BlueHorseshoe

Let's just pause here and examine what you're saying . . . Supposing these traders hold three positions a day and, on average, they have two wins and one loss per day. Each trade has a $200 stop and a $200 target, let's say - does that sound fair? So they make $200 per day on average. Their risk per trade is $200, and they apply the "2% rule", so they have a $10k account. After a week they will have made 200 x 5 = $1000. After just ten weeks, they will therefore have doubled the account to $20k. After a full year they will have made a 520% return on their account if they mantain the same position size as when they started. If they double their position size when they double their account they will have made a return of $310,000 after 50 weeks, or 1600%. I imagine that after about ten years they would own most of the world's wealth. Not bad for "only" a few hundred per day! Just in case anyone is thinking "ah, but they'd have to withdraw most of what they made to pay the bills so they couldn't compound it" you need to think again - if I could give you a system such as this are you seriously telling me that you couldn't find a way to support yourself without the trading income for 10 weeks while you doubled your account? With such a guarantee of future wealth anyone sane would go and live in a freakin dumpster for a few weeks! It's disappointing to see people come here over and over again peddling the same worn out spiele . . . BlueHorseshoe

A less efficient way to do this than the one ZDO suggests would just be to reference your signal timeframe for your conditions and a lower timeframe for the entry. So, if you're trading from 15 minute bars you could place orders on a 1 minute timeframe chart. BlueHorseshoe

Hello, I think it's just poor wording on behalf of TS. It means you can only use it within a strategy (whether running in sim or live mode) and not within an indicator, paintbar study etc . . . Slightly weirder are a set of reserved words that most commonly refer to the output of a strategy but can only be used outside of the strategy in indicators etc - never understood that! Hope that helps. BlueHorseshoe

That's a fantastically thoughtful and helpful response - thanks ZDO! BlueHorseshoe

Tradestation is great, but I don't think any kind of demo is available, is it? On a side note, why do you say EL is a combo of Pascal and C#? I'm not challenging you - I just didn't know that . . . BlueHorseshoe

Hi Roger, Apparently Border Collies are the best for trading with because they're used to running rings around sheep! :rofl: You might also be interested in my G.R.A.B. Index - details here. BlueHorseshoe

I thought the riddle was really pretty good, myself, although you could have given up the answer weeks ago Did anyone know that Poddles are rated the second most intelligent dog breed, scoring above German Shepherds? Now there's a suprising fact for you . . . BlueHorseshoe

Having systems that 'learn' from and adapt to changes in market behaviour seems like a great idea, but . . . If a system is too receptive it learns too readily and curve-fits to inconsequential noise in the price data. If a system seeks to avoid this by using large data samples to make robust generalised inferences about market behaviour, there is a risk that its resistance to change will seriously lag any significant but vital shift in behaviour. There is necessarily a "sweet spot" between the two that indicates the optimal learning rate for the system. However, to find this "sweet spot" one is faced with the precise same problem that I identified above. To mediate between the curve-fitted solution and the lagging solution a new variable must be introduced and it too must be mediated with some criterion for optimisation . . . I can't find anything in the machine learning literature that I've read that suggests a viable way out of this catch-22. Does anybody have any suggestions? BlueHorseshoe

CONTRADICTION!!! :helloooo: The big money is rubbish and consistently underperforms . . . So you ride their coat tails expecting to outperform them? Doesn't really make sense when you think about it, does it? More vendor hyperbole, and the same flawed argument we've all heard a thousand times before. There are billion dollar funds that are more nimble than any trader on TL will ever be. If you can consistently outperform an average return of 20-30% with limited risk then you would have absolutely no need to be running a coaching operation and dealing with the endless hassles of marketing and trading to the general public - money from willing investors would be pouring in . . . Nice try. BlueHorseshoe

Hi Gekko, I agree with what you're saying about success - a lot of it comes back to your own goals versus other people's expectations. I am particularly poor at managing other people's expectations and envious of those who can take a "f**k you" attitude and be genuine about it. With regard to your example, if the $250k salary bloke starts trading with all of last years income and makes $1k a week, then that's a 20% return on capital, which is pretty damn good! I used to like to throw the old "those who can . . . " quote about as a teenager (it's an especially fun thing to say to a teacher, believe me!). Is it GB Shaw? Anyway, then I thought about it. Just because a high school physics teacher doesn't work at CERN doesn't mean they don't know a damn sight more about physics than a sixteen year old. I come here to try and offer help (teach?) those who might know less than I do, and to learn from those who know more than I do. The way I figure it, the more time I spend explaining the diference between an SMA and an EMA to complete newbies, the more time people like MysticForex or MightyMouse posting above will have to spend helping me. There are probably no real 'experts' in trading, and you don't have to be an expert to help someone out - you just need to have a bit more knowledge or experience than they do. BlueHorseshoe

Hi JMB, As a general rule, all else being equal, you'll find that the larger your stop-loss the higher your win-rate. A risk/reward ratio of 10:1 where the stop is ten times the size of the target can push a win rate well above 80%. The converse is also true - as the ratio swings the other way then the win rate will typically decrease below 50%. Consider Trend Followers who have very small stops and (theoretically) very large profit targets. They have a low win rate. Now consider scalping strategies that will risk a full point to make a single tick. They have a high win rate. Correctly compared (disregarding trading frequency etc) neither strategy is inherently more likely to be profitable than the other. So inverse risk/reward ratios are not in my opinion a good reason not to trade binary options. BlueHorseshoe

