On Volatility versus Risk I commented on Risk and Volatility, without
however mentioning practical activities related to the direct use of metrics
for risk.
Market risk, which is the possibility of financial loss due
to the typical ups and downs of the market, should, as the intuition suggests,
reflect the degree of oscillation that the asset provides.
The simplest metric, known and used for this purpose is the
standard deviation. The standard deviation has inaccuracies in quantifying
the risk, because strictly financial returns do not follow a normal
distribution, which is the necessary premise to calculate the standard deviation. The
metric of the standard deviation necessarily produce a bias, always deviated
from the true value.
The biggest problem with the standard deviation, however, is
not the premise that it performs when considering the normal data. The standard deviation by itself does not present the risk
management of an efficient way, because it still lacks the statistical scenario
was considered.
Value-at-Risk
The methodology Value-at-Risk (VaR) fills this gap. It
was developed in order to produce management reports more objective and
intuitive. His metric is well established in statistical terms and
reflects exactly what one would expect from a market risk measure. The VaR
results in a net asset value, or a possibility of financial loss. This
possibility of financial loss is tied to a time interval and statistical
significance. So we say: The portfolio has the possibility of losing $X on
Y days (or hours) with a statistical significance of Z%.
Financial processes, however, do not follow a behavior said
homoscedastic or constant over time but heteroscedastic. The volatility
can undergo dramatic changes as in times of crisis. These phenomena can cause a sharp drop in equity of a
particular institution leading to bankruptcy in many cases.
Stress
An analysis complementary to VaR, called Stress analysis is
used to simulate situations of panic and financial crises. These
simulations analyze the tails of the distributions of returns and evaluate the
portfolio in more extreme scenarios. It is possible to simulate historical
crises, specific situations and scenarios to highly pessimistic and
uncorrelated.
It is necessary to understand the main feature of the market
is the strong stochastic process which asset prices are submitted. Unexpected situations are possible and even expected, and it
is impossible to predict turbulence and crisis situations with sufficient
precision. The Financial Risk Management, however, has enough tools to
evaluate and quantify the risks and fluctuations to which the portfolio is
submitted.
