For a change, I'm rather surprised by ET. The bulk of the advice given to you has been on the money. I retired young and have traded for 10+ years in order to augment various income streams in order to pay the bills. I never made a lot but achieved my goal of not tapping into my nest egg, except for an occasional big ticket cash purchases like a new car. I investigated every trading technique that I could understand (TASC, SFO, Active Trader, Futures mag, etc.). Many were over my head. Two years ago I found a silver bullet and that worked for 22 months. 55 straight winning weeks and no serious drawdowns. Six figure gains two years running. For some, that may be pocket change but for a non professional retail investor/trader like me, that's a realized fantasy. If I wasn't making good money, I would have needed an intervention. 6-1/2 adrenaline rush filled trading hours a day and another 4-6 for data maintenance, record keeping, next day prep work, etc.. I couldn't wait for the next day. I hated weekends. It was the most fun I ever had with my clothes on I'm not bragging. Additional financial security is a dream come true. Along the way I surmised that it would end when all of the bailout and stimulus money had its effect. And it did. And now I feel like the Maytag repairman, waitng for the phone to ring or for the next financial crisis. I have no clue if your silver bullet will last. But your have to assume that it won't. Make hay while the sun shines (your swing trading) but don't quit the day job until you can afford to live on whatever your non trading income stream will be. There will be lean times. Me? While hoping my silver bullet works again, I'm looking for another one. Maybe lightning strikes twice?
SM, I've waited a bit to make a written response to you regarding this issue. I think you've gotten many good responses here and I'm glad you're getting a bit of a reality check. There are several issues with the model you have. Foremost, let me say there are 100's of ways to make money trading. Edges exist everywhere and many are very reliable and will never go away. With a few years of commitment and research you will find these edges. Once you get past the systemic analysis learning curve, many more edges will be revealed to you. That said, I didn't want to be harsh when we spoke, but, you do not have an edge. Your model is where thousands of retail traders start - i.e. indicator based MACD + Stoch. crosses ... yada yada. Market edges do not exist in these indicators alone. Had you actually done the research and testing I suggested, you would have seen the fact that such simplistic rules may produce good results at random periods of time. To prove my point, I can show you an "edge" based on random +/- ticks that will produce incredible results 30% of the time when starting with a ceratain random seed variable. This system is literally random at its heart and given the user's sample space, it may appear to be the holy grail. At this point, I know you want to make all this money and its tempting to give it a shot. I wish you would reconsider and do the required leg work. Building proper models takes a hell'va lot of time and testing. If you do chose to go ahead with what you have now, you're better off enjoying that money at the casino... I say this because if you truly were to approach this endeavour as a business - you would not be risking your hard earned money on something that is "iffy" at best, a smoke and mirror show at worst. You're falling into the trap that many fall into, you're not seeing the big picture and the long term potential side effects of blowing your load before you have enough experience. Get the right tools, learn proper testing procedure (Pardo has some great books). Learn everything you can about random walk and binomial distributions and the apply what you've learned watching markets the last 5 years or so. There are no shortcuts and there is no "get rich quick" and there is no holy grail in this business. Above all treat this like a long term business plan - NOT like something you can just toss your cash at to see what happens. Anything less is failure lurking behind every corner. Mike
Mike, I did. I did appreciate your feedback. And I did sign up with stockfetcher.com and taught myself backtesting. I then started tweaking. I swapped out some of those indicators, and also added in bollinger bands to better assist me in seeing the turn. I then back tested it against some of the indexes themselves and it worked great. Made 16% in two weeks, which I think is pretty good considering how compounding works. This is without exposing myself to any great drawdown risk because I'm only using ETFs. It has not given me an incorrect answer since October 23rd and correctly identified two market turns on the days they happened so I could change my holdings. A few things you don't know about me is that I have been trading for the better part of the last year. I haven't blown out my account and I have make hundreds of trades and even a little money using my old system. My old system is now used to supplyment my "silver bullet". While I may not have the number of years under my belt that many have, I think my knowledge is in the middle in the pack because I've put a heck of a lot of effort into learning, and math is my strong suit. I am an engineer, and