Disclaimer: This dissertation has been written by a student and is not an example of our professional work, which you can see examples of here.

Any opinions, findings, conclusions, or recommendations expressed in this dissertation are those of the authors and do not necessarily reflect the views of UKDiss.com.

Effect of High Frequency Trading on Stock Markets

Info: 7926 words (32 pages) Dissertation
Published: 9th Dec 2019

Reference this

Tagged: Finance


HFT: High Frequency Trading/ Trader

AT:Algorithmic Trading/Trader

NASDAQ:National Association of Securities Dealers Automated Quotation

TWAP:Time weighted average price

VWAP: Volume weighted average price

BPS:basis points

ECN:Electronic communication network

SEC:securities and exchange commission

NYSE:New York Stock Exchange















Since I was in high school I have always felt that there was something different about trading, it was an unknown activity that most of my colleagues did not even know that existed. It is seen as an activity that only privileged ones may have access to. Nothing far from reality but that does not mean that we have to be uninformed about the importance of the probably wealthiest business in the world and its implications. If we go deeper into stock markets, we get to a “vanguard” way of trading. I am talking about HFT.

Algorithmic Trading and HFT are a common currency in the stock exchanges markets of US, Europe and Asia, absorbing a 60% of the total volume. This figure involves several different classes of assets and market players (from investment banks to hedge funds).

The trend indicates that in the future this volume will keep on increasing, leaving behind traditional trading and moving forward to the so-called financial Skynet, where robots will operate among them, removing the human component out of the equation. Human participation will barely be dedicated to commercial aspect and maintenance of the system.

This paper aims to give out an amount of consistent background information about what HFT is, to do so, we will focus on different definitions, strategies, regulatory framework…

This paper seeks to give my opinion about where the future of trading will be driven to, how market quality is affected by algos, I will also seek to build a consistent opinion about the need (or not) of a regulation of this activity and the importance of speed cost transactions.



1.       Definition about HFT :

1.1   Definition and related concepts.

1.1.1) HFT Definition

HFT is an instrument used by institutional traders in order to maximize profits. These systems observe different parameters and operate in concordance to these parameters. Operations are executed automatically without the intervention of humans and it may execute thousands of operations in 1 second. It usually applies DMA or SA technologies for order routing. We may say that usually HFT is based on pre-established trading decisions. All these decisions are mainly used by professional and institutional traders since individual traders cannot control these systems and do not have enough buying power to get these algos. HFT is based in real-time information and under certain parameters it automatically submits and executes the order. No human intervention is required but to keep the systems updated or to introduce modifications in order to adapt to market conditions. The goal of HFT is to submit multiple buying/selling orders in order to get small profits per order. They do not seek to achieve great sums in each operation, they just seek to get a few cents at each order but placing millions of orders every single day, what at the end of the day turns out to be a fortune. Quick submission and cancellation of orders is absolutely necessary in order to get these small profits per trade. What makes the difference in getting more or less profits are high speed access to the market, what drives high speed access to the market is proximity to financial data centers and so for institutional companies pay enormous amounts of money to get the fastest road from their location to the financial center. Low-latency requirements are required in order to get effective HFT systems, and of course we need a liquid market in order to place the orders, otherwise will be you who is creating the market and no counterparty will be against you.

HFT have been evidenced to have positive effects on market quality and not so many negative side effects, but as stated by Biais and Foucault (2014) HFT can affect market quality if the own a large fraction of the market volume. In most of the research papers published by different authors they all reach to the same conclusion. HFT activity varies depending on the securities’ market (large caps vs small caps) and HFT account a significant part of market activity (⅓-⅔).

1.1.2) Related concepts

Market making refers to the strategy of quoting a simultaneous buy and sell limit order (quote) for a financial instrument in order to profit from the bid-ask spread.

Market making strategies can be either imposed by regulatory entities or voluntarily. To better understand the definition, machines use strategies in order to provide liquidity and price stability to certain secondary markets in which these two aspects may not be guaranteed. To do this, they play with the differences between “bid” or buying prices and “ask” or selling prices. There is always an existing spread between market price and offered price and is there where market makers play their role.


