Coming soon: free P123 membership + Designer terms chages

Announcements
101

Andreas, Stitts, Konstantin,

there are indeed not many great R2Gs available right now. When I apply my screening filter ($500k bottom liquidity and 50% annualized excess combined), there are only 45 models out there today that qualify. Of those, only 25 have actually any excess return since launch!

I look forward to the new R2G layout and in particular the 3yr rolling backtest which will expose the weakness of quite a few models. I hope that Sharpe will then also be calculated based on the rolling backtest and thus eliminate the market timing cosmetics so frequently applied on p123.

102

Steve, [quote]
… “I think your situation and environment was much different …” The situation and environment is exactly the same …
[/quote] Now I see, the situation ir really the same and no wonder the results and outcomes as well.

[quote]
… "But I still can’t get how do designers expect to compete with non diversified models priced >2 times than the market. " I think we need to see what 123 comes up with. Beyond that, our ability to bring in new P123 members (which means more R2G subscribers), depends on streamlining the signup process. … Steve
[/quote] P123, apparently, already came up … a lot of times. But as I can see from the thread linked above, there no single step from designers made. Just wondering what kind of investment market customer are designers targeting with “streamlining the signup process”?

[quote]
… in-sample is not “performance” in any way shape or form. It bears no comparison to OOS. Please eliminate it from your arguments. Then we can focus on focus. …
[/quote] Steve, take a note, I am expressing my personal opinion, based on my experience. This analysis work for me well and I don’t care how to call things or is it academically proved. Nevertheless I am open to learn something new and ready to here from you arguments, pros and cons. I propose to do this in another thread as this is nor the main point nor directly related. I will stick with key point that I can’t get how do designers expect to compete with non diversified models priced >2 times than the market?

sevensisters, no wonder you came to the same outcomes on different segment. At the moment I would suggest to ordinary investor (not a trader) stay with free models. At least they will get what expect for the competitive price.

103

“…I am expressing my personal opinion, based on my experience. This analysis work for me well…”

Konstantin - The problem with in-sample testing is that the designer can run and re-run backtests until results are extremely optimal. But once a model is launched, it is cast in stone, with the exception that one could keep cancelling their systems and re-launching them. This does happen of course, but is presumably much less frequent than the optimization process. So the difference is fundamental, post-launch is a record of the designer’s performance, pre-launch is nothing of the sort. Is post-launch performance a predictor of future performance? Absolutely not. And neither is in-sample testing. But in any case, it is not correct to say that a model is “underperforming post-launch”. OK?

“P123, apparently, already came up …” The only thing they have come up with to date is assign different categories. While I originally thought it was an interesting concept, I eventually started by-passing the categories and started looking at all models. I think most other people do as well.

"Just wondering what kind of investment market customer are designers targeting with “streamlining the signup process”?

DO you know what a Long Sales Letter is? You have undoubtedly seen one but may not be familiar with the term. Open up any spam-like EMAIL that comes to your in-box that has a link to a sales pitch. Every one of these has the same format, its called a Long Sales Letter. You will notice that there are no ways out even if you try to close the window (there will be a pop-up window). There are no miscellaneous links, no side bars with non-related material, you will keep reading and reading until your time is fully invested, you finally get to the price, usually with a special deal, discounted price, etc. Now, I don’t know about you, but I find these sales pitches very annoying. But statistics show that they work better than any other format. So people use it. Now I’m not saying we should adopt this approach but there are some techniques that we can take advantage of.

For example, I have a website and I want to market my R2G systems from it. Do you think it is logical to sell a reader on the model and price, only to send them off to a P123 registration page, where they find out they are going to pay more money? Then once they register for a free trial they are free to look around at everything and anything, forgetting about the original reason they came to P123. By the time they finish their free trial, they see how complex P123 and wonder if it is worth the price, again the original reason is long in the past. There is no incentive for me, as a marketer, to participate in bringing new people to P123 or R2G for that matter, it is just a waste of time.

So what needs to happen once the reader is “sold” on a model at my site, is to put the blinders on him/her so he doesn’t stray from the path. Take him directly to the model and have them sign up then and there. Don’t let him look around, don’t give options, or the sale will be lost.

This has nothing to do with what the target investor is, whether it makes sense to offer him a free membership, it is a sales technique.

