Algorithmic Trading and its Impact on the Market

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume:09Issue: 08| Aug 2022 www.irjet.net p-ISSN:2395-0072

Algorithmic Trading and its Impact on the Market

Ashutosh Agarwal1 , Dr. Padmashree T2

Department of Information Science, RV College Of Engineering, Bangalore-560059 ***

Abstract Institutional investors currently use algorithmic trading as one of the most popular and developing trading strategies on the Indian stock market. It is a form of trading where systems are programmed with predetermined rules and instructions to execute transactions at a high rate of speed and accuracy that is hard for human traders to achieve manually. Many retail traders and market regulators have opposed algorithmic trading because of its quick execution. Although algorithmic trading has taken a battering for unintended volatility and blocking market quality by adding large volume at specific levels in accordance with their system or strategy, the evidence relevant to its drawbacks has not yet been established. This paper moves in the right way by supporting algorithmic trading and giving it the credit it deserves for raising market quality. This study analyses the National Stock Exchange's (NSE) stock market to directly identify algorithmic trading. It then aims to identify the primary benefits of algorithmic trading and express the rationale behind its expansion not just in India but alsoonthe international market.

Keywords Algorithmic trading, Trading Strategies, Computer based trading, Institutional investors.

I. INTRODUCTION

Therulesofconventionalbrokinghavechangedasa result of algorithmic trading. Trading software must be properly understood bytradersinorder forthem to implement and backtest their strategies while also remaining competitive on exchanges where large volumes are being transacted usingcomplexalgorithms.Utilizingcutting-edgetechnology forstockexchangetradingisknownasalgorithmictrading. Trading entails creating pre-established rules in order to executethedealtoreapresultsquickly,somethingahuman tradercouldnotdomanually.

Because there is little need for a human trader to be involved in thiskind ofsystem, decisionsaremade quickly andprecisely.Becauseofthis,thealgorithmisabletoprofit from any market chances well before a trader can even recognise them. It is merely a technique for lowering the price, impact on the market, and risk of slippages. Pension funds, investment banks, hedge funds, and mutual funds frequently use it because these institutional traders must place big volume orders in markets where the size cannot be supported in its entirety at once and must be divided

into smaller portions. Over the past few of years, its popularityhasgrowngradually.

Thetechnicalrequirementsforalgorithmictradingareas follows

• Python programming skills, paid programmers, or pre-made trading software are required to programmethenecessarytradingstrategy.

• Fast network connectivity and availability of topnotch trading platforms are required for order placement.

• Access to market data streams via a reputable broker, which the algorithm will watch for opportunitiestoenterorexitatransaction.

• Beforeastrategyisimplementedonlivemarkets,it shouldbepossibletobacktestittogaininsightinto its fundamental metrics, such as drawdown, the sharperatio,theaveragerisk-to-rewardratio,etc.

• Depending on how intricate the algorithm is that has been implemented rules are, historical data is availableforbacktesting.

II. RELATED WORK

According to Hendershott and Riordan (2018), algorithmic tradingusescheapliquidity (i.e.,whenthe spreadissmall) and supplies expensive liquidity. They contend that algorithmic trading systems are more likely to execute trades when spreads are narrow because they are less inclined to place additional buy/sell orders or even to cancelexistingones.

WuandSiwasarit(2019)proposetheOrderImbalance(OI) indicatorasametrictoidentifypricediscoveryinemerging markets, and they also indicate the non-spontaneous creation of market efficiency. According to the authors' findings, one may anticipate that the order to trade ratio willgodownwhenspreadsarebiggerandupwhenspreads arenarrow.

According to studies, order expectancy activities are connected to High Frequency Trading firms. Hirschey (2013) discovers that high-frequency traders' aggressive selling behaviours are typically followed by those of nonhigh-frequency traders, and the trend lasts for up to 10 minutes. The author comes to the conclusion that the

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume:09Issue: 08| Aug 2022 www.irjet.net p-ISSN:2395-0072

occurrence was caused by high-frequency traders' order anticipationtactics.

According to Breckenfelder's (2013) analysis of the NASDAQOMXS30index,thecompetitionamongHFTfirms leads to an increase in trades that use more liquidity as indicatedbytheAmihudilliquidityratio,whichdepletesthe market'sliquidity.

According to Brogaard, Hendershott, and Riordan (2014a), aggressive high-frequency traders prefer to trade in the direction of long-term trends rather than against them, whichincreasesoverallpricingefficiency.

Inthreeforeignexchangemarkets theeuro-dollar,dollaryen, and euro-yen algorithmic trading is linked to improved price efficiency, as determined by the frequency oftrianglearbitragepossibilitiesandtheautocorrelationof highfrequencyreturns,accordingtoChaboudetal.(2014).