Smart Order Routing (SOR) systems may possible to have access to a large amount of financial pools in order to detect the best order routing destination and make order execution more effective. It acts like a scanner of real-time prices in order to detect the best bid and ask prices for a specific operation. The following figure illustrates how this order is realized for a specific order.

In order to get the best real-time prices first, once the order has been received, the routing system of the investment company, has to scan the different execution centers for their order book situations. Afterwards,the system has to incorporate a model that enables to calculate the total execution price of trades in different markets including applicable trading, clearing and settlement fees or even taxes, i.e. the explicit costs dimension (Domowitz 2002).

1.2  Historical background of trading.

Securities’ trading has suffered long evolution since the beginning of the 18th century when the commodities market started to take place in Wall Street. At the beginning and as is obvious, volume of traded securities was really small. Once the trading has been evolving, traders they start to adapt to market conditions and they stand in the trading floors searching for the right counterparty that backs their trade.  In the 70’s the NASD started to develop its computer-assisted market making system for AQ, what formed what nowadays is known as the NASDAQ. However, across the pond it was not until the 80’s that first computer-assisted equities launched their services, and it was not until the 90’s that securities trading was ordered in entirely automated exchanges. Therefore, as the electrification of securities markets was taking place, orders do not longer need to be transmitted from hand to hand until the physical trader who is in Wall Street, with this trading implementation the market is decentralized as you can have access from any point of the globe. Investors just need an online broker to place their orders and an internet connection to be able to have real time data. If we focus now on the investors’ side, we see how it has also suffered an evolution. Electronic trading processes have taken over control, as IT has developed its presence and now we can find automation at every single stage of the process. In the 90’s a big boost for Electronic trading was established with the introduction of ECN’s. This ECN’s allow trading of financial securities out of the habitual exchanges. Since that time companies have really invested into ECNs. By doing this they benefit from greater speed, lower cost, reduction of manual errors and consequently greater use of high frequency trading.  With the beginning of the 21st century the gain of popularity suffered by HFT has been exponential, especially for the “reduction of spreads”.

Before this implementation took place, minimum spread between ask and bid prices were 1/6th of a dollar, while after this implementation they were reduced to 1 cent. Of course, this has a particular effect on HFT since traders that before searched for profit in this sixth of a dollar now have to search for new ways of exploiting the margins between ask and bid, and the main alternative is algorithmic trading. All this, all along with the improvement of high speed technology acted as a promoter for HFT using.


As we can see in the following chart, HFT revenues have declined from more than 7bn in 2009 to barelz 1 million and a half in 2014. This may be caused for the incredible competition that exists in this market. Speed makes once again the difference, and slower traders have been pushed out of the market, what explain this difference.

We can see as well how the share of HFT in total equity trading has positively increased since 2005 where was almost inexistent until reaching a consistent 40% of the market share in Europe and 60% in US in 2010. However, after the crisis of 2009 these figures have started to reduce until 2014. This may be cause for the increasing cost of infrastructures for HFT machines as well as the rising of trading platforms which are not like the usual ones, the so-called alternative platforms. In the future, I think that the upcoming regulation will act as a curb for the HFT.

2.      HFTrading Strategies.

2.1  Scope of HFT strategies

HFT is not the same as AT, in fact is a subcategory of AT. However, not all the algorithmic strategies are HFT strategies. We will focus only in HFT since AT is out of our scope. HFT strategies can be subdivided in 4 main categories depending on its generation. This classification was made on its time by Almgren and also contains some information from Jonhson (2010). First generation algorithms focus only on benchmarks that are based on market generated data (e.g. VWAP) and are independent from the actual order and the order book situation at order arrival, while the second generation tries to define the benchmark based on the individual order and to handle the trade-off between market impact and timing risk.

Third generation algorithms are furthermore able to adapt to their own performance during executions. A fourth generation – that is not included in the Almgren (2009) classification – consists of so called newsreader algorithms.

2.2   First generation Algos

Participation rate:

-They are predefined to participate up to an X % of a market. They reflect the real current market volume in their order as they are trading market volume.