Steve

104

Steve, once again I see there was misunderstanding, probably because of my poor English, sorry for that.

For sure, there are little future predictability in OOS data, but we should deal with it. For sure, there is much less future predictability, if not zero, in in-sample data. So for not over-optimized models with consistent in-sample/OOS there is relatively a little bit more in-sample future predictability than for curve-fitted models failing OOS. And still pros and newbies look at past results, no matter in-sample or OOS.

I agree past results do not predicts much of the future performance. My point was opposite, that inconsistency in in-sample vs. OOS data most probably predicts poor model future performance and such a models metrics should be taken critically if not totally ignored. The point of my statement was to show how many models just falling out of investor interest, no matter what was in-sample or OOS performance. Generally this is statement what to await from the model: at lest half of return and twice of DD. While this can be true for high alpha models, for more realistic alpha models I’ve used 25% OOS/in-sample deviation limit. And we still on the bull market/QE/low rates.

Having in mind what designers asking P123 for I think this is mostly designers who shaping P123 development and results are OK. Any wish list is full of internal functionality issues and rarely marketing proposals. I feel there is not much understanding in designers community what is the cost of marketing efforts to results in low quantity and quality supply environment.

Price and quality are the kings of supply if seeking for steady/quality business growth. For sure you can always shape the market for your product, but if product is poor, this will cost you too much. Not in marketing campaign, but in customers leaving you constantly and instantly. So your marketing efforts becomes your main value adding activity and you don’t have resources for and don’t need to develop product. This is OK for current market, nothing personal, just business. I see this is not the way P123 is willing to go. I do support this. Is it OK for designers? All the designers are seeking for realistic steady wealth grow for their own and, strangely, eager to offer just opposite to the market.

105

“inconsistency in in-sample vs. OOS data most probably predicts poor model future performance and such a models metrics should be taken critically if not totally ignored”

Every designer backtests and optimize their systems differently. In some cases, designers don’t expect the same level of performance between in-sample and OOS. I don’t believe your argument is valid and we are better off not seeing back-test data. Seeing backtest results leads to dangerous manipulation of results as we have seen already.

Marketing - most similar companies spend 90% on marketing. From what I can tell, P123 doesn’t spend anything. I remember seeing “Advanced Get” multi-day training sessions in an Ottawa hotel. I think Stockopedia is doing similar things. I get EMAILs from VectorVest and Zacks regularly, along with pamphlets in the regular mail. The point I’m trying to make is that if P123 wants to put R2G designers to work, or alternatively affiliates, bringing in new members then they need to streamline the signup process. What we have now is not worth the effort, at least not for me. And I believe R2G is saturated. We can only sell to other existing P123 members for so long. We are preaching to the converted.

Steve

106

I disagree about in-sample data not being a good predictor of future performance. If the experiment was done in a scientific manner it can help predict future results. Saying experimentation has no bearing on the future is like saying an apple will not fall to the floor if you drop it even though it has done so the last 100 times you dropped it and not much has changed. If the experiment is not done in a scientific manner that is when it may not have much predictive power. It is incorrect to say it has no bearing just like it is incorrect to say it has. It is all about the specifics.

107

SZ - To make a blanket statement that a model is “under-performing” presumes that one understands the context of the in-sample test. Only the designer knows how the model was developed and the backtest strategy. i.e. the context. For anyone else, they can come to no conclusions.

Steve

108

Steve, SZ - this is never ending discussion, because there is no border line what is optimized and what is not. Investing into world stocks already optimization cutting off bonds. There could be discussion that is reasonable optimization and that is not. Again, no agreement will be achieved as this is very subjective. In mine point of view if model OOS is somewhat inconsistent with in-sample during more or less consistent environment than system most probably is over-optimized. If you think my argument is invalid do ignore it as I ignore systems based on my arguments! Any way you are free to argue that there are a lot of R2Gs for a reasonable price on P123 to offer to wider market.

Steve, you are right, only designer knows the details, investor just presume in-sample consistency with OOS. BTW I never seen disclaimer “designers don’t expect the same level of performance between in-sample and OOS” in the model description. It will be interesting marketing experiment. I am curious, how do you expect to market your customers without in-sample and <3 years OOS presentation?

