Beruflich Dokumente
Kultur Dokumente
23
27
31
SR =
r rf
In other words, Sharpe ratio can be interpreted as the excess return per unit volatility
(risk). Under the same amount of risk, a higher excess return is more preferable. In
other words, we prefer to have a higher Sharpe ratio. If we assume a linear
relationship between return and risk, Sharpe ratio is simply the gradient of the Return
vs. Risk graph, where risk here is characterized by volatility:
VaR r z ,
Where Z~N(0,1), r is the mean return and is the return volatility.
CVaR r
e 2
2
Example 2 (E) Based on the closing prices of MTR Corporation (0066) in 2013, find
the values of the abovementioned risk measures on returns (use Hang Seng Index as
the benchmark if necessary.): (a) Volatility, (b) Sharpe Ratio (let , (c) 5% Value at
Risk (VaR0.05), (d) 1% Conditional Value at Risk (CVaR0.01).
Example 3 (E) Based on the simulation in Chapter 5 Example 4, find the values of
the above risk measure on returns: (a) Volatility, (b) Sharpe Ratio (let rf =2.3%), (c)
1% Value at Risk (VaR0.01), (d) 2.5% Conditional Value at Risk (CVaR0.025).
6.7 Durations & Modified Duration
Unlike the previous risk measures, duration is a risk measure exclusively for fixed
income securities such as bond.
The timing of future payments is important in fixed income securities. In this
section we will be discussing some risk measures concerning the timing of the future
payments.
A. Duration
Let R1, R2,..,Rn be a series of payments made at the times 1,2,.,n. Then the
tv R
t
t 1
n
v R
t
t 1
Noted that:
(1) It can be proved (by differentiation) that the duration d is a decreasing function
of i. Therefore, duration decreases as interest rate increases.
(2) If there is only one future payment, then d is the point in time at which that
payment is made.
B. Modified Duration
We now consider the relationship between the rate of change in the present value of a
series of future payments as the rate of interest changes. Let this present value be
denoted by P(i), i.e.
n
t 1
t 1
P(i) vt Rt (1 i) t Rt
P ' (i )
[P (i i ) P (i )] / i
v
P( i )
P( i )
where i is the change in interest rate.
The term P(i) measures the rate of change in the present value of the payments as the
rate of interest changes. Dividing by P(i) express this instantaneous rate of change
in units independent of the size of the present value itself. The minus sign is
necessary to make v positive, since P(i) is negative.
d
or d (1 i)v , i.e. modified duration is equal to
1 i
duration divided by 1 + i. Since i here is per period interest rate, we have a very
d v
if interest is compounded continuously!
The normal probability plot is a graphical technique for normality testing which
assesses whether or not a data set is approximately normally distributed.
We first rank the data, say x1 < x2 <....<xn (i.e. xi is the ith ranked value). Then the data
are plotted against a theoretical normal distribution (i.e. we generate the corresponding
y-values, y1 < y 2 <....< yn such that Y~N(,2)) in such a way that the points should
form an approximate straight line. Departures from this straight line indicate departures
from normality. We can also plot two lines: (1) X versus Z and (2) Y versus Z. Since
Y~N(,2) and Z~N(0,1), the second one is a straight line. Again, it indicates
departures from normality if the first line departs from the second one.
Example 5(E) Based on the closing prices of Hong Kong Stock Exchange (0388) in
2015, construct two normal probability plots to assess, by visual inspection, whether
the returns are normally distributed.