I have an MBA, and both of those in concert should/did help with my learning curve. While off the shelf indicators may seem pedestrian to you, they can be the basis of great trading. But you have to set them up right, understand their limitations, and supplement their shortcomings with other indicators. Getting into specifics, it was my idea to share my algorithm with you in return for some techical assistance. But you only ran it once and then recommended that I learn coding myself. What the hell, man? I had other ideas to improve it. You were correct that it didn't work well over all timeframes, but there was no follow through tweaking. And now you come here after all this time and tell me that my idea is going to fail without seeing the benefit of that tweaking or cash management anywhere close to how I had envisioned it? I hope you don't think it will work because I'd prefer you don't use it...you didn't put in the hours. But I'll tell you one more thing. You can talk of random walks and binomial distributions (and yes...I have studied them...and programmed them...I'm an engineer also), and yet it seemed you couldn't understand that something doesn't have to work across all of time...it only has to work long enough to get a fellow where he wants to be in life. Never-the-less, I'm extremely grateful that you did that because I could see in black and white that I did get a buy signal at generally the right times over a 6 month span. I'll post a picture of my "smoothed" buy signals so you can see how much better it is now. The perhaps you can post one of your own. SM
SM, My wife has mentioned that I sometimes sound a bit harsh with my writing style - I apologize if I came across that way. My intention was not to be harsh, but rather present an opinion that you can take however you wish. I've been in this business a while and seen quite a few edges, some good and some bad and everything else in between. I don't copy other people's work and I certainly didn't want to discourage you - I simply lost interest in the concept because I've seen 100's of permutations of the same exact thing. I'm at a point in my career where I need new and different intellectual challenges to keep me interested, otherwise I prefer to spend my limited free time doing things I find more enjoyable. The technical assistance I offered was a recommendation to doing quick and easy backtests - it was not meant to offer/develop the strategy details, that takes a lot of work. I ran your ideas once in an attempt to show you how easy it was to do. The serious development is up to you - test, evaluate, test again, possibly trade live, retest, again again until you see exactly what you're looking for, the process is highly personal, takes a lot of time and its something everyone should do for a few years IMO. I just don't have the inclination to start that process with any indicator based model. I probably come across as pompous or set in my ways... well, that's ok by me. The reason I said what I said in a public forum is that ego - i.e. thinking that one is ahead of the curve is one of the (seven? - there are likely 100's) deadly sins of trading. You are never ahead of the curve. Thinking the way you are thinking per the first few posts of this thread sounds to me like you're not quite grounded in reality. One unproven model does not a successful trader make. 1000's of trades + years of real money track record *may* equate to future success. Again, how do you know the difference? Time and experience... Let examine some of your statements: 1. The Random Walk idea was meant to point out the fact that given your sample space, you do not know whether or not your model is seeing a "lucky" streak or not. That said, what is your sample space and how many trades? For a swing trading model, with a relatively low trade frequency anything under 10 years is a blip on the radar. 2. YES of course it needs to work over all of time - you need to test bear cycles, bull cycles, expanding vola, contracting vola... what if the next 10 years are contracting vola? Can your model deal with that? Can your committment and personal finances deal with that? Do you know what it will behave like under contracting vola - what if it fails? If there are periods that the model is not working, then why not? Figure it out and change your method. Swing methods with fewer trades are inherently less robust and more susceptible to fundamental changes in market behavior. Having it work "long enough" is playing russian roulette IMO. 3. Being an engineer and an MBA doesn't mean anything when you place a trade. You think the market cares abuot how much effort you put in or how good your model is supposed to behave? Nope. What you can control is that you did all the homework, extra credit assignments, read every bit of material and made a valid assessment of your worst case potential risk. Making the leap from a initial model to quiting your day job is a bit premature. From the sound of it you've just started your homework. Would you like me to post a model backtest or a real live trading record? A few of my models aren't backtestable, but here is one that is, its a simple index vola setup. going ten years back. I've been trading this model live for about 6 years now. It took about 1.5 years to get right... Mike
Mike, You did come off sounding a little harsh and I responded to that. Maybe I was a little over the top. A couple of points for you to consider. I started this thread because I knew that I was getting carried away in my own mind on flights of fancy. I'm aware this may not work. Forgot to say...Books? All I have is the internet and MASTER SWING TRADER. I've got a few other crappy books that can't hold a candle the Master Swing Trader book. I think we are coming from opposite ends of the spectrum regarding what we're looking for in a model. I mean no disrespect, but honestly, I don't want a model that works for 6 years. I'm looking for something that works right now, and works so well, that I can chuck it all in a year or so. I get that your perspective is different, but perhaps there are two classes of models. 1. A model that works all the time, and gives you returns commiserate with an "all weather" model. 2. A model that works only part of the time, but kicks butt in the right environment. I'm looking for "Type 2". And when it stops working, then hopefully I'll come up with another Type 2 until the first one starts working again. So I disagree that a model needs to work all the time...it doesn't. Ever see an amphibious car? They work in the water and on the road. But they excel at neither. We're in a trending market...I need something for that. Later, I'll worry about chop. No offense, but I don't want to be here in 6 years talking about my model except to say Hello to friends and read the economics. My intention is to find something that makes money now, and move that into more conservative holdings. If I see a window of opportunity, I don't avoid it because it won't always be there...I just jump through it. FWIW, I did backtest it over few decades. Since I'm trading indexes, that wasn't hard to do. My method doesn't work well in a sideways market at all, but I kind of expected that since it's designed to anticipate trends. It was also reiterating buy orders but you have to know that the results from a reiterated buy order shouldn't be quantified when you evaluate performance...otherwise you get double-counting errors and the average "trend ride" gets shorter, skewing results. FWIW, since I use ETFs without leveraging my account beyond the inherent leverage in the ETFs, I'm not too worried about my account blowing out. If I was doing options or even shorting, I'd worry, but I'm finding that just buying leveraged ETFs give me good enough returns without taking on risk. Oh, and its worth saying that this isn't budgeted money...I live off a paycheck. Yes, I'm fully aware that the market doesn't know I'm an engineer and an MBA. Heh, most people don't even know I'm an MBA, because it really only helps me in my day job when I'm marketing (sad to say). But I'm trying to point out that some people can try to "crack the code" their whole life (DECADES OF EXPERIENCE) and never figure it out. Some can figure it out right away. I've been actively swing trading for almost a year, but I've followed stocks and market behavior for 30 years. I'm willing to bet that success in this field is most likely correlated more to intelligence than years of trading and it probably helps if you have a lot of both, but a lot of either will probably do it. I don't believe that I have to put in another 2 or 3 years before I can even hope to find something that works...honestly, it doesn't seem all complex...just find the right combination, backtest it, tweak it, rinse and repeat until you find something that works. In summary, yes...there are thousands of ways to do it. I think a noob can come in here, put some ideas together, and come up with something that makes money fast enough to achieve his life goals before it stops working. If thats never happened before, who knows...maybe I'll be the first one. SM
Here is one of my graphs for FAS. I also have one for when the price action tops You can see the little green arrows at the bottom of the graph saying when to buy. Hopefully the graph isn't so busy you can't see the price action. You will note that there are re-iterated "buy" signals. The idea is that if you're in one fund (like FAS) and it reiterates a buy signal, you ignore it and continue to hold until you get sell signal or topping signal. What you need to walk away with is that the first (i.e, not reiterated buy signals) occur at good locations. Not too shabby for pedestrian indicators. It is how I'm using them that matters. It works well for a trending market, but not so well for chop. SM
Therein lies the rub. Trend indicators work well in trending markets not choppy markets. Oscillators don't do well in trending markets. At times, shorter periods work better. At other times, longer. Use shorter, have more whipsaws. Use longer, have more lag. Backtest and optimize for ideal periodicity and you'll find that going forward it won't work and what will work will be a different number. Believe everything that indicators reflect and you'll have nasty some surprises. Eg, MACD buy signal below zero can be false but not above zero (and vice versa for sell signal). The point? Nice chart. There are times your canned indicators are going to perform extremely well. At other times, they'll whack/erode you. Prove you can do it in all kinds of markets and you'll have something. Until then, it's a finite sample and means nothing.
