FNCE30007 Lecture Notes - Lecture 6: Elwing, Fumi-E, Louisiana State University

• School: University of Melbourne
• Department: Finance
• Course: FNCE30007
• Professor: Neal Galpin

Document Summary

Lawoflargenumbers sampling error central limit theorem stole where r ni the randomvariable stock pricepaths lymeration. Then discount at mile free rate lymerate random variables nl rt 02t therandomvariableslike compoundwittant ratestofind st slit. Montecarlovaluationcomputes an estimateofthepricebyusing togenerate netofpossiblefuturevaluesofunderlying asset st fromwhich the derivative"spayoff xi is computedforeachsm. tlxmibmd risk neutraldistributionofunderlying asset f which are thendiscounted wrongexpert daveragedtoobtainan estimateoftheprice simulation lsmt t. It rt allows you to time the maximum 1 minimum of astockpumice afloatinglookbackputpaysoff masdsmasost 03 5 man st wont care scenario s man occurs at contract end hayaffeo. Depends on n changing n will give a differentestimate. With an infinitenumberof draws the montecarloestimateequalsthe trueprice needmany stockfumiepathsmuchthatthesampleapproximates the true distributionof st well. Howmanypaths are needed depends on theaffiliation approximating stock pricedistribution m enlsa t t r o"s et. Samplingerrod if we fixthenumberofpathsdrepeatthemcsimulation we would generate a different netofstockprices therefore adifferentoptionprice estimate differentestimates sampling error rangeofestimates variability.