First Written Note. TraderEX Lab Hans Jakob Collett Humlevik and Søren Oscar Hellenes ID: nhh3

Size: px

Start display at page:

Download "First Written Note. TraderEX Lab Hans Jakob Collett Humlevik and Søren Oscar Hellenes ID: nhh3"

Frederick Cunningham
5 years ago
Views:

1 First Written Note TraderEX Lab Hans Jakob Collett Humlevik and Søren Oscar Hellenes ID: nhh3 Trading setup TraderEx simulated a continuous order- driven market. Orders are kept by a limit- order book that ensures price and time priority. The traders could place limit- orders and market orders. Market orders executes instantly at the lowest offer or highest bid in the order book. Limit- orders are kept in the order book until a bid or offer is placed with the matching price. The students where separated into three different types of uninformed traders: buyers, sellers and proprietary traders. The machine orders where statistically generated to reflect three different motives for trading. These are as listed in Schwartz 1 : information motives, liquidity motives and momentum motives. The session was simulated as a normal trading day from 09:30 to 16:00. But the real time was about 45 minutes. As figure 1 in the appendix shows traders where evaluated along four performance measures: risk limits, net sell/buy target, profits and VWAP. The four performance measurers are listed below. P&L: Aggregated net profit or loss, realized and unrealized, through the simulation. Profit is positively correlated with the total score. VWAP: The Volume- Weighted Average Price is the average stock price up to a certain time in the trading day. It is calculated as!"#$%!"#$%!"#$%$!"#$%!"#$%&!"#$%$!"!"#$%! 1 Schwartz, RA, GM Sipress and BW Weber, Mastering the Art of Equity Trading through Simulation, The TraderEx Course, Wiley, 2010 parge 46. 1

2 Risk: The risk bound was 300 units. That means that you should not deviate more than 300 units from the amount of stocks needed to be sold evenly each minute to reach you goal. If you deviate more, your risk is higher than preferred, and you will get a reduction in your score. Higher risk gives lower score. Job Completion: Measures how close you are to your goal at the end of the trading day. The closer the goal the higher score. Task In the simulation we were given the role selling 2500 units ( shares), with the risk bound of 300. In addition to obtaining the highest profit possible. The last two are paradoxical, and have different weightings. You may have to give up the risk limit to pursuit profit or opposite. Expectations Having some knowledge of fast moving markets we had an expectation of a great amount of order activity during the day. In the preparations of the simulation we also had some experience with the TraderEx software and the book Mastering the Art of Equity Trading Through Simulation (hereby referred to as Schwartz). From this starting point we knew that the market would not follow a random walk pattern 2 in the short term and that trends differing from the equilibrium price could occur. We did not have any information regarding the skills and behavior of the other students. However, knowing that the informed traders would trade when the market price differ from the equilibrium price, we discussed the possibility to anticipate the equilibrium value by following the informed traders. At the same time we knew that both the liquidity and momentum traders could create and build upon trends that do not follow the equilibrium price. Not knowing the 2 Schwartz, RA, GM Sipress and BW Weber, Mastering the Art of Equity Trading through Simulation, The TraderEx Course, Wiley, 2010 parge 46 2

3 number of different traders and their roles we were not able to anticipate the equilibrium price. Approach After receiving our order to sell 2500 units, having a long position of 2500, we decided that the performance indicators given in figure 1 should guide our approach. Given that the job completion and risk counted for 60 percent of the final score, we figured that selling all units within the risk limit of 300 would be our main task. Bearing this in mind we made a schedule with a given amount of units to sell per hour. This schedule would spared our trades evenly and guide us to a flat position of zero units at the end of the day. However, having a long position we also considered the natural strategy of selling at the highest price possible to gain our client alpha. If the market prices where to rise significantly, we agreed that it would be profitable to leave our risk and sell a greater amount than our schedule indicated. The strategy of trading at the top of the order book, when the market price were above the equilibrium price, would only be an opportunity if we where able to return to our initial risk. Considering the goal of ending at a flat position and the risk of trading at the top of the book, we decided to follow the original schedule but with an eye open for profit opportunities. Having settled our main approach we knew that balancing between limit and market orders would be of importance. Considering the goal of receiving the best price possible we assumed that placing a number of limit orders to execute on a ricing trend would be beneficial. However, not knowing how the market would evolve, placing limit orders would make it more difficult to follow our schedule of selling a given amount of units per hour. While expecting the limit orders to be executed on informed traders market orders as the market moved upwards we decided to mainly use limit orders. Bearing this in mind we would use market orders to adjust our risk line. In short our approach would be to place limit orders to be executed on a rising trend and complement with market orders to account for the risk limit. 3

