量化投资模型——高频交易做市模型相关（Avellaneda ＆ Stoikov’s）研究解读＆代码资源

**

写在前面，强烈需要同好来交流，如果你也很感兴趣，并且想一起做点什么，请通过我的邮箱来联系博主！

**
A comprehensive guide to Avellaneda & Stoikov’s market-making strategy
AS模型其实有一个前提假设不是很成立，就是说就是as模型那个强假设，假设股票是没有趋势的，因为它的一个股价的变动是符合sigma的，没有drift。
拿来做数字货币的：
https://github.com/camandrewz/CamCapital
Resources Used:

High-frequency trading in a limit order book (Marco Avellaneda & Sasha Stoikov)
(https://people.orie.cornell.edu/sfs33/LimitOrderBook.pdf)
Optimal High-Frequency Market Making (Takahiro Fushimi, Christian Gonz ́alez Rojas, & Molly Herman)
(http://stanford.edu/class/msande448/2018/Final/Reports/gr5.pdf)
https://www.reddit.com/r/highfreqtrading
https://www.reddit.com/r/algotrading
https://docs.ftx.us
https://fix8.org

https://github.com/hello2all/gamma-ray
Gamma Ray is a high frequnecy market maker trading bot for crypto currency markets written in C++. It is able to consistantly update maker orders under 100 micro seconds.

https://github.com/im1235/ISAC
Optimal control of risk aversion in Avellaneda Stoikov high frequency market making model with Soft Actor Critic reinforcement learning

https://github.com/ondra-novak/mmbot
mmbot
Market Making trading bot for cryptomarkets

https://github.com/sstoikov/microprice
An estimator of the fair price, given the state of the order book.
https://github.com/grayvalley/microprice-calibration
The Micro-Price (Stoikov, 2017) is a high frequency estimator of future prices. (See: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2970694)

此外，关于一个在做市策略里面常用的一个指标，到达率，笔者和小伙伴讨论了一个关于A股上的情况，如下图所示，横轴是挂单价和中间价的距离，y轴是10分钟内击穿的概率，可以看到A股基本上3，4个跳价的到达率就很低了；从目前来看，我觉得逐笔没那么重要了lol 之前还在想a股会hit很深，要用千档，说到底，就是大单很少。a股用算法拆单很熟练了，除了散户很少一个大单把价格hit很深。可能在特定的标的work，比如散户参与多的资产上面