Welcome to my journal. Each day I’ll post my daily statement and chart showing executed trades for my trading on MES (Micro ES futures). I’m UK based, and my broker of choice for micro futures is AMP Global. I’ve deposited just over $1000 ($1064.47 to be exact due to currency conversion). I’ll be trading one micro contract with that balance for now, attracting only $40 margin. Yes, I’m a little fish, in a big pond. But I am sure there are plenty of guys and gals like me here on this forum, so let’s see what we can do with a grand. Some background. I’m predominantly a data & software guy, but love trading. So this is not a full time gig for me, but maybe someday, who knows. I also trade FX, currently only GBPUSD, for which I have a suite of algos that do my bidding, all custom developed by me. Also working on forward testing a DAX algo. Hobby trader? Yes, I get to combine my software, data and trading interests. My MES trading here will also be handled by an algo I’ve written, running on MT5 with the aforementioned AMP Global. The algo has been put together over the last week and is released into the wild today 20th October 2020, to start this journal. The algo is fairly basic; but I’ve found simple is best most of the time, it’s not over complicated, nor over-fitted. It trades the MES 5m chart looking for mean reversion opportunities. We should see on average 5 trades a day. Given I am UK based, I run the algo 1300 GMT to 1800 GMT covering pre-US open through the first 5 hours of the US trading day. A word on risk. I know all the rules, 1% risk per trade, blah, blah. It’s my money, and it’s my journal, so I get to decide the risk, agreed? Okay then, risk is 3% per trade. Why? Well, the algo has a stop loss of 24 ticks, and a take profit of 16 ticks. The maths gurus among you will know that’s a Risk:Reward of 1.5:1, or a cash risk per trade of $30, which is 3% of my $1000 starting bank. And given I cannot trade any smaller than 1 micro lot, it all kinda works out nice and even. The pro’s among you are probably shaking your heads, “here’s another guy with a negative RR and way too much risk per trade”. That’s fine. Like I say, it’s my journal, and my money, if you don’t agree, just watch and wait for your “we told you so moment”. Otherwise, enjoy the ride. TraderDaveUK
Actually, let me give your question a little more attention. There are 2 scenarios I'd see the algo handling. 1. I am in a trade at the time of a sudden move. It either stops me out, or hits my target, albeit subject to slippage depending on the move. 2. I am not in a trade, in which case, a sudden shift from the mean will offer the algo multiple opportunities as and when price reverts, some of which may not be successful. Either way, my back test has not highlighted any particular problem with these moves, but hey, via this journal, you'll get to see exactly what happens!
i think this is demo or backtesting i tell you why: the fills are often at the very top and bottom of the bar, that is indicative of using limit orders with an automated system. the system will go back in time and fit the fill to the data, when backtested. in real trading the fills will not occur as they do in real time. hope i'm wrong but that's what i see. also for a trade to enter and exit on the same bar is highly suspect as in backtesting there is no way to know if the high or the low was made first. so again the platform will pick best case scenario and fit the trade in. i see one of those on your chart.
Hello MarkBrown. So, I post one day of results, that show a profit, and it must be demo or back-testing? I do like a healthy dose of scepticism to start the day But seriously, it's a live account with AMP Global, I've blacked out the account number and some personals: And here's the AMP statement for yesterday: For what it's worth, I am not here to convince the world that I am some master trader. As I noted in my first post, I am a hobby trader, and this account is simply a way to hopefully show a return - not a living - can also be made on a small account. As to your point about the trade that entered and exited on the same bar, it's entirely valid. An MT5 algo evaluates on every tick. I include code in the algo to evaluate only on bar opening, i.e. evaluate and execute an order once per bar. You are referring to this bar: Which is this trade: The algo entered right on the open of the bar at 1500 (chart time). The open of the bar was 3436, the order was filled at 3435.5, 2 ticks off the open. The bar high to low range is 42 ticks. The algo has a stop of 24 ticks, so stops out within the 5 minute bar. I cannot be any clearer than that. Hope that addresses your points. I've tried to be as open as clear as I can here.
Good to see a software person start a journal. Wondering what you hope to accomplish by it. Just posting trades will result in yet another wasted thread with no learning whatsoever. You don't have to reveal your top-secret alpha generating secret sauce of course but here are a few suggestions if you want the community to engage and learn in the process: 1. What is the real world rationale for the algo. 2. Why do you think it generates alpha. You mentioned mean reversion but that generates a lot more questions. Why this particular set of risk vs. reward. 3. What problems did you run into while implementing this and how did you solve them. 4. Did you do any back testing that leads you to believe this will be profitable. 5. What sort of back testing did you do. 6. What type of data did you use and for how long. 7. What were the performance results of the back test. 8. What do you need to see in live testing order to continue using it or abandon it. 9. Are there any kind of volatility filters that you applied during your back test. 10. What would you do if you start seeing results that go outside of the bounds of the back test. 11. Have you tried testing during different periods in the market and found the results to be affected. Such as testing during US quiet periods for instance. etc. etc. It would be a fun exercise to engage in as a group and try to brain storm some answers. I know the fear of revealing too much prevents people from getting the full benefit from engagement. But I don't believe any solid idea based of market fundamentals can be affected by others gaming it. The more an idea is based on market fundamentals, the more it will benefit from increased participation. Good luck with your algo. all the same,