BlueHorseshoe

Hi, The yen is generally a difficult market to trade . . . Does it have clear sustained trends, with obvious pullbacks for low risk entries? No. Does it enter sustained long term trends that last for years? Maybe, but you'd most likely get chopped up on your entry. Is it obviously mean-reverting? No. The closest that I think that I could get to describing its behaviour is to say that it trends in "deep" channels. By "deep" I mean that the distance between the top and the bottom of the channel tends to be quite large. Your best approach is probably therefore to trade off the top of a downward trending channel, and the bottom of an upward trending channel. This seems to be what you're doing, so assuming that you apply appropriate position sizing it's hard to see you getting into too much difficulty over the long run. Hope that's helpful, Bluehorseshoe

When "supply gains the upper hand" at a supply line, then surely price has found support there? It seems mostly a semantic argument, perhaps. Or maybe you're emphasizing the fact that you're not trying to predict price with any of these lines? Predicting price behaviour and predicting the levels at which we need to pay close attention to price behaviour in real time, however, are two different things in my opinion. BlueHorseshoe

It's very simple. Let's consider where the euro might find support next Tuesday. I can look at a couple of identical historical lows, join them together, point my horizontal line into the future, and tell you where that line will be on Tuesday. Or, I can look at a couple of non-identical historical lows that I feel are significant, join them together, point my diagonal line into the future, and tell you where that line will be on Tuesday. Ditto a fib line. Ditto a monthly pivot. Now, can you tell me where your MA will be next Tuesday? No. That's why it can't provide support - because nobody knows where it will be ahead of time. I can tell you quite clearly - buy at the 50% fib from these two data points next Tuesday. I can't tell you to buy at the 200SMA because the data that will be required to calculate it doesn't exist in advance. None of that should be read as an endorsement of any of the methods above - I'm just highlighting the differences. The first bunch are all linear extrapolations; the MA isn't. BlueHorseshoe

That seems to me to be like saying "given the many ways one can fire a laser at the moon, it's hardly precise". The point is that once you decide in which way to fire the laser or draw the trendline, then it is very precise. The line might have no value for trading, but it's precise. An average isn't - it's just an average, a neat way of generalising what's happened. If you want to buy a hell of a lot of something, like an institution might, then you might have a fixed price at which you're prepared to purchase it. You could have a buy limit at that price level. If you buy some, and then the market rallies away from your fixed level and you decide that it isn't coming back, then you might adjust the fixed level at which you're prepared to buy. If the market heads north again, you may decide to adjust again. Each rally away from your entry being roughly equal in price and time, the chances are that your second adjustment will be similar to your first. That means that I can connect your first and second prices levels and project forward to estimate your third - if you want to buy the market again, this is where you are most likely to do so. Of course, maybe they don't want to buy any more and the level has become meaningless. And obviously the market involves far more than one large institutional participant. So the point of the line isn't to give buyers or seller a "reason" to step in - they'll have their own reasons - the point is to locate the price level where they are most likely to do so. Whether or not they are then present at this level should tell you something about supply and demand. You use diagonal trendlines (supply and demand lines) on your own charts - what function do you feel that they perform? You don't use MAs - why not? How do you think that the two differ? Thanks BlueHorseshoe

In so far as all of the above are just guidelines, I agree with you. However, I think that lumping trendlines and MAs together in the same basket is misguided. An average (or anything similar) is the result of a basic statistical process enacted upon all data. A trendline is the result of a human (or algorithmic) process enacted upon key events within that data. Either may or may not be significant, depending on how you use them, but they shouldn't be treated the same. An MA is a far more general measure than a trendline. Consider the following scenario: You fire a ball bearing down a narrow sided chute (you're a pinball wizard!). As it ricochates of the sides of the chute you connect these points and plot support and resistance trendlines. Every half a second, you also note the ball's spacial location, and you plot an X period average of this. The average will run aproximately down the middle of the chute; the trendlines to the (in this case absolute) extremes. This is a poor example. The closest you'll get to the physically adamant sides of the chute in reality are the limit moves of the market you trade. However, it should illustrate my point . . . A trendline offers a precise location at which buyers ( or sellers ) should step in. An MA offers a generalised location at which price has historically traded. The obvious amateur question is "which is better?" I would suggest that the precision of thge trendline and the vaguery of the MA are directly proportional to the edge that can be found around each. As for a horizontal line vs a diagonal support or resistance line (or as Db would call them, a Supply or Demand line), then that depends on what the market is doing. If a rally is clearly in place then the price at which buyers are willing to trade will naturally be higher than it was X periods before; if the market is rangebound then a horizontal line should be as exact as anything more complicated. What does all of this amount to? An ability to diferentiate between a trending or a consolidating market early in the session might be the difference between profit and loss for many daytraders. And for newer traders, in drastically simplified terms, a tendline should be held accountable almost to the tick, whereas an MA is a far more general measuremenent, and should be given some leeway. BlueHorseshoe

Somehow I stumbled across this - http://www.guardian.co.uk/commentisfree/joris-luyendijk-banking-blog/2012/apr/30/interdealer-broker-voices-of-finance - absolutely no precise practical value, but interesting nonetheless . . . BlueHorseshoe

Bluehorseshoe

I disagree with this actually. I think the OP "cherry picks" trending markets to try and demonstrate the uselessness of such indicators when they are employed to try and pick tops and bottoms in longer term trends. I regard this as a mis-use of such indicators (or certainly one that defies common sense). BlueHorseshoe

Hi Db, Thanks for your reply. I would argue that if one is trading completely mechanically then these two things are essentially the same. What's more, if I could show that something rule-based worked very consistently it wouldn't really matter whether I personally was able to trade it effectively (I could be a complete psychological wreck who screws everything up). If you're really interested then surely you can follow what I describe in realtime on a chart - all you need to do is have a glance at the end of each day and note any entries or exits. This is very effective here when the market trends clearer (exactly the same could be said for what my chart shows), and is obviously a much cleaner and more cost efficient way to trade. But what happens when the trend is less obvious and the market becomes "choppy"? The mean reverting tendency of the ES is very well documented, and easily test-able. My experience is that what you show (a form of breakout/trend following) breaks down more significantly in trendless markets than does trading pullbacks. This is because the OB/OS part is more forgiving when one judges the trend incorrectly. See the chart. A good example of this is the trade we discussed in the Wykoff thread. With the benefit of hindsight, you called the trend correctly, I called it incorrectly (so much for MAs!). However, whether I took the long or the short signals during those couple of months didn't really matter - I would have made money either way. Times of trend change are when this approach most notably fails. I'm not trying to encourage/discourage anyone from whatever works for them, and personally I would trade without any indicators to support trading decisions if I thought that I could; what I don't like is when people get up on a pedestal (like the OP) and condemn something as useless just because they don't understand how to use it. BlueHorseshoe

Hi Db, My post above explains one way in which oscillators can be helpful. I suspect that the reason you don't pull up a chart and look at this yourself is that you've already decided that this sort of thing doesn't work . . . As for realtime - when I traded like this and tried to discuss a position with you in real time earlier this year, you didn't seem too interested. The chart below shows what happens when something like this works out well. If the trend is up buy OS pullbacks, if the trend is down sell OB rallies. If you can do this without an oscillator to define OB/OS or an indicator to define trend, then that's all good too. BlueHorseshoe

I like to print out a screenshot of the equity curve and then hold a ruler up against it - if I can tell the EC from the ruler then something's very wrong and I have to start over BlueHorseshoe

Greg, all your article has suceeded in doing here is demonstrating that you have little idea of what you're talking about . . . On the chart you show GOOG is clearly in an uptrend (whether you use MAs, trendlines, swing charting - nobody would really argue with this), so why on earth would you try and use an OB reading to short it, hmm? What you need to do is use an OS reading to BUY THE PULLBACKS. I've marked these on a chart below, along with failed rallies in a downtrend, which is where you short on OB. You'll also need an oscillator with a responsive setting - a 2-Period RSI or a 6-Period CCI, for instance. There you go - you just learned how to use oscillators to trade with the trend - and I didn't charge you a thing! Regards, BlueHorseshoe

Hi, You could always take the 'market scan' type approach - just sit down and look at what's moving on the day - if the eurozone is collapsing then @EC might be going places (is this the same contract that you guys all call 6E - never quite been sure?), if half the Persian gulf is aflame, then perhaps @CL, and if there's major collapse of the banking infrastructure, then @ES. And some days are just quiet all round, so maybe you had better avoid these if your strategy requires volatility. BlueHorseshoe

Hi Wrbtrader, Quite the opposite (confusion was my fault on re-reading) - I used to 'swing trade' (two, three, four day holds) - nicely profitable over 10 months, but rather a rollercoaster on a small account. I'm now trying to develop the methods I used into an approach for daytrading. The 'pullback' scenario I described was hypothetical, not a reference to what the ES is currently doing. As for Colonel B, I just wanted to know whether he was talking claptrap or whether there was more substance to his claims - given his detailed and very helpful post, the latter would seem to be the case. It works for him, so it's probably worth my time to investigate it further. BlueHorseshoe

Trading markets that are near perfectly correlated in sync is just going to tell you the same thing at the same time twice over. If one market 'leads' the other then there might be an edge to exploit in the lagging market (one side of an stat-arb opportunity, in fact). However, you talk about trading one or the other, so I don't think you're talking about this. Please could you provide some more explanation? The ES is retracing, say, and I want to buy this pullback because I think the trend will resume - what are the 10 year notes going to tell me that will help me to time my entry, or decide whether or not to trade? Cheers, BlueHorseshoe

The circumstances associated with finding a strategy that loses 90% and finding a strategy that wins 90% are logically identical. Remember also, it's just as easy to curve-fit a strategy for poor performance as it is to curve fit a strategy for best performance. The un-robustness of an over-fitted bad strategy and the robustness of an over-fitted good strategy are also logically identical. And finally, just 'flipping' the entries and exits idn't that straightforward, due to costs. Rather than find a strategy that loses 90% and then reverse, a more practical approach may be to find a general market tendency description that is untrue. BlueHorseshoe

Money Never Sleeps - the Wall Street sequel? BlueHorseshoe

Hi Jack, There are strategies that show a 90% historical win rate, but there are also many good reasons not to get too hung up on them: 1. They trade very infrequently (because they're always waiting for a 'sure thing'). This means that you'll be trying to make a decent dollar return from a scant few trades, which means that you'd need to load the boat for each one. It's not just about expectancy, but also opportunity (the frequency with which a system trades, allowing you to exercise its edge). 2. For the reason given in 1 above, historical performance is likely derived from a statistically insignificant set of trade results. 3. Win rate is an almost meaningless measure of system performance. What if you only had a 10% win rate but your winners were 20 times larger than your losers? You'd make money (indeed, that's roughly how many trend following funds operate, trading for outsized gains on outlying moves). The only way in which win rate is really relevant is from a psychological perspective - it will need to be reasonably high if you struggle to deal with losses. There's nothing wrong with this - I can't really handle anything with less than a 50% win rate, even though it might make money, so I deliberately focus on things with higher win rates. At the same time, I realise this has nothing at all to do with performance. Hope that's helpful to you, BlueHorseshoe

Hi Ingot, This looks interesting. Here are a few thoughts I had: You mention 800 pips down to recent support. Given that you're in a downward trending channel, and that the last two thrusts took out prior support by about 100%, will prior support be your profit target? There was loss of momentum between the previous two down waves, does this concern you? Other participants may define the top of the channel slightly differently, and supply may not enter precisely where you have drawn your line - how much room will you give it to move against your entry? The ascent into resistance has been more agressive than the prior upswings - is this a consideration, and is it an indication that this market is more or less likely to roll over? What's happening in other Euro pairs - is there resistance close by in other charts? How bearish are you really - if you get stopped out shorting the diagonal, would you short again at the horizontal around 1.3060 level? Please don't let any of this influence you in any significant way - you're trading it, not me! BlueHorseshoe

Hi, Remember that 'speculator' is a broad term. For instance, suppose my company makes picture frames and I need to import some lumber to make my products, then I might have to do business in another currency. If I think that this is cheaper to do now than it will be in a week's time when I actually need the lumber, then I might ask my bank to perform the currency conversion today. Effectively I am speculating, but it's rather different to how a retail forex trader speculates. It's also different to how a 'commercial' trades. If I was worried about changes in the price of the lumber itself, then I might hedge with lumber futures - this would imply that I maybe "know" something about it's value beyond publicly known information that a speculator has access to, because I am an end user within that industry (my friend at the sawmill gave me a tip-off!) . . . In the case of the currency swap then I am as in-the-dark as anyone else, and I am "speculating", despite the commercial nature of my interest. Of course, the transaction will be carried out by my bank. They might just charge me their markup and exchange the currency. But they might also think that my speculation about price is incorrect, and take the other side of my position. If they do this based upon some sort of technical information then they're most probably speculating, but if they do it because they "know" something (their friend at the treasury gave them a tip-off!), then they're acting more as a commercial. So these terms are all fairly fluid really . . . Hope that's helpful, BlueHorseshoe

Nope, but it still amounts to advertising, whether or not you benefit from it. And we only have your word that it's not an affiliate link . . . On a different note - the blog is really cool. I agree with you on stops for most types of trading, and have argued this point, combined with massive de-leveraging, here on the forum in the past (nobody ever wants to hear it though ). It would be great if you found the time to contribute here on TL from time to time. Here in the UK we have cookies with Smarties in - they're pretty nice! BlueHorseshoe

The reason that you have no ads on your blog is because Wordpress don't allow them - you couldn't put them there even if you wanted to . . . BlueHorseshoe

Chris, PM me and I'll give you some advice . . . BlueHorseshoe

That's simply not true, Andi. You have a link from your blog to the Xber9Trends website who, I can't help but notice, operate a rather juicy affiliate program . . . If I start looking at my cookies, will I find that I was tracked to that site? Would you care to retract your statement above, or maybe remove the direct link from your website? BlueHorseshoe

Let's just be very clear on this: I have total respect for DB, his great commitment to the threads he runs, and his willingness to share his knowledge and experience. I regularly read his threads and have found them helpful - and I agree with the vast majority of what he says - if I didn't then why else would I be here on this thread asking his opinions and advice? All of this, though, doesn't mean that I can't disagree with specific things he might claim. I am sure that DB doesn't wish to have his every utterance glorified as some kind of guru, and appreciates that anyone with intelligence will engage critically with what he posts. BlueHorseshoe