BlueHorseshoe

Just because past futures resembled past pasts does not mean to say that future futures will . . . The problem of Inductive Reasoning . . . I had Scottish porridge oats for breakfast this morning (as might have the Scottish philosopher David Hume, responsible for one of the most thorough treatments of induction). They've never poisoned me in the past, but who is to say that one day oats won't just suddenly become poisonous to the human body? Seems implausible, but it certainly isn't logically impossible. We all act as though inductive principles are totally valid, even though we know they are not, because they are the best tool that we have for dealing with the world around us. Quant trading is no exception. BlueHorseshoe

Hello, The option you are using to post is intended for developers to post finished versions of indicators. Posts such as yours should be made under the 'TS Custom Programming' section. Printing a label is fairly straightforward . . . however TS is a little clunky in how these are updated and displayed, so if this is plotted on your chart too frequently then your screen will end up a horrendous mess. Also, if you're going to be using TS a lot then it pays to learn a little EasyLanguage, so I'll give you some code below that prints text and you can have a go at tweaking it to suit your goals. Keep posting your adapted code back here with any problems you have and I will help you to get it to how you want it. BlueHorseshoe // HiLo Text // By FiveV January, 2009 // re: TS thread 85427 Input: leftStren (4), rightStren (4), UpAbove (.50), DnBelow (.50), TextColor (red), Decimal (2), LeftorRight (0 {minus is left, plus is right}); Var: TextIDHi (0), TextIDLo (0) ; If pivotHighVS(1,H,leftStren,rightStren,rightStren+1) <> -1 then begin TextIDHi = Text_New(Date,calctime(Time,LeftorRight),H[rightStren]+UpAbove,numtostr(High[rightStren],decimal)); Text_SetStyle(TextIDHi,1,1); Text_SetColor(TextIDHi,TextColor); end; If pivotLowVS(1,L,leftStren,rightStren,rightStren+1) <> -1 then begin TextIDLo = Text_New(Date,calctime(Time,LeftorRight),L[rightStren]-DnBelow,numtostr(Low[rightStren],decimal)); Text_SetStyle(TextIDLo,1,0); Text_SetColor(TextIDLo,TextColor); end;

Yes there is. Right click on the subpane chart. Select 'format symbol'. Go to the 'scaling' tab. Under 'Axis' on the 'Sub-graph' dropdown select 'Hidden'. Click 'OK' and . . . the sub-chart should vanish. Regards, BlueHorseshoe

Hi SCTS, Try the following code: inputs:starttime(0835),x(5); vars:count(0),upperbound(0),lowerbound(0); if t=starttime then count=1 else count=count+1; if count=x then begin upperbound=highest(h,x); lowerbound=lowest(l,x); end; plot1[x-1](upperbound); plot2[x-1](lowerbound); There are two inputs that you will need to manage. The first is starttime - you'll need to enter the closing time of the first bar of your 'day' for whatever timeframe you use for charting. For the ES this would be 0835 exchange time for a five minute chart, for example. If you only ever look at cash session data for stocks then there is a simpler and more efficient way to achieve this - just let me know and I'll tweak it. The second input is X - as defined in your post. This is just the number of bars into the session from which the channel is calculated. I have attached a screenshot to show how this looks on my chart (with X=3). Any problems, just give me a shout! BlueHorseshoe

Attached is a screenshot and 10yr equity curve for pullback trading BSX, hopefully to try and inspire you Remember, the MA trend filter is near historic optimal (50 periods), so your actual performance would not have been quite this good. On the other hand, this is long only - if you short rallies in downtrends you can probably double your trading opportunities. Kind regards, BlueHorseshoe

A funny coincidence - I have just been watching seals hunting emperor penguins on the 'Frozen Seas' episode of Blue Planet If the feather doesn't move often enough though, then what's the point? If the frequency isn't higher than EOD trading, then why spend all day in front of a computer? BlueHorseshoe

Try something like: If marketposition=0 then sellshort next bar at 1.334; When you have an existing short position then marketposition=-1 and when you have an open long position then marketposition=1. This is regardless of the number of shares/contracts you hold. When you are flat market position=0. Note that both this method and UrmaBlume's will permit a new position to be initiated if the prior position has been exited and then price trades back down to 1.334. You may want to limit yourself to one entry per day at this level, in which case your code might look like this: If entriestoday=0 then sellshort next bar at 1.334; There is no need to state a condition for marketposition because if there have been no entries you must already be flat (unless you're holding positions overnight). If you allow a specified number of positions each day - say four - then you'd need: If marketposition=0 and entriestoday=4 then sellshort next bar at 1.334; I see no reason for the "begin . . . end" loop in Urma's code, a specific price would normally be specified, and "this bar" can be used instead of "at market", so this could be edited to something like: If current shares=0 and close<=1.334 then buy this bar; Most likely the difference may be because Urma is using the MultiCharts version of EasyLanguage. Hope that helps. Kind regards, BlueHorseshoe

