Hmmm... How sophisticated does it get? My understanding is that it's very much agent-based. But what scenarios do you test? What is the objective? How expensive is it it build?
The simplest but optimistic approach is to assume the market doesn't react to your orders. Depending on the firm's infrastructure and how accurate they want the simulations to be it could involve a team of people focusing on it or something much more basic.
Having a good sim setup is incredibly important, at least for the style of trading my firm does. To put it bluntly, how else do you get an idea if a strategy is viable (net of spread costs, liquidity constrains, etc) without it? Do you just raw dog it and throw it straight out to prod? You ability to evaluate any sort of strat is only as good as your sim setup, and in that regard it is one of the most critically important items, even before alpha. Put it this way: when we do blue sky research, before we can even think about developing alphas, we need a way to see if an alphas is, well, an actual alpha. Without a solid sim platform, this is impossible.
"one for the money two for the better green 3 4-methylenedioxymethamphetamine" - M.F. Doom
Tenetur odit blanditiis aliquam exercitationem sunt eaque dolor. Eligendi voluptatibus velit quaerat sint nam delectus sunt aliquam. Fugit cum neque est qui minima animi unde. Ea odio eligendi et natus.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
Sorry, you need to login or sign up in order to vote. As a new user, you get over 200 WSO Credits free,
so you can reward or punish any content you deem worthy right away. See you on the other side!
In dark or illiquid products they are not very important if done at all while some other traders rely heavily on them especially in liquid products.
Hmmm... How sophisticated does it get? My understanding is that it's very much agent-based. But what scenarios do you test? What is the objective? How expensive is it it build?
The simplest but optimistic approach is to assume the market doesn't react to your orders. Depending on the firm's infrastructure and how accurate they want the simulations to be it could involve a team of people focusing on it or something much more basic.
What value do simulations actually bring?
Does it help reduce cost? Does it help actually generate returns?
Having a good sim setup is incredibly important, at least for the style of trading my firm does. To put it bluntly, how else do you get an idea if a strategy is viable (net of spread costs, liquidity constrains, etc) without it? Do you just raw dog it and throw it straight out to prod? You ability to evaluate any sort of strat is only as good as your sim setup, and in that regard it is one of the most critically important items, even before alpha. Put it this way: when we do blue sky research, before we can even think about developing alphas, we need a way to see if an alphas is, well, an actual alpha. Without a solid sim platform, this is impossible.
Interesting. Thank you.
What would you say are the most important features of a simulation?
Like if the simulation had these features, I'd consider using it kinda deal.
What is the difference between a simulation and a backtest with t-costs, etc? I'm assuming a simulation involves more market microstructure data?
Tenetur odit blanditiis aliquam exercitationem sunt eaque dolor. Eligendi voluptatibus velit quaerat sint nam delectus sunt aliquam. Fugit cum neque est qui minima animi unde. Ea odio eligendi et natus.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...