I start this log by going back a few months and describe in a few parts how I arrived to today. This is part one. In September 2016 I received my copy of Robert Carver's book "Systematic Trading: A unique new method for designing trading and investing systems". I had meanwhile read most of his blog and also his trading results as described on this forum (@globalarbtrader , see here: https://www.elitetrader.com/et/threads/fully-automated-futures-trading.289589/). One of the things I learned from the book was how to manage risk in relation to the account size and how to select the appropriate instruments based on this analysis. After reading the book I was convinced that I wanted to make my version of the described system in software. However, I was not able to understand the provided Python sample code properly. Somehow I don't particularly like the Python language. Instead I implemented my system in java as I have some prior experience with it and know how to write a simple program to connect with Interactive Brokers' TWS. I have an IB account. In parallel I also tried some paper trading of futures contracts as I had no experience with those instruments. After reading the book it took me two weeks to write the software, followed by one month paper trading to get the bugs out, and to get a feel for how the system behaves. Then I started live trading mid October, with a small account (< 25 k USD). Because my account size is small can only a limited number of instruments be tracked. Those that are within reach of my account can be used to open a position. I don't bother with tracking instruments that are far outside my reach. Of course did I expect the account to grow, so I started monitoring a few instruments which were just outside reach but of which I assumed that they would get within reach in not a too distant future. I start the program once per trading day and close it some 18 hours later. Instruments I track are in Korea, Germany and the USA, which makes that the program needs to be active for so many hours. Both during and after running it am I able to adjust certain parameter settings. Most of the time is the program dormant, it comes to live a few moments per day. For each of the instruments which it tracks it evaluates near the end of the regular trading hours whether trading is required. It uses the intraday price at that moment as "end-of-day" and combines that with historical close prices in order to make the calculations and reach a conclusion. The forecast value is calculated based on a trend following method and a breakout/momentum method. A few variants are used for each of these. The value volatility of the instrument is also calculated and compared to the total account value (net liquidation value). Both items are then combined to reach a conclusion on the desired position size and trading is done if the desired size is not equal to the actual position size. Trading is thus limited to once per day for each instrument. The terminology "forecast value" and "value volatility" should be read in the way as it is defined in the book. Disclaimer: I have no affiliation with the author or publisher of the mentioned book. I receive no compensation for reporting my results here. Your mileage may vary. This blog is not intended as trading advice to buy or sell certain products. It is purely for entertainment purposes. You know, all the usual disclaimers....
This is part two. To be able to trade futures is a funded account required. However, the cash is not used for these positions, only the margin it provides is required as collateral. I wanted to use the available cash to get a bit more return. So I developed a derivative program, based on the basic propositions of the book. It uses a list of approximately 40 ETFs and selects the ones with the highest forecast. The cash is used to open a position in these "top performers". The size of each position is adjusted for the value volatility of these ETFs, and 80% of the total account size. The positions are therefore not equal in size. I manually run this program once per trading day. It uses end-of-day close prices and does not include the intraday price at the moment I run the program. Trading is only done if a ticker enters the list of top performers or if a position gets closed. At that time are all other positions rebalanced, based on their value volatility and the account size. This keeps the trading costs low. The ETFs I use are a combination of various classes, such as US industry sectors, equity indices of various countries, currencies, bonds, commodities, etc. From this description is it obvious that this is a long-only strategy. It could be expanded into a system with both long and short positions, by using the ETFs with the most positive forecast for a long position, and using ETFs with the most negative forecast for a short position. However, not all ETFs are available for short selling at IB at low costs. Until now have I not been motivated enough to try and develop such an expanded system. I started running a first version of this program in October, shortly after I started live trading of the futures. Besides finding the top performers every day it also rebalanced daily the positions based on their price volatility. After a month I concluded that this daily rebalancing is costly and doesn't add to profitability. So I removed this. The current version of this system runs since mid November 2016. It provides thus far a positive contribution to the account size. In other words: it is profitable until now. However, holding these stock positions requires some margin, implying that less margin remains for the futures positions. Another risk is that the cash is invested in stock which results in value fluctuations of the account. This fluctuation adds to the fluctuation caused by the futures trading system, thus the overall account value volatility is increased. These two aspects cause cross-coupling between the futures trading system and the ETF stock trading system.
So, what have I achieved by the end of 2016? The account net liquidation value went up by 24% since the start in mid October. Not much happened in the first few weeks but the account size grew in November and December. During these two months I had to adapt a few times the weight percentages of the futures instruments because their value volatility fluctuated (and my account size changed). Meaning that they were sometimes within reach of my account and sometimes not. After I stopped the daily ETF rebalancing the ETF stock system turned out to be very stable, trading only three times since mid November. Thus far has it delivered a positive contribution to the growth of the account net liquidation value. And the unrealized profit (UPnL, profit of the current open positions) is also positive. Although it does give a contribution, most of the account growth is caused by the futures. Now that I have described how I arrived at this point is it my intention to write here every now and then my progress. And the experiences I come across while trading.
I can't say whether it is "most" or not. Don't forget that I only started by mid October. The system uses both trend following as well as momentum/breakout trading rules.
Market showed a great trend in Nov/Dec and any trend following system will make money. However, trend following/momentum strategy will loss money in side way market. Unfortunately the market is random for day trading when you add ALL of them together, and all the trend follower or revert to mean traders will be break even in long term, but they all loss eventually due to commission and slippage. My point is you have to be careful not to come to conclusion too early that your system is profitable.
If someone believes such...wouldn't it seem obvious that a trend trader should only trade markets that are trending instead of staying with the same market that will have those durations of sideway markets ? Simply, a trend trader would trade different markets and not get married to one market especially when its going sideways. I wonder about this because I actually saw someone trade the same market for 12 years and although profitable by points...he was actually a losing trader when commission and slippage was added in. Thus, it would make sense for him to find a different broker with better commission rates or don't trade the same market especially when its in a sideways price action. I'm just thinking out loud on this because I came across a blog recently with similar like thought as you just expressed and it got me thinking about what could be a solution.
I think he is doing more than fine and by trading ETFs instead of futures, so less to none leverage, I wish more younger traders did this. Let him build up account over $100k then slide some into futures for more bang for the margin. Problem with some ETFs is lack of volume. Futures where margins are less than $1500 are best for younger trader, they move less but on small accounts move well enough till trader gains knowledge.
Thank you for your warning. It is not my intention to make any definitive statement on whether the system is profitable or not. Let's see what the future brings.
Are you referring to me? Most of the volatility in my account is coming from the futures that I'm holding. The ETFs I'm holding generate less fluctuation in the account liquidation value. ETF liquidity is not really an issue for me as I only selected the "mainstream" ones.