Welcome to the new Traders Laboratory! Please bear with us as we finish the migration over the next few days. If you find any issues, want to leave feedback, get in touch with us, or offer suggestions please post to the Support forum here.
BlueHorseshoe
-
Content Count
1399 -
Joined
-
Last visited
Everything posted by BlueHorseshoe
-
Hi Brolizzz, I'm afraid I'm not qualified to offer you any specific trading advice or tell you which derivatives you should be trading. My recommendation (to you and all other traders) is simply that you ensure that you are fully cognizant of how the execuition foibles of a particular derivative impact upon your strategy's perfomance. You should be able to put a pretty certain figure to this, not a broad estimate. For an example of what failure to fill limit orders can do to a strategy, take a quick look at the perfomance reports in post #50 here: http://www.traderslaboratory.com/forums/day-trading-scalping/13811-daytraders-do-you-know-your-enemy-7.html Hope that's helpful, BlueHorseshoe
-
I guess C# etc are more acceptable if the OP was ever to look for professional programming work within the hedge fund industry? Perhaps it would also be interesting to have a thread detailing things that actually cannot be implemented in EasyLanguage. BlueHorseshoe
-
Hi UB, Thanks for an interesting post. In your opinion, given the caveats you discuss, is the "answer" to adapt and develop neural nets with these pitfalls firmly in mind, or to look for different approaches altogether? Would you be able to list some of the AI processes you have experimented with? Cheers, BlueHorseshoe -
Hi Syswizard, If I had a definitive answer to that then I would probably be a far richer man! However, I focus on minimizing the number of optimisable parameters within a strategy, by having parameters that produce acceptable returns regardless of value, and by employing fully recursive statistical measures rather than parameter dependent indicators. Another interesting thing to study is the predictability of shifts in the optimal parameters over time (for instance, is a more significant jump in the optimum more or less robust than a steady shift in the optimum?). The general concensus amongst quantitative traders and mathematicians seems to be that optimisation is a good thing, whereas over-optimisation isn't - an easier concept to advocate than to put into practice in my opinion. Eckhart refers to the 'optimisation paradox', whereby a historical optimum will likely perform slightly worse than the future optimum, but far better than a random value. I do think that Neural Networks, at least in the forms I am familiar with, carry significant risk of curve-fitting. BlueHorseshoe
-
Hi Perrin, Glad you found the book enjoyable - it was recommended to me by someone else on TL, but I forget who. Some of the strategies, as you probably gathered, are a product of machine learning, although obviously these algorithms have to be created by a human and then given a selection of data to mine. I think that HFT replacing Pit Traders pretty much sums it up, and I'm sure pit traders probably used to play games with one another in just the same way. I also agree about swing trading; in fact, I think that the rise of HFT has probably made the markets more suitable for swing trader. Though algorithms also occupy this space, the advantage of speed becomes irrelevant - trading from daily bars all that really matters is the effectiveness of the strategy. Or that's my take on it. BlueHorseshoe
-
I know what you're getting at, however . . . One might well ask "where's the chaos in the highly systematical approach of mechanical trend followers?" And yet a chaotic event of the kind described by Mandlebrot is nothing more than an 'outlier' in statistical terms - precisely the type of event from which true trend followers seek to profit. My point is that "chaotic" market structures (or non-structures) can be exploited through "non-chaotic" (ie systematic) trading approaches. As far as Bill Williams is concerned, I have no idea who he is or of what his approach consists . . . BlueHorseshoe
-
Theoretically this is certainly true (and with probable lower volatility of returns as well). But in practice . . . ? As someone pointed out earlier, for some it will be the case, and for others not. Lots of different people participate and turn a profit in different ways. BlueHorseshoe
-
Yes, but for some reason it's not considered as glamourous as the cut and thrust of trying to "scalp" the es etc, so the swing traders don't generate much air-time. If you haven't found it already, take a look at the thread 'Optiontimer's Project'. BlueHorseshoe
-
Is this a statement or a question? If it is a statement then it is wrong. The CME electronic trading system is a FIFO order book for matching bids and offers. The criteria according to which orders can be placed in the book are extremely limited, and all other order types that you see offered by by brokers are simply held on the broker's own server until the additional criteria are met, and then forwarded to the exchange in one of the acceptable order forms. A market maker derives his or her profit by taking the other side of an order, typically with the aim of profiting from the spread. The CME server simply matches orders, none of which are the exchange's own orders, and generates it's income through execution fees. I hope that's helpful to you. BlueHorseshoe
-
Hi Alex, Position within the numeric series is identified in square brackets after the value. Take the close of the bar for example. The current close is c[0], although if no square brackets follow the value then it is automatically assumed to be that of the current bar, so you will seldom or never see c[0] written, as just stating c is the same. The close one bar prior to the current close would be identified as c[1], and then c[2], and so on. Here is an example of code that buys when the 20sma average five bars prior has crossed above the 100sma five bars back: If average(c,20)[5] crosses over average(c,100)[5] then Buy this bar; If you wanted to ensure that the averages have not crossed back in the bars that followed, then you can either set up a variable that switches itself on and off when a crossover occurs, or you can repeat the process above to get something like the following: Vars: myAVG(0), myAVGslow(0); myAVG=Average(c,20); myAVGslow=Average(c,100); If myAVG[5] crosses over myAVGslow[5] and myAVG[4]>myAVGslow[4] and myAVG[3]>myAVGslow[3] and myAVG[2]>myAVGslow[2] and myAVG[1]>myAVGslow[1] and myAVG>myAVGslow then Buy this bar; I hope that answers your question. BlueHorseshoe
-
"Caution". BlueHorseshoe
-
There are very few vendors of educational products for traders based here in the UK. The two who run the larger and more successful businesses have long since learned that their profits lie in converting the non-trading public, and not in trying to jag forums full of those who have already discovered the limitations of trying to buy such an education ( or are jaded and cynical old trolls like myself ). If Knowledge to Action and Trading College can work this out Alan, why can't you? BlueHorseshoe
-
You'll have to pay for that - it's why he's got a big golden C under his name. Isn't that right Alan? BlueHorseshoe
-
Introduce Yourself Here - Don't Be Shy!!
BlueHorseshoe replied to trading4life's topic in Beginners Forum
Welcome to the forum, Jennifer, and a happy new year to you also! BlueHorseshoe -
Hi Tradeforever, Not all fx brokers make their money from the spread (or so I have been informed) - some just charge the very small spread of the underlying market and then a fixed commission per trade rather than 'marking up' the spread. As another alternative, you can trade many major currency pairs as futures. These are exchange traded and therefore well regulated, and although you'll have to pay commission the spread will be just like other liquid futures contracts (seldom more than a single tick). Also, you'll be able to see volume, which you can't in the interbank market. If you trade currency futures then the conflict of interest issue with brokers trading against you disappears. Hope that helps. BlueHorseshoe
-
Hi Splint, Glad that helped! I didn't want to mention that the EMA function you were attempting to call was unusual, as I assumed that the function must just have a different name within MultiCharts. If it is the same as in TradeStation then it's as follows: XAverage(NumericSeries,NumericSimple); Or as a specific example: XAverage(Close,20); Note that there is no 'displace' input required - it's just the time series value and the length. If you can't get it working with this information then try giving Tams a nudge to revisit the thread, as he/she is a MultiCharts user. Regards, BlueHorseshoe
-