Most of those companies sells crap, if you are emotionally OK with this, then why not to go to one of those 90% marketing budget company? True, this is a trend of the last decades, but the question is what do YOU want? You want your clients get wealthy and made you too or you want collect the fees? I never shared this and had no willing to but… I have a feeling there are a lot of designers that are unhappy with results they get from their models performance wise (nevertheless there are a lot of opposite statements) and feel they could never sell it as a quality product, so there is temptation to put the value on marketing as others do. I am OK with this, investing never been easy. But, once again, is it reasonable to force P123 ideological changes instead of just changing platform in 5 min.?

I don’t think P123 is willing to put designer at work, this should be mutual interest. This is not P123 core business and you pay membership fees. P123 can help and I think they are. Ask yourself, will you invest into promotion of R2Gs as they are now? Me absolutely not. Some, like Georg, see the reason and do marketing himself with 20% of total of all the P123 designers income. You, Steve, not. If you don’t believe in your product (without marketing efforts), why P123 should?

I am not sure of P123 saturation. I think P123 is very open community and signup process is a peanuts comparing to what efforts investor should make to get into model investing. I agree P123 is hard to understand and this is not about website structure or tools. This is about models offered. The more alpha model is chasing the more it was optimized so the more things to investigate to understand the risk and objectives. So signup streamlining is not the main bottleneck. You can lead the horse to water but you cannot make it drink. And you can’t keep you clients focused on your solutions while there are competing ones, signup streamlining is only a temporary solution in this case.

What I would like to emphasis is current R2Gs not only off competition by the price, but mostly by solution quality offered in part of diversification. Actually only few R2Gs (mostly ETF based) for sale could compete on diversification with what market is offering at the moment. That is OK, portfolios was never meant to be that much diversified. The Books were! But we can’t sell the books now. And functionality is very limited. Nothing new here, this was discussed already. I think this could be one of the priority step for P123 going wider market in line with free membership and separate sales website introduction.

109

SZ, Statisticians, All,

It seems that some kinds of optimization cannot make your ports worse (on average). Consider the example of starting with a ranking system that has 5 factors with randomly selected weights. Then you use Steve’s spreadsheet to optimized the weights and your return increases.

Is the new weighting likely to be better, worse, or the same going forward? I think the argument that the new system may not be much better could be correct but it would be difficult to argue that a randomly selected weighting is the best system and could not be improved upon. The only question would be: what are the chances that my method finds one of the better weightings? It could be as bad as a coin flip but I do not see how the odds could be worse. So I would say that the optimized system is likely (certainly no guarantee) to be the same or improved. It may be reasonable to use a random or an equally weighted system as your best prediction of future returns. I get that.

Edit: I really get that, thinking about it. In fact, it may be reasonable to get the average of many randomly selected weightings to get the best (conservative) estimate of future results.

I think this would not apply to market-timing. There is the potential that the market-timing model has no real cause-and-effect relationship and that it just keeps you out of a market that is generally rising. “First do no harm.”

I’m probably stating the obvious but it seems that optimization gets lumped together in a non-scientific way as SZ suggests.

Would love to get any ideas or corrections to improve my methods.

110

“if model OOS is somewhat inconsistent with in-sample during more or less consistent environment than system most probably is over-optimized.”

Kumar - the point is that in-sample is not performance and it should not be displayed as such. It is a period of optimization. ALL designers optimize whether they believe they are or not. How a model is optimized is up to the designer and no one should be presume how it is done or how representative it is of OOS performance.

I could argue that SZ’s SZS Large Cap S&P Defensive II: Focused Conservative Allocation is not consistent with in-sample data. It is performing well beyond what is suggested by in-sample data. The stone was dropped, it missed the earth and the stone is being sling-slot out into space. I guess that this model should be avoided right? After all it is inconsistent. It is designed around the healthcare sector and intermediate bonds, bubble city. Is that fair?

“how do you expect to market your customers without in-sample and <3 years OOS presentation?”

Quite frankly, I would rather wait three years rather than see customers come and leave because they lost their money.

“Most of those companies sells crap”

I would be hard-pressed to say that advanced Get or Stockopedia are crap.

“like Georg, see the reason and do marketing himself with 20% of total of all the P123 designers income.”

