BlueHorseshoe

I think what you describe is pretty much what MysticForex's method aims to exploit, isn't it? The concept would be to buy a short term downtrend/breakdown, buying a position from those who are shorting this short term downtrend, in favour of the longer term uptrend. This means trading contrary to those with short term outlook (typically retail traders and those who are easily shaken out of positions) and aligning oneself with heavily capitalised institutions who have moved the market to where it is by accumulating inventory (positions which they will typically add to as the short-term outlook traders are shorting). Buying pullbacks in trends works. BlueHorseshoe

I'm looking to combine multiple entry/exit methods, each currently programmed as individual strategies, into one strategy. Each source strategy has it's own specific exits (profit targets, stop-losses). How can I prevent one source strategy's exit order (eg 'setstoploss(500)' ) from exiting a position generated by an entry order to which it is not applicable? Is there an easy way to link marketposition with the specific entry order that caused it? I do have a method of doing this, but once again it's a very convoluted approach and I was wondering if there is a simpler method? Many thanks, BlueHorseshoe

How do you determine the long-term trend on a daily price chart? Do you currently consider the S&Ps to be in an uptrend? Do you employ long-term trend filters in your trading?

Identifying trends and entering on pullbacks is a sound methodology and works well in most markets (especially the stock indices) - I never understand why more people don't trade like this. As others have pointed out, correctly determining both the trend and an overbought/oversold condition are the two main difficulties here (they're also the only real input variables, which is good because it reduces the risk of curve-fitting). For currencies the trend filter (SMA) needs to be appropriately selected - because these markets often produce strong trending moves, either a very short (eg 30) sma or a longer (eg 200 as suggested by MysticForex) SMA works well. In the indices you'd want something inbetween (eg 80sma). Also, using the directional slope of the SMA often produces much better results than a price/sma relative measure. For identifying short term overbought/oversold conditions nothing but the most recent past matters - shorten the lookback for oscillators such as the stochastics, rsi, cci, to between 2-5 periods. There is absolutely no edge in a 14-period stochastic. If a 2-period stochastic sounds an unlikely proposition then go to the data and test what I'm saying in the market you want to trade. Hope that helps. BlueHorseshoe

This is the sort of very generalised advice that I was cautioning against in my previous post in this thread (see first page). There are plenty of ways to trade profitably using all sorts of R:R combos, and risk does not always have to be significantly less than reward. Also, I am assuming that when you say '3 to 1' risk to reward you actually mean '1 to 3' risk to reward, whereby 1 unit is risked for a 3 unit reward (or alternatively a '3 to 1' reward to risk)? While it is true that a 1:3 risk reward ratio would yield a profit if you were successful in 1 out of 3 trades, a 6:1 risk reward ratio would yield the same profit if you were successful in 7 out of 8 trades. You don't find many traders who will risk 6 units to make 1 (if you were scalping the ES with a $200 profit target, would you want your stop-loss $1200 away?) . . . But then you also won't find many traders who turn a consistent profit. R:R ratios are a complex function of trading strategies; more often than not they are something that will arise organically from the development process and shouldn't need to be over-thought. Risk reward ratios cannot be prescribed for all strategies or traders in all markets in the way that you seem to be implying in your post. BlueHorseshoe

Although I also don't have a fixed r:r ratio at the time of entry, one thing that I always find useful when testing a strategy is to use a 1:1 ratio and then optimise this single value. If the optimal value doesn't produce impressive results then it is unlikely that there is any edge in the entry. It's a bit like saying 'if I buy when this signal occurs and I allow the market to move an equal amount N in each direction, then if the entry signal has an edge I ought to get better results than a coin toss for some value of N . . .' This doesn't mean to say that the strategy couldn't be profitable as the edge can obviously come from an exit, but it is always useful to know from what element(s) of the strategy the edge is derived. It is often argued that focussing on entries rather than exits is a sign of an amateur ("the exit is where the money is made"). The same authors often promote the idea of a very small r:r ratio (Van Tharp would be a classic example, who is pretty insistent that the reward must be many multiples of the initial risk). The positive expectancy of the strategy I trade is derived mostly from the entry signals, however, and from an adaptive exit that calls for greater risk than reward. This flies in the face of popular advice. So I think that better advice than the overly prescriptive approach that many authors give would be something like: Know your strategy. Understand from which elements it derives a positive expectancy. Work to maximise the dollar output from this edge through intelligent position sizing. Hope that's helpful, BlueHorseshoe

I don't have a fixed R/R Ratio, although my average loss is almost twice my average win. The probability of the tp being hit is complex and in some cases will be indeterminate. One of the key factors involved is obviously whether a stop loss is also used. If no stop loss is used, the probability of a small target being hit sooner is greater, whether the probability of it being hit at some time is greater I'm not so sure. Whether the target is above or below current price is also a theoretical factor, as price is bounded below but not above. BlueHorseshoe

