Are there any successful system traders out there? I would like to hear from automated traders who have been running profitable systems for more than a couple of years. Maybe they can advise on some of the pitfalls they encountered? I look at my strategies and constantly think either they are too simple to work or over-fitted and so forth. I have been running live for few months now and so far small profit but there have been many times where i wanted to override a trade. Sometimes i think maybe this is all a big waste of time and unless you have real edge like HFT or arbitrage then odds are this is random walk. I still have my doubts about my systems. Maybe someone successful can share their story. Would love to hear some success stories for inspiration.
Even shiller p/e is curve-fitted, that should tell you something. the walk forward since publication of shiller doesnt match up at all with backtested history. FAIL.
No I mean maybe it is possible, but lets just say the guy wold sold the idea of shiller made more money than the guy using shiller for their investment decisions.
I'm still trading my system(s) manually to make sure they perform correctly, but I find the largest problems I have run into are in regards to slippage. A good example is if you have a system that closes on the close of the last bar of the day .... getting this fill price in realtime almost never seems to happen. That being said, my system typically performs better without my intervention. Sometimes I'm right, but I find more often then not I end up chasing price in the wrong direction in an attempt to get a better fill price.
Why don't you just trade it automated and then keep note of live fills vs what the sim fills would be?? Also, you have to expect slippage unless it is a limit order. You could try entering a market order 10 seconds before last bar end of day closes, still you will eat 1 tick slippage. So far we have no successful automated traders post, this is disheartening
I personally have a rule that if I can't do it in real time then I'm not willing to trade the system. I find that by trading systems manually first it allows me to test and modify them before incurring large losses or to find small improvements to increase the success/profitability. An example of this is that recently I was getting stopped out only to watch the price bounce right afterwards. I increased the stop amount and although the max draw down is larger (still within a tolerable amount) the win rate jumped up by 12%. I could have found this by testing every combination of parameters but that would be curve fitting and I try to avoid that. When designing automated systems we are using historical data for backtesting, this isn't the same as trading live. Just because a bar hits your fill price, doesn't mean you got filled. I find it's best to use a tick above the fill price as the criteria. Also if using indicators for calculations then the value are changing in real time and can result in false signals. Bollinger bands are bad for this. They work great in hindsight, but trying to decide realtime leads to a lot of false signals in some systems. I know some people trade automated systems quite successfully, just like some people trade price action and some trade moving averages and some trade combinations of techniques. Everything works sometimes and everything fails sometimes, but to me it seems like the most successful traders are the ones that are confident in their system. Trading is hard, and that goes for all types of trading.
You can't just go ahead and change stop price after being stopped one time. That means your strategy for stops was useless and not backtested thoroughly or you had no faith in it. Also, seems like you have a fixed stop amount which will probably fail because i find most dynamic stops are better than static ones. With static stops it is super easy to over-fit the data - anyone can move the stop a couple of points to miss the big losing outlier trades in backtesting. In backtesting you can use software like Multicharts to assume you only get filled if price trades through your price so that assumption is false. You can also add slippage and commission into you backtesting. I don't use indicators so i would not know how they function in real time. I think if you have any manual system that has not been backetsted then you are probably going to fail unless you have actually traded live with it for more than a year. Still no successful traders willing to chime in????
Perhaps all you can do is run your stuff in real life (sounds like you've done that so you're a step ahead already), then see if it does what you expected it to. If not, then you've got to think hard about why. If you start interfering manually with a fully automated system, you're doing greybox trading, which is fine in and of itself, but keep in mind that you may be violating the assumptions on which your system was built. Only you can answer that. The "secular trend" is one of ever-increasing competition. "Everybody's" code is getting better over time, or else they've quit. In fact, many have quit/are quitting, as it's not getting any easier, on average. So who wins in an ever-more competitive short-term trading environment? The answer is longer-term investors and the companies they invest in -- it's an increasingly efficient market. For better and for worse, that means that trading gets harder in the short term, unless you are a broker and have your own clients to trade against or whose order flow you can sell to "wholesalers", which (unfortunately in my opinion) is legal in many markets, such as US equities.