Georg was doing his stuff long before he came to P123 and brought investors with him.

“I think P123 is very open community and signup process is a peanuts comparing to what efforts investor should make to get into model investing.”

Its a pretty safe bet to say that less than 1% of investors that come to P123 stay longer than their free trial. Less than 5% return after signing up for a free trial.

Take care
Steve

111

[quote]
“if model OOS is somewhat inconsistent with in-sample during more or less consistent environment than system most probably is over-optimized.”

Kumar - the point is that in-sample is not performance and it should not be displayed as such. It is a period of optimization. …
[/quote] Steve, in-sample IS performance and absolutely should be displayed! It is performance of period of optimization and have it’s value not only for designer.

[quote]
… How a model is optimized is up to the designer and no one should be presume how it is done or how representative it is of OOS performance. …
[/quote] I do strongly believe your customers will disagree with such a statement.

[quote]
… I could argue that SZ’s SZS Large Cap S&P Defensive II: Focused Conservative Allocation is not consistent with in-sample data. It is performing well beyond what is suggested by in-sample data. …
[/quote]Correct. [quote]
… I guess that this model should be avoided right? After all it is inconsistent. It is designed around the healthcare sector and intermediate bonds, bubble city. …
[/quote]Absolutely correct until investor do clearly understand what was driven out-performance and risk connected. Just like Cherrypicking the Blue Chips - Standard, Buffett and Undervalued Blue Chip Stocks due to not including market timer into in-sample performance and Keating’s Ultra Defensive Portfolio build upon TLT. All out-performing by more than 40%. [quote]
… Is that fair? …
[/quote]You asking investor?

[quote]
… “how do you expect to market your customers without in-sample and <3 years OOS presentation?”

Quite frankly, I would rather wait three years rather than see customers come and leave because they lost their money. …
[/quote] Good answer. Am I, investor, interested in just 3 years OOS? Absolutely NO. Know what will be next? I will ask for 10 years OOS and then I will say it is past and it still not predicting the future, so I probably will no buy in. Are you ready to wait? See Steve, people always find reason if they want/need to. And this is not a joke!

[quote]
… “Most of those companies sells crap”

I would be hard-pressed to say that advanced Get or Stockopedia are crap. …
[/quote]I said “most”, I don’t know what those companies sell but I came out of production, costs and sales, so I strongly believe in what I have said.

[quote]
… “like Georg, see the reason and do marketing himself with 20% of total of all the P123 designers income.”

Georg was doing his stuff long before he came to P123 and brought investors with him. …
[/quote]??? So what? They don’t need signup streamline? So it is possible? Isn’t it you, Steve, who was trading for decades previously? Isn’t it what Marc Gerstein constantly asking designers for? I have come to P123 through the Georg, but I am not his investor and do not think his imarketsignals.com is strong and expensive ($ wise) marketing tool.

[quote]
… “I think P123 is very open community and signup process is a peanuts comparing to what efforts investor should make to get into model investing.”

Its a pretty safe bet to say that less than 1% of investors that come to P123 stay longer than their free trial. Less than 5% return after signing up for a free trial. … Steve
[/quote]Nevertheless this is not a stats, so what? What is different from the market? Steve this is a new normal for modern market based on 90% marketing cost. And still you stating “stay longer than their free trial” and “after signing up for a free trial”. So customers pass the signup and quit only after. You just acknowledge the problem is not a signup!

Note. For all interested in discussion about Optimization Marc Gerstein just opened separate thread Pitfalls of Optimization.

112

Konstantin - I’m pretty tired of this thread so I stop here. My final comments are -
(1) Sorry but in-sample testing is not “performance”
(2) Designers have the ability to explain what testing they have done and why (on their own website or drop-box). However, subscribers should not be given the ability to examine in-sample data “out of context”, presented as if it is performance.
(3) High marketing cost is not a “new normal”, this has always been. The statistics I gave are real, from my experience as a P123 affiliate. In any case, it isn’t what you or I believe, I am simply telling P123 that if they want me to try to bring in new customers then they have to make the effort so that it is possible. This means streamlining the process. I’ve told them how to do it and why. The rest is for them to decide.

Steve

113

All:

I wonder how much designers are making from selling systems vs. using them.  Clearly R2G's has a brought a new element into P123.  Seems like some R2Ger's are just full of themselves and think P123 should cater to their, and only their, whims.

Bill

114

Steve, as you can see from my last over-quotation I am tired as well discussing themes that are minor to my initial arguments.

In context of the market designers and P123 targeting with free subscription most R2Gs are strongly:

  1. over-priced
  2. under-diversified
  3. inconsistent (there meant to be)
  4. under-presented

And most of the points are about designers, not P123. Having this in mind I see no reason for P123 to do any marketing efforts and see a real danger for Investor membership sales drawdown without any sustainable growth in R2Gs sales introducing free membership at the moment.

I would suggest P123 to introduce Books for sales as early as possible. This will involve strong Book functionality upgrade project which will require a long enough testing period before could be introduced to the market for sale.

Steve, as for your list:

  1. Call it whatever you like. [quote]
    Commission Rule 4.41(b)(1)(I) hypothetical or simulated performance results have certain inherent limitations.
    [/quote]2. How many time Marc Gerstein asked the designers to be active and explain their R2Gs.
  2. Marketing and crap is two sides of the same coin. Georg experience negates all your statistics and so your proposals for P123. You negated you own proposals!

Steve, I really appreciate your input into community and P123 development, but this time I see no arguments for streamline you are proposing.

I am not a professional and not very experienced investor. But P123 in some instances looks very unprofessional and mostly due to designers activity. It is no wonder for me R2G sales are not good and I see it mostly up to designers to change this. And I wish P123 could stop for a while, look back and do polish everything what is already done for reputation and credibility sake.

115

Konstantin - the way I see it we are going around in circles.

“Commission Rule 4.41(b)(1)(I) hypothetical or simulated performance results have certain inherent limitations.”

This refers to the OOS “performance”, nothing to do with backtest, which is not “performance”. R2G is not adhering to industry norms, consider mutual funds which have to have one year of pricing data prior to presenting performance results. And that is real listed prices, not backtest. I’m trying to think of another (any other) platform that presents backtest in the fashion we do.

As for pricing, I can go to Metastock and get a black-box indicator for $100-$500 per month, with no description of how it works. The pricing here is nothing except an opinion.

Steve

116

Steve, as soon, as …[quote]
Commission Rule 4.41(b)(1)(I) hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may have under-compensated or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses. There have been no promises, guarantees or warranties suggesting that any trading will result in a profit or will not result in a loss.
[/quote] … we are really going circles and even not the ones I would like to go.

If you would like to get those platforms list just Google for quote stated above.

What is the average real invested capital with Metastok? <$1K. How do I know? Most reference I saw to Metastock was done by teenagers stating they are doing 1000% monthly. That way results justifies the price. In my notation this is crap. Nevertheless there are lot lower prices on the market and I put it on the table earlier in this thread. Still everyone is free to choose the platform he need and like.

The pricing here is a Shakespeare’s to be, or not to be?

117

Konstantin - The rule you are quoting is referring to two different things. First:

“hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may have under-compensated or over-compensated for the impact, if any, of certain market factors, such as lack of liquidity.”

Basically what this says is that R2G OOS performance is not an actual performance record (it is simulated performance). Real world factors such as liquidity (and slippage) are not accounted for.

The second part has nothing to do with simulated performance:

Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profits or losses. There have been no promises, guarantees or warranties suggesting that any trading will result in a profit or will not result in a loss.

This says that in addition to the impact of liquidity/slippage, simulated trading programs are also designed with the benefit of hindsight. They go on to state that no representation is being made that any account will or is likely to achieve profits or losses. Simulated trading programs are what we call in-sample testing. Simulated performance and simulated trading programs are two distinct entities. The problem is that some people here are treating backtesting as performance. The SEC doesn’t, nor should they.

Take care,
Steve

118

Steve, you are very creative person :slight_smile: So do we all here :slight_smile:

Following your logic, OOS as not real trading data should not be called performance too. The same way, if I am trading this model, my personal trading data would be different when other investors, so should not be called performance too. So that are you proposing to call the performance?

I understand your point, only objective real 100% repeatable data should be called performance. The problem, there is no and never will be such a data. So other problem - there we can state data is reasonably objective and call it performance? Still decision will be subjective and no consensus available, our discussion is an evidence.