Using data from 30 stocks from the NASDAQ-OMX Stockholm, Hagströmer and Nordén (2013) find evidence thatmarket-makingHFTactionslowershort-termvolatility (measuredbyone-minutemidpointquotechanges).

According to Boehmer, Fong, and Wu's (2015) global survey of 42 stock 11 markets, short-term volatility (as determined by standardised intraday price ranges) grows as algorithmic trading activity intensifies. Additionally, the authorspointoutthattheriseinvolatilitycannotbelinked to quicker price discovery or algorithmic traders' propensityfortradinginunpredictablemarkets.

Accordingtostudies,orderanticipationactivitiesarelinked to HFT enterprises. Hirschey (2013) discovers that nonhigh-frequency traders typically follow the aggressive selling activity of high-frequency traders, and the trend lasts for up to five minutes. The author comes to the conclusion that order anticipation tactics used by highfrequency traders are to blame for the phenomena. The Emini S&P 500 futures market is another place where highfrequency traders may use order anticipation tactics, accordingtoClark-Joseph(2013).

III.METHODOLOGY

Five-stepofimplementingalgorithmictradingstrategies:

A. Choose the genre/strategies paradigm.

Theparadigmforthestrategyischosenasthefirstphase. It can be an execution-based, hedging, market-making, arbitrage-based, or alpha-generating approach. Let's use pair trading as an example of statistical arbitrage that is both market neutral (Beta neutral) and produces alpha, or profitsregardlessofmarketvolatility.

B. Decide on statistical significance

You can select the specific equities you want to trade usingavisualcorrelationorthemarketoutlook(inthecase ofpairtradingstrategy).Verifythestatisticalsignificanceof the approach for the chosen equities. Check for cointegration of the chosen pairs, for instance, in the case of pairtrading.

C. Create a trading model

Now,programmethelogicthatwillserveasthebasisfor ones strategy's buy/sell signals. When trading pairs, look for "mean reversion"; compute the pair's spread's z-score; and generate buy/sell signals when you anticipate a mean reversion. The "Stop Loss" and "Profit Taking" conditions shouldbechosen.

D. Hitting or quoting technique

Determining whether the tactic will be "quoting" or "hitting" is crucial. The degree to which your plan will be aggressive or passive is largely determined by your executionapproach.

E. Optimisation & Backtesting

Back-testing the technique serves as a vital tool for estimatingtheproposedhypothesis'performancebasedon past evidence. If performance data and backtest results support the hypothesis, a strategy is said to be good. As a result, it's crucial to select historical data that has enough data points. In order to cover a variety of market circumstances, this will produce a sufficient number of sample trades (at least 100 trades) (bullish, bearish etc.). Makecareful to account for brokerageandslippage fees as well. While backtesting, you mightstill need to make some approximationstogetmorerealisticresults.Forinstance,it can be challenging to determine when you get a fill while backtesting quotation tactics. Therefore, it is customary to presumethatthepositionsarefilledatthepriceofthemost recentdeal.

IV.RESULTS

Table 1 Comparison Algo TradingInIndian Stock Market

TheBenefitsofAlgorithmicTrading

Frequ ency Perc ent

Valid Perce nt

Cumulati ve Percent

LessRoomforError 18 36.036.0 36.0 WhileTrading,It SeparatesSignificantand IrrelevantFactors. 14 28.028.0 64.0

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume:09Issue: 08| Aug 2022 www.irjet.net p-ISSN:2395-0072

DeterminesTheIdeal MarketandTimefor Trading. 14 28.028.0 92.0

TradingIsMadePossible onAVarietyofMarkets. 4 8.0 8.0 100.0

According to Table 1, algorithmic trading is preferred in the stock market because it decreases the possibility of error because everything is carried out in accordance with set procedures and no psychological factors can influence hastydecisions.Nearly36%ofthosesurveyedselectedthis choice. Less room for error exists in algorithmic trading because instructions are delivered based on mathematical models. Thus, all orders will be placed in accordance with the pre-established guidelines, preventing mistakes that couldleadtosignificantlosses.

Algorithmic trading reduces market noise since it just follows the data that it receives as input and places orders based entirely on this data. When trading manually, it's possible to be swayed by the media and any tips you may pickuponthestreet.

Therighttimingmust bepresentforalgorithmictrading tobehelpfulintrading.Basedontherulesthatarefedinto thesystem,italsochoosestheoptimalmarkettotradeon.

