So I went back in looking at your rules on this system as I am trying to start with some basic guidelines for a range trading system. I tried to think about what each part is telling me. I didnt post on the thread because I didnt know if I was way off. Buy the close of this bar when: 1. it is above the 50th percentile of the past 200 days' closes. ( Above the 50th dictates an uptrend bias) 2. it is below the 10th percentile of the last 15 days' closes. (signals short term oversold for the pullback) 3. It is the third consecutive close down. (signal very short term oversold to enter with minimal DD?) The exit is for good RR? ----------------------------------- With these concepts in mind I am trying to figure out a range system and here are some thoughts 1. Within 70% of last 60-90 days price movement 2. above the 10 percentile of last 5 days. (signal too many buyers in range) 3. 3rd consecutive close up without violating Resistance. Very shorterm excess buyers Sell on close of this 3rd day Exit at bottom of range pending the bottom of the range and top have a 1:3 RR minimum These are some jumping off points. Any help is greatly appreciated. I have that Larry Connors book on the way.
Sorry to have been ignoring this thread... I'll get to some of the questions here over the weekend. I got a little caught up posting on BoWo's thread (yeah I know...), but seriously I think we're having some constructive conversation over there... and some things that are relevant to this thread. I don't want to double post everything, so I'd encourage you to go back a few days and check the posts in that thread. (Betcha can't believe I just plugged the BWolinsky Trading thread, huh?)
Thanks. Are you still going to post on this thread? Took a look at BoWo and your info over there is interesting too.
Yes! I promise. I spent most of the weekend in the office so posting here is kinda down the list of things to do. Looks like the BoWo thread may have been a bust... I thought the guy was smarter than he seems to be and I thought we were making progress, but alas... it appears I was mistaken. I am going to package up the posts I made over there and repost them here because there are good lessons there that speak directly to the point of this thread (how to evaluate trading systems.) Maybe I will take a few mins and go move them over here now...
I believe there are more intelligent, interested people in this thread than the BoWo trading thread... in fact I find the environment here to be significantly different than it was several years ago. There do seem to be more people here who realize that trading is very difficult and a long, hard road to possible (uncertain) success. Several years ago I remember these boards having a lot of people with a real sense of entitlement ("You owe it to me to share your super secret trading method.") That seems to be gone and I'm encouraged by the level of interaction we've had here... so let's go on a bit. I did post some on BoWo's thread. To summarize, he has a system built on the ratio of QID / QLD. Now, the irony is the kid doesn't understand basic math, but there really is a possible trade with the leveraged instruments that involves basically being short because the rebalance eventually drives them toward zero. (see FAZ.) This is a trade idea we looked at, but basically decided there wasn't enough juice in it to justify deploying capital there. Here is part of a post I made in that forum trying to explain the flaws in his system idea. The kid couldn't understand it, but there are some ideas that are interesting for this thread perhaps: ======================= I made the claim that the QID QLD "Pairs" approach is profoundly flawed. Rather than leaving that hanging where it might look like a mean-spirited attack, let me explain what I mean. I also haven't taken the time to wade through the OP's self serving drivel to really carefully understand the system. Let me say what I understand it to be, and why that approach is doomed... perhaps I am mistaken in my assumptions (which would be my fault for not reading carefully enough).. then my conclusions would also be wrong. I believe this is not pairs trading at all because it only trades one leg of QID QLD when he believes the market is overbought / oversold. Thus, there is no ratio relationship or reduction of market risk like one would normally expect from pairs trading. It is a simple overbought / oversold system using leveraged products. And therein lies the problem. These leveraged products are designed to reflect X times the 1 day return of the underlying. Without going into a math lesson, this means they "reset" each night because they are rebalanced. Tomorrow's QLD is a significantly different instrument than todays! Any analyses based on levels, averages, etc are not valid from day to day on these products because you are comparing apples to oranges. It is possible to construct a mathematical simulation showing how these double and triple leveraged products seem to flex, but it is illusion and does not offer profitable trading opportunities. It would be easy for someone using primitive retail system development tools (Tradestation, wealth lab, etc) to be deceived if they really don't understand the mechanics at work here. I believe this is what happened to the OP due to lack of experience, which is certainly in itself nothing to be ashamed of. We all started with no experience but I find people who are successful in this business maintain a pretty consistent air of humility that is quite at odds with BoWo's chest thumping. The OP should also check out the settlement prices he is using since these products do not settle like regular stocks and that procedure has changed and evolved as they became more popular. In other words, he is unlikely to be able to execute at his backtested prices because the lookback data won't carry forward. We could also discuss the logic of presenting levels for a stock to the 10th decimal place, but that's just a matter of experience. I could go on, but I know some people are reading this thread and wanted to offer them a bigger picture perspective on the "system" that is being discussed. The rebalance issue seems like a simple one, but it is at the very core of the concept... and sadly, that concept is profoundly flawed. Shame the OP will never read this since he has me on ignore, but I'm offering this as a public service message... it's not for him any more than last night's stats lesson was for him... if someone else reading this can carry a lesson away then my time was well spent. You are correct, the idea in a pairs trade is to sell something overvalued and buy something undervalued. You are also correct there is usually, but not always, the expectation that some directional market risk will be eliminated from the process of buying / selling. You are completely wrong and show your inexperience in saying that daily pairs trading is more efficient than intraday. I realize this is a limitation of your testing environment and lack of access to real data (bid/ask spreads historical lookback). The central idea with pairs trading is that instruments "should" (or do) trade in a normal relationship, however you define that. Departures from that normal relationship are trading opportunities for pairs models. These departures happen more often and with more predictability intraday. There are more trading opportunities and your risk is better controlled. It is very true that execution (which is a learned skill) plays a big part of intraday pairs trading and that costs will represent a much larger percentage of your P&L, but that is true of any intraday trading. If you do decide to investigate these intraday models I will caution you that you must work off bid / ask spreads. There are many intraday pairs that might look profitable to you using WealthLab, but the spread completely accounts for the total potential profit in the trade (thus, no trade.) Now, here's where you need to pay attention because here's why your QID / QLD system will not work. You may use the f-word at me now, you may pretend you didn't read this, but 2 or 3 years down the road come back and re-read this. Ready? -In a pairs system we look at the ratio between two securities. We find some kind of normal value and recognize trading opportunities when there are departures from that normal value. -Your system uses double and double inverse QQQQ's. These instruments do a pretty good job (very small tracking error (look it up, probably in that stack of CFA books you should have studied...) of doing what they are designed to do which is to replicate twice and twice the inverse one day percentage return of the underlying. -Thus, the ratio relationship is completely deterministic. Simple math (I'm not going to do it for you here) will show you that this ratio will flex according to the size and direction of the moves of the underlying. There is no rubber band... there is no flexing around value. You can build a simulation in Matlab (or Excel if it's all you've got) and see how this works... and then verify that the QID QLD spread behaves as the simulation would predict. -Any optimization is basically asking the question: "Over the last two years what size pullbacks should I have been buying and selling in the Q's." You will get a good answer that will make your system results look great, but it is a function of the natural math of this flexing. You do not understand this. The parameters you optimize will have absolutely no carry forward because there is no actual pairs relationship. How do I know this? I have never thought of this as a trading idea until last winter when a smart kid interviewing for a summer internship brought an idea very similar to your QID / QLD pairs system to us. (Wolinksy... you weren't the first to think of this lol.) in talking through it in the interview, I was able to lead the kid to see the error in his thinking in 3 or 4 minutes... like I said he was a smart kid. Afterward, I found the idea intriguing enough that we did a little simulation and testing to verify no trade. It is easy to be fooled by relationships like this, especially when you don't have a lot of experience with these models and actual trading. I'm not criticizing you for making this mistake, but open your mind and consider carefully what I'm saying here. If you have questions, please ask... I will do my best to help you understand and to answer any questions. I'm certainly not an expert at pairs trading. yes I have some pairs models that make pretty consistent money, but it's not what I would consider a passion or an area of expertise. Still, this is pretty simple.
And then I posted real distributions from several of my systems: The attached Excel file has 4 worksheets: Pairs is a daily pairs model that holds from 2 - 20 days. These are returns for the whole system which trades several different stock and ETF pairs. System1 is a non-discretionary system that I trade. System2 is a system one of our trainees developed. BoWo is the list of pair trades from his backtest that Mr Wolinksy posted earlier to this thread. I also attached a page of summary statistics and histograms of returns, with kernel density plots and normal curves for comparison for each of the systems.
After BoWo posted some (fumbling attempts at) analysis, I posted the following explanations: Ok thanks for your work Beau. Let me share a few thoughts and give you some background information on the systems. I have to tell you up front that I am a little limited in what I can say about a few of these because they trade some things that may not be super liquid. We have trouble executing some of these at sizes bigger than $250K or so, and our execution is very, very, very good.... it's just that the markets are so thin we have to be a little careful. For that reason I can discuss concepts and ideas behind the systems but I may refuse to disclose some specifics. Pairs. Is a straight up pairs system traded on stocks. We did a lot of development work with end of day data and found it was impossible to replicate those trades with real money. This is one reason I am so critical of the QID/QLD system... you're backtesting to settlement prices which are prices you can't reliably execute. If you don't understand why, then you need to understand market structure and procedures more. The way we found to get around this was to cut the day into a few big pieces and use those prices for backtesting and execution. As you can see, the system works pretty well. As you know, I'm not a big fan of most of the system stats you throw around here from your WL output, but one that I do consider is simply the ratio of standard deviation of trade returns to average trade return. The bigger this number is, the "more wild" the system is and it provides a quick way to compare across similar systems. This is one area in which this system excels -- for the size of its average return, its standard deviation is very small. Going forward in this post I will sometimes quote BoWo: >It does not appear that it has fat tails, more than that it almost seems like you're pairs model is a bit concerned about "preserving capital." A little loosening of your autostops criteria should more than double your trade profit, but will adversely effect your win percentage in all likelihood. There are no stops with this system. I am very opposed to optimizing or doing multiple runs of system tests. In this system we had a basic idea ( = take the spread ratio, draw a moving average, draw some bands a few standard deviations around the average and make some trading rules.), coded that idea and saw if it worked. This is the essence of simplicity. Rather than optimizing parameters, you would find that most of our systems don't show that much of a difference with different input values, which is something we really strive for. That, to me, is a sign of a robust system. >Your pairs model has "a lot of small profits", but you need a more concise ability to predict market direction, and the way I do that is by various lengths of linear regression predicted values and by testing statistical significance. One of the things we have worked very hard on is trying to NOT predict market direction and this pairs model is the ultimate expression of that. We have looked at linear regression channels and lines and found them to have no predictive value except in a few special isolated circumstances. We currently run no systems that use this tool, but it's certainly possible you may have found some utility. I would be very suspicious of any pairs model that uses regression channels unless it is an "outperformance" type of pair. (As an aside and a huge overgeneralization pairs models break down into trending or mean reversion pairs with the latter being much more common. The pairs system I posted is a classic mean reversion system.) I saw a post where you mentioned using R-squared as an indicator in your pairs system... we also did this and more advanced versions of goodness of fit tests and found that they did not help performance. With these models you just gotta take your trades and take your chances and if you have a good model the numbers work out after a large number of trades.
System 1 is the S&P Adds Deletes that started this thread off. >Bowo says:The kurtosis and skew doesn't matter to me in this case, because there is such a wide disparity in the maximum and minimum values that it would suggest that there is inherently a lot more risk with this system. Yes. The huge losses and wins would definitely get my attention... in fact my first inclination would be to throw the system out. However, I would encourage you to look at the distribution and the histogram with outliers removed (quick and dirty: just drop the 4-5 biggest and smallest values from the list I sent.) If you do this you will see that the system appears to have a core of fairly stable returns. Depending how you do this analysis you'd probably find a more or less normal curve with a slight negative average return (yes, negative). You cannot really exclude outliers because they account for all of the risk and all of the performance in this system, but it leads us to a more clear understanding which in this case is that the system seems to have a core of a large number of trades that don't do anything at all and then a handful of big winners and losers. To date, the winners have FAR outweighed the losers, but I realize going forward there could be a stunning loser that changes this because of the structure of the system. Can you imagine shorting a stock, buying the S&P against it, the company is taken over at a huge premium so you have a 400% loss on your short, and the market goes down 5% over the same timeframe just to add a little cosmc F*CK YOU at the same time? Yeah... it certainly could happen. It's unlikely in this case because of what the companies are (much more likely with some POS little biotech or something), but the point is that it's possible. You're correct in your assessment of risk here, which is why I think a key is to just not trade this very big. If I have say $10M to trade with then I'm not going to be doing these trades more than $200k / side at max. I would scale that to any account size... so if you have $10k in your trading account then you're probably just buying an odd lot of some of these trades... that's fine.. trading small is a legitimate way to reduce risk. (Think anti-leverage) >Sadly, I can always say, and I think anyone else can, too, that having a trade lose 88% of it's value is unacceptable. You need to add a stop around 30-40%. Please do me the favor of updating that and showing the change in the distribution and performance summary. Well, it's not unacceptable, but it's not good. It is needed for this system... it's the only way to capture the big profits as well... adding a stop as you suggest would make the system unprofitable. System2 is a system developed by one of our trainees with less than a year's experience. Let me digress a little bit. I have had much better experience training people with no finance background to be successful traders. Why? I think because you have some classes, you study for some tests and take some post grad courses and you think the world works in a certain way. (For instance, consider the industry's focus on returns vs a benchmark. The main reason for this I believe is that big institutional money can't consistently beat the market because they ARE the market.) As long as a person has the ability for critical thought, is a cynic, is of above average intelligence and is passionate about trading, it can be done. The thing that slows us down a bit is that sometimes it's a bit annoying to have to teach basic math courses, but that's also an opportunity imo. Many people learn techniques in school but we focus on concepts... that avoids some problems. So what you have here is a system that looks for distortions near the beginning of the day and holds the trade, at most, to close. The trading universe is large cap stocks, but the points where we execute sometimes don't have a whole lot of two sided liquidity, so even this is a challenge. I consider this the best of the systems I have posted because of the stability of the returns, because it is intraday (=limited risk and more efficient use of capital), and because of the fundamental soundness of the idea. Win ratio is another statistic I'm not overly focused on... I have some systems that do well that have less than 20% win ratio. One especially attractive thing I see in the distribution of returns is the rather long right tail and truncated left tail. In a backtest, this would make me suspicious but this is what we're striving for. To see it in actual returns means you're doing a lot of things right. We probably could significantly improve performance of this system by holding overnight, but I don't want the extra risk. One thing you can consider is to look at two measures: average yesterday's close to today's open return and average today's open to today's close return. When you see more and more of the moves are happening overnight, that's another kind of risk you need to be aware of. (Can certainly represent opportunity too.) I won't comment too much on the BoWo system on the sheet. I have serious reservations (from hearing you talk) about the process you followed... When I see lots of data points at the top and bottom of a backtest distribution it is a sign of possible overoptimization--you may simply have plugged for the best combination of parameters that give you a lot of big winners and cluster a lot of losers in the same area on the downside. So this concerns me. You seemed to think that there are a lot of similarities between my pairs distribution and yours. Without getting too caught up on statistical tests, it's possible to ask the question like this: Assume that there is a giant bag of all possible trades for the system. My distribution is a random sample of the trades in that bag. Understanding that your sample is smaller, what are the chances that your trades are another random sample from that same basket? The answer is this case is EXTREMELY unlikely... you can see you have a lot of high and low clusters that dont exist in mine... so it suggests to me that there is a different process at work behind the two systems. (FYI, the K-S test would be appropriate here.)
Nice, good to have you back over here. Capturing the drag on the lev etfs was one of the first ideas i looked at...ran a bunch of sims with different drift, vol, and serial corr scenarios. Looked at taking advantage of some different option constructs, but basically came to the same conclusion as you. I'll see if I can dig up that sheet and post the results just for sake of comparison.
Talon, I have been following this thread for quite some time and have wanted to chime in but haven't had the courage. But this thread is why I have spent too many hours on this site. That aside I wanted to share a trade idea I had a made some money on. You addressed this earlier and I believe I understand the 2x/3x etf's. To keep it simple, I shorted equal dollar amounts of FAS and FAZ and rebalanced after a 5 to 10% move in either direction. (I would either short more or buy back). As you said, these etf's will approach 0. I realized this early on. Would you mind elaborating a bit more on why this startegy can't work in the long run? My thought was that in a strong bull market (like we've seen) you will lose money and thus why I stopped. (I felt like I was constantly chasing the big market moves). I'm not sure if I'm making sense here but I think you get my drift. Thanks for doing this