Research On Application Of Algorithmic Trading Based On Risk Parity

In the a.ctual tra.ding process,many institutional investors need to trade large orders.During the order process.considering market,liquidity and market shock costs.invest.ors began to use algorithmic trading to hide t.heir trading processes and dismantle large orders.The order is placed in small orders churing a period.Algorithmic trading was first proposed in the 1980s.After decades of develop-ment.it is mainly divided into three generations.The first generation algorithmic trading strategy considers miniunizing market impact costs through historical in-formation.The most typical is the Volume Weighted Average Price(VWAP)algorithm.This type of algorithm is now widely used in the market.The second-generation algorithm trading strategy is represented by the implementation of the implementation of the Shortfall Algorithm(IS).which introduces the time impact cost of transactions on the basis of the first-generation algorithm.The third-generation algorithmic trading strategy only deepens and improves on the basis of the second-generation algorithm,considering dynamic adjustment strate-gies.The current scholars mainly study two directions,one is how to optimize the volume predietion model in the VWAP algorithm,and the other direction is how to optimize the price change model int the IS algorithm.But whether it is a traditional model or an optimized model,in the actual application process.it is easy to canse excessive losses in some extreme cases.Therefore.in this artiecle.we consider ihtroducing a new risk characterization model to reuce this loss.There are currently three types of evaluation eriteria that are popular.One is optimization expectations.The first-generation algorithmic trading strategies mainly use this evaluation criterion to characterize the market impact cost of the strategy.The second is to use the mean-variance or utility function.The second-generation algorithmic trading strategy mainly uses this evaluation cri-terion.This type of risk characterization model can take into account market shock costs and time shock costs,but it is easily affected by extreme risks.The third is the risk parity criterion,which aims to solve the equity risk concentration problem in a standard balanced portfolio like 60/40.to ensure that the risk of each asset is completely average,and no asset class can dominate the volatility of the portfolio.This paper considers introducing the risk parity model into the IS strategy,which can disperse the time shock cost in the transaction process and reduce the impact of extreme situations on transaction costs.In order to verify the effectiveness of the IS strategy under the risk parity model,we also compare the model under the risk parity with the traditional VWAP model and IS model to verify the effectiveness of the model.We mainly study the application of the model in the Chinese market.Taking into account the liquidity of the market,we take stock index data of the Shanghai 50ETF.We take the minute-level data from March 2016 to September 2019 for verification.Through inspection and analysis,we first obtained that the VWAP algorithm is a special type of IS algorithm.Second,by comparing the returns of different strategies in extreme situations,we find that IS strategies under the risk parity model can more effectively reduce transaction costs.By comparing the return and return variance of the IS algorithm under risk parity with other algorithm transactions from 2017 to 2019,we found that.the IS strategy under the risk parity model is less risky than the other strategies when the returns are almost consistent.Many show that the risk parity model can effectively reduce the time shock cost of transactions.