The capability of running numerous strategies simultaneouslyistheprimaryandmostcrucialjustification for use in the future. Keeping track of more than three to four methods manually isn't possible, however algorithmic trading enables traders to use as many strategies as their capitalwillsupport. Table 2 Comparison ofAlgorithms inStock Market in India

Statistics,Descriptive

ExecutionReliability 504.66 .479

BeingAbletoBackTest 504.64 .485

SpeedandSecrecy 504.64 .485

PriceIncreases 504.62 .530

%Commissions 504.62 .530

MeetsPre-TradeExpectations 504.60 .606

Table 2 shows that the consistency of execution is the main advantage of using algorithmic trading. This is because orders are carried out in milliseconds, and soon nanoseconds,thankstoadvancesintechnology.

The use of previous data for back-testing algorithmic trading is another benefit. One can use this to increase confidenceintheirsystemanddetermineitsprofitability.

Another benefit of algorithmic trading is understanding the changes in price. Applications for algorithmic trading are readily available, and numerous trading firms offer software solutions for algorithmic trading. They receive technical charts from these software programmes that depict marketpricechanges.Thismakesitpossible for the investortocomprehendthemarket'smovement.

Based on a mean score of 4.60, algorithmic trading enables customization, pre-trade estimate matching, and softwareprogrammeeaseofuse.

Oneofthekeyadvantagesofalgorithmictradingisthatit does away with the influence of psychological barriers on one'sdecision-making.Thetraderdoesnotneedtobecome scared or greedy because the execution of orders is based on prearranged instructions based on mathematical models.

V. ADVANTAGES

SomeoftheproposedbenefitsoftheAlgorithmicTrading areenvisionedasfollows:

A. Lower Prices

Since algorithms are more economical for lowmaintenance trades, sales desk headcounts have changed and been reducedasa result.One key advance in reducing trading costs is the ability to send orders to exchanges electronically without going via brokers. Automation has alsoimprovedback-officetasksandpost-tradeserviceslike clearingandsettlement.

B. Enables Better Share Pricing and Liquidity

Broker-dealers frequently utilise algorithms to match buy and sell orders without disclosing quotes. Broker Algorithms in a sense enable increased liquidity, price on shares for clients, and larger commissions to brokers by limitinginformationleakageandacceptingboththebidand offersidesofatrade.

C. An Algorithm Can Analyse and Respond to News More Quickly Than a Human Trader.

For instance, an algorithm could notify a trader whenever information on business X is released and if the stock of that firm increases or decreases in value by, say, 1%overthecourseoffiveminutes.Forinstance,clientscan leverage live news content from Reuters News Scope Realtime product to power automated trading and react to market-moving events as they happen. To aid automated trading, each news item is electronically "meta-tagged" to

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 152

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

identify industries, specific businesses, stories, or certain piecesofdata.

D. To Keep an Eye on Risk Situations and Act Quickly When Necessary

Algorithms can automatically hedge a position if a parameter like Value-at-Risk (VaR) is surpassed by using real-timeanalyticstocontinuouslyrecalculateVaR.

E. Address Issues with Regulatory Compliance

It is crucial to follow the law, yet doing so is becoming moredifficult due to increasinglystrict restrictions. Future businesses will increasingly use cutting-edge algorithmic trading technology to address concerns with regulatory compliance.

F. Automated Monitoring

To check for tendencies of misuse in algo trading, regulators could automate surveillance. However, the absence of experienced personnel, adequate IT resources, and the restricted availability of automated surveillance toolsforalgotradesmakesupervisiontechnicallydifficult

G. Better Price Offered with Little Impact on the Market

Themajorityofalgorithms(whichfallunderthecategory of execution algorithms) must offer the best execution pricing for clients placing significant market orders. To lessen the impact of a large transaction on the market, executionalgorithmssplitlarge ordersintosmallerorders.

VI. CONCLUSION

One can see algorithmic trading as the future of stock markettradingmethod.Fortradersmanaginghighvolumes and different strategies, it is advantageous. Retail traders, however, cannot afford the cost of running an algorithm, anditisevenmoredifficultforthemtoquitorenterusinga similar approach despite the fact that the algorithm is considerablyfasterandmoreaccurate.SimilartohowSEBI has controlled these trading activities, adequate measures should be made to ensure that they are beneficial to all classes of investors and that, to avoid inequality, trading can be carried out concurrently and at the same time whereverthetradersare.

VII.REFERENCES

[1] M. Jahja, D. Farrow, R. Rosenfeld and R. J. Tibshirani, "Kalman filter sensor fusion and constrained regression: Equivalences and insights",33rd ConferenceonNeuralInformationProcessingSystems (NeurIPS2019),2019.

