Maybe you could publish the specific code that you use to generate the brownian data and to price the corresponding options and someone might help you find a bug. If everything was correct in your model, then the black-scholes formula would calculate fair (or risk-neutral) price of options, which can't be exploited in any way, the same thing as you can't exploit perfect brownian motion stock prices. On top of that you can apply any kind of risk management, money management, whatever, it can't change the final outcome, which should always be very close 0. For example the famous Kelly for optimal bet size only works if you have an edge in your game. Just to be clear, I am not saying your system can't work against the real data. The real market data can for give you an edge that we are all looking for. The black swan events that happen in real life are just an example of real market diverting from perfect random walk. But if you claim your system profits from randomly generated data, that can only happen due to a flaw in your data, or bug in your code or due to curve fitting when not enough testing data is applied.
I don't know what you think of GBM-generated data or what your reservation is against it, but it is not some wild random numbers; it is much like real stock prices, just simulated, ie. generated. So, for me and the system GBM reflects real stock prices. Otherwise it wouldn't make any sense. There are many academic research papers working exactly with the same tool, ie. GBM. GBM data can give the same edge like real market data as there is principially no difference. And: black swan events aren't that important for this. It is also a two-edged sword: it can also be an advantage for the system as it uses both Calls and Puts... In a previous posting I already mentioned: the generated data can afterwards be verified by calculating the observed volatility (ie. the stddev). It has to match the input parameter historical volatility. And guess what: it does. Such tests have already been done long ago here; the generator works correct. Here's also a good article on GBM with source code: https://mhittesdorf.wordpress.com/2...-asset-prices-with-geometric-brownian-motion/ My system uses a similar approach using "Leptokurtic Model of Asset Returns" by using the t-distribution. And here's also a 276-page book or paper titled "Monte Carlo Simulation With Java and C++" http://www.javaquant.net/books/MCBook-1.2.pdf
It's impossible to simulate real market data correctly without knowing all of at least the most important algos responsible for determining the major turns. Hopefully this will make sense and you will use it to go on to build something great based on real data, because from my pov debating the validity of this would be absurd. Not just u botpro, but anyone who can't immediately recognize the difference should be rethinking just exactly what it is that creates real data. I assure you there's no number generator that can simulate real data without knowing the real forces driving price. Anyone who has every even glimpsed the algos in action in the market will agree with this. SMH.
To anyone who comes across this thread, please do NOT give 'botpro' your money. IMO you'd be better off taking it to the craps tables. (Assuming 'botpro' isn't just a troll, which wouldn't surprise me.) 'botpro', IMO, you are on the wrong track, theoretically and practically. Theoretically, because you seem to think you've designed a "perpetual motion machine" insofar as you're simulating data and the strategy simultaneously (and like any perpetual motion machine, it will fall straight out of the sky once it gets out into the real world); practically, because there are lots of "gotchas" when it comes to a real-world market like options (fragmentation, internalization, dividends, trade busts, ......). IMO, developing a strategy based on "simulated data" is a total waste of time; maybe some academics can get away with a trick like that, but if you try it in the real world, you're probably just going to throw money into a fiery furnace. You clearly haven't done your homework yet. If you want to blow out an account with your "system", please do it with your own money rather than someone else's.
I think you are in the wrong thread here. This is a commerical thread, I pay for this, so please refrain. Ok, you have made your warning, and it will stay here for all to judge. But such unsubstantiated accusations are not welcome, please respect this.
Algos create the real data. Without knowing those algos u can NOT simulate real data. When u say your simulated data is not the wild random numbers, but rather it looks just like real data because it's continuous it's still random data from the pov it's not driven by the algos that create the real data. If u don't get what a huge difference that makes u will literally need luck because otherwise u will spend the rest of your life without any chance of ever finding what drives price.
The algo has already been posted here many times. It is the same stochastic differential equation as on this wiki page: https://en.wikipedia.org/wiki/Geometric_Brownian_motion It is about modelling the reality, not creating the reality.