Thanks - I see now what you mean by 'streaky'. I have read a little of Taleb, who thinks that the best way to understand pricing is through 'jump diffusion' models, which essentially result in skewed gaussian distributions with extreme kurtosis and fat tails - now there's a mouthful! I've often wondered what would happen if I were to throw every single futures instrument together into one single 'synthetic' market - this new market ought to be far more random, and therefore ought to be more normally distributed. Unfortunately I wouldn't have the capital to trade it! Thanks, BlueHorseshoe

Hi BlowFish, Could you expand a little on what you meant by the above please? Thanks BlueHorseshoe

Call back in a few days and let us know how that's working out for you? BlueHorseshoe

I'm sure you don't really want me to give you a crash-course in propositional logic but . . . There are infinite things that are generally thought to be state-able with absolute certainty and they form the set of propositions called tautologies. Each is a propositional construct that is true under any possible boolean valuation of its variables. Some classic examples (I'll give you five!) pulled from Wikipedia would be: ("A or not A"), the law of the excluded middle. This formula has only one propositional variable, A. Any valuation for this formula must, by definition, assign A one of the truth values true or false, and assign A the other truth value. ("if A implies B then not-B implies not-A", and vice versa), which expresses the law of contraposition. ("if not both A and B, then either not-A or not-B", and vice versa), which is known as de Morgan's law. ("if A implies B and B implies C, then A implies C"), which is the principle known as syllogism. (if at least one of A or B is true, and each implies C, then C must be true as well), which is the principle known as proof by cases. The law of the excluded middle allows you to know that if the market closed higher today then it didn't not close higher today. Useful, eh? Unfortunately tautologies exhibit an absolute degree of redundancy . . . rather like so many of my posts on this forum BlueHorseshoe

A certain type of fund will, I think. But there are also plenty of funds that trade intraday. Nevertheless, I think that your explanation is the correct one for increased volume. An interesting study would actually be a comparrisson of opening and closing volume. BlueHorseshoe

Thanks - that's a really useful (though blindingly obvious) point that I hadn't considered before. In theory one could attempt to systematically converge an 'optimistic' target on the closing price as the session progressed. Cheers, BlueHorseshoe

This idea has occurred to me before, but I don't understand options well enough to begin testing it. I'd also wondered abut the VIX . . .

Ha! I was just about to go on youtube and find the Seinfeld parody to post! BlueHorseshoe

I wasn't drawing a conclusion about the success of all hedgefunds. I was drawing a conclusion about the success of the successful ones. My original post suggested that this success ran to billions, and your response denied that this was the case, and suggested that this success was largely the result of accounting manipulation. Unless I have misunderstood what you have written completely . . . BlueHorseshoe

Hi BlowFish, Yes, this is my thinking. It wouldn't be everyones but, in terms of my earlier posts, this is where I chose to position myself within a field of probabilities. When I started trading then I was willing to risk losing the entire account if the market was 'unlucky'. If in ten year's time I have several million in the account, then I won't be willing to lose it all. So I will become (systematically, according to an precise rule) more risk averse as my capital grows. As MightyMouse correctly points out, this is not a choice that can be mathematically rationalised - it's an emotive (for want of a better word) decision. The important point for me is that I made this decision early on, built it into my system as a fixed rule, and am prepared to live with the consequences as they develop - in other words, it's not 'emotive' in the sense that I have an emotional response and act upon it with each individual turn of my equity curve. Incidentally, the money management algorithm I use is just an 'off-the-shelf' one (I even cribbed the TS code from a newsletter), and there are plenty of great books on this topic (Ralph Vance for instance). BlueHorseshoe

If all five positions went against you you would lose 10%. So, let's say you have an account of $100 and all five positions get stopped out. You've lost $10. Is $10 really a big deal to you? I can't imagine so. I'd happily bet $10 on the flip of a coin. It's a gamble, with a very small amount of money. In money management terms risking $10 of a $100 account and risking $10000 of a $100000 account are the same. But is anyone going to pretend that their emotional reaction to them would be the same? A 10% loss isn't ugly to you, Obsidian; rather, a 10% of your current account equity is ugly. There's a difference. BlueHorseshoe

[ heaves big sigh ] . . . Yes, of course I do. That's what the article that I was referring to is about. You can find it here: http://www.traderslaboratory.com/forums/general-discussion/12949-blowing-up.html Nicholas Taleb could also have gone bust. But neither of them could have gone bust at the same time, or in the same manner. Taleb could only ever have gone broke very slowly. Although it is inevitable that both of them would lose everything if they continued trading long enough, the circumstances of this would necessarily be wildly different due to where each trader had chosen to position themselves within a field of probabilities. Someone who risks 100% of their account on a single trade is simply chosing to position themselves at an extreme within the probability field. As one set of outcomes grows less likely, another grows more certain. The trader who positions themselves at the opposite end of this field, risking just 1%, can still go bust, though only very slowly (like Taleb). Both traders are still gambling - they're each just choosing the way in which they can go broke (and as a collorary of this, the way in which they can make money). It's a bit like having to chose a method of execution; you can be shot in the head point blank or you can be slowly starved to death, but either way you're going to end up dead. Jeez, what a morbid little rant that turned into . . . BlueHorseshoe

