BlueHorseshoe

Hey there, 'Position' is an easylanguage reserved word - you can't make it a variable within your code (in just the same way as you can't make 'Buy' or 'High' a variable). Change this to something like MyPos and you will be fine (I just tested it!). Kind regards, BlueHorseshoe

Maybe you should look at Hydrogen Futures? If there was any simple way to predict these things then the world would be a different place. The very fact that they are able to unfold when they do is defined by a quality of unexpectedness. You could look at Global Macro approaches, but don't use leverage, be prepared to take some heat, and admit when you are wrong. When it goes your way (the underlying event driving price, not just the trade) then scale into it. Or you can try and profit from the failure of others to correctly price the probability of these things occurring - take a look at what Nassim Taleb did with options. Don't get drawn in by some nonsense prognosticator of doom. There's always one out there. Might be something to do with a fib level or might be something to do with rising interest rates . . . don't get dragged into it. These people sell doom because . . . well, because it sells! Kind regards, BlueHorseshoe

I guess you could even do this simultaneously in numerous stocks to try and remain delta neutral to any sudden sector or index shift? For the OP: if what Patuca is discussing is of interest you might like to do a bit of research into the 'SOES Bandits'. I'd also recommend Scott Paterson's book 'Dark Pools'. BlueHorseshoe

Hello, Sorry for the delayed response! The Bid & Ask are always "passive" orders, but they're not necessarily Market Maker's orders. You too can post a limit order and be passive - just wait to see if it gets filled. Where your orders sit in the queue will depend on the contract/security and the exchange it trades on. Stocks on the NYSE will, I think, still have Market Makers (known on that exchange as 'specialists'). They probably get priority on fills (ie their passive limit order is moved to the front of the queue regardless of when they join), I can't remember for certain though. An instrument like the e-mini S&P futures contract, which trades on the CME, is operates on a FIFO basis ('first in, first out'), which means no matter who you are you join the back of the queue at the time you join. The only way to become front of the queue is if the orders ahead of yours are either cancelled or filled. I am not certain, but I think I recall that there are still (electronic, and probably HFT) a form of Market Maker for these futures products as well, called "designated liquidity providers". They get perks (rebates, waived exchange fees etc), but not priority in the queue. With regard to wide spreads . . . the spread is only as wide as you and other traders make it. If you post inside the spread you will narrow it. In very liquid instruments the spread is kept at the minimum price increment of one tick or one cent at all but the most volatile of times. I can count on one hand the number of times I've seen a two tick spread in the ES last more than the blink of an eye. Finally, all various forms of Market Maker have specific mandates to work with. A common requirement is that they maintain both a bid and ask at all times (needn't be the best bid and ask though). Another is simply that account for a certain volume of trading per day. Market makers and market-making algorithms from electronic trading firms may still be present when the Bid and Ask are wide - they just don't see any need to help narrow it by posting inside it. Remember, they make their money from the spread, so they want it wide. If you have any more questions then let me know and I will do my best to answer. Kind regards, BlueHorseshoe

Hi Johni, There are lots of ways you could do this. It's late at night, and this is the first way that springs into my tired brain (might not be the best): If x<30 then begin buy "EntryA" 1 contract this bar; countA=1; end; if countA=1 and countA[4]=1 and countA[5]=0 then begin sell "EntryA" 1 contract this bar; countA=0; end; If x<30 and x[1]<30 then begin buy "EntryB" 1 contract this bar; countB=1; end; if countB=1 and countB[4]=1 and countB[5]=0 then begin sell "EntryB" 1 contract this bar; countB=0; end; You can suss out EntryC for yourself from this. You'll also need to set up the count variables. Give it a go and let me know how you get on! BlueHorseshoe

How much you paying, Wiz? BlueHorseshoe

Maybe try using your 4:1 of durum and soft wheat, but ensure the latter is '00' ground. You could see if you can track down this UK series from a few years back: Simply Italian - Episode Guide - Channel 4 It claims to be about Italian cuisine more broadly, but it was entirely about pasta making. And the presenter is somewhat 'easy on the eye' BlueHorseshoe

Hi Urma, I wasn't challenging the quality of the recommendation, nor looking for an argument. Best wishes, BlueHorseshoe

Hi Urma, I don't doubt that - but if you post such a singular endorsement it's likely just to be taken down by the moderators. As soon as you acknowledge that he is one option amongst many, you're given much more general information and then stating a preference. As well as this, isn't it always good to point out to the OP that EL is relatively simple to learn to code (at the level the OP currently requires) and will serve as an invaluable tool in their development as a trader, allowing them to test the validity of their ideas before committing money to a strategy? Maybe I should stop interfering Kind regards, BlueHorseshoe