TWAP Algorithms

-TWAP algorithms get a huge order to place in the market and instead of placing it all at a time in a market, what would cause price inefficiencies that would probably get against them, they divide the order into equal slices and place it at different time intervals. This way, HFT avoid situations like the one below.

For example, an algorithm receiving an order of 4000 shares to execute in the Eurostoxx market, in order to fully execute the order at the same time, what it does is take a time lapse, in this case, 6-7 seconds and it starts placing small slices of 29 shares at each time. This way, the price does not suffer a huge change as seen in the previous image and you prevent other market players to “identify” you.

The VWAP is the average value of a stock traded in a given time, usually a day. Traders use this method to ensure that their stock purchases are in line with the average price in the market. Trades are weighted on their size, so it’s obvious to say that larger trades will have a greater impact than small ones. To estimate the price, VWAP algorithms are based on historical data to estimate the volume patterns.

2.3   Second generation algos

The second generation algos try to minimize implementation shortfall. The current price/midpoint at thte time of arrival of an order serves as a benchmark, which shall be met or outperformed. They try to minimize the market impact of a large order taking into account potential negative price movements during the execution process.

2.4   Third generation algos

These algorithms re-evaluate and adapt their execution schedule during the execution period, making them adaptive to changing market conditions and reflecting gains/losses in the execution period by a more/less aggressive execution schedule.


As defined before, HFT strategies have the characteristics of relying on computers to execute the strategies and try to be the fastest actors on the market. Everybody talks about speed when talking about HFT, but why is so important speed for HFT?

Basically is like air for human beings. Without speed HFT would be nothing. They are fed of speed. Speed makes the difference between market actors and is the main reason why it is so important. If you have access to information that only you have, because of the fact of being faster than the competence, then, you have a competitive advantage that you can turn into millions of € a day. Apart from this, HFT strategies are based in really short periods of time, where the winner takes it all. There is no space for slow computers in the stock market. The fastest investor grabs all the profit from a market opportunity and the rest fall apart.

To be fast is crucial to have a low latency in the market, what we mean by latency and that has been previously mentioned is that how much time does it take you (as a computer) to get information from a trading platform, process it and take an action to it by sending orders back to the market. This latency will make the difference between a winner and a loser. As Hasbrouck and Saar say in their study (2012), there are super-fast traders that lasts about 2-3 milliseconds to react to a market update. This is 150 times faster than a blinking human eye.

What makes the difference between 2-3 milliseconds and 5 milliseconds may seem as something totally imperceptible, but is not like this at all. Huge companies like Morgan Stanley, greatest hedge funds of the greatest financial centers, they pay millions of dollar to be strategically co-located. What I mean by strategically co-located is to be the closest to exchange servers. By placing your services closer to the exchanges’ services you are reducing distance to the place where orders are exchanged, and so for you are reducing latency and increasing the speed of your transactions, that ultimately will lead to an increase of profits. That is why Michael Lewis, as he illustrates in his book “Flash Boys” makes the emphasis on speed to reduce the latency to become profitable. He states that people who created the network between exchange centers and financial centers literally did ANYTHING to reduce the speed of transaction in 17 milliseconds. If they had to cross a mountain they did, if they had to buy a property, they did, and everything because they knew that speed makes the difference. This is something unbelievable to small and medium investors’ eyes, but something that makes the difference among large companies.

As Bruno Biais and Thierry Foucault state in their article “HFT and Market Quality”, HFT strategies are heterogeneous not only depending on their generation as stated before, but also on the goal of the strategy.

We can differentiate 5 great blocks of strategies to which trading speed matters.

1.1. HFT Market Making: Market-makers provide liquidity to other traders. So for, HFT firstly place limit orders that provide liquidity to other traders. In fact, some evidences have been found concerning this information. “Jovanonvic and Menkveld after studying a HFT market maker in the Dutch stock market they came to the result that a single HFT was providing a 78% of the market liquidity on the operations in which he was taking part. That is to say that ¾ of the quantity traded between it and other traders was provided by it. Some more evidences have been found concerning the market making strategies. That is to say, Brogaard (2011b) who evidenced that some HFT act as market makers by using the so-called price reversal strategy, that states that they buy when prices are declining and they sell when prices are increasing in the last short periods of time.