Quote from Smart Money: honestly, I don't want a model that works for 6 years. I'm looking for something that works right now, and works so well, that I can chuck it all in a year or so. This does not exist, but almost every newer trader chases after them 700% annual, rapid retirement, plop into a spreadsheet and figure out how it gorws at 30% monthly - "Holy grail". The higher the return, the riskier it is. The higher the Risk of Ruin. A trader seeks more to avoid blowout than to retire that quickly I get that your perspective is different, but perhaps there are two classes of models. 1. A model that works all the time, and gives you returns commiserate with an "all weather" model. 2. A model that works only part of the time, but kicks butt in the right environment. No, there are at least 3. 97+% are in a third category - "dang, I have spent years, thousands of $$$, a lot of energy, and it still doesn't work in realtime..." FWIW, I did backtest it over few decades. Backtesting does not really mean much. Rigorous, longterm forward testing, with Even then, you are usually left with nothing worthwhile. FWIW, since I use ETFs without leveraging my account beyond the inherent leverage in the ETFs, I'm not too worried about my account blowing out. Then you are not going to get someting that lets you "chuck it all in a year or two" Yes, I'm fully aware that the market doesn't know I'm an engineer and an MBA. Heh, most people don't even know I'm an MBA, because it really only helps me in my day job when I'm marketing (sad to say). But I'm trying to point out that some people can try to "crack the code" their whole life (DECADES OF EXPERIENCE) and never figure it out. Some can figure it out right away. I've been actively swing trading for almost a year, but I've followed stocks and market behavior for 30 years. I'm willing to bet that success in this field is most likely correlated more to intelligence than years of trading It is a lot more correlated to tripping over 1+ serious edges, which few traders will achieve, regardless of how much intelligence they have. And serious learning about money management. and it probably helps if you have a lot of both, but a lot of either will probably do it. I don't believe that I have to put in another 2 or 3 years before I can even hope to find something that works...honestly, it doesn't seem all complex... Thoroughly wrong. Most never get there, even after huge amounts of effort, research just find the right combination, backtest it, tweak it, rinse and repeat until you find something that works. Thoroughly wrong. Most never get there, even after huge amounts of effort, research In summary, yes...there are thousands of ways to do it. I think a noob can come in here, put some ideas together, and come up with something that makes money fast enough to achieve his life goals before it stops working. You would be very incorrect. It is a belief, but rarely a reality. If thats never happened before, who knows...maybe I'll be the first one. Now you are getting it...
Alright guys, thanks for the advice, the reality checks, etc. I think from a broad overview, there are only so many parameters associated with stocks. I mean, you have your change in volume, change in price, change in average price, ending close, change from open to close, etc.,etc. The point I'm driving at is that there are a quite a few aspects of stocks to try to build an indicator out of, but in the end, there are only so many permutations of these different aspects. Unless you build an indicator from scratch that tracks some type of parameter that isn't typically tracked, I believe that every parameter is already out there in readily available indicators and it is up to us to figure out what kind of combination to put them in, and when to pay attention to one more than another. Maybe there is nothing new under the sun. I can say this. I had basketful of a short ETF but on Friday, my system told me to sell it off and go long. Still happy! SM
Interesting thread.... If you are confident of your system and have some money in the Bank in case of emergency, then go full time. But you should also have a backup plan that after loosing x% of the original money, you should quit full time and go back to work. Have a good one.