While JAM is a very well respected company, it is worth noting that there are other programmers out there You'll find a list of those endorsed by TS (including Miller) on their website. Whoever you use, I would expect to pay a minimum of around $90 per hour . . . Kind regards, BlueHorseshoe

I know that, but sometimes it's useful to have someone remind me of the distinction! Thanks, BlueHorseshoe

Maybe they'll make more money now they're upfront about their revenue stream? BlueHorseshoe

Hi Niko, You seem to be talking about two different types of fear: fear of missing out (greed) and fear of losing money and being wrong. Not sure how helpful the distinction is, but they may result in different behaviour? In the extreme instances, poor traders may chase a market that is running away without them, and run away from a market that is moving adversely against them (the old "I'll just move my stop and give it another couple of ticks" routine). As soon as you break volume down into trades at bid and ask, you can see a glimpse of what might be going on there (both these types of traders are likely to use market orders to enter and exit), but you can never be certain whether those trades are mostly establishing or liquidating positions. Hope some of those thoughts are helpful to you. Regards, BlueHorseshoe

Hello, I've just watched your video, and I have to say that I actually quite like this idea, and am about to devote an hour or two to testing it. All good approaches seem to come from a simple starting point such as this. I'll just note that the "raw probabilities" you display don't necessarily translate into a workable strategy - something may well happen 75% of the time, but what damage does the remaining 25% do to your balance? And of that 75%, what kind of adverse excursion do you have to sit through to net a profitable outcome? All food for thought . . . Regards, BlueHorseshoe

That's literally the bottom line, really - if you can make it work then it works, and why even worry about HFT? You can bet that the HFT algos and anyone else who profits in any other timeframe patiently awaits their opportunities, so you can only do the same. The only reason I have ever concerned myself with them is due to the fact that the types of short-term opportunity I have identified have given such insufficient margin for error that market microstructure can erode 100% of the gains . . . But that's not relevant to anyone who can identify better opportunities intraday than I have been able to BlueHorseshoe

I agree. It's difficult to know what goes on (beyond what you read, which may be lies and disinformation), but is that any different to in the past? How many people off the floor in the 80s knew what traders in the pits really did, or what strategies they used? How many people know that information even today? Ultimately you've just got to trade with the information you have, or not trade at all . . . And though you may not know how you will fair in the future, you can at least know how you would have faired in the past, HFTs and all else included. BlueHorseshoe

Is it a wash though? Or can it cause you to be adversely selected a higher percentage of the time? Consider this - you want to buy if price falls to X. You place a limit order there. The HFT is also bullish, and they have a limit order ahead of yours in the queue. The first order trades at X. Neither of you is filled. But the HFT suddenly gets spooked and recognises that price has some way still to fall. It pulls it's order for 1000 contracts. A seller crosses the spread and hits your bid, and price begins to fall . . . Who knows what the HFT "saw"? But even if you saw the same thing, you couldn't have pulled your order as fast as they did. The speed with which they are able to abandon a market has caused you to be adversely selected for a losing trade. The alternative is that the HFT remains convinced that X will be a bottom, and leaves its order in situ. Five hundred sellers hit the bid, you're not filled, and price rallies. Want to chase it with a market order? Well then a different type of HFT algo will be waiting to game you . . . This isn't even about the HFT knowing anything you don't or couldn't know - it's about how fast they are able to respond to that information. I reckon that if you can implement your strategy entirely with market orders (and those such as stops which translate into them) and make money, then that probably means you're outside the range where HFT can significantly impact your trading. Regards, BlueHorseshoe nb. Having just read your other post: if your average hold time is 120 seconds and losing a tick here and there isn't enough to damage your profitability, then I am rather impressed and more than a little envious

But how would they know? When they suddenly have to transact at worse prices, how would they know whether that was the result of HFT impact on a market? BlueHorseshoe

Yes, you can - prorealtime has a very user friendly and easy programming language - teach yourself. I started out with this platform years ago before moving on to tradestation, and it was very straighforward. Alternatively, you can pay me lots and lots of money to re-acquaint myself with the language and write the code BlueHorseshoe

Moving on to Market Making . . . The market maker attempts to earn the spread on each trade. In doing so, they act as counterparty to both Active Buyers and Active Sellers, and therefore facilitate trade, literally "making" the market. As a market maker you will typically act passively, having both buy and sell orders in the market at any one time. Ideally what you want is for an Active Seller to match your bid at 98 (making you long), and then an Active Buyer to match your ask at 101 (making you flat). This round trip just earned you the spread - 101-98=3 ticks profit. Market making has always been a lot more complicated. You might have orders in the market in hundreds of stocks at any one time. Your risk becomes more complicated as soon as you hold a position (you want to complete a round trip as quickly as possible). So you might begin to skew, or "lean" your book, having slightly more sell orders than buy orders, for example. Also, having access to the information about the willingness of informed Active market participants, you might begin to develop a slightly longer term directional view of the market, causing you to lean your book further to take advantage of this. Much of what HFTs do is just a super high speed version of this, in which computers identify mathematically optimal scenarios based on available information, and can adjust your book in milliseconds as this information changes. Being computers, they can also do this across thousands of markets, internationally, at once. Market makers in stocks were typically guys with computers (not HFT) on the floors of stock exchanges. Market makers in options and futures were typically the guys shouting in the pits of futures exchanges. They both did a similar thing. And both have now largely been replaced by HFT. Hope that all makes sense - please ask any questions! BlueHorseshoe

Hi Mastertonster, You have the basic idea, so answering your question is easy enough . . . Orders can be divided into two types: passive and active. Passive orders are limit orders. They're placed on the order book at the exchange, and will sit their until an Active order fills them. Think of Passive traders as "patient" traders. Active traders are "impatient" - they want their order filled now! They use Market orders. They just agree to buy at the lowest price that a Passive trader is willing to sell at (the Ask). Orders that are 'stops' are Active: until price reaches the stop level they are stored on your broker's servers, and then they become a Market order. Here's an example . . . The highest bid is 98. The lowest ask is 101. The gap between them is known as the Spread. If you want your Buy order filled right now, who do you trade with? The best available counterparty is the person who is willing to sell at 101. So you must cross the spread and buy from them at 101, paying 3 ticks more. If you try to Buy 50 shares and the Passive sellers at 101 are only offering to sell 20, then you will become a counterparty to the Passive sellers at 102, and so on . . . If the number of Passive sellers is low then your order will be filled at increasingly higher prices - this is known as "slippage". If you're willing to be patient, then you can simply join the queue of people willing to buy at 98 by placing a limit order at that price. What you're hoping is that some impatient seller will cross the spread and start selling to the Passive limit order buyers at 98. However, this may never happen, or there might be so many people ahead of you in the queue that the Active buyers are exhausted before you find a counterparty. If you're Passive you know the exact price at which your trade will execute, but you don't know whether it will happen or not. If you're Active then you can be certain that your trade will execute, but you cannot be sure at what price. BlueHorseshoe

High Frequency Trading Hails Its First Billionaire | Michael Shedlock | FINANCIAL SENSE BlueHorseshoe

Here's what you're up against: Watch The Banned HFT Spoofing Algo In Action | Zero Hedge . . . Although Coscia wasn't a massively well capitalised fund; he was a lone bandit with a computer in the basement, which makes him a hero not a villain as far as I'm concerned! BlueHorseshoe

Hi Flimbo, It's a good question. Everyone will have their own experience and their own answer, but I personally never had any success with it. I tried both the "stare at the T&S for hours and it will make sense" approach, and I have also spent a lot of time creating programs that attempted to make sense of the depth of book, and volumes crossing the spread with orders ("market delta"). I certainly don't think it's possible to compete with the HFTs for true scalping in any liquid futures market, even if you automate what you're doing. In a less liquid market - a single stock for instance, the tape may still provide an advantage? There's also the question of what you want to do . . . Say you're trading breakouts in the ES, entering a position intraday and then exiting at the close - being able to read market activity around breakout levels to gauge interest from buyers and sellers to identify optimal times to enter a position might be possible with practice. Trying to make this market by reading the tape, trading for single tick profits, I don't think will be. At the end of the day, most of the information that you will need to make valid suppositions about other participant's actions is not made available. You either need a very clever mathematical way to estimate it, or to devise an approach that makes better use of the information that you do have. Finally, the following thread might be of interest to you: http://www.traderslaboratory.com/forums/day-trading-scalping/13811-daytraders-do-you-know-your-enemy.html Hope that's helpful! BlueHorseshoe

SetExitOnClose will work fine in sim; in real trading you will need a separate exit instruction as UrmaBlume suggests. I would use: if t=1355 then begin sell this bar; buytocover this bar; end; The explanation from the EL dictionary is copied below. Kind regards, BlueHorseshoe SetExitOnClose is a built-in stop reserved word used to place an order to exit all shares or contracts in all positions on the close of the last bar of the trading session on an intra-day chart. For historical simulations, SetExitOnClose generates a market order on the bar close event of the last intra-day bar for each day in the chart. When used in an automated strategy placing real-world orders, SetExitOnClose generates a limit order into the post market trading session, (if one exists), otherwise a market order is generated for the open of the next regular session day. SetExitOnClose by default uses the closing session time specified by the custom session settings of the chart. SetExitOnClose can only be used in a Strategy. When automating a strategy, you will not want to use SetExitOnClose as a way to make sure that all of your positions are closed prior to the regular session end time. To do this, you will need to generate a closing order prior to the last bar in the chart.