"Trade Tiny" - I agree with. Greedy people with short-term outlooks tend to over-leverage. However, I don't agree that risk management and position sizing are intertwined. Risk management (knowing where a stop-loss, a profit target, etc are placed) is part of developing a trading strategy. Money management is all about getting the most out of that strategy (once you have it) through position sizing. Think of it this way: risk management (entry, exit, stoploss, breakeven, trail, profit target, risk reward ratio etc) is like crafting a weighted dice; money management is all about knowing how often to roll the dice to maximise profits without the risk of going bust. With poor money management you can go bust even though your risk management is impeccable. With great money management you can avoid going bust even though your risk management is terrible. BlueHorseshoe

HFTs have always traded against one another, I think, but a drop in volume would certainly be difficult for them, as you suggest. While they could adapt to pretty much any other type of change in the behaviour of a market's participants, if there's nobody there to trade with then it's got to affect their performance. Discussing HFT in relation to the ES is a little tricky - due to the nature of the CME order book it's not immediately obvious how any substantial advantage could be gained through speed. Or not obvious to me. BlueHorseshoe

Here's a link to an interesting paper on HFT and methods: Overview of High Frequency Trading BlueHorseshoe

Hi Colonel BS, You have me on 'ignore', I figure, so it doesn't matter what I call you . . . Unfortunately 'positive thinking' won't help you much when you come up against the cold, hard facts of the algo-dominated reality the daytrader must inhabit. I know it's an asset in pretty much every other vocation, but not this one. Doesn't matter how much positive energy you have, if you go one way and the market repeatedly goes the other you're screwed . . . Write down all your trades this month. Now add the value of two ticks per trade to your net profit (that's $25 per round trip in the ES). Does that look any better? That's what your profits would look like if you didn't have to keep paying the spread because you could know when you'll get filled on a limit and when a market order is the only way to get set. Traders with decent winning strategies systematically underestimate the impact of spread and commission on their performance (post #50). BlueHorseshoe

Careful now . . . Money Management (ie Position Sizing) and Risk Management are not the same thing. BlueHorseshoe

Hello, Mandelbrot's fractal theory might cause you to assess probabilities using power laws of exponentiation rather than assuming a Gaussian distribution of outcomes (as one would expect flipping a coin as the article describes). You would price an Option completely differently to the prevalent Merton-Scholes model, for instance. You wouldn't rely on linear regression methods and concepts like GARCH. You would anticipate the occurrence of 'outliers' with far greater frequency than that predicted by a bell curve approach. And you wouldn't work with an artificially imposed two sigma threshold such as that imposed by Bollinger Bands. Hope that helps. BlueHorseshoe

There are many variations of this type of approach (see Optiontimer's thread, for example) and very good reasons to believe that trading pullbacks in trends can be a great tack. If you want to trade in this way you have two distinct tasks to undertake (which is really far less than many other strategies demand), both of which are more difficult than they sound - here they are: 1. You need a method to identify whether the market is trending and, if so, whether the trend is up or down. 2. You need an objective method to identify a pullback, and you need to know how deep a pullback needs to be to provide you with an optimal entry. In my opinion one of the keys (and it's market specific) is knowing whether to 'buy into' a pullback using a limit order, or to await a 'confirmation' thrust back in the direction of the underlying trend as is shown in the examples above (where a buy order is presumably placed at the prior bar high). I hope this post is of help to someone. BlueHorseshoe

There is absolutely no reason (apart for the intellectual limits and programming capabilities of the typical commercial EA vendor) why a strategy can't be made to adapt to changing market conditions. From the A.I. sophistication of genetic algorithms and neural nets to the simple regime switches and volatility filters, there are countless ways to do this . . . If you invest the time to learn about such approaches then find a programmer to implement them for you, you'll probably find you have a far better strategy in hand than anything you can buy from a commercial vendor, and for less expense. Hope that helps. BlueHorseshoe

"Not until you crawl over here and kiss my feet and beg . . ." :rofl: BlueHorseshoe

Michael Durbin - High Frequency Trading The nuts and bolts of HFT infrastructure, and a glimpse of strategies. BlueHorseshoe

ZDO's advice here is all very generous, but I think the value of this particular point is massively under-emphasized. One way to see exactly how it affects an equity curve for something with fixed contract sizes is to test using 1000 x equity, and then divide the trade-by-trade results by 1000. This will simulate the outcome you would get if you were able to trade fractions (thousandths) of a contract. BlueHorseshoe

I'm probably stating the bleeding obvious here, but have you read 'Dark Pools'? One of the issues is that HFT is an umbrella term, and the only thing that all such operations have in common is the frequency with which they trade. Some use traditional strategies such as stat arb but at lightning speeds, where others exploit new loopholes that are only available with the advantage of speed (and often strong relationships with the exchange), relating to latency, server jamming, and gaming the order book. There's another good book written by someone who worked at Citadel setting up the HFT infrastructure, but I can't remember the title or author - I'll try and post this tomorrow for you. BlueHorseshoe

I hoped this thread would be a fun place to come for a bit of entertainment when people have five minutes to fill, and not another TL warzone . . . BlueHorseshoe

I'm not so sure about that - I believe "retail" accounts for, at best, a fraction of trading volume, and always has done. If you can identify that there is a difference, why cant you define it and adapt to profit from it? BlueHorseshoe