Hi Splint, 1. The 'end' command is used to terminate and exit a loop - there is no loop within your code so the final line can be removed. 2. The error on line 7 isn't really the semicolon at the end as your code editor (and mine!) suggests - it's the brackets. Count the '(' brackets and there are 5. Count the ')' brackets and there are only 3. The number of each type of bracket here needs to be equal before you can terminate the line with a semicolon. 3. One set of brackets on line 7 is redundant - there is no need to bracket everything to the right of the = sign, so you can lose the first bracket. 4. You can create a new variable representing the average to simplify line 7. 5. Depending on your desired outcome the Print command may not have been used correctly. Here is your code with the basic issues corrected: Inputs: Price(c), Length(100), Displace(0) ; Variables: Gradient(0); Gradient=((MovAvgExponential(Price,Length,Displace)-(MovAvgExponential(Price,Length,Displace)[1]))/2); Plot1(Gradient); Print(Gradient); And here is your code with my additional suggestions and simplifications implemented: Inputs: Price(c), Length(100), Displace(0); Variables: myAVG(0), Gradient(0); myAVG=MovAvgExponential(Price,Length,Displace); Gradient=(myAVG-myAVG[1])/2; Plot1(Gradient); Print(Gradient); Hope that helps! BlueHorseshoe
-
If you don't want to create a new variable as Jeff suggests, just add the condition in each time for each entry/exit - the code will be less efficient, but if you're fairly new to coding it might make it a little easier to follow the logic when you review the code. You could also try the following, as the results may suprise you: Input: n(0); If t>=n and t<=n+100 then // insert entry logic ; The start time has now become an input, with the end time one hour later. By optimising the input n in Tradestation and then viewing the optimisation results (on the initial spreadsheet select the column you're interested in - profit factor, say - and then hit the plot as a graph icon) you will be able to see which hour(s) of the day produce the best return for your strategy. Hope that's helpful. BlueHorseshoe
-
Hi Tams, The number has not been altered, but is the number meaningful in and of itself, or can it only signify in the context of a time series? Selecting data points as you describe does alter the time series. BlueHorseshoe
-
Swing Trading Futures
BlueHorseshoe replied to karoshiman's topic in Swing Trading and Position Trading
Hi Karoshiman, I'm assuming that you have read Brandt's book? If not, then I recommend it - very practical and realistic account of how small gains are made over the long haul, rather than the usual fantastical results and dubious conclusions. Plus he has a long and auditable track record as a CTA. Good luck with the sugar position - I'm sure your victory will be sweet! BlueHorseshoe -
Hi Dude, Thanks for the link to the correlation table - might have saved me some work! The chart correlation I posted was over 5 days, I think, but it doesn't make a great deal of difference. I used a dollar index tracking ETF as the link with Yen was more implicit in Alan's statement than the relationship with the dollar itself. Alan is quite unlikely to re-visit the thread, I expect, so it's probably somewhat a moot point . . . BlueHorseshoe
-
Hi TLadmin, What was it about this thread that caused you to post again after an 8 month absence? Unfortunately there are not "two sides" to this argument. There is just simple fact (the fact being whether the two instruments are or are not correlated). As far as "benefit to users and attracting new users" goes, I would suggest (based on my experiences of TL) that your users respond better to objective factual content than they do to anecdotal nonsense of the kind that the OP was offering. We're in familiar vendor territory here - asked to provide any kind of evidence for the claims being made, the OP flees the thread and goes to troll for customers elsewhere. I could be very wrong, but I would guess that most of TL's income comes via advertising from major brokers etc, and not from small-time vendors with dated websites. Might the best way to encourage and maintain traffic here would therefore be to encourage and support quality content, and not the sort of unsupportable claptrap that is being pedalled in this thread? Of course, that's all just my opinion BlueHorseshoe
-
You might want to take a look at the chart below - the S&P and the Dollar certainly don't seem to be correlated . . . Or have I got my charts upside down? Do you have any other pearls of wisdom that you'd like to share with us Alan? BlueHorseshoe
-
Hi Alan, Can you provide some sort of stat to support this belief? Thanks, BlueHorseshoe
-
Hi Predictor, I realise that it is a side issue to the main dispute over this company's claims. With regard to the fill simulation methods you mention in NinjaTrader, I have found them to be woefully inaccurate (they're also present in TradeStation). Having said which, NT's method for volume may be more sophisticated than TS's. Thanks, BlueHorseshoe