks', in some cases they are in fact uncertainties.
With the US government shutdown and an approaching debt ceiling, what is the
probability of the US defaulting, and how severe would be the consequences? The
answer to the latter is likely "massive" and, largely as a result,
the answer to the former is probably "tiny". But what do you get when
you multiply "massive" by "tiny"? Using the
'probability-weighted scenarios' approach above, both the probabilities and the
outcomes somehow have to be quantified. And with as many moving parts as either
a nation's economy or its government have, such estimation is almost impossible
to do with any real confidence.
It's not just politics. Another hot
topic at various junctures has been terrorism. What is the probability of a
terrorist attack, and (aside from the human consequences) what would be the
extent of the economic or market impact? Historical data are useless if you
think there has been a structural shift in the world (as after 9/11).
What about natural disasters? Logic
dictates than almost all financial markets should price in some chance (however
minute) of the 'big one' hitting Tokyo or Los Angeles. When, where exactly, how
big and how markets would respond in each case, is anyone's guess.
Investor decision making in reality
In practice, uncertainty tends to be
viewed in a binary fashion, with market participants considering the extreme
scenarios and looking at limit-case payoffs in each.
Tail 'risks' are often priced
according to sentiment. For example, when sentiment is positive, a
low-probability, high impact negative outcome might be treated as negligible,
while when sentiment is bad, its probability might be treated as implausibly
high. This can be interpreted as a very significant 'risk' premium, due to the
probability of the outcome occurring being unknown or unquantifiable.
Of course, while implying probabilities
from prices is useful analytically, it should not be assumed that pricing
reflects pure expectations. Aside from the treatment of tail risks mentioned
above, there are many other considerations such as carry, positioning, flow,
second-guessing of central banks and other policymakers and so on.
And moreover, there is more to
consider in markets than valuation alone.
Where the risk vs uncertainty
distinction becomes crucial
To reiterate: When you can make
estimates, you are dealing with risk. When you are best guessing, you are
dealing with uncertainty.
You may find that you're asked to
discuss the implications of an election, a US government shutdown, the future
of the Euro, a natural disaster or a terrorist attack investor decision making,
or simply answer a 'logic' question relating to uncertain outcomes. Being able
to call upon this type of analysis this in your investment banking interview
(especially for a financial markets position) should help you considerably.
While you don't need to be able to repeat the entire discussion, understanding
and articulating these sorts of concepts is exactly what you should be looking
to do.
