Thanks for your theory contribution, and I want to know how to extend this excellent model to couples of alpha not only just one alpha like book imbalance
Please do more videos like this! I am interested in your paper High-frequency trading in a limit order book (2008) and how that can be implemented in code. Would love to see a video like that :) Cheers!
@@polmaksim avellaneda is the name! We are running the first edition of the Hummingbot BotCamp now and one of the students wants to implement the micro price. I think that could be and easy task because we keep a local copy of the ob updated by websockets and we have methods to query them in cython so it’s very fast!
@@cardosofede Thank you for the answer. I am using Hummingbot and experimenting with strategies. Avellaneda is one of my favorite, but was not looking under the hood yet. 🙂
Hi Sasha, very nice work! One quick question, one potential issue of the estimation is that it is relying on at least one day of data to estimate. Say I deployed an alpha which consumes the microprice, how do you bootstrap the G1, Q, and R matrices? Are you just using the value obtained by using the previous days?
Thank you for the video! Since micro-price lives between mid-price and weighted mid-price, I wonder if it will pass to just approximate the micro-price by taking the average of mid-price and weighted mid-price?
This will depend on the asset you’re looking at, sometimes it will be closer to mid price (for small tick stocks) sometimes closer to weighted mid (for large tick stocks), sometimes right in the middle.
@@sstoikovwould it be effective to define a weighted average of the mid price and weighted mid price with the weights depending on the stock’s tick size?
Please do more videos. It is so hard to find cool stuff like this amid all the junk in you-tube.
Actively using your reservation / fair price theory on cross-exchange market making, thank you for the contribution.
Great to hear! I have a new paper called “the microstructure of cointegrated assets”. Might be useful for you!
Hi Paul! I trade cross exchange marker making with hummingbot! Would you mind explaining me how you use the fair price in xemm?
wow marvellous video. many many thanks!
Many thanks Sasha!! great video!!..
Awesome Sasha!
Thanks for your theory contribution, and I want to know how to extend this excellent model to couples of alpha not only just one alpha like book imbalance
Very inspiring and informative! Do you think a similar idea can be applied to trade on OTC markets?
Please do more videos like this! I am interested in your paper High-frequency trading in a limit order book (2008) and how that can be implemented in code. Would love to see a video like that :) Cheers!
I am also working on that so feel free to reach out
we have this strategy implemented in Hummingbot!
@@cardosofede Could you please tell, what is the name of this strategy in Hummingbot? Thank you.
@@polmaksim avellaneda is the name! We are running the first edition of the Hummingbot BotCamp now and one of the students wants to implement the micro price. I think that could be and easy task because we keep a local copy of the ob updated by websockets and we have methods to query them in cython so it’s very fast!
@@cardosofede Thank you for the answer. I am using Hummingbot and experimenting with strategies. Avellaneda is one of my favorite, but was not looking under the hood yet. 🙂
Hi Sasha, very nice work! One quick question, one potential issue of the estimation is that it is relying on at least one day of data to estimate. Say I deployed an alpha which consumes the microprice, how do you bootstrap the G1, Q, and R matrices? Are you just using the value obtained by using the previous days?
Thank you for the video! Since micro-price lives between mid-price and weighted mid-price, I wonder if it will pass to just approximate the micro-price by taking the average of mid-price and weighted mid-price?
This will depend on the asset you’re looking at, sometimes it will be closer to mid price (for small tick stocks) sometimes closer to weighted mid (for large tick stocks), sometimes right in the middle.
@@sstoikovwould it be effective to define a weighted average of the mid price and weighted mid price with the weights depending on the stock’s tick size?
@@superman39756 great idea - if you have some good data maybe we could work on it together
Hft neuromarket-making (liquid stones)