Speed is really important for these strategies because it allows market makers to rapidly react to increases of market illiquidity in punctual spots. For instance, if a large order of stocks is absorbed by a market, the liquidity levels of this market are going to disappear at a given price (best quotes), so for, the bid-ask spread of the stock may widen. This increase of the spread between buying (ask) and selling (bid) price may turn into an opportunity for new spreads, that only the fastest traders of the market will be able to submit new limit orders within the new spread. Apart from this, the possibility to act quickly to unexpected news of the market reduces the risk of being picked off the market. If a terrorist attack takes places and a “black swan” occurs, have no doubt that small investors orders will not be the first ones to be sold in the market. Indeed, having fast access to the market may allow you to quickly cancel your orders before the price limits disappear under your price levels and you will be able to save a vital money that most of the investors will not.

1.2. High frequency arbitrage

HFT is also used to trade arbitrage opportunities. Arbitrage takes place when some related assets prices suffer a deviation. Let’s say that there is a stock quoting at the NYSE at the price of 30 and its future quoting at the futures’ market of Chicago. If for whatever reason, the price of one becomes cheaper than the other, trader could simultaneously short the most expensive of both and buying the cheapest, and profit from the difference. Most of the pure arbitrages opportunities only last a couple of seconds, apart from this, arbitrage opportunities are almost riskless opportunities to make a profit so the first detector of the market will take it all, leaving behind the slow traders. That is why submitting transactions’ speed is so important. According to Foucault, Kozhan and Tham, and based on the study in a triangular arbitrage opportunity (three markets), the market opportunities take place every day several times, but the average lasting time of the opportunity is less than 1 second. The fact that arbitrage opportunities take place in the market and are so short lived suggests that is not very useful to correct this inefficiency by using HFT strategies. For instance, an opportunity that lasts 20 miliseconds instead of 2 seconds does not seem to be something really valuable for society.

1.3. Discretional strategies

Taking discretional strategies consists on submitting an order in one asset before a major change in price takes place. It is not exactly like arbitrage where you had to buy multiple contracts either​in the short or long directions, here you need some signals that help you forecasting the price. These signals may vary it’s nature from futures (Menkveld 2011) (Zhang 2012) to news. They all have in common that fast access to this signals is considered to be useful as their reaction time is vital to profit from these signals, even if these signals are already public.

Broogard, hendershott, and Riordan (2012) found that HFT react to macro-economic announcements as well. As stated before, this announcement is only valuable by the time it is not public and not everybody has access to the information, so that is why they have to be exploited very quickly. Based on Scholtus et al 2012 scientifically research, he stated that for his sample (707 macro-economic announcements) a delay of reaction of 300 milliseconds reduces the returns on a strategy exploiting the informational content of the announcement by 0,44 basis points

1.4. Structural strategies.

As SEC states, these strategies operate specific characteristics of market structures, so depending on the nature of the market the profits may vary. For example, HFT exploit the fact that today’s market is highly fragmented. That is due to the fact that trades for a stock are not centralized and so generates arbitrage opportunities. As Bruno Biais and Foucalt (2014) state, HFT profit from speed to apply these strategies. For instance, a HFT observes that 4 trading platforms showing real-time price. 3 of them update their quotes at the same time, but however, there is one platform which refresh its prices 0,3 seconds later than the others. HFT are able to identify this opportunity, to react to it, and take profit, all in these 0,3 seconds. That is why speed matters for this kind of strategies.

1.5. Manipulation

A worry that most of the people have is that HFT make use of their fast access to the market to manipulate prices. What they do to create this manipulation is that they buy a large amount of stocks, they expect people to herd behave and wrongly jump into the rally, amplifying the movement, that is to say, reaching higher prices than they bought, and then, selling their stocks at inflated prices. This operation can be exactly executed the other way round.  HFT traders are better at these strategies since either are them the ones creating the fake rallies that right after unwind themselves, or either they can quickly react to others momentum ignition strategies.