Firstly, what do you mean by scalper? Assuming you mean buying at bid and selling at ask to make the spread, then there is an advantage if you can do it profitably. The advantage is that your equity curve will be made up of lots and lots of smaller gains and losses making for a much smoother ride (which has significant implications for position sizing). The key concept here has been termed "expectatunity" - a portmanteau of "expectancy" and "opportunity". It is fine to have a small expectancy or edge if you have the opportunity to apply it with sufficient frequency. Think of casinos - the house edge is tiny, but as long as they get enough gamblers through the doors it stacks up to huge profits (hence all the loss-leaders like free food!). In this sense, Scalping is "better" than all other trading, with the exception of true arbitrage. Unfortunately true scalping, which was once the preserve of floor traders and market makers, has now become a HFT dominated game, and is therefore pretty much inaccessible to the retail trader. This leaves daytrading . . . Finding profitable daytrading strategies (great entries and exits) is easy . . . finding ones that comfortably overcome costs is definitely not. Regards, BlueHorseshoe

One of the Turtles - Curtis Faith - wrote a very good book describing the methodology and the issues surrounding it. I would definitely recommend this (much better than the Covel books on the subject!). Regards, BlueHorseshoe

Hi, What you describe sounds promising, and as long as you don't have unrealistic expectations of performance it ought to work well enough. I would caution you not to get bogged down in the use of too many indicators/oscillators (they're mostly overly complex manipulations of price data that tell you nothing and exhibit serious lag). In terms of the more technical, mechanical aspect to your approach, you only need to do two things: Identify the long term trend. A simple moving average is perfectly adequate for this purpose - I have yet to find anything that clearly beats it. What setting for the moving average? I would recommend that you optimise this (it's just one parameter), and then keep on re-optimising with each new potential trading opportunity. You'll usually underperform the historical optimum, but outperform what you would have achieved with no optimisation at all. Identify the pullbacks in the trend. To do this, you just need to compare aspects of prices over the last couple of bars. Something as simple as "three consecutive down closes in an uptrend" can work, but I recommend a 2 period RSI. If you look at the calculation of this (you'll find the formula on Wikipedia), you'll see that rather than being some complicated nonsense, in the 2 period form it becomes a very simple price comparison. For each of the above, you could then apply a discretionary overlay based on your knowledge of the industry. You'd simply want a fundamental (or insider!?!?!) answer to each of the following questions: Is the long term trend really valid, or just a price excursion that you can't find any real fundamental justification for? Is the pullback just a brief and minor correction before the long term trend resumes, or is it the result of something much more significant that may begin a major trend change for that particular stock? Hope that helps! BlueHorseshoe

A Money Management / Position Sizing idea for random trading - hopefully of interest to those reading this thread: http://www.traderslaboratory.com/forums/money-management/17381-random-trading-natural-selection.html#post186339 BlueHorseshoe

Hi, Thanks for reading. I'm not sure about my statement either - in fact I'm not sure about the whole argument. One simple problem with what I say there is that days with outlying ranges tend to be down days (crash scenarios) in certain markets. Another problem is that there are very good fundamental reasons for expecting certain markets such as SPY to exhibit a long term uptrend (caused by smaller but more frequent up days). One possible solution to this is to "tune" the degree to which randomness dictates trading decisions (i.e. skew the distribution of outcomes or "weight the coin"). Another is get closer to the noise by applying the concept using smaller timeframes. Both of the above obviously entail new problems all of their own. As soon as you apply the concept I describe in the PDF though, does any of the above matter? Some instances of the strategy will benefit from the skew, whereas others won't; the former enjoy increased position size to generate a net profit. Hope I've understood you correctly. BlueHorseshoe

Attached is a PDF containing my thoughts about random trading, largely inspired by Mystic's thread and podcast interview. It would be great to hear people's responses, especially if you can spot flaws in my argument. BlueHorseshoe Random Trading and Natural Selection.pdf

Good listening - thanks! BlueHorseshoe

No. This is already the "mini" contract, remember, and has a much smaller tick value than the big S&P contract. One possibility (depending on what you're doing, where you live, and definitely not for daytrading) is to use spreadbetting or CFDs to trade in smaller increments. Another option, once again depending on what you're doing, would be an ETF such as SPY. Some of the currency futures contracts are much smaller than others, and it is possible to get "micro" accounts that allow you to trade smaller lots on the spot FX market. Hope that's helpful, BlueHorseshoe

Has anyone done any investigative work around this as a specific entry concept - i.e. "this is where I predict a large number of sell stops must lie based on other participants having established a long position due to the preceding price action, therefore if price moves to this area I will try to scalp around it" . . . ? I suppose the first thing to do would be to identify the entry methodologies most often used by losing traders, and then to look for a failure of these that would result in the stop being hit. Volume at bid/ask would perhaps give an indication of uptake on any particular instance of a signal and how many contracts were "trapped". All this being said, we're essentially talking about retail traders here, and they're such a minority of market volume you have to wonder whether where they place their stops is entirely diluted by commercial volume . . . BlueHorseshoe

I don't subscribe to the notion of "big fish" substantially manipulating liquid markets as Leecifer seems to suggest . . . but as far as trading approaches go point 8 seems to me to be just about the best advice that anyone could give you. BlueHorseshoe

That answers my question very thoroughly - thanks! Regards, BlueHorseshoe

My argument is probably pretty unclear from the start, as everything you say above is obviously correct. My point is that, depending on the trading style the 2% rule is open to debate, and that it is more open to debate where no leverage is used, as a 100% price decline is necessary before a position is unavoidably closed out, whereas when leverage is used there is less ability to quantify risk one a trader's own terms and greater likelihood that this will be determined by exchange/broker and margin requirements. If you're unleveraged, you can trade without a stop with little chance of losing all your equity. If you're leveraged you can't. None of this is anything that won't already be obvious to yourself, but might help a newer trader in thinking about risk and how/who defines it. BlueHorseshoe

I haven't misunderstood the 2% rule. You have introduced another (non-linear) element - the notion of a stop-loss. Removing that (complicating) addition to what we are discussing, then the maximum loss you can occur is entryprice*bigpointvalue. If I have 10k in my account today and GLD trades at 100, then I can buy 100 shares of GLD. My 'risk' on the trade is 100% (crazy) . . . or is it? Say that my holding time is restricted to one day, and I know that the average one day decline in GLD is far less than 1%, and with low deviation from this mean. Is the risk to my account really 10k? In actual, literal terms it is. But in reality, a 100% single day price decline would be required to realise this 10k loss. As soon as you introduce leverage, the situation obviously changes. If you introduce a hard stop to limit losses, then the situation becomes as you describe, but this is because (a) you have restricted the loss that can be incurred, rather than the market restricting it, (b) the level of leverage is incorporated into the calculation you make. On a slightly different but related note, if you take two strategies with identical single contract dollar outcomes over a period, but with different %win rate profiles, then the application of the 2% rule can significantly alter dollar outcomes when applied alongside a fixed fractional type position sizing formula. Consider the following with fixed-fractional: - 2% + 2% - 2% + 2% - 2% + 15% Versus: + 2% + 2% - 2% - 2% - 2% + 15% Happy to talk about this further, but maybe it should all be moved across to a MM thread as it's diverging from the topic of this thread? BlueHorseshoe

Hi Karoshiman, I wasn't offended, and the first sentence of my post was sincere, not sarcastic. By "getting smart" I actually meant it - to set all this up within a trading platform takes more than a little effort. I very much agree - I think that daytrade999 needs to work out some concrete answers before taking up any idea that has been introduced in this thread. Cheers, BlueHorseshoe

Hello, I'd rather not - I'm no expert on any of this. If you hang around long enough you'll pick things like this up. There are regular posters here who have managed funds of considerable size and worked "in the industry". They would be able to give you far more useful and reliable information than I would. There are also books that touch on this kind of thing - anyone any good suggestion? BlueHorseshoe

If you wanted to get really smart (referring to my original example), rather than use SPX, you'd use 'KaroshimansSPX' - a bespoke index calculated by netting the price changes of A and B off the SPX. Otherwise . . . Suppose A and SPX exhibit strong positive correlation at all times. A certain amount of the reason why SPX behaves as A behaves is because A is a component of SPX. Though it may be true, the information you have is therefore part tautology, and so not very useful! You'd also want to know exactly how the index is weighted. What happens if you take the APPL component out of QQQ, for example? If you want to discuss any of the other implications you have in mind, please go ahead. Best wishes, BlueHorseshoe

I am not a "pro" - just a retail trader like yourself who treats trading a small account like an enjoyable and stimulating hobby. I make no distinction between Friday and any other day of the week, but then I'm holding positions for longer than a day. It is totally up to you what special day-of-week considerations you may have - but you should make sure you have some quantitative research to support them - how have employment numbers or air strikes typically impacted upon your strategy's performance in the past? The "2% of capital" rule comes from managed futures, and if you're trading un-leveraged stocks it's not entirely relevant. Do you think that Buffet only invests 2% and leaves the rest in cash? * Say you use all your available capital tomorrow to buy shares of GLD. Now, is 100% of your account really at risk? Is the value of gold actually going to fall to zero at any time soon? The only scenarios in which this might happen are "end of the world" type scenarios in which the value of your shares will be the least of your worries. Kind regards, BlueHorseshoe * I believe this is one of the reasons fund managers often use leveraged products - they have a fraction of their capital funding a postion but still get the same bang for their buck as with an un-leveraged asset, and meanwhile the rest sits in t-bills etc bolstering their returns by a few extra percent each year.

I have read the book and played around with some of the concepts and my conclusion was not. Most of the concepts are volatility-breakout type things, not dissimilar to some of the Larry Williams approaches of old - not generally something that works in many of todays markets. Others may have a different opinion. Kind regards, BlueHorseshoe ps If anyone wants a PDF copy then give me a shout!