I'm with you there, just waiting for a signal to rebuy my small physical gold. I like to turn it over anyways for the cc rebates so any market timing benefit is just a perk.
Traditional TA type analysis is kind of buggered, the reason being is because too many people know about it. Would not surprise me if gold drops thru support, screws the majority, then bounces. My opinion atm, no buy signal quite yet but havent looked yet into more detail, will do that later today.
After running today's data... ES, NQ, metals are still saying short. Bonds were mixed yesterday, back on sell now. NG has been on buy for a few days now, let's see if it can take off Monday. Forex, I plan to go short AUD/USD on Sunday night. I'm short GBP/USD, long CAD/JPY, and short USD/CAD, all still look good. I'm long EUR/USD but after todays data its kind of iffy. I think I will hold it until Monday close and see how things look. Things get weird with taking multiple positions, with different pairs cancelling each other partially out. I'm not going to worry about it and just keep taking trades, trusting that the net position is correct based on my model. For stocks, I'm still working on my trade plan. I'm torn right now between entering at open vs using various methods to set a limit order. I need to run more data through before I decide on a course of action.
So I spent all day today finishing the process of moving all my calculations/data gathering and wrangling from excel to python. I discovered that I had a bug in my excel stuff that did indeed give me a bad signal on EUR/USD. No big deal but it is early by a day best case. That's why I'm keeping it nice and small until I have good experience with what I am doing. It is amazing though to replace 30 minutes a day of copy and pasting and fill down and right with a keystroke. Why did I wait so long to learn programming lol. The other trades look good somehow a bug had made its way in to just one of the rows on my excel files.
When I was algo writing/trading a few years back, I set my algo computer on idle while I went to work in the morning. There was a bug in the coding which backtesting hadn't picked up and it placed a buy long trade after I had walked out the door. The long trade was at the very top of the day and from there on the futures market went down. Just to add more pain to the equation, as I was leaving work to come home, I received a long phone call in the car park which held me up another hour. Came home to the disaster, lost substantial dollars.
I'm pretty paranoid about that sort of thing. That's a big reason I'm starting with forex, in order to trade as small as possible while I learn. I closed the mistaken EUR/USD position and just got short AUD/USD as well. Along with learning to program right now, I am still kind of getting the hang of trading this style. I think I am starting to get a better understanding of how to interpret my data. I should be holding my winning trades a little longer going forward as I get a better feel of how a good trade should develop. They seem to having something of a life cycle barring any shock event. The amplitude and duration of that cycle is of course completely unpredictable IMO and can only be monitored and guessed at in real time. Open positions: Long CADJPY Short AUDUSD, USDCAD, GBPUSD
I was attempting a fully auto system, the coding gets quite extensive and it's amazing how many coding bugs you'll discover, sometimes backtesting just doesn't reveal them. On hindsight, if I were ever to go down this track again (which I doubt), I would create a semi auto algo, one which supplied signals but wouldn't be part of placing the trades. Maybe one which would sound off audible alerts only. The reason I'm no longer going to, is because from Australia, IB is the only one suitable platform for this type thing that I am aware of and there is a snowball's chance in hell I would use IB again.
Right now my system consists of some extensive data analysis performed nightly in Python, from which I make subjective decisions on my trades. The algos I've worked on are intended to get me better prices than a market at open order. I'm not sure they provide any value and right now I'm just entering at market as soon as I pick a position. I don't even look to see where the product is trading. I'm really busy with my real job right now but when I have time I intend to work a lot more on that.
I'm doing something very similar, except I'm still forward testing and not taking real trades yet. My system is objective; if a signal is generated then I take the trade and hold for 10 days in the test phase. So far the results have been very promising, but I might continue this process for a few more weeks until I'm entirely comfortable with the system.