Before moving on and studying the effects of HFT on market quality we will focus on the profitability of HFT, what will help us to better understand how profits are shared between market participants, whether if HFT profits have declined since more and more HFT are arriving to the trading desk, and understanding the repartition of benefits between HFT and non HFT will help us to forecast the effects of HFT on market quality. It is also interesting to understand the size of HFT profits and how a tax can affect their activity.

Following the evidences that Hendershott et al found in his study, profits of HFT are 1.14$ per 10000$ traded. Here, we have to make the difference between market orders and limit orders. Profit on market orders are higher than those ones reached with limit orders. Baron also did a more exhaustive research on this topic. Indeed, he used data from CME in Chicago in August 2010. For his study he reached to the same point, small profits per trade. (0.77$) and market orders strategies more profitable than limit orders strategies. With the reliable data collected by Baron et al, more conclusions can be taken from this study. HFT that use market orders predominantly earn profits while HFT that use limit orders do not always get profits, what it means that when trade against HFT relying on market orders, they are losing. We cannot forget that HFT can obtain losses in unexpected market situations where the algos cannot anticipate or follow the market movements.

2. How the market quality is affected since the use of machines?

In theory, applying what we have seen until now we can say that HFT provide liquidity at a lower price thanks to its low latency. Thanks to this HFT market makers should improve market liquidity and increase trader’s welfare by reducing intermediation costs. However, on the other hand, HFT have faster access to market information what leads to adverse selection problems since slow traders will systematically lose to HFT. Following Biais, Foucault and Moinas (2012) where welfare of adverse selection is studied, both costs and benefits are identified. On the one hand, it raises the possibility that investors may find a profitable trading opportunity. On the other hand, it increases the possibility of adverse selection generation between actors, reason why many traders decide to reduce their operations when a large part of the market is being provided by HFT. When firms decide to invest in HFT they are not taking into account the social externality it may generate to the market, they are not assuming it, they are simply comparing the investment they are about to do with the potential profits it may provide to the firm. Apart from this a company may pay great sums of money only not to be overtaken in terms of speed and keep on being competitive face to other firms, what leads to extremely high investments on HFT that are far distant from what would be socially optimal.

HFT increase price discovery by operating the mispricing of the market. Let’s say that if HFT trade faster on information that other traders they are contributing to increase price discovery since they are pushing the speed at which information is translated into prices.

Dougast and Foucault state that there is a trade-off between speed and accuracy. As we all know, news is not always right, and sometimes investors that react too fast to news are not making the price more efficient but the only thing they are doing is creating “smoke” in the price. HFT have also been accused of taking part in several flash crashes of the market. What we understand by a flash crash is a violent movement of the price in one direction. HFT may take part in these events since they all react at the same time to news, sending buying or selling orders to a same market and reducing the liquidity of itself. Of course, if all the market orders of the HFT are accepted, the price has only a possible direction… So it is possible to state that HFT alter the volatility of the asset in which they are investing.

2.2 Endogenity issues.

It is true to state that volatility affects HFT activity, but it is also interesting to say that volatility is affected by HFT.

In theory as Biais, Foucault and Moinas (2013) state, a raise of the uncertainty about an asset’ price, increases the possibilities of investing in HFT. Brogaard (2011) as well in his research found that Nasdaq activity was affected by the news release. So, we can consider that news, being a source of volatility have a direct relation with the activity of HFTs

As volatility, HFT and liquidity have a direct relation of endogeneity, the established correlations between these 3 do not have to be seen as alterations of market quality. One way to avoid these endogeneity issues is to use variables that affect HFT but does not affect the other variables. For instance, reduction of latency or implement co location services may seem as ideas to affect HFT but not the other variables.




HFT usually place their market buying orders just before the price increases and they place their market selling orders just before the price drops. This is consistent with the idea that HFT possess information that most of the investors do not. However, for limit orders they act inversely, they sell contracts when the price is falling and they buy when the price is increasing. Hendershott and Riordan found that algorithmic market orders have greater impact on the market that individual market orders.