4 Trading history Figure 2 shows that the Figure 2 trading day started with a price of 19,8. Shortly after is rose to 20,1 and continued rising to 20,4 at 10:57. At this point the price started rising more rapidly to a peak of 24,3 at 11:57. The price now fell slightly before it cracked down to 21,4 at 11:59. From 12:00 to 12:53 the trading price remained relatively stable at 21,1-21,4 before it started falling to 18,4 at 13:29. Here the price remained until 14:10 when it rose to a small peak of 19,7 at 14:23. The trading price fell back to 18,1 at 14:33. In the time period from to 15:48 the market was volatile. After 15:48 there was a fall to 13,2 at 15:50 before it rose back to 19 at 15:55. From 15:55 to 15:59 the price fell again to 15,6 ending the day with one trade at 11,2 and one at 1 at 15:59:56. Performance Our trades are shown in figure 3, which contains our sell and buy price, as well as the PStar given from the dataset Fundamental value. (From time 8406 to and after we did not initiate any trades). The inventory of our units can also be seen from the risk plot in figure 4 in the appendix. We started of by selling at the Figure 3 Price Our Trades Time Price sell Price buy PStar 4

5 price of 20,1 and 20,2 to reduce our inventory. However, as the bids started to rise at 10:57 (time 5612 in figure 3) we left our initial strategy and started selling more than the schedule indicated. At 11:03 we had sold 2080 units. The increase in price happened more rapidly than we expected and we were caught off guard. At this point we where concerned by the level of our inventory and that the risk was higher than recommended. We therefor decided not to initiate more sell orders until the risk was at a more responsible level. However, as the price peaked at 11:56 we considered the bids to be at a point where we could increase our average price. Again we left our strategy and sold another 150 units. Being in this position we considered to change our strategy and buy some units to reduce the risk. As the price fell at 13:09 (time in figure 3) we therefore started buying a total of 1001 units at the average price of 19,7. We now had an inventory of 1229 units and the risk was again at a responsible level. Knowing that the price earlier had topped at 24,3 we wanted to wait for the order book to push prices upwards before we started selling again. Waiting for a rising price we where again caught off guard and the time ran away from us. As the price fell to a day- low of 13,2 at 15:50 we knew that we would miss our goal of selling 2500 units. We ended the day with an inventory of 1229 units. Evaluation of performance The numbers marked in blue in figure 5 in the appendix shows our performance numbers. As mentioned we had an inventory of 1229 units, 1271 away from our target. At the end of the day we ended up with a profit of 2548, 757 realized and 1791 unrealized. Our average sell price was 20,46, which is 0,1 over the VWAP of 20,36. This is manly due to the steady selling in the beginning of the day and the units sold at the day- high of 24,3. Ending up with a risk of 495 is because of our large number of sales before 11:00 and not being able to reduce our inventory after 13:30. 5

6 Studying figure 5 and figure 1 we can estimate at what point our score was affected positively and negatively. Ending up with an average sell- price above VWAP and a profit of 2548 we affected the performance score positively. However, not being able to sell all units and holding a higher than recommended risk most of the day, our score was affected negatively. With risk and job completion counting for 60 percent of the total score we were not able to compete within the highest part of our group. As figure 5 shows nhh1 and nhh4 managed to combine the task of selling more units and gaining a better sell price then us. Reflections After analyzing our trading day it is clear that we could improve in several aspects. As the evaluation of our performance indicates we did not pay close enough attention to the measurements of job completion and risk. Having an initial approach that took account for the target of selling 2500 units within the risk limit, we deviated from our strategy several times. When the order book moved up towards the day high of 24,3 we sold 2000 units and left our strategy before 11:00. If selling at the rising price was a good decision or not is hard to say with no further calculation. It is possible to calculate the profit gained from selling at a high price versus leaving our risk limit. However, due to the restrictions of this text we will not do this calculation here. Being inexperienced traders we bought a too large volume that led us to deviate from the upper risk barrier. We had now changed strategy several times and where too focused on gaining alpha. When price never rose to what we estimated as a good price we were paralyzed and realized our poor position to late. Figure 6 shows our sell and buy limit orders. From here it is clear that we were to optimistic and placed orders far over the bid price. At this point we should have dropped our chase for profit, counting for 10 percent of the score, and been more disciplined. If we rather followed our initial strategy and sold units at an even rate, we would have reduced our risk and completed our job. 6

7 Figure 6 Looking back at our actions we have leaned that a good trader is a patient trader. We will take this with us to the next trading session. Before the trading starts we will create a more sophisticated strategy that take account for several different outcomes. Being more disciplined and trading at an even rate will be beneficial. Trading large volumes at a rising trend might be accepted given that you are able to adapt your strategy. Having more experience with how the market works we will also consider the possibility to affect the trends. Balancing limit orders and market orders we might be able to create trends by placing large limit orders and trade on the trends using market orders. 7

8 Appendix Figure 1 Figure 4 8

9 Figure 5 Sources: Harris, Larry, Trading and Exchanges, Market Microstructure for Practitioners, Oxford University Press, Schwartz, RA, GM Sipress and BW Weber, Mastering the Art of Equity Trading through Simulation, The TraderEx Course, Wiley,

TraderEx Self-Paced Tutorial and Case

Background to: TraderEx Self-Paced Tutorial and Case Securities Trading TraderEx LLC, July 2011 Trading in financial markets involves the conversion of an investment decision into a desired portfolio position.