Hi Trader1, Thanks for your reply. I don't risk 100% of my account on each trade. I'll go so far as to reveal a piece of information that I consider quite personal: I currently risk 3.7% of my account per trade. That number was larger when I first began, and will steadily fall as the account grows. If one day I am lucky enough to have several million in the account, then I would be risking less than 1% per trade. These figures are a product of the specific money management algorithm that I have chosen to use. It's great that you view your trading as a business - I'm sure most of us agree that this is the best way to approach it for a whole variety of reasons. I view my trading as a business also. However, a business is still a gamble . . . Shell Oil is a business. Look what happened to them two summers back when they had their 'little' accident. I know someone who made millions in the eighties through a series of successful businesses. Now they're broke, because they kept on 'being in business'. Business has risks. Most new businesses go bust. A business is a gamble. My whole point in contributing to this thread was to point out that the whole 2% risk per trade rule is complete arbitrary nonsense for most traders on here, and shouldn't be taken as gospel. My secondary point was that any sort of risk calculation, though helpful, is ultimately meaningless. If you keep on playing the game long enough no matter how you manage risk, then you will go broke - it's a mathematical certainty. BlueHorseshoe

Yes. And when I try to build a rocket then I am subject to precisely the same laws of physics as the guys at NASA. Now who do you think is going to build the better spaceship, me or NASA? Difficult though it can be to have to acknowledge it, there are people (and organisations) out there who are better at what we do than we are. Vastly better. Almost infinitely better. They pretty much walk on water. And the funds they manage make billions of dollars. I have no doubt that they also employ crafty practices to smooth or bolster their published returns. But to suppose that their entire success is some Maddoff-like edifice of deceit seems, to me, to be a bitter and unworthy stance. BlueHorseshoe

I'm not really interested in 'risk of ruin'. I'm not interested in 'ruin' in terms of my trading account. That money is there because it is money that I can afford to lose. If the account is 'ruined' then I won't be. Many people will say that this is pretty much gambling. I'm under no illusion that trading is anything other than a structured form of gambling in which the trader positions themselves in relation to a range of probabilities (of which risk of ruin is just one). How a trader choses to do this is up to them alone. I am quite comfortable risking the money in my trading account against my perceived probability of return on it. Yesterday someone posted a great link to an article about Taleb and Niederhoffer. Either could go bust; the former could only go bust by slowly 'bleeding to death', whereas the latter could only go bust by 'blowing up'. This was where each had chosen to position themselves within a field of probabilities. It doesn't matter what fancy money management algorithm you use (imagine how sophisticated LTCM's risk management must have been, for example), the only way to avoid the risk of ruining your account is to stop trading it, and if you continue trading, then the only way to avoid the risk of ruin to your finances as a whole is to ensure that you only trade with money that you can afford to lose. BlueHorseshoe

Hello, I appear to have started a discussion about the ways in which CTAs market their funds. This wasn't my intention, and it's also a complete deviation from the topic of the thread, not to mention rather dull, so I don't want to make a long deal out of this but . . . MightyMouse, are you saying that a fund such as, say, Citadel, doesn't make its money out of stochastic pricing models? BlueHorseshoe

I personally find it fascinating that these traders continue to thrive. Whilst quant funds and HFTs continue to spend millions of dollars developing systems built on assumptions of stochastic price behaviour (and of course make billions of dollars from these models) . . . There are a handful of these guys who simply sit in the markets and wait with infinite patience for the outliers that they know will surely arrive. Whilst I agree that many technical elements of trend following aren't really adaptable to what any of us here do (unless you happen to be managing several billion dollars across a fifty market portfolio!), I think there are certain more abstract lessons to be learnt from the trend followers: - Traders need to have an overall generalised view, or 'philosophy', of the markets - without this kind of belief system traders have no way in which to structure what they do with any consistency. This perspective needs to be derived from some basic 'truth' that transcends the specifics of any strategy that may aim to capture it. - A trader needs to be comfortable with the ultimate uncertainty of the process they are engaged in and realise that even the most consistent types of market behavior will necessarily produce exceptions. BlueHorseshoe

That's two posts from you today and, as usual, my day just got a whole lot more interesting! Incidentally, one of the better volatility stops I have tested was derived from code created by Michael Bryant in a Breakout Futures article, although I think he intended the function to be used for an entirely different purpose. Nevertheless, I agree that volatility stops can be tricky, and that trailing stops are seldom beneficial. Thanks again, BlueHorseshoe