This system was inspired by my last system, Trading Idea #024, which is a mean-reversion system on a monthly timeframe. As I built and tested that system, I wondered what would happen I just took a one-bar trade on a monthly chart based on 2-period RSI stochastic coming out of oversold territory (over 10%). In other words, a bounce from oversold.
The general thinking here is that as price moves out of oversold, we are just jumping on the momentum train and picking up a quick and easy win. The reasoning behind the monthly timeframe, as I have stated elsewhere, has to do with the holding period constraints (minimum 30 days) with which I must comply.
One month, in and out. That’s it! I will be working with equities, as I do not have enough historical data yet to try this on futures.
Phase 1: Plan & Design
1. Trading Idea
The big idea is to capture the first month’s move and then get out:
- 2-period RSI for entries
- Long entries only
- Entry on the first day of the month
- Exit on the first day of the next month
Here are the entries:
- RSI(2) has moved out of oversold, using the 10% line as our oversold threshold
- Enter next bar at market
2. System Definition
RSI, 2 period
|Risk_Capital(2500)||Integer||Yes||Amount of capital required for each trade|
|Leverage(4)||Integer||Yes||Multiplier for risk capital|
|Stop_Loss_Offset(.02)||Integer||Yes||Amount subtracted for the latest monthly Low price for setting our stop order|
- Risk_Capital = $2,500 per symbol (purchasing power = $10,000)
- Number of shares = Risk_Capital / Latest Close Price * Leverage (rounded down)
- RSI >= RSI_Oversold and
- RSI [1 bar ago] < RSI_Oversold
Set stop to $.02 below latest low
- Exit long position after 1 bar
3. Performance Objectives
The system will meet the following objectives:
|System Type (trend, mean-reversion, day, swing, etc.)||Mean-reversion|
|Risk of Ruin||0%|
|Win Percent||> 50%|
|Max Drawdown %||< 35%|
|Profit/Drawdown Ratio||> 2.0|
This idea is S.M.A.R.T.: Specific, Measurable, Achievable, Realistic, Time-bound
4. Market Selection
|Equities||Random equities from S&P||AAPL, ABT, ADBE, ADI, ADM, ADP, ADSK, AEE, AEP, AES, AFL, AIG, ALL, AMAT, AMGN, AON, APA, APD, AVY, AXP, AZO, BA, BAC, BAX, BBY, BDX, BEN, BK, BLL, BMY, BSX||I selected all symbols starting with ‘A’ and ‘B’ that have been part of the S&P 500 since 2000.|
Chart Type, Timeframe, Session, Time Zone:
|Chart Type||Regular Candlestick||Charting is only useful for validating entry and exit signals|
|Timeframe / Interval(s)||Monthyl|
Phase 2: Build
5. Manual Test
Pass. The handful of random instruments I selected seemed to work just fine.
It does not get much simpler than this.
7. Unit Test
Phase 3: Test
We are not optimizing any parameters for this system.
9. Walk-Forward Analysis
For this system, I performed walk-forward analysis by placing all symbols in a portfolio and then run a backtest, from January 1999 to December 2020. The reason for the long backtest period is two-fold:
- I am working with monthly charts, which means fewer signals
- The more data used for testing, the better
Personally, I would like even more data than that, but that would complicate things too much, given the amount of time I have to develop and analyze this system. Below is a table of the symbols tested. I selected only those with a profit factor of greater than 1.5, which leaves me with 14 to advance to incubation.
Here are some additional thoughts:
- Profit Factor for this set of instruments, through December 2020, is 1.14. As a portfolio of random symbols, this system does not perform well.
- Average number of trades per symbol: 12, which is less than one trade per year
- Highest number of open positions: 20; this would require $50k in capital ($2,500 * 20), which will have a bearing on Monte Carlo analysis
10. Monte Carlo Simulation
There were not enough trades for some of these symbols (as few as 2 trades in 20 years) to perform a reliable Monte Carlo analysis, so I took all trades for those 14 symbols, treated it as a portfolio, and ran the Monte Carlo program. I used a starting equity of $50k, since that is maximum amount I need available for trading. Here are the results:
Monte Carlo fails because the Return to Drawdown ratio is less than 2.0. Even though it failed, I will look at the incubation period (2021) for academic purposes.
Although the system has failed, I took the 14 symbols and put them in incubation. Here are the resulting trades:
There was not a lot of trades produced by this system and it performed poorly. This would not have passed incubation.
Phase 4: Deploy
12. Production / Portfolio Assignment
We did not make it this far.
Trading System Result: FAIL
Notes and Commentary
The system failed, but I cannot know until I try it. I have built several systems like this, where I am bouncing out of oversold using a very short RSI. This system has some problems, but here are my top three:
- Too much capital required at certain times
- Low profitability
- Infrequency of trades
This is what I have to deal with on monthly timeframe. I think that maybe, just maybe, mean reversion is less meaningful at this level.
Continuous Improvement Department:
Overall, this system idea is okay, but not worth it. Here are some thoughts about what I can do with what I learned:
- Step away from the quick exit and try longer holding periods
- Filter for symbols with only a positive earnings-per-share (EPS)
I am finding it difficult to get creative in the longer timeframes, but I have some thoughts around breakout systems.
This is my last system for 2021. I have built and tested 25 systems and have had 4 pass the criteria for a potential winning algorithmic trading system. This is a 16% success rate, if you are keeping score. I expected much less, to be honest. Regardless, I will keep building these systems and publishing as often as possible. I suggest subscribing to my newsletter to get the latest system ideas in your inbox as I publish. As always, thank you for reading and following along on my journey. Do you have a trading idea that you want me to put through our process to see if it would make a valid trading system? Leave a comment below and we can discuss.
- J. Welles Wilder, New Concepts in Technical Trading Systems
- Relative Strength Index (RSI) – Investopedia
Did you like what you read? Do you want to see more? Subscribe now and receive our email twice per month, with a new Trading System Idea and other fun stuff.
Don’t worry, we will never, ever, ever sell, overuse, or donate your email address. Promise.