Hi All, I figured Id say hi and introduce myself since this would be the sub forum Ill spend most of my time in. A little history, Ive traded off and on for about 8 years, never consistently profitable. I have finally put my pride away and accepted that trading without a system with strict rules my personality will make my account implode. I tend to have good ideas, and good predictions but its only a matter of time until my demons come out (ADD). ADD is tied to impulsiveness, impulsiveness does not do well in trading. However, ADD people do very well with routine. A system gives me that and gives me some ease in my mind when I trade this way. Having said that I have developed an initial Strategy and have made it more complex tweaking with parameters. Its been very interesting to figure out the reason profits/ wins/ drawdowns etc. move when variables are changed in different ways. But there is one I found that cant explain and was seeing if there may see someone who could come up with something. Simply, my system is a more elaborate buying/ shorting of big daily movers. I first back tested for 5 years with great results. 1st year a small gain and then large yearly gains after that. I then backed it out to 10 years and something interesting occurred. The first 5 years were pretty much flat. After a little investigating it seemed that my criteria was hardly triggering any trades during those first 5 years. 80% of the trades for the last decade occurred in the last 4 years. Anybody have any insight? Has the US stock market turned more volatile overall in the last 5 years with bigger daily swings?
I'm not an idiot and thats the first thing I did, the VIX isn't everything. Backtesting isn't that simple. But I appreciate your extremely charming hypothesis to a new poster....that post count must make you feel virtually very big and powerful. If anyone else would like to have a logical discussion and toss out some theories I'm all ears. If it helps there was one small spike in activity after the 2008 crash and thats about before the last 3 years.
I can only offer a hunch based on everything I've read. First, I read that many traders after 2007/2008 didn't do so well.... something to do with HFTs entering the scene. I have also read that as of late, many traders are finding it difficult to make money and are hurting. EPrado, a member on here just today described how things used to be much easier and those who survived had to adapt. So my hunch is that backtesting the exact same strategy with the same parameters for 10 years isn't the best thing to do. By tweaking some variables, I imagine that the first 5 years of the back test would be better than the last five, and vice versa. Because trading seems to be evolving so quickly, I just don't think that one set of variables can be expected to work so well year after year, especially these days with so much automation. I don't think a strategy has to be changed, but the values of the variables have to be tweaked.
The thought about HFT crossed my mind but I wasnt sure if that causes larger short term moves or smoothes them out. I think I recall reading somewhere that HFT has been smoothing the market out at a high level, maybe not day to day. (Minus the oops Flash Crashes) Admittedly I dont know much about HFT. Ok so a followup question to your thought, at what point do you say "Ok, this isnt a drawdawn, my system needs to be updated" My drawdown is a little high at 25% but Im ok with it. If it hits say 30% and out of that statistical range is that when you reevaluate? Overall, at least its pulling a good return on the recent data.
You drawdown could be 50%...but that's not (that) bad -- if you have upsides of 150%+ ...that's what you need to establish. that kind of thinking or formula. What you don't want, or want to be doing...is essentially...just betting on a color on the roulette wheel or playing blackjack.
I wish I could help with your further questions but its too far out of my league. I don't understand though how a person can have such a big drawdown. I mean take for example a day trader who puts on 3-5 trades a day. If on an average day he makes 1k, and perhaps win 3 out of 5 days, then on the 4 day is BE, and on the 5th day maybe loses 1k. He should be up 2k per week. Sure he can have a series of bad trades in a row, the number of which should be related to his win %, but given that he puts on 4-5 trades a day, that is 20-25 trades per week. To have a really bad week, he would have to lose almost every single trade! Now I take it that if most are swing trading, risking 2% of their capital on every trade and have 5 or 6 bad trades in a row, then I can see how a drawdown could be 10%.... but anything more than this just tells me that the method needs work.
Do you mind posting more stats? It's hard to even try to guess, otherwise. I'll help stir up some ideas for you.. 5 yrs vs 10 yrs - How many signals were there? - How many were long vs short? - What do the curves look like? - avg gain vs avg loss? - max gain vs max loss? - if you plotted the losses, what would that look like? - what about the gains? - what about for only shorts? - longs?
I believe its because I had to increase my % of equity per trade as it does not generate a whole lot of trades even though they are low risk, high reward. This causes wilder swings. Ideally, I run this system and develop a second later that also generates low risk trades and run both at the same time so I get more trades. Ill be able to calculate the combined stats of running both (Or maybe this is considered just one system with multiple strategies ?)