Boehmer, Fong and Wu (2012) take a sample for 39 countries. This sample compiles about 12,800 stocks. A negative correlation is evidenced between the amount of messages normalized by volume in each stock and the autocorrelation of stock returns (in absolute value) at the 30 minutes’ horizon, which they interpret as a measure of price inefficiency. Of course, as explained above, due to endogeneity issues, such correlation cannot be interpreted in causal terms.


In different studies, like the one realized by Hendershott, Jones and Menkveld is studied how the liquidity provision is affected by the use of HFT. They use the rate of electronic message as a proxy for the amount of AT taking place. To them, it is important how to standardize the increasing traffic volume of electronic messages. In order to better understand the conclusions, they reached is important to define what is an Autoquote: Autoquote was an important innovation for algorithmic traders because an automated quote update could provide more immediate feedback about the potential terms of trade. This speedup of a few seconds would provide critical new information to algorithms, but would be unlikely to directly affect the trading behavior of slower-reacting humans. Autoquote allowed algorithmic liquidity suppliers to, say, quickly notice an abnormally wide inside quote and provide liquidity accordingly via a limit order.

We will focus on the conclusions of the study and not so many on the methodology provided to get to these conclusions. In the sample of these authors is stated that the technological revolution that financial markets have suffered all along these past years has fully changed the way stocks are traded. Many financial corporations now trade using HFT. In their 5-year sample (2002-2006) AT has gained a great importance and market liquidity has increased, as well.  It is stated that AR reduces the execution costs and produces an augmentation in the quotes’ information. In the same way, market makers, that is to say, liquidity providers will see how their revenue increase as well since with the use of AT the costs of executing an order have declined. Linkages between different markets have also been improved since the using of HFT.  As the price of stocks have become more informative and a fastest access to it has been granted, liquidity and efficiency in prices will rise.  In the same way, Herdershott et al state that some standard measures that determine the market liquidity increase after the introduction of Autoquote for large caps. However, it does not have the same effect on small caps, since any big effect is stated after the introduction of Autoquote for this market.

This figure shows the number of electronic messages transmitted per minute by a proxy used in the sample of Hendershott, Jones and Menkveld (2011). We can see how the increase of this message (what may be translated as an increase of the AT activity) is exponential as the entrance of new actors in the market takes place.


From the theorical point of view, there is not a clear definition about if HFT is a positive or negative contributor to volatility. As Biais & Foucault say, in the short term, the volatility reflects the impact of liquidity demand. If we follow the evidences that HFT increases the level of liquidity in the markets, therefore, HFT should reduce the volatility of prices, at least, in the short term. However, if we state that HFT have been contributors to some of the toughest market impacts of last years as it was the “flash crash” * of 2010, then in this case, we can state that HFT increased the volatility of the markets. Following the line of the theoric aspects, the scientifically evidences reach to the same conclusions. Chaboud et al (2009) they realized an economic study in which they measured the impact of volatility in three currency pairs. Using the data available on EBS (one of the two main electronic platforms in the currency market) they can identify whether if an order has been manually submitted, or on the other hand, has been automatically placed. After using different OLS regression using the data they possessed they got to the conclusion that there exists a positive relationship between the volatility of these three pairs of currencies and the measure they used for AT. However, this relationship only states that algorithmic traders seem to be more active in high volatility days. Using a different approach, the same authors find a negative relationship between AT and volatility, in contrast with the results obtained using the OLS regression. Other authors like Haasbrouck and Saar (2012) use different methodologies for several markets and they reach to the same conclusion. They find a negative relationship between AT and volatility. Apart from these authors Brogaard (2011) “shows that stocks in which short-sales by HFTs are the most affected (relative to other stocks) experience a relatively bigger increase in volatility (measured over different time intervals)”. Once again, this evidence is coherent with the negative relationship between HFT and volatility.

However, we find an evidence that disagrees with these negative relationships between our two variables. Boehmer Fong and Wu (2012) after using different proxies for volatility they find a positive relationship between our two values.