Volume:09Issue: 08| Aug 2022 www.irjet.net p-ISSN:2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 153

[2] J. Durbin and S. J. Koopman, "Time Series Analysis by StateSpaceMethods",OxfordStatisticalScienceSeries Oxford,2001.

[3] E. Chan, Algorithmic Trading: Winning Strategies and Their Rationale, Hoboken, New Jersey:John Wiley and Sons,2013.

[4] K.Longmore, KalmanFilter Example: PairsTradingin R.RobotWealth,September2019

[5] M. E. Thompson and A. Thavaneswaran, "Filtering via estimating functions", Applied Mathematics Letters, vol.2,no.5,pp.6167,1999.

[6] M. E. Thompson, "Dynamic data science and official statistics", The Canadian Journal of Statistics, vol. 46, no.1,pp.10-23,2018.

[7] A. Thavaneswaran and M. E. Thompson, "Nonnormal filtering via estimating functions", N. Balakrishnan editor Aspects of Probability and Statistics,pp.173-183, 2019.

[8] A. Thavaneswaran, N. Ravishanker and Y. Liang, "Generalized duration models and optimal estimation using estimating functions", Annals of the Institute of Statistical Mathematics, vol. 67, no. 1, pp. 129156, 2015.

[9] A. Thavaneswaran, A. Paseka and J. Frank, "Generalized Value at Risk Forecasting", Communications in Statistics Theory and Methods,pp.1-8,2019.

[10] A. Arratia, Computational Finance: An Introductory Coursewith,R.AtlantisPress,2014.

[11] Á.Cartea,S.Jaimungal andJ.Penalva,Algorithmicand High-Frequency Trading (Mathematics Finance and Risk),CambridgeUniversityPress,2015.

[12] C. Conlan, Automated Trading with R: Quantitative ResearchandPlatformDevelopment,Apress,2016.

[13] A. Thavaneswaran, Y. Liang, Z. Zhu and R. K. Thulasiram, "Novel Data Driven Fuzzy Algorithmic Volatility Forecasting Models with Applications to Algorithmic Trading", proceeding of IEEE International Conference on Fuzzy Systems (FUZZ-IEEE),July,2020.

[14] E. G. Gatev, W. N. Goetzmann and K. G. Rouwenhorst, Pairs trading: Performance of a relative-value arbitragerule.ReviewofFinancialStudies,vol.19,no. 3,pp.797-827,2006.

[15] R. J. Elliott, J. van der Hoek and W. P. Malcolm, "Pairs trading", Quantitative Finance, vol. 5, pp. 271-276, 2005.

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume:09Issue: 08| Aug 2022 www.irjet.net p-ISSN:2395-0072

[16] C.Ed.deMoura,A.H.PizzingaandJ.P.Zubelli,"Apairs trading strategy based on linear state space models and the kalman filter", Quantitative Finance, vol. 16, pp.1559-1573,2016.

[17] Y.Liang, A. Thavaneswaran, N. Yu, M. E. Hoque and R. K. Thulasiram, "Dynamic data science applications in optimal profit algorithmic trading", proceedings (workshop) of IEEE 44th Annual Computers Software and Applications Conference (COMPSAC 2020), pp. 1294-1299,2020.

[18] Y.Liang,A.Thavaneswaran,A.Paseka,Z.ZhuandR.K. Thulasiram, "A novel data-driven algorithmic trading strategy using joint forecasts of volatility and stock price", proceedings (Symposia) of IEEE 44th Annual Computers Software and Applications Conference (COMPSAC 2020),pp.293-302,2020.

[19] A. Chaboud, "Rise of the Machines: Algorithmic TradingintheForeignExchangeMarket", Int'l Finance Discussion Papers Board of Governors of the Federal Reserve System,Oct.2009

[20] T. Hendershott, C. M. Jones and A. J. Menkveld, "Does Algorithmic Trading Improve Liquidity?", J. Finance, pp.1-33,Feb.2011.

[21] R. K. Narang, Inside the Black Box: The Simple Truth AboutQuantitativeTrading,WileyFinance,2009.

[22] M. H. Pesaran, D. E. Giles and, "Predictability of Asset Returns and the Efficient Market Hypothesis" in HandbookofEmpiricalEconomicsandFinance,Taylor &Francis,pp.281-312,2010.

[23] M. A. Kaboudan, "Genetic Programming Prediction of Stock Prices", J. Computational Economics, pp. 207236,Dec.2000.

[24] J. M. Hill, "Alpha as a Net Zero-Sum Game", J. Portfolio Management,vol.32,no.4,pp.24-32,2006.

[25] R.KissellandR.Malamut,"AlgorithmDecisionMaking Framework", J. Trading,vol.1,no.1,pp.10,2006.

2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal