|Tools, Techniques, and Training for Nirvana Platforms|
|OmniTrader, VisualTrader, OmniFunds & OmniVest|
|Proverbs 19:21 (ESV) ... Many are the plans in the mind of a man, but it is the purpose of the Lord that will stand.|
| MTV Scripts (for OmniVest, etc)|
USA: GA, Lawrenceville
|The info in this initial post is generally correct, but since the "offering" has been significantly simplified and made more cost-effective since this post was made, I recommend that you use this post as "backstory reference", and focus mainly on the posts that follow. |
There was a huge amount of information posted about Scripts prior to the B1 release, but the "streamlining & simplification" had such a large impact on Scripts, that I have archived the old posts and will be starting afresh.
Much of the story is the same, regarding their many uses. You do NOT need to be a "programmer" to use MTV Scripts ... you can just copy and paste, or maybe make very small and simple modifications to them. MTV Scripts are *very* long formulae that exactly duplicate the Trend and Volatility Market State output shown on the screen, *for a particular pattern of inputs*. That is, a "formula snapshot" is available for almost every pattern that the MTV plugins can model ... and that formula (aka Script) is the means by which you can apply MTV filtering and ranking, etc to your OmniVest applications. So ... anyone that wants to "try out" MTV in OmniVest will need to know about, and eventually own, some MTV Scripts. The Scripts also can be used in OT/OV/OF as a "surrogate" for inactive Subscriptions, but that's not their main purpose.
From a conceptual and pricing standpoint, Scripts come in two broad categories: Guru (aka Library) Scripts, and Expert Scripts. There are 5*5*5= 125 Guru-Library Patterns, each of which has Scripts to calculate the Trend and for Volatility States related to those input-Patterns. Any MTV plugin (Edge, Flex, Powr) can generate these 125 patterns (Edge provides a "Master" subset of 15, unless it's upgraded to Plus). The Guru Library Scripts do a good job of covering the entire "spectrum" of the 2.2 million possible MTV patterns, in 125 manageable and understandable steps. So, that's where most people should start.
Expert Scripts use the same formula-engine, but offer deeper Pattern-resolution to echo the ten fine-tuning Expert inputs in MTVpowr (4 Experts for Guru #1, and 3 apiece for Guru #2 & #3). So, if a purely Guru-defined Pattern is seen (via the MTVpowr plugin) to benefit from some Expert-input refinements, the Script-creation engine precisely incorporates those tweaks. Since there are 2.2 million variations, requests for Expert Scripts must be individually processed, making them a bit more expensive than Guru Scripts.
One of the benefits of the B1-Guru-streamlining has been *price reductions* for MTV Guru-Library Scripts (~20% less than Expert Scripts), and the creation of a significant incremental-Script-purchase discounting schedule, which applies to all types of MTV Scripts, offering up to 80% discounts. By far, the "best deal" for Scripts comes with joining the Directors Group, which provides the full Guru-Library of 500 "Scaling" Scripts, plus free upgrades to all MTV products, for life (product must initially be purchased). Click here for more info about the DG.
The discounting schedule is no longer a meaningful consideration, since I've offered a huge discounted package (saves over 90% off the onesy-twosy cost) ... see the next post!
Scripts are simple to order ... just find the MTV Pattern that you like by using one of the MTV plugins, then a popup Menu item shows you the list price and tells you how to order the Script (via email). Normally turn-around is just a few hours.
Scripts can be used in OmniVest as Condition-Filters for Strategies or Dynamic Lists or Portfolio Balancer, OR can be used for ranking of DL's or evaluation functions in Portfolio Wizard. A single "Scaling" Script can be easily tweaked to be used in every one of those applications.
Scripts come in two forms ... Scaling and Boolean. Although the Boolean forms were introduced first, the Scaling forms are more versatile, if you don't mind typing such things as "0 <" or "0 >" in front of the long formula when you use it for T/F Conditions. Therefore I feel that purchasing Scaling Scripts is the "wisest" approach.
With the highly-discounted simplified approach, the spotlight is almost wholly on the versatile Scaling Scripts ... Boolean can still be ordered but are not individually as versatile.
When a Script is ordered, it is delivered as a Word "docx" file (for various reasons, PDF's don't work out as well). The top of the file fully documents the parameters that define the Pattern, as well as many of the "derived" values used within the calculations (ie, some clues as to how MTV is "interpreting" your input-Pattern ... click here for a video that explains how MTV "thinks").
The informational header is followed by concise Instructions for applying the Script in various ways. After that, there is a simple title for the Script type, followed by its long Script formula that can be more than a page long in some cases. If more than one Script is ordered, then a blank line separates them ... Scaling scripts can be ordered as a single T|V or both T&V, and Boolean scripts come in T|V 6-packs or T&V 12-packs.
All Scripts are available, and most often ordered, for a "Current Symbol" application. That is, the script works on whatever Symbol the program (OV/OT,etc) tells it to. But almost all of the Scripts are also available for "Alternate Symbol" application. This is extremely powerful for multi-tier filtering ... checking a symbol's Industry average(s) &/or multiple Indexes. The Alternate Symbol Script formula is quite a bit longer than it's sibling for Current Symbol, since it uses GetClose("SPY") instead of C (and Open, High, Low). On account of that, a few of the Guru-Library Scripts can't be provided in AltSym form ... they would exceed the 32,767-char max imposed by Windows.
USA: GA, Lawrenceville
|GREAT PACKAGE-DEAL for MTV SCRIPTS! |
There's been plenty of time now to let the Scripts issue simmer, and see what level of interest there has been by MTV users to getting Scripts on a "onesy-twosy" basis, using the pricing provided in the OmniScript Ordering popup panel available in all the MTV versions. Aside from an "all-in" interest from those that joined the Directors Group, there have been very few orders for individual Scripts.
Part of the issue may be lack of clear explanations as to how to choose and use MTV Scripts ... good news! ... the posts immediately following this one give detailed instructions about how to do exactly that, using OT and OmniVest.
So ... it's time to simplify the playing field. I've made some decisions about this:
1. Primarily make Scripts available *only* for the 125 permutations of the Gurus, without including impacts of the MTVpowr Experts (2.2 million pattern-permutations = 8.8 million Scaling Scripts, not counting Boolean) ... *unless* someone wants to purchase a *large* number of Scripts at once which use Expert overrides. Reason for this is the time-hassle of creating the customized individual packages for each instance. I'd originally hoped to automate this on a website but my website programmer sort of lost interest.
2. THE GOOD NEWS: I've already set up a very special-deal 500-Scripts-package that provides four Scripts for each of the 125 Guru-permutations that are available in MTVpowr, MTVflex, and (enhanced) MTVedge. For each Guru-pattern (three Gurus each varying from 1-5), the package provides "Scaling" formulae (more versatile than Boolean) for: Trend-score for Current Sym, Volat-score for Current Sym, Trend-score for Alternate-Sym, and Volat-score for Alternate-Sym. "Alternate-Sym" defaults to "SPY" but it's easy to do a Find/Replace to put in any symbol you want as the basis. I originally offered that package of 125*4= 500 scripts *only* to people who joined the Directors Group ... which is still open to accepting new members, and offers a lot of other benefits documented here:
... the additional price to join the Directors Group (beyond purchase of MTVpowr + lifetime subscription) was $1920
I've decided to formally offer that 500-script library to folks who *don't* want to join the Directors Group, even if they *don't* want to buy the MTV indicator DLL at all ... for a one-time package-price of $1600 for the 500 Scripts, which is about $6.50 per pair of Scripts ... about *one-twelfth* of the normal $75-per-pair price. The normal pricing is still detailed in the "ordering popup" in MTV ... there used to be a big discount table for people who bought a lot of them, and so on ... but all that has been shelved due to its complexity and lack of sustained interest by users in buying them "piecemeal".
If you are interested in buying one or more of the four 125-Script subsets contained in the Guru Library for $600 apiece, Click here for explanation of options, and which you "need" for OmniVest and for OT/VT platforms.
If someone buys the script-library, then they'd presumably want to at least VIEW on an OT chart what each script behaves like, to help choose which to use in OVest. The script can be plotted in OT using the Quick-Indicators tool ... but the graph would be just a simple line with none of the color variations, etc that make MTV states easy to understand. So, I recommend to folks who want to "play with" different scripts *visually* in OT/VT, to get MTVflex (exposes the three Gurus but no Experts) *without* the extra-cost Subscription. Once one of the 125 guru-patterns has been chosen to test further, that Script can be used in a Strategy Filter Block as a conditional rule. OR, a Subscription for Flex can be activated, and a call to the indicator will work in the FB (easier and offers StratWiz testing).
I know that's a lot to swallow, but MTV is so versatile that it naturally has nuances. It has various levels (Edge, Flex, Powr), plus option to activate the Return-value via Subscription, to provide a range of prices to fit different traders' needs and pocketbooks.
HAVING SAID ALL THAT ... if someone prefers to buy MTV (any version) w/ or w/o a subscription to try it out, and if they are able to "home-in" on a few guru-patterns that they are confident will fit their needs ... then I'll be happy to sell those Guru-library pattern Scaling-scripts in pairs (Trnd+Volat) for $75, or $40 individually, per the MTV popup ordering form. (Other options in the popup, for Boolean &/or Expert-input tuned patterns, are negotiably-available in quantity).
SUMMARY: If you're serious about applying MTV to OmniVest for a variety of strategies, then I strongly recommend the $1600 500-script package-deal. Once and done, for a very low price per Script. Keep in mind that a single guru-pattern will *not* be ideal for all different kinds of strategies. That's why MTV exists with parameters (Guru and Expert) ... so it can adapt pretty much universally to any strategy.
Email me if you're interested! ... And be sure to review the "how-to" info in the following posts!
USA: GA, Lawrenceville
|After learning about the MTV-Scripts package-deal above, one of our Forum members contacted me with three excellent "application" questions. He is a very skilled OmniVest user, actively engaged with the newest development, and wants to use MTV Scripts with Elite Trader to create uploadable custom OT Strategies to be used as Elite OV Systems. |
1. How do I find the right script for a particular purpose? Say I want to upload NSP-41 and improve it's CAR/MDD ratio.? Surely It's possible to identify a subset of the 500 that are good candidates for this purpose?
2. Assuming that the script is intended for use as a filter, is it better used as a Filter Block in an OmniVest Strategy or as a Filter within a Dynamic List?
3. Is the script package useful for the purpose of staying in RTM trades longer than they would normally stay in? Obviously a trailing stop might work, but there are disadvantages, like turning a winning trade into a loser?
The answers that I've provided are deliberately fairly generic, applying also to OT/VT implementation, but have OmniVest as the primary "target" ... the answers to #1 and #2 each have fairly long 2-part answers, so I'm breaking this up into several posts. I hope all of you find this "how to" helpful, and from it, begin to better understand how powerful MTV Scripts are, for use with OmniVest.
USA: GA, Lawrenceville
Absolutely it's possible ... easy in fact ... MTV has been "designed" to facilitate that. There are three ways to do this, but I'll address only two of them since the third involves a *lot* of cut and paste hassling-around.
a. Open a blank black-background chart template (I provide a nice one for free :~), and set it to view a year of daily bars. If you have a Strategy in mind, then activate it *without* MTV filters, and keep the summary vote-line visible below the chart with MTV, so you can see the "raw" entries and exits. Now, add five instances of MTVflex to the chart (each in its own little Indicator Pane). Set the Guru-parameters for those five to five related variants, so that the indicators are generally similar but show the nuances of the Guru that you are varying.
b. Since the Gurus: StateLnthShortToLong, ResponseQuik2Even, VoteMethodAndThresh are generally-speaking in the order of Most to Least impact, then the most efficient approach is to make your first five variants on the chart deal with the Most impactful Guru first. So, set Responses and VoteMethod each to the middle value of 3 for all the indicator-instances, and make each individual instance use a different StateLnth input of 1,2,3,4,5. StateLnth in simple terms varies the MTV-logic from patterns appropriate for short-term trades to long-term trades (much, much more info about this is in MTV Help)
c. Visually examine how the Green MTV bars line up with (successful) Long trades, and the Red with Short Trades ... trades in either direction that Enter outside of their complementary colors would be blocked by MTV. Digest how the different variants 1-5 "fit" with your "raw Strategy" voteline ... make some notes to yourself maybe ... and REPEAT this visual filtering for a variety of Symbols that are representative of those which you will be applying the filter to for live trading. Be sure to check out symbols with varying volatility, and those that are in clear uptrends, downtrends, and swing-states during that 12-month period. Home in on one of the five choices ... or maybe 2-3 of them if it's unclear as yet.
d. Now, save that chart template with an appropriate name (VaryStateLnth maybe), and save it again for the next round with a new name (VaryResponseSL#, where # is the "best" setting you came up with in step c above. Once that new-name chart is in place, edit the settings of the five MTV instances so that the first Guru is set to "#" (from step c) for all of them, the third Guru remains set at 3 from before, and the second Response Guru has values from 1-5.
e. Repeat the step c process for this new, more refined set of choices. It's more likely now that you might find more than just one Response-# example that seems "pretty good" for your Strategy ... but be sure to check a goodly number of symbols as mentioned above before you drive on. Make notes as you go! If you pick 2-3 different Response#'s, then note which seem most and least attractive.
f. Probably you can guess what happens now ... first, SaveAs the template with a new name for the next round (VaryMethState#Resp#, using the #'s discovered in steps c and e above). Modify all five MTV indic's to use the same State# and Resp#, and vary the VoteMethod Guru input from 1-5. Note: the VoteMethod shifts the MTV pattern around in less "predictable" ways from one input-# pattern to the next ... don't be surprised that the gradation is not as smooth as what you saw in steps c and e.
g. Repeat step c/f process again ... the choices will have become more difficult, so note-taking is essential at this point. It's likely that you can rule OUT two or maybe three of the five variants by the end of your multi-symbol comparisons.
h. Now, if you picked more than one "viable" Response setting in step e, repeat steps f+g for each different Response#, and take notes. (Hopefully you'll only need to do one variant for this step).
i. Finally, if you picked more than one viable StateLnth setting in step c, repeat steps d-g for each different selected StateLnth#, and take notes. (Hopefully you won't need to do this step)
j. VOILA! you've now intelligently and carefully discovered a "handful" of the 125 different Guru-patterns that appear to work best for that Strategy. Now, dig out the script for each combo (identified by those three Guru categories, as #1-#5 for each), and TEST it ... either with a "comparable" System in OmniVest, or by adding it to a new Filter Block for the OT/VT Strategy that you've been looking at on the voteline. The OT "analysis" run, &/or PortSim, will give you lots of statistical metrics to evaluate it, for a variety of Symbols &/or Time-Windows.
... on to the second part of the answer to question #1, in the next post ...
USA: GA, Lawrenceville
|Continuing with the answers ... this describes the second way of dealing with Q#1 - much faster and easier: |
Strategy Wizard Method:
This method requires that you have an active Subscription for your MTVflex (or MTVpowr or enhanced MTVedge). It costs a bit more but you'll see from the notes below that it saves a LOT of time.
a. Set up the "representative" Strategy in OT that "most closely matches" the OmniVest System that you plan to use. Note that most OT canned Strategies differ to some degree from their OmniVest counterparts. Barry can provide an "approximate equivalence" table if you need it. I have one, but it's not up to date. By "set up", I mean to activate that Strategy in OT, and modify its Trade Plan to use MOO orders for entry and exit, and possibly change the System block to only create Long or Short signals rather than Both ... again, to match what you want to do in OmniVest (many OV Systems do not offer all three options).
b. Add a FilterBlock to the Strategy, where you will apply the MTV filters. This will use formulae with *replaceable parameters* that you define on the upper-right side (more later on that). Add and Enable a pair of S+L filters in the top-left panel ... call them MTV Guru Variants or something like that. Then ...
In the Long filter pane, paste in this formula: MTVflex( 0,0, 333,SLen,Resp,Meth, 777,0,0,-2) > 1
In the Short filter pane, paste in this formula: MTVflex( 0,0, 333,SLen,Resp,Meth, 777,0,0,-2) < -1
... these formulae let Long signals pass if the MTV-pattern has green bars, and lets Short signals pass if the MTV-pattern has red bars.
c. In the top right of the Filter Block, Add three variables to the Parameters pane, matching the three names you used in the L+S filter-formulae in step b, and set the Min,Max,Incr,Def values as ...
SLen 1,5,1,3 ... Resp 1,5,1,3 ... Meth 1,5,1,3 ... *none* should have the "Opti" box checked.
Set Consensus to All, leave the Symbol & Timeframe alone (for now), and in Performace, set Metric to Nbars and Number=5, with Use Reversing Signals checked. Optimize metric does not matter (really, the other stuff in this section really doesn't either :~)
d. On the top left, check Enable, and do NOT check Optimize. Click OK, then click Save As and give this variant of the Strategy a new name (maybe with MTV added to it).
e. Test the new Strategy by doing an Analysis run, for whatever Symbol List and time Window you like (picking those is another really long discussion that I'll skip for now). If there are odd results, go back and check your work ... maybe run both the new and canned versions of the Strat together to see how they differ. The values used for the three parameters in the Filter Block will be the defaults, all = 3.
f. Now you're set to go! Open up Strategy Wizard (BUY it if you haven't got it yet!) ... and select the MTV-filtered Strategy to test. Set it for Daily timeframe (only), and Exhaustive tests, using the Performance Report for simplicity (if you prefer, use Portfolio Simulation with Fixed $ Amount, but I recommend Performance mode). Name the Output file (MTVwithStratName.txt for example), and UNcheck "Append to file". Click Save As just to make sure it's initialized.
g. Using the Add button, create three iterative rules for this SW run (review SW Help Manual if you don't know how to do this ... it's too long to describe here) ... the final result should look something like this:
Filter\MTV Guru Variants\SLen\Default: 1-5(1)
Filter\MTV Guru Variants\Resp\Default: 1-5(1)
Filter\MTV Guru Variants\Meth\Default: 1-5(1)
... this should be pretty self-explanatory since you went thru the FB creation process above ... I hope :~)
h. SaveAs again, just for grins ... then click RUN, and take some time off. It will be running your MTV'd Strat on all the symbols in the FL, for all the bars defined in Test Settings, for 125 different MTV Guru variations. This could take several hours, depending on your PC speed and #sym's and #bars. Relax! Have Fun!
i. When the SW run is complete, you should look at the Output panel, focusing on a few of the metrics that are most important to you. I recommend that you pay most attention to Number of Trades, Hit Rate, and Profit Per Trade ... most of the other stuff is either less valuable, or could be misleading for our purposes here. Use the Sort Column feature to figure out which combinations of SLen, Resp & Guru provide the best response. Write those rows down! You might also use the 2D or 3D charts to help find selections that are the least "spikey", and so on. Again, a full tutorial on using SW is out of scope here. The manual helps with this :~)
j. Once you have a collection of maybe 3-7 Guru-patterns that seem best statistically, I strongly advise that you VIEW them on a chart, using the Five-MTV-Instances-On-One-Chart setup that the Visual Method describes, but with the five variants being the "best five" you found from the SW run. Check out a lot of symbols, in varying time windows. This will help you gain confidence and better understanding of what you've homed-in on, and might help you narrow your choices.
k. That's it ... copy the matching Guru-pattern MTV Scripts into OmiVest for use with the System that best matches that Strategy. Note that this can be done in two ways, which I'll discuss in the next section ...
USA: GA, Lawrenceville
|Continuing with the answers ... this describes alternatives of how to apply it in OmniVest: |
Dynamic List Approach:
These comments could apply to both OmniVest and OmniTrader, since DL's are possible in both. I'll use OV as the basis here. Keep in mind that OmniVest has extra fees if you build a big collection of DL's ... but afaik, there is no additional cost if the DL is part of an Elite Trader upload. So, each user can evaluate that tradeoff for themselves. Briefly: I believe that using the MTV filter selected using either the Visual or StratWiz methods discussed earlier as rules for a Dynamic List is usually a *better* approach than putting the MTV Script in a Filter Block, especially if the "starting population" is large, and the other filters (volume, price, etc) don't cull it down much. This has a particularly significant impact on OT efficiency ... using a DL will likely result in much shorter Analysis runs than using an MTV filter block with a *big* list of symbols. Either method will produce the same net results in OT. BUT ... there is a BIG difference in OmniVest.
a. Use the OmniVest Help to learn about creating Dynamic Lists (this is not a tutorial for that!). What you need to do is to paste the OScript formula into the DL filter-rule pane, and add a simple prefix to it that defines what State you want the DL symbols to be filtered for. If you want to filter for Longs, then prefix the MTV Script with:
1 < (... MTV script here ... )
... and if you want to filter for Shorts, prefix the MTV Script with:
-1 > (... MTV script here ... )
... Note that it's wise to add the parentheses before and after the long script, but probably not necessary
... By this time you should have realized that if you want to trade both Long and Short, you will need to set up TWO Dynamic Lists ... and create two separate OmniVest Strategies each for the appropriate direction.
b. If you want, you can use a different nuance of the same MTV filter to help with DL ranking, though personally I think this is overkill at this stage. To do that, keep in mind that the MTV Script will return higher values for stronger, more mature trends ... 4 means a very strong, well established trend while 2 means a the beginning of a trend, or a fairly "weak" trend. The filter is set up to capture all of those ... if you want, you can just paste in the MTV script formula, without adding any suffixes, to the Ranking field and set it for Decreasing if you want to bias choices for mature/strong trends, or for Increasing if you want "new" trends, especially ones with a "gradual' start that *might* offer you a more attractive (earlier) entry. Up to you - these are generalizations.
c. For those who want to "get fancy", you might choose to combine two MTV Scripts into a single DL filter, using "(1 < ... one Script ...) OR (1 < ... another Script...)" - in this case, adding the parentheses (as well as the threshold# and comparison >< sign) is essential. WHY would you want to do this? Well, if your Visual or StratWiz searching discovered TWO MTV Guru-patterns that seemed equally useful, then by including both of them here with an OR connector, you'll be presented with Symbols that satisfy EITHER of them. If you do this, then I *would* advise you to include both MTV Scripts in the Ranking ... ADD them together so that you get the best of both worlds: (...one Script...) + (...another Script...)
d. Taking the "fancy" down a different path, you might want to also incorporate an MTV *Volatility* criteria in the DL rules. The prior instructions for finding an appropriate pattern for Trend are essentially the same for Volatility, but SHOULD (ideally) be done SEPARATELY. That is, the Guru-pattern that works best for a given Strategy regarding the MTV-Trend will most likely be somewhat different than the Guru-pattern that works best with that Strategy regarding MTV-Volatility. There are a few things that you must set up differently ... the main one is that the uses of the MTVflex indicator in the test filter block should have a "-3" as the last input instead of "-2". Also, it's hard to generalize what volatility-state is "best" for a given strategy ... sometimes Calm might be best for Entry, but other Strat's might work better with Wild. Since those states (blue histogram bars) tend to vary more frequently than Trend states, the search for "best patterns" is best done using the StratWiz method. When doing this, CHANGE the filter block formula so the "> 1" and "< -1" suffixes are " = Vscore" and " <= -Vscore" instead. Add that new Vscore parameter to the FB list, varying it from 0 to 4 by 1's, with default =2. And, add that extra iterative rule to the StratWiz run, "...\Vscore\Default=0,4(1)". I realize this is a condensed description, but if you don't really follow it, then you probably should not be trying to do this nuance, at least not at first :~). Once you determine the "ideal" Volat MTV Script, you can combine it with the ideal Trend MTV Script using: ( 1 < ...MTV Trend Script ...) AND ( Vscor# = ...MTV Volat Script ...)
NOTE: the Vscor# comes from your review of the SW runs ... and what you learn there might suggest that you should use >= or <= instead of =. Again, if you don't follow this ... just skip it for now.
e. And of course for you wild-eyed experimenters, the "fancy" can be a variety of combinations of AND / OR clauses, mixing Trend and Volat rules. There is one annoying gotcha though ... if your MTV scripts collectively have more than about 32,000 characters total, then OV probably won't be able to evaluate them, due to Windows internal limitations on string length. And... btw... early testing of MTV Scripts showed that DESPITE their ridiculous-seeming length ... THEY RUN EFFICIENTLY. So, don't worry about bogging down the OV servers or getting a cautionary call from N staff.
... next post will discuss how to use MTV Scripts for OmniVest Strategy Filters ...
USA: GA, Lawrenceville
|Continuing with the answers ... this describes the other OmniVest-use alternative: |
FIRST, read the prior description re DL implementation, and it's pros & cons. As before, the comments here could apply to both OmniVest and OmniTrader, since Filter Blocks are available in both. I'll use OV as the basis here. Briefly: Either method will produce the same net results in OT. BUT ... there is a BIG difference in OmniVest ... since when using a Filter with a Strategy, you MUST assign a "reference Symbol" with it. That is, when the Filter for the Strategy is evaluated, whether or not it Enters the trade for a given Symbol on the list is evaluated with regard to the MTV state of the REFERENCE symbol (ie the MTV state does not vary for different symbols in the test-list). So, you might use "SPY" as the reference symbol ... and if the general market (as measured by SPY) is MTV-trending up, then all Long Entries for all Symbols in the list will "pass" the Strategy filter.
IMPORTANT note: if you believe that the SPY aka "general market" Trend (&/or Volatility) state is relevant to the viability of Entry on a given bar for the symbols in your list, AND you believe that the MTV-Trend/Volat state for THAT SYMBOL is ALSO important, then you can use BOTH the Dynamic List method AND the Strategy Filter method in combination. But note that the Guru-pattern you choose for the Symbol-by-Symbol DL filtering most likely will be somewhat different than the Guru-pattern you choose for the Reference-Symbol Strategy-filtering. That is what the "steps" below address ...
a. First ... the way that you apply the MTV Script in the Strategy Filter input-field is *just the same* as the examples and "fancy" variations described for the Dynamic List implementation above, so please read that section to find out how the syntax works. Remember ... Long and Short might have different Guru patterns ... and Volatility might be different too ... and the Reference Symbol might call for a different set than Symbol-by-Symbol. SO ... you need to redo the best-Guru-pattern selection procedure discussed earlier (Visual or StratWiz), locking in the MTV evaluation for your desired Reference Symbol, but evaluating the results across multiple "tradable" Symbols in your OT focus-list, etc.
b. To modify the Visual Search method, simply specify a "Base Symbol" in the bottom left of the Indicator Parameters pane that matches the Reference Symbol you plan to use in OmniVest. This will make the MTV indicators change a lot from before ... but as you click through different FL symbols, the MTV indicators will NOT change, since they are locked in to SPY (or whatever you chose). The selection procedure described earlier is the same. When you get to the final stage where you are putting the MTV script into the OT strategy's Filter Block, simply *specify the same Reference Symbol* in the "Symbol & Timeframe" section. To do this, click the elipses (...) button on the far right to bring up a panel that allows you to enter the Reference Symbol (click the radio button to select it instead of Current). Don't change the timeframe!
c. To modify the StratWiz Search method, you simply apply the same Reference-Symbol specification in the Filter Block as described at the end of step b above ... the rest of the procedure is unchanged.
d. The "elephant in the room" for both of these methods is "what Reference Symbol should I use". SPY, DJI, maybe an MG-Group or Industry Average? Clearly, you can manually iterate through different explicit choices using Visual or StratWiz methods described in b+c above. However, if you have an active Subscription with MTVflex (or MTVpowr), then it's possible to use the second "Alternate Symbol" MTV input in conjunction with StratWiz explicit-list iteration, to check a large number of different Reference Symbols as an additional iterative loop around the SW Runs described earlier. Use MTV Help (for that second input) to learn how to pinpoint what integer-input values relate to any of a very wide range (nearly 500) of different potential Reference symbols. This again is a big topic ... I'm just giving you methodology here.
... next email will deal with the final question ...
USA: GA, Lawrenceville
|Continuing with the answers ... this describes the other OmniVest-use alternative: |
Using MTV to Improve Exits:
This is an easy answer for standard OmniVest use ... the Exit rules are locked in by the canned System Trade Plans, and users are not allowed to modify that, for RTM or any other canned OV Systems. Period.
However, for Elite Trader use, you can create a custom OT Strategy with whatever TradePlan you want, and upload it to OmniVest. With the exception of the MOO order-type limitation, this opens up a huge variety of possibilities for Trade Management (which in my opinion is very important to trading success, especially for non-optimized/curve-fit Strategies).
It is possible to set up MTV in its current form to be used as an Exit (or as an Entry, for that matter). A really good implementation involves either a "shell" OLang Stop or System routine, OR an upgrade-revision to MTV (which is in the works but will be a while coming). For now, you can do rudimentary Exit control in a Trade Plan, using MTV, by setting up an OmniScript-Condition tied to the Exit MOO Order. That condition would simply query the MTV-Trend-state of the symbol in trade, and exit the trade when the Trend-state becomes "contrary" to healthy continuation of the trade. For some strategies, with some MTV-Guru patterns, this might be (for a Long Trade) when the Trend State stops being green (ie it's yellow or red). In MTV Script-ese, for a Trend Script, this would be: 1 >= (...MTV script ...)
Similarly, to exit a Short trade when MTV turns flat or long (yellow or green), the MTV Script-ese would be:
-1 <= (... MTV script...)
(note that the MTV values of -1, 0, +1 mean the same thing for MTV Scripts, which do *not* model the "momentum" effect discussed in MTV Help)
Please note that I have *NOT* tested this in OT ... there are certain major distinctions in how OT processes Stops differently than Indicators and Systems ... but I *strongly suspect* that the MTV Scripts will work fine in this form. However, if you choose to use an MTV indicator-call (with active subscription), there is a 50/50 chance that some issues might occur. HOWEVER, since (for now) the MTV Indicator "code" is not allowed by Nirvana to be uploaded via Elite Trader, that's a moot point (uploads cannot include DLL's for "arguable" reasons ... sigh).
Of course, if you use MTV for this purpose (creating an exit condition), then you will need to utilize a Visual search OR StratWiz method to determine the best Guru-pattern settings to use for the Exit. Quite likely different from those you used to improve Entries via the Filter Block (or Dynamic List). Make sure that you first sort out those other MTV choices, and apply them to your custom Strategy, before doing the search for the best Exit pattern.
SUMMARY: Using ELITE TRADER, you CAN use MTV Scripts to define an Exit condition for your Trade Plan, in addition to other conditions if you'd like.
USA: GA, Lawrenceville
|And ... drum roll ... that same "expert" OmniVest user asked another question ... |
Using MTV Scripts in OmniVest ProTools
It's been quite a while since I looked at PW or PB, but I do recall considering this question when I first created the scripts and the answer is YES ... MTV Scripts can be used in those tools.
As I remember, there are places where the user can input a Boolean-T/F OmniScript formula, and other places where the input-OmniScript must resolve as a scalar (a number).
All you need to remember is that if you want the Script to produce a Boolean-T/F result, then you need to prefix it with something like "1 < (... MTV Script to identify Long trend...)" or "-1 > (... MTV Script to identify Short trend...)". The instructions I provided earlier in this thread discuss how either of these can be "fancified".
If the input field needs a scalar (numeric) value, just remember that all MTV Scaling Scripts (such as the ones in the Full Guru Library package) return integer values between -4 and +4 (ie nine possibilities: negative=Short trend or Calm volat; positive=Long trend or Wild volat). The integer "steps" should be very useful for ProTools analysis, since they are naturally in convenient "bins".
BOTTOM LINE re MTV Scripts: they can be used anywhere in OT/VT or OmniVest that an OmniScript formula is used.
USA: GA, Lawrenceville
It's understandable why folks ask the questions ... they want a shortcut-solution "panacea" without doing any gruntwork to learn the best answer.
The reason that I'm so stubborn about not giving out "shortcut solutions" as to what MTV settings to use with a given (class of) Strategy / Instrument / Timeframe is that, when I tried to guess in the past then subsequently tested the guesses (against the full Guru-pattern spectrum), I discovered that my guesses usually were not optimal and sometimes not even in the ballpark.
I will qualify that a bit ... if I, as the world's expert on MTV, want to apply MTV to a Strategy which is totally comprised of my own custom-OLang routines which I, as the world's expert on those Entry/Exit routines, fully understand, then the likelihood of my guessing a decent "range" of MTV patterns to try out (or *not* to try out) is not so bad. But, sad to say, most people aren't in-depth experts either on MTV or on Nirvana routines. (The good news is that I do provide VERY extensive doc's re how MTV "thinks" ... so at least my clients have a reasonable-path to becoming MTV experts :~)
But 99% of the time, when folks ask me that kind of question, it's in relation to using MTV with a Strategy that Nirvana created all or most of the components for (such as RTM entry logic). I am *not even close* to being an expert on the logic used by those canned Nirvana components ... and in the case of RTM, my own limited testing has shown at least some of the many RTM entries to be "oddly placed" in light of the "training video" descriptions of how they are supposed to be determined ... which makes me even more skittish about commenting on them.
Regarding the second question (what portion of the 500 might be in the ballpark) ... since there are quite a few Nirvana RTM strats, and there are a lot of instruments and timeframes they can be used with ... then a wild guess would be that probably somewhere between half and a third of the 125 Guru patterns (or 500 Scripts) might be potentially useful. NOTE: I'd probably guess the same thing about any class of strategies, not just RTM's ... and the 1/2-1/3 collection that would be potentially useful for one class would *almost certainly* differ from the 1/2-1/3 for another class ... though it's also likely there would be some overlap.
I realize that answer isn't what folks want to hear, but it's the truth. Fortunately, there is a straightforward, reliable path to knowing the answer, which does not require you to own the Guru Library to find out ... if you own any of the MTV routines, with active Subscription ... you can find out quickly. (remember ... you can *rent* any of those routines with everything activated for just $50/month):
USEFUL SOLUTION METHOD:
I've provided detailed step by step instructions about how to use MTV in a Strat, with StratWiz, to find those "likely candidates" very quickly. Click here and carefully review the thread (and please note that it's in the TT forum Section and Room which one would naturally go to for that info :~)
EVEN IF you for some reason "don't like the idea" of doing it this way, I strongly suggest you just give it a try anyways. By going through the process, you'll better appreciate (from seeing the SW results):
A. How flexible MTV really is ... helping out virtually any Strategy, and
B. Why I'm so adamant about answering the "which Guru pattern should I use" in such an annoyingly-obtuse way ;~)
I hope this helps!
USA: GA, Lawrenceville
Updated clarification about pricing, content, availability, use, and flexibility of "subsets" of the 500 Scripts in the Guru Library package:
The 500 scripts are in four categories of 125 each ... which differ substantially ... the "125" is from 5 LengthGuru * 5 ResponseGuru * 5 VoteMethGuru combinations:
1. 125 scripts for Current symbol, evaluating MTV Trend
2. 125 scripts for Current symbol, evaluating MTV Volatility
3. 125 scripts for Alternate (proxy) symbol, evaluating MTV Trend
4. 125 scripts for Alternate (proxy) symbol, evaluating MTV Trend
A. You'll *normally use* 1 & 2 for use with Strategy Lab, Portfolio Builder and Dynamic Lists (though you could combine them with 3 & 4 if you want)
B. You'll be able to use *any* of the four for Portfolio Wizard or Portfolio Balancer Evaluation Functions, in OV/OT-like combinations.
All four categories have significant value, used individually or in combination. To eval each symbol on its own, use 1 & 2. To evaluate one or more an Index or Sector, Industry or Group *in combination with* the eval of a Symbol, then use 3 & 4.
Alternate Symbol Scripts:
a. These have many, many component elements of the form 'WMA(GetClose("SPY"),5)' where the Current Symbol Scripts would use the simpler 'WMA(C,5)'. So, the Alternate Symbol Scripts are much longer (more characters) and in a few rare cases are not avail since they hit the Microsoft max limit.
b. It's expected that the trader will use Find/Replace (in any text editor or word processing package) to change "SPY" to whatever Index or MG### they want to target. In OT/VT, the generic '$Sector', '$Group', '$SubGroup' can be used to automatically reference the MG### Average that the Current Symbol is a member of.
c. It's also possible to use 'Symbol' instead of "SPY" in the Alternate Symbol Scripts to make them functionally identical to the Current Symbol Scripts (except the few that are too long).
d. Syntax: in notes b & c above, when using any of the choices surrounded by 'single' quotes, you must *remove* those quotes. For note a, when substituting different actual symbol-names, the "double" quotes are required in the formula.
If you want to purchase *only* the scripts in some but not all of those four categories, they cost $600 apiece.
|Owner of site: Jim Dean -- Forum content is confidential, and may not be distributed without written permission.|