So we cannot state that exists a universal truth, and a real relationship between these two values, but different points of view can be clearly defined when we approach to volatility.

*Flash crash is known as the brief period that took place on May 6th where extremely volatility was the main starring of this period. That day the Dow Jones fell by almost 1000 points in barely seconds, what was translated as the greatest intraday loss for the index


We say that a regulatory intervention is needed when the activity of a group of market actors produces negative effects on the activity of a third party.  Bruno Biais et Foucault (2014) show that HFT may create congestion externalities comparative to the access to exchanges’ trading platforms or to market information. They state as well, that HFT may cause negative effects for third parties. Cases of adverse selection have been identified since the use of computerized trading. As shown before, and as Biais, Foucault and Moinas (2012) show, the fact that traditional traders have a slower reaction to changes than HFT, is per se a cause of adverse selection for traders.  Apart from this, traders can be trapped in a race where they overinvest in technology in order to reduce latency and increases the possibility of trading with a better informed investor.  Last, but not least, HFT might have a crowd out effect on extremely slow market makers.



Taxation: One way to restrain the unreasonable expenses in technology to develop the HFT servers is by making less attractive the investments on it, and that is what government tries to do by taxing the investments in this field. For instance, we can account the tax applied by NYSE Euronext. This tax is applied to all the orders on the market when the order has a order to execution ratio greater than 100. What it intends to do this tax is to identify HFT which usually have a greater ratio than this one, and this way discourage the use of HFT.

Hangströmer and Nordén (2012) evidenced this theorical aspect. They observed that HFT market makers most of the times have a larger order to execution ratio than other type of market participants. It makes sense from the point that they are revising, cancelling and re-submitting orders in order to ward off adverse execution. If authorities really want to discourage the use of HFT they have to tax marketable orders. These orders are the type of orders that computers use and it has been evidenced that they use this type of order to exploit the information advantages they have.

Market tools: It is fair to say that fast access to the information is what drives traders to invest in HFT technologies since they have evidenced that some differences of milliseconds are what makes the difference between winning or losing. So using market mechanisms to reduce this adverse selection that leads to overinvestment seems like a good idea to Biais & Foucault (2014). Many of the readers may think that those mechanisms should appear automatically, but unfortunately there exists some market fails that forces the mechanisms not to appear as it had to. For example, there may exists some barriers for slow traders, and new market structure does not make easy the adoption of new trading tools by exchange.

Minimum exposure time: Under this idea, all the orders coming from HFT could not be cancelled before X time. That means that HFT activity would be highly affected, since most of the HFT strategies profit from their speed to make profit. Consequently, this would have an immediate effect on liquidity provision by investment firms. We have seen how HFT are excellent liquidity providers that benefit from speed to efficiently provide the market of this liquidity. So for, I don’t think that the best idea to limit or control HFT liquidity is just to cut down the activity of liquidity providers that so beneficial are for the right working of markets.


Regulation: The EC has taken into account the activity of HFT into the new MIFID II Regulation. The commission proposal states that any person above a certain quantitative threshold is automatically seen as an investment firm. Like this, the investment firm would automatically be object of vigilance. Minimum requirements in terms of capital and management would have to be respected, as well. This way, systemic risk creation would be mitigated.

Where will financial markets be driven in the future?

With all the literature read until now and all the information collected I can state that markets

will be computer driven in the future for a great part of the market. It is true that new

regulations are coming to the trading floor and we cannot still define what the behavior of the

main actors will be. In my opinion, new markets will be exploited by HFT, markets like bonds

will be highly exploited by these actors that expand its boundaries. Even if in the last years

the HFT have seen how their revenues have declined we can state a fundamental reason for it:

markets need of volatility to earn profits, and after the crisis the markets have suffered a

declining of the volatility. I am sure that when the next economic crisis come, they will

increase their year revenues since the volume in these periods tend to skyrocket. Invesments in HFT will keep on taking place more and more since it seen as “the holy grail” by people, but the fact that the pie is getting is smaller more eaters are invited to the party means that there will be a moment where there will be too much eaters and only the most efficient ones remain. This situation, all along with the new regulations make me think that HFT will take a considerable share of the market but will not be absolutely dominated by them as it’s been stated by some specialists.


To get all the data that I have collected and then written in this essay, I have read documents,

either from my tutor who gave me some resources, either from google scholar where I have

really got lots of information related to the topic. I have chosen this methodology because I

feel that is the easiest one for me, that I am not an expert in the topic, and also for the people.

Probably If I had done surveys or I had collected data from any other channel I would not

have got the same qualitative resources that I have got.



After writing this essay some conclusions come to my mind. First one is that the rise of

machines is only a consequence of the evolution of the stocks markets. New possibilities have

come to the trading floor and with the implementation of technologies changes in the way of

trading is a reality. It is understandable that if traders want to reduce costs and want to earn

greater profits, which at the end is the main goal of a speculator, we can understand the rising

of this methodology of trading. In the same way, it has been evidenced that the effects of HFT

in the market liquidity, are mostly positives. In the latest publications they all agree that

positively contribute to price formation and market quality.

Imposing taxes on these institutions that are considered to have a high order to execution ratio

or a high traffic of messages does not seem like to be the perfect option. Basically, because it

would affect market making strategies. Another option would be to equal forces, that is to say

to cross HFT orders between HFT traders, and to trade slow traders’ orders between slow

traders only. That seems to be really difficult since I do not know how a server could detect where orders do come from and how these orders can only be cross between equals.

Probably the creation of platforms that would allow the entrance only to slow traders seems like a better idea, but I am not really convinced until what point these platforms would be unbreakable, and how the activities of the main markets would be affected after the expulsion of slow traders. It is true, that at least, the adverse selection problem would be cut down, and slow traders would be more protected against the market. I am not sure until which point the reduction of investments by HFT would be something real, as these few HFT traders would always seek to be the fastest of the market.

In the same way, considering the high impact that HFT have in the market, I see as something coherent that institutions realize stress tests to the trading companies. Following the line, that we do not know how the future will be under these regulations seems logical to first do some pilot experiments that allow the regulators determine whether the effects of the regulation may be positive or not.


From the academic perspective, is clear that I will not be able to provide to the academic community the quality that a researcher may provide. However, I have tried to recap the most important aspects about the impact that HFT may have in the stock markets. I do not know whether if it is the best approach the one that I have chosen or not, but I have tried to give my best to provide a quality essay. In terms of knowledge, it is clear, that even if I knew what was HFT I had no idea about all the impacts that it has in the market, I did not even have idea that there existed so many HFT strategies. It has been a real pleasure to write this essay because I have really felt that I was learning a lot of new concepts that I have never heard before. I am sure as well, that as this is not the most known topic for an essay, I like to think that not so many students have written the same as I have.

I have to say that there exists a wide number of options to choose if we want to talk about HFT, I have decided to cover these concepts, which to me, they seem to be the most important ones, and they cover a great part of the literature available.

I have to say that I think that there is still room for improvement when talking about HFT, there should exist any type of market regulation that works in order to avoid flash crashes like the ones in 2010. There are situations that are clearly generated by HFT and that traditional traders cannot control, when they play such an important role as HFT do. So, an implementation of a limit number of sell within a small period of time should be implemented in order to avoid these situations. I think that it would also be important to do analyses at companies’ level, to approach to them and see what are the real issues for midcaps and large caps, like this we will see the difference between different type of market players. Once we have identified the problems, try to find measures that adapt both mid and large caps. Sometimes we only focus on the profitability side and we do not have to forget the responsible side.

Last but not least, it has to be said that market regulators should take into account all the benefits that HFT brings to the floor and they have to try to reduce inequalities while not affecting to all these positive outputs that HFT brings.







Cite This Work

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Related Services

View all

Related Content

All Tags

Content relating to: "Finance"

Finance is a field of study involving matters of the management, and creation, of money and investments including the dynamics of assets and liabilities, under conditions of uncertainty and risk.

Related Articles

DMCA / Removal Request

If you are the original writer of this dissertation and no longer wish to have your work published on the UKDiss.com website then please: