The most active is being SI (as usually, since a while). On the strategy side development, i am also working on a modification of my current entry oscillator indicator to tune better the entries. I am also reducing the "wrapping" space, in order to avoid large runaways (still possible with current settings). I found pretty useful the first hitting time formulas of the various GBM variants. In a while i will begin overlaying new intruments (or entire folios) to "cover" all trading range. <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3144370" width="1200" /> Tom
If I understand your latest mods: Select a folio of instruments matched in terms of volatility and speed. Allow entries to slide with market direction. Tighten the range in order to trigger earlier reversals. Can you not reduce the individual instrument scalp size and keep a large folio take profit target to accomplish the same? ie. set the instrument scalp to 8x spread and folio profit to $500?
Hi PocketChange, yes you are right. I think that the continuous observation of the bot at work seems to suggest that it may be useful to group instruments together according to some key features related to the general dynamics. As one would not clearly mix micro with mini, similarly within the minis themselves, different instruments seems to have pretty different dynamics. Actual values, scalp size and folio target profit should not be too important, as long as they are within reasonable ranges. I think that they key point for relatively large investors is the possibility to "cover" the whole trading ranges, in order to have a sort of "self hedging" effect caused by the multiple folios active on the same account with entries in all different directions. Clearly, this way, within the natural market fluctuation, the whole movements should be eventually turned into profits (as each player is intrinsecally profitable). Possibly strandled position (runaways) should be temporarily compensated by the instruments more active in capturing the smaller fluctuations. A good area of research could also be the "optimal" rules for "folio overlay activation". Tom
This time the folio had some initial dd caused by strong moves of SI. After all it turned out to be a pretty bad folio, due to the eccessive speed of SI wrt to the other components. Also the logic of the "clone instrument" i think requires a precise assessment through simulation and can't be really deal with with pure "intuition". In fact, playing now discretionarily with "clones" i am having the impression that those in profit should be closed, to allow for the original logic to work on retracements. Letting multiple clones open, i suspect, can increase "stagnation" time. This is something i need to look into with simulation. I think i need to code an additional automatic logic to possibly open clone instruments which will capture smaller timeframe fluctuations, and then shut them down. Tom <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3146378" width="1200" />
I have been thinking about the "cloning" question, and i think that a reasonable way to employ them is probably creating them on <b>sideways</b> instrument to increase gains. In case an instrument begins to run, it is probably useful to manually close all the clones (close+manual mode) and leave the original running (or else it's like having the disadvantage of highly correlated instruments). Then when the original is near to end of run (<b>measure exhaustion</b>) they could be renabled (auto), making sure they take the <b>same direction of the original</b> (to scalp and overload the retracement) this could be useful, as often a sideways behavior can last for many days, so it can be useful to <b>overload</b> the fluctuating instruments. After all it appears that a larger folio (with many sideways instruments) helps a lot to quicky realize a lot of market "noise" (easy profits) [clearly, this needs very good capital]. Here is a folio of 23 instruments. I have added a "nighty" style (see picture) [on suggestion by a smart reader]. <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3153447" width="1200" /> Now i am working on complete <b>stop removal</b>, by introducing <b> entry semphores</b>. This should reduce significantly drawdowns, by at least keeping unchanged the performances. Tom
Hi there Restarted trading the big folio. A friend has jockingly called it on a skype chat the "kitchen folio" in reference it has nearly all in it ;-))) Well it's just for test purposes. Besides, a large folio is certainly good for capturing as much noise as possible (if one has the capital ;-)). After a digression on some "cosmetic" changes and new features(see the new "night" style) suggested by a new friend (investor/trader, who is possibly preparing to start real trading too) i am now working again on further strategy research, and always testing new ideas and variants. This is becoming a sort of huge "collaborative" effort and i really thank all helping so much. (Particulary, J... from Canada ;-)) for his invaluable suggestions, and who has even written an entire large application just for testing purposes of the bot execution). <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3157962" width="1200" /> Tom
This week had some larger dd (60K), but nothing worrying for a 23 folio. SI crazy as always ruled the folio. Anyway the investment turned into a good profit, with a realized over 35K. Had a few of new strategic ideas, to steal some % point of efficiency and i am currently implementing them. I have another instance running which is also doing good (currently 26K in less than 17 hours): will show in a next post (the strategy is a newer version). <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3167992" width="1200" />
Here is the newer instance (in the "night" view). Luckyest performers were SI, CL, HO, CHF, QM. <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3168729" width="1200" />
And these are the trades of the incredibly active SI. <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3168735" width="1200" /> Even if this has been named the "kitchen sink folio" by my smart canadian friend, it's undoubtable that a large folio is much fun to trade. And trading multiple overlaid folio instances even more fun (clearly having the necessary Ms to trade it). Small folios are more subject to a phenomenon of stagnation, which is not a pleasant thing. I am currently attempting to make the game tigher, while retaining the same level of general efficiency. Not an easy task, it seems ;-) Tom
hi During the holidays i have been working on strategy ideas. I have found a <b>new exit rule</b> which seems to increase performances in simulation. Also it feels better psicologically too, as it works on an individual basis, taking care of each position, instead of relying on a retracement to avg. It gives a different type of feel and look. Here is an example with SI (last few hours). I will be probably including this new rule in the next strategy versions. (I am still checking variations ...) <img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=3172057" width="1200" /> Tom