Again, following your logic, if in-sample is not performance and OOS is not actually performance, than both are comparable as non-performance. My point was in comparison. Your - what to call performance. I have done comparison and NOT requiring you to agree. I understand all the subjectivity of my act. Do you understand and agree on the subjectivity of yours? Do you understand that nor do I nor others will never ever agree with you just because … this is ridiculous philosophical discussion about nothing having no direct relation to what we are doing here.

Once again, I totally don’t care how to call it. More important this is more or less comparable data and shows models inconsistency. And here we go again. :slight_smile:

119

Konstantin - you quoted the code, not me. There is no creativity here.

“I understand your point, only objective real 100% repeatable data should be called performance.”

It has nothing to do with repeatable data. Simulated performance is when the stocks are identified/trades picked but not actually purchased. But the trades are recorded as if it were in your account. Actual performance is when the trades are physically made. In both cases picks are made in real time or going forward if you wish. There is no repeatability or historical bias in either case. There is no subjectivity involved in this activity or in my interpretation of the code.

The reason I am here on this thread is to make sure that you understand the difference between back-test and OOS, and even if you don’t wish to understand, that others are not misled. There are many ways in which back-test is performed and used. As many ways as there are users. Some people have indicated that they only care about one wee speck of time in 2013 when interest rates were rising. They are making an educated guess about the future of interest rates. Some only care about the last five years, and optimize on the recent past, presuming it is more relevant than older time periods. Some optimize the heck out of their system, with the understanding that future results will not be comparable with their optimized back-test, but will be better than the benchmark, provided they are using factors drawn from good science. Believe it or not, different people have different ways of designing their systems. Not everyone is providing backtest data meant to be representative of future performance. That is forced upon us by the manner in which systems are presented.

And in addition to all this, to suggest that results should be similar provided that the environment is the same, well guess what? We are in unknown territory, so your concept is flawed right out of the starting block.

So if you are under the impression that comparison of in-sample versus OOS is the best way of determining a good or bad model, then that is your choice. But don’t lead others into believing it is the right choice. I have seen no evidence of anyone being able to predict which models will perform best in the future. The best we can do is state our strategy and let our track record (simulated performance) speak for itself.

Steve

120

Steve, there are very little of absolutely black or white in this world. Mostly shades of gray. The problem with our discussion is that you trying to interpret my arguments as extreme black or white while I never stated that. That’s how our discussion burned out.

From your last explanations I see the difference between in-sample, OOS and actual trades performances you outlining. Who ever disagreed? This is not the point of our discussion. The point is comparability of whose different type of data. Different by its nature, by its probability to be a proxy to the real past and possible future they are still trades performances and as such is comparable. Who ever stated this is absolutely correct, precise, “results should be similar”, “lead others into believing it is the right choice”, “is the best way of determining a good or bad model” and so on? This looks like Don Quixote fights windmills that he imagines to be enemies.

The only my disagreement is your statement in-sample data could not be called performance. While I do understand your point to emphasis the difference between in-sample and OOS data (which no one in this discussion questioned!) this way it is ridiculous not to call the trade results no matter of what nature the performance. If this is the main point of you than “let’s agree to disagree” and stop here.

I see the disagreement is much deeper.

[quote]
Some optimize the heck out of their system, with the understanding that future results will not be comparable with their optimized back-test, but will be better than the benchmark, provided they are using factors drawn from good science. Believe it or not, different people have different ways of designing their systems. Not everyone is providing backtest data meant to be representative of future performance. That is forced upon us by the manner in which systems are presented.
[/quote]Somehow I managed to step into designers territory - the model quality. In fact I am not so advanced in model design to discuss this theme deeply but I do believe in case no back-test data would be presented by P123 there would be still near zero R2G sales as of today and back-test presentation will be the most requested feature. I do understand how designers feel about it. It is OK to ignore my arguments about models inconsistency. I don’t feel any pain about it. All I want is to focus designers attention how current offers misses the market expectations. Otherwise you need another argument for poor R2G sales besides anecdotal P123 marketing efforts lack. Anyway, no matter what I said, most probably things will not change and Mr.Market will judge.

P.S. Creativity is not in the fact but in its interpretation. :wink: