An experiential learning module has been introduced to the University of Dayton Business School for the fall 2009 term: Managing a small currency portfolio around market trends. This new module dovetails with the Davis Center for Portfolio Management, which focuses on the management of a stock portfolio, and will be managed out of the Hanley Group Derivatives Trading Center. The module will be managed along the lines of “Adding Girth To Your Profits” (Futures, December 2008). The girth model is a trend-following model incorporating volatility, momentum and velocity.
There are significant operational risks to managing a forex portfolio in an educational setting, even using a non-discretionary model. Foremost, the students are novices to the commodity markets, the method of the model management itself, and the trading software used. Therefore, the first semester is an implementation of the model using a simulation platform, and focuses on the methodologies of the markets and the nuances of the model.
Only if the students demonstrate the discipline and the desire to continue on a team-based approach to managing a forex portfolio will real money be used. However, the simulation will look and feel like a real portfolio; thus, any significant operational risks will be brought to light without the danger of a large monetary loss.
THE TRADING METHOD
Students at the University of Dayton will manage a euro portfolio around the four-hour currency cycle. They will use the 10-period exponential moving average (EMA) and the 20-period EMA as their proxy for market trend. The model will enter new trades only on an EMA cross (for example, if the 10-period EMA crosses above the 20-period EMA, the model will go long).
We will calculate girth at the close of each four-hour candlestick. We will use girth as an early close indication to both long and short positions. Existing trades will be closed when the girth indicator decreases below a specified threshold, indicating that the trend may reverse.
Reviewing “Early out” shows how the model picked a strong trend upward, and indicated an early close to the upside trade in late May. However, the students would still be watching the girth number at the four-hour intervals. We saw a limited cross to the downside and then a resumption of the uptrend. The students would be watching the four-hour girth threshold and the trend indicators and would have entered another long position.
The graph shows how the trends and the girth indicator are used to preserve profits pending a reversal in trend. During the late May move from 1.34 to 1.44, two long positions were taken. They were both entered at the time of the EMA cross. Both trades were exited prior to the EMA cross, when the girth figure crossed below the girth threshold. Typically, early exit due to decreasing girth results in a more favorable profit position than that taken if the trader simply waited for an exit on the EMA cross to the downside.
On or around the second week of classes, following training in market methodology and the use of a disciplined trading plan, the students will open one simulation account for the management of the girth model. Working with novice traders has necessitated having, at most, one open position at a time in the portfolio.
The girth model presented in the December 2008 article also has been simplified to its base form: using girth only to indicate early exit, not using girth to add positions in a strong up or downtrend. Teaching novice traders a strict control on data input, trade entry, trade exit and passing the book to the next group watching the next four-hour time block will be difficult enough. In other words, the operational risk of managing the simplified early exit position is great. Additionally, we do expect human error.
The experiential learning skills being taught in this class go beyond forex trading methods (both fundamental and technical), but also cover the skills involved in the running of a small fund: front, middle and back office skills coupled with trading teamwork. If the process runs properly, the use of simulation money will be irrelevant to the learning experience.
Each class will have approximately 25 members. Because there are six four-hour candlesticks, the class is split into seven teams of three to four members each. Six of the teams are responsible for data collection, trade initiation and trade exit; the seventh team supervises the other six execution teams and performs back-office functions such as trade documentation, profit and loss reporting and data analysis for risk management and preparation of investor reports. The Hanley Group Derivatives Trading Center is the hypothetical owner of the portfolio, and the student teams furnish the center with daily and monthly profit/loss reports.
Operationally, each student stays on one team for one week, and then assignments are rolled forward: the executing 3 a.m. team becomes the executing 7 a.m. team, and the executing 11 p.m. team becomes the management team. Each of the executing teams is responsible for one of the day’s four-hour candlesticks.
At the close of its candlestick, the specified executing team inputs that candlestick’s data into the spreadsheet model. Based on the continuously updated girth threshold (as monitored by the professor and overseen by the management team), the model then indicates one of three options to the executing team: close the current trade, enter a new trade or take no action.
In the event of no action, the student running the model updates the model and e-mails it to all members of his team and to the subsequent executing team in the candlestick time frame. In the event that trade entry or exit is indicated, the executing student immediately notifies another member of his team, who independently enters the data to confirm the action. After confirmation, the first student takes the indicated action and e-mails the updated model to his and the subsequent executing team.
It is obvious from the chart in “Rolling forward” that the executing 3 a.m. team may bear the most executing operational risk due to the time of the candlestick close and the sleep cycle of most students. Further, the forex market trades continuously from Sunday late afternoon until Friday late afternoon, so there is no market close during which to discover and to correct errors through the week.
Another significant operational risk is the passing of a contaminated book: a model that has errors in it. To mitigate this risk, each morning the management team confirms the accuracy of all the data entered into the model over the last 24 hours, prepares a daily profit/loss report that combines both cash and mark-to-market positions and reports any anomalies to the fund’s faculty supervisor and
The managing team also is responsible for preparing weekly reports summarizing the fund’s closed trades, floating P/L and other parameters of the fund’s performance, including any operational risk occurrences or human errors.
At the end of New York trading on the last trading day of each month, the fund closes any open position and goes to cash. The managing team is responsible for this action and for the preparation of the end-of-month performance and investor return reports. The end-of-month liquidated positions may be re-established at the beginning of the new month next trading day, if girth and trends dictate. These steps have been implemented to mitigate any issue of an operationally contaminated book.
Risk management is largely intrinsic to the girth model, as the model never enters trades against the four-hour trend. The model instead is traded with the trend, after the trend is established, and then looks to make an early exit from
The girth model works best when there are strong trends during the four-hour time frame. The model may churn or yield no profits if the market is not trending during the four-hour time frame. However, the trader will never be able to hold a long position against a short trend; thus, it inherently limits losses. Further, each trade will be entered with a hard stop loss, which is determined by the volatility of the market. If normal market historical volatility dictates 80 pips of noise in the market, then an 80-pip stop loss is set. However, if historical volatility has increased two fold, then the stop loss will be set higher than 80 pips as a factor of volatility.
Implementing the girth model also gives students a chance to calculate value at risk (VaR) and to monitor VaR over a floating P/L position. Of course, the actual calculation of VaR needs to be taught before such an undertaking ensures.
A strong sense of discipline will also be instilled via reporting, which will be performed on a daily, weekly, monthly and end-of-semester basis. Basic reports will be posted to the Web at the same frequency. Students will be required to produce updated analyses of volatility, VaR, average trade length, P/L (cash and floating), maximum drawdowns and any deviations from the trading model. At the semester’s end, the entire group will issue a written report to the center summarizing the fund’s performance. The group will prepare a professional road show presentation, comparing the fund’s performance to various benchmarks, and evaluating correlations between the fund’s performance and that of other investment types.
There is a lot to learn and a lot to implement before the girth model can run smoothly. The process has a large risk of human and operational error, even after the requisite knowledge base is obtained. For these reasons, running the model in simulation to build experience is prudent. The move to trading real money depends on how well the simulation goes.
Leslie K. McNew is a clinical professor of finance and visiting director at the Hanley Group Derivatives Trading Center, Department of Economics and Finance at the University of Dayton Business School. E-mail her at email@example.com . Dr. Tup Ingram is a private trader. E-mail him at firstname.lastname@example.org .