15.11.2014 Views

Risk Management and Value Creation in ... - Arabictrader.com

Risk Management and Value Creation in ... - Arabictrader.com

Risk Management and Value Creation in ... - Arabictrader.com

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

188 RISK MANAGEMENT AND VALUE CREATION IN FINANCIAL INSTITUTIONS<br />

ity of their occurrence. Therefore, VaR has evolved as the st<strong>and</strong>ard methodology<br />

for measur<strong>in</strong>g market risk. It is def<strong>in</strong>ed as the loss to the portfolio<br />

due to an adverse <strong>and</strong> unexpected move <strong>in</strong> one or more market risk factors<br />

that is only exceeded with a given probability α% over a predeterm<strong>in</strong>ed<br />

period of time. In the case of trad<strong>in</strong>g units, the time <strong>in</strong>terval typically chosen<br />

is a one-day (hold<strong>in</strong>g) period. VaR is therefore the α-quantile of the cumulative<br />

probability distribution of the value changes <strong>in</strong> the portfolio over the<br />

measurement period H, <strong>and</strong> α is typically 2.5% or 1%. Hence, there is a<br />

(1 – α)% probability that the critical threshold loss will not be exceeded. 242<br />

Mathematically, we can express the probability that a loss will exceed<br />

VaR over a predeterm<strong>in</strong>ed period of time <strong>and</strong> that this will occur less than<br />

α% of the time as:<br />

p( [ ∆VH – E( RH) ]> VaRH) ≤ α% ⇔ f ( RH) dRH<br />

≤ α % (5.27)<br />

where p = Probability<br />

∆V H<br />

= Change <strong>in</strong> the portfolio value V over period H, which<br />

equals R H<br />

, the return of the portfolio over the same<br />

time horizon<br />

E(R H<br />

) = Expected (or mean) return of the portfolio over time<br />

horizon H<br />

VaR H<br />

= <strong>Value</strong> at risk for period H, which is a negative number<br />

α = Confidence level for not exceed<strong>in</strong>g threshold VaR H<br />

f(R H<br />

) = Assumed distribution of the portfolio returns over time<br />

horizon H<br />

Note that this def<strong>in</strong>ition so far does not make any assumptions about<br />

the distribution function of the value changes. We assume now that the (daily)<br />

returns are (as mostly assumed <strong>in</strong> modern f<strong>in</strong>ance theory <strong>and</strong> especially for<br />

tradable market <strong>in</strong>struments) normally distributed 243 with mean return µ<br />

<strong>and</strong> st<strong>and</strong>ard deviation of the return σ R<br />

, that is f(R) ~ N(µ; σ R<br />

). We can then<br />

rearrange terms 244 <strong>and</strong> eventually derive the dollar amount for VaR as: 245<br />

VaR = ([Φ -1 (1 – α)⋅σ R<br />

]–µ) ⋅ V = (c⋅σ R<br />

– µ) ⋅ V (5.28)<br />

where Φ -1 = Inverse st<strong>and</strong>ard normal cumulative density function<br />

VaR<br />

∫<br />

– ∞<br />

242 See Stulz (2000), pp. 4-9–4-10, Hirschbeck (1998), p. 143, Dowd (1998), p. 41,<br />

Jorion (1997), Guldimann et al. (1995).<br />

243 See Dowd (1998), p. 42.<br />

244 By us<strong>in</strong>g the property that we can transform R to a st<strong>and</strong>ard normal variable by<br />

the follow<strong>in</strong>g transformation [(R–µ)/σ].<br />

245 See Hirschbeck (1998), p. 154.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!