Hi all,i am working on a report about limitations of the Black-Scholes approach, and I am currently focusing on the real-world impossibility of continuous reajustments on the replicating portfolio.I mean when a trader sells, for example, a call option he cashes the premium from the buyer. With this amount he creates a portfolio buying delta underlying and putting the extra cash amount on the money market to earn the riskfree rate. This portfolio is reajusted continuously as the option's delta is moving. And this way the option is perfectly replicated: at the end the traders position (= replicating portfolio - option) PnL is zero.As it is not possible to continuously reajust, I am trying to compute the extra cost generated by a discrete reasjustment strategy of the portfolio.For example after selling a 100 days ATM Call option, reajusting once a day. What is the expected extra cost (or loss) associated with this strategy ?(working under the Black-Scholes asumptions so volatility is constant and option's underlying price follows a GBM)I have already set up a Monte Carlo simulation to obtain numerical estimations of this extra cost, and I would like to know if it is possible to obtain thoses results mathematicaly, I mean via stochastic calculus tools.Ideas or references are really welcome. (I guess it is not straightforward to solve).

Eqn (1) from Derman gives an approx. error. There is a fair amount of literature on this topic which googling will find.regards,

BS with transaction cost is a (reasonably) well research subject...

Leland has some paper on this. M