BlueHorseshoe

Hi, Thanks for the links. I have already seen 'Floored'. Although it was fascinating, it focussed more on the traders themselves, exchange culture, and emotional responses to the decline of floor trading, I thought. There was never really any 'technical' detail about what they did in, say, the 80s, and exactly how Globex later impacted on it. Rightly so, as that obviously wasn't what the film was intended to to document and it would have made it boring to most people, but hopefully you get my point. Thanks BlueHorseshoe

As a VERY general rule, the higher the frequency with which you trade, the fewer markets you need to trade. Floor traders used to make fortunes scalping just a single market; a trend following fund holding positions for months or years will often trade upwards of a hundred markets. Ultimately though, it's up to you. Have a play around trading in sim - how does it feel if you try and follow gold, crude, the ES, and the Euro all at once? What about if you just trade wheat? You'll quickly discover your own preferences if you're new to trading - just try not to lose lots of money doing so. Hope that's helpful. BlueHorseshoe

I'm afraid that I'm probably guilty of dragging this thread off topic by mentioning HFT . . . Several of you have obviously worked on the floors of exchanges scalping various instruments. It would be great if someone could give a more detailed account of what scalping used to entail in the pits - when markets became more electronic, what were the signs you noticed as a scalper? I'd also be fascinated to read an account from anyone else like ZDO who actively scalps today using a retail trading platform. Why do you think that a small handful of traders are able to do this, while a whole bunch of others (like me!) just don't get it? BlueHorseshoe

Hi SIUYA, One of the attractions from my reading was that the KF calculation is recursive, and I'm always looking for ways to remove (easily curve-fitted) inputs such as MA length. Unfortunately, upon further reading it would seem that the Kalman Filter also has an input. Rather than controlling a lookback length as with an MA, the input adjusts the sensitivity to 'noise'. So the benefit that I imagined existed in this respect isn't there. My enquiries are for a slightly different reason . . . I'm fed up with hard stops and I'm looking for other ways to manage risk. As you know, I considered options. I'm now looking at the possibility of incorporating elements of stat arb, and Kalman Filters are one of the two common procedures used in calculating an instrument's beta. But that's something for another thread. I have been able to piece together a Kalman Filter from the formulas and information I've found online - I'll share the code for this on this thread shortly. I guess I was hoping that this thread might elicit input from someone with experience of pairs trading . . . Cheers, BlueHorseshoe

Hi Db, Please could you share a little info about the order types you would use for entries and exits? Thanks, BlueHorseshoe

Hi ZDO, Thanks for an informative post. What do you trade, and what is a typical adverse excursion for you when trading for a one tick profit? BlueHorseshoe

I don't think that HFTs conduct StatArb. Pairs often have a half-life of weeks or months; HFTs are in and out of positions within seconds. And surely trading inside the bid ask spread is pretty much scalping, no? Most of the approaches you list are market neutral - it is reasonably well documented that HFTs frequently experience a volatility of returns that at least partially preclude this. However, enough has probably been said here about HFT - the topic was manual scalping afterall. BlueHorseshoe

Maybe the HFTs don't scalp futures. Who really knows? Maybe all that short term volume is arb against baskets and ETFs. Like you, I have learned to stear clear and not waste my time searching for these opportunities. For anyone who's interested in manual scalping, then it would be intresting to try and find out whether 'The Flipper' is still flourishing (I believe he was eventually unmasked as 'Paul Rotter', if I remember rightly). BlueHorseshoe

I suspect you may be missing something. Unfortunately I am also missing the same thing, so I can't tell you what it is. The reply from Windsurfer gives a good idea though. Incidentally, I spent a lot of time at the start of the year trying to develop a procedure for seeding the order book with limits pre-market and then sweeping them before they were hit if they were not required. In other words, a way to pretty much ensure a fill at limit. I did not succeed. BlueHorseshoe

There's another book that follows pretty much the same format as the MW books which you might enjoy if you haven't already read it. Nowhere near as good, but still well worth reading. More of a focus on successful retail traders rather than hedgefund superstars, although for some the interview is clearly an exercise in publicising a fund launch. Real People; Real Traders - Adrienne Laris Toghraie and Murray Ruggiero One author is a psychologist/trading coach, while the other is a system developer. BlueHorseshoe Real Traders, Real People: How People Like You are Making it in the Markets: Amazon.co.uk: Adrienne Tograhie, Murray Ruggiero: Books

You could start putting money aside until you have enough in an account (that you can afford to lose). There will still be plenty of trading opportunities in one or two years time, so there's really no rush. Whereas on the other hand, if you do rush, and in some way overtrade, then you probably won't be able to exploit whatever opportunities exist further down the line. Having said all that, the Predictor's recommendation that you investigate futures is, depending on your trading style, probably a good one. Hope that's helpful, BlueHorseshoe

As others have hinted, I think there was a natural progression from traditional market-making and pit trading into HFT. Many of the original HFT firms were controlled or backed by former successful floor traders. Manual scalping may still be possible for the skilled and well trained stock trader; for trading a liquid futures market I would expect the process would now need to be machine-implemented. Just for starters, you'd need a naked access zero-latency data feed and then intelligent algorithms to publish filtered tick streams minimizing flutter etc. There seems to be a general idea (although unless you know someone involved in HFT there would be no way to verify this) that scalping strategies operated by such firms employ modelling techniques, usually cribbed from non-financial disciplines, to measure trade intensity and search for liquidity. I've attached a paper that might be of interest. BlueHorseshoe Particle Filtering.pdf

Almost a net profit of $400,000 then. So the next question is - are CFTC fines tax-deductable? BlueHorseshoe

Hi Mike, Would be great if you found the time to stick around on TL. Regards, BlueHorseshoe

There doesn't appear to be a thread anywhere on TL discussing Kalman Filters - obviously there should be! Here's a link to an Ernie Chan post giving a description of the potential utility for Kalman Filters in linear regression models, and a more general post from another blog: http://epchan.blogspot.co.uk/2011/04/many-facets-of-linear-regression.html http://intelligenttradingtech.blogspot.co.uk/2010/05/kalman-filter-for-financial-time-series.html Does anyone have any knowledge, experience, or thoughts on how to apply Kalman Filters? Thanks BlueHorseshoe

MMS reported me and the CFTC stripped me of the awards . . . Easy come, easy go . . .

Someone's been reading Northrop Frye . . . BlueHorseshoe

A few weeks ago another TL user suggested to me that I would do well to focus my attention more on 'tactics' than on 'strategies'. Reading an old floor trader book the other day, an identical point was made by the author. I was wondering what others on here make of this distinction? Do you have 'tactics' that you deploy in your trading, and how do they differ from the 'strategies' you use? What do you think the term 'tactic' means in trading? BlueHorseshoe

Hi Curtis, I was intending to initiate a thread on this topic within the next couple of days - hopefully this will draw out info from others on here who might have more experience with pairs/spreads/statarb in futures. BlueHorseshoe

Thanks to everyone for their replies in this thread. It has been extremely helpful to me and is exactly the sort of reason that TL is so worthwhile. Hopefully my posts are ocassionally equally helpful to others here. Options clearly aren't the answer to the problem I was trying to solve here (indeed, there may be no answer - limited risk and unlimited upside is doubtless an expectation that borders on the unrealistic). They seem unsuited to directional strategies, and the element of uncertainty in their pricing seems no less severe than the element of uncertainty in the directional behaviour of the futures position whose risk I was seeking to modify. Cheers, BlueHorseshoe

Not necessarily . . . If they were all following an identical rule-set in the same market, then whether there were 1 or 10,000 traders and accounts becomes irrelevant; they'd all be doing exactly the same thing. If the trade decisions were poor then the result would simply equate to a series of larger positions. Assuming that this wasn't the case, then there is the possibility that some were notionally profitable and there was a theoretical edge to the way they were trading, but that in practice this edge was not large enough to overcome costs. The fact that they appear to have been daytrading makes this more likely. One thing that I find interesting in the report is the repeated reference to adverts on CraigsList. Clearly this company's marketing was targeted more at non-traders seeking some sort of employment, as opposed to struggling/failing traders. Another thing I find interesting is the difference between people's naivety and willingness to believe something, and their willingness to part with significant sums of money based on this belief. When I was completely new to trading I was certainly naive enough to believe that there was great value in the teachings of certain vendors, and yet I was never naive enough to part with thousands of dollars for tuition of any kind. BlueHorseshoe

I have spent a lot of time backtesting with price behaviour around daily pivots and have been unable to find any kind of edge with them whatsoever. This doesn't mean that there isn't one, of course - I could have missed something that others are aware of. But I can certainly say with confidence that I wouldn't place any kind of emphasis on pivot levels. I hope that's helpful to you. BlueHorseshoe

If you're certain it's a good trade, you could always try shorting with CFDs or Spreadbetting. My guess is that, unless you're wanting to take a pretty significant position, you'll be able to short with a company like IG and they'll bucket the trade and not worry themselves about shorting the underlying. I'm not honestly convinced that they accurately track the liquidity of smallcap stocks anyway, so their spread may not properly reflect their own difficulty in shorting. Hope that's helpful. BlueHorseshoe

Hi Karoshiman, Thanks for your reply. I was very unclear of what my intentions were in my first post (I think I mentioned improving profitability, which is completely wrong!). My intention is to find a way of managing and quantifying risk that does not entail the disadvantages of using a hard stop in the underlying. I am not usually the beneficiary of large moves in my trading - when I am it is normally by 'fluke'. My trades are typically held for three to five days and take advantage of whatever movement takes place within that time period - in the ES this is generally about 15-20 points. From what you are saying, it seems that an Out of the Money option would not be sufficiently responsive. At what point does an option become fully responsive to movement in the underlying - ie at what point does it move totally in tandem with the underlying? There is a reason why option strategies exist... because options are mostly used for very specific trading purposes other than pure directional trading, which can be done with futures as well and with a lot less complexity. I have no doubt that this is probably true, and that I am hoping for something that cannot actually be achieved with options. I'd like to be at least reasonably clear about why, however, before abandoning this line investigation. I am not a big fan of sim trading. But in this case, I highly recommend to sim trade options for a while parallel to your current trading and see how prices react and get a feel for it. This sounds like it might be a good idea, as it I think it is this practical understanding that is lacking (I have already read about the theory of options pricing), and obviously I can't afford to gain this experience with real money. Any suggestions on where to start with a sim account, given that I typically only trade one or two times per month? Many thanks, BlueHorseshoe

Hi SIUYA, Thanks for the detailed reply. This quickly starts to get complicated, and Karoshiman makes a good point below - Options generally have their own strategies because they are not useful for simple directional trading. But I'd like to keep working through this until I fully understand why. A lot will decide on how you wish to manage the options. Do you wish to just get an equivalent exposure at the time. (if so then you modify the number of options purchased depending on the delta), or do you want to risk a certain amount of premium in absolute $ amounts, or do you wish to get enough exposure so that if the market collapses then your exposure is a certain amount? I suppose I want to replicate what I can do with futures, but with a known and limited risk at the time of entry and thereafter. In other words, I want a method of limiting risk without the possibility of being stopped out. I fully realise that this may be ludicrously idealistic expectation of what can be done with options, however! Can we try to work through a specific example (though it needn't be in a precise fashion) so that I can see how this can or can't be done? In the case of my chart above, the daily position summary was roughly as follows: 15/06/12 - Sell short on close @1337.5 18/06/12 - Market closed @ 1341 - Unrealised P/L = -$175 19/06/12 - Market closed @ 1350 - Unrealised P/L = -$625 20/06/12 - Market closed @ 1351 - Unrealised P/L = -$675 21/06/12 - Buy to cover on close @ 1319.00 - Realised P/L = +$925 This is obviously based on a contract trading at $50 per point. How could I best have gone about replicating the above position with options to achieve the same dollar outcome? What might my costs have been, and how would I have been able to calculate these? To make a profit equivalent to $50 per point in the underlying, would my premium be $5, $50, $500? If on 20th June, had the ES closed at 1300, how could I calculate what my profit would have been with the same options scenario at that time? eg; with a ten tick down move in the underlying, you should expect to see a 5 tick up move in a put that has a delta 50 (or 0.5 = 50%), a 3 tick up move in the delta 30 put. So to replicate the price movement of a futures position precisely, I would need to purchase a Delta 100 Put? Does such a thing exist? When you compare ticks above, does a tick in the option price have the same value as a tick in the underlying (ie does a single ES option move in $12.50 tick increments the same as the ES)? As options are priced (see above) with a volatility and time component, then usually yes. Sometimes no as it depends on , how long it takes to occur, and what the implied volatilities do between purchase and sale. I understand this, but I don't understand how it can be managed in such a way as to make the option tradeable in any quantifiable manner. This surely adds in a host of additional conditions that must be met in order to be profitable - "not only must the market trade in the direction you predict, but also within a certain amount of time, and in a certain way"? It seems that I could make what with a futures contract would be a profitable trade, but the profit I would be likely to achieve with futures would be totally indeterminate? Thank you for your patience! BlueHorseshoe