StockFetcher Forums · Filter Exchange · A NEW ^VIX TRADING SYSTEM<< 1 ... 23 24 25 26 27 ... 40 >>Post Follow-up
Kevin_in_GA
3,955 posts
msg #132184
Ignore Kevin_in_GA
10/28/2016 8:33:38 AM

We might have different settings - I notice that your "PF" column value is 10 while mine is 5. Also, There was no Long Buy signaled on 10/20 in any of my nightly signals, or I would have recorded it in this thread. Odd - not really sure why this happens but my guess is it settings-driven.

gmg733
391 posts
msg #132194
Ignore gmg733
10/28/2016 12:28:32 PM

This can be one of a few issues. I've ran into signals that show up one day, gone the next and then shows up again. Or a symbol just shows up without a buy.

For the data in SS, you can download every night, but as Pete told me prices can settle after a close, this is why I run my filters at 11p CST. Also, there are price artifact settings and in the import prices, you might have options set different than Kevins. I use import prices and symbols, but there is an option to clear the previous prices. Also look at the 'Import Current Prices'.

I've asked Pete about these anomalies and price data does change and settle. For instance see this note on the feature above:

Import Current Prices: When checked, prices from the most current day will additionally be imported. Normally, prices are imported only from the archived database of prices from Yahoo. However, it can take 8 or more hours for Yahoo to process the current day's prices and move them into the archived database. By checking this box, however, current prices can be imported earlier than would otherwise be the case. In fact, prices can be imported numerous times during the trading day, with each import updating the prices as-of the time of the import.

It isn't perfect but the benefits far out way a few nagging issues.

nibor100
16 posts
msg #132223
Ignore nibor100
10/29/2016 10:10:46 AM

Kevin,

Yesterday I received my 1st email of forum posts and as a new subscriber I have a few questions regarding your system characteristics:

a. Your first set of results/stats is entitled 'Fixed Trade Equity-Portofolio Size 5' but I believe you only ran the 20 strategies on a single symbol VIX, so what does Portfolio Size 5 refer to?

b. In the stats there is a phrase I'm not familiar with "Longest Flat period 48, could you explain what this measures?

c. Further below in the section "And the absolute Number of Open...." ; you mention 20 slices and 5 units. Is a unit equivalent to a slice of investment funds?

d. In your #4 Vix long filter you use expressions pdi and mdi which I can't find defined anywhere in the usage guide, indicator list or the forum historical posts. Are they shortcuts for +DI and -DI? If so, where do I find a list of valid shortcuts/abbreviations?

Thanks,
Ed S.



Kevin_in_GA
3,955 posts
msg #132227
Ignore Kevin_in_GA
modified
10/29/2016 10:01:42 PM

SIGNALS FOR MONDAY 10/31/16

NO NEW SIGNALS - HOLD ALL CURRENT POSITIONS.

OPEN TRADESDate Opened^VIX Opening ValueDate Closed^VIX Closing ValueNet Gain/Loss (%)
LONG ^VIX10/26/201613.66 - - - - -16.19+18.52%
SHORT^VIX10/27/201614.37 - - - - -16.19-12.67
SHORT^VIX10/28/201615.67 - - - - -16.19-3.32%
SHORT^VIX10/28/201615.67 - - - - -16.19-3.32%



CLOSED TRADESDate Opened^VIX Opening ValueDate Closed^VIX Closing ValueNet Gain/Loss (%)
SHORT^VIX08/29/201614.0709/08/201611.76 +16.42%
LONG ^VIX09/06/201612.5409/12/201620.13+61.66%
LONG ^VIX09/09/201612.4409/12/201620.13+60.37%
LONG ^VIX09/13/201615.9809/14/201617.63+10.32%
SHORT ^VIX09/12/201620.1309/19/201615.16+24.61%
SHORT ^VIX09/09/201612.5209/22/201613.41-7.28%
SHORT ^VIX09/12/201620.1309/22/201613.41+33.32%
SHORT ^VIX09/14/201617.6309/23/201612.02+31.74%
LONG ^VIX10/03/201613.7510/05/201613.54-1.67%
LONG ^VIX09/27/201613.3610/07/201613.50+0.90%
LONG ^VIX10/11/201613.7110/12/201615.53+12.96%
LONG ^VIX09/28/201612.9010/12/201615.53+20.05%
LONG ^VIX10/06/201613.1110/12/201615.53+18.13%
SHORT ^VIX10/12/201615.5310/19/201614.95+3.73%
SHORT ^VIX10/12/201615.5310/19/201614.95+3.73%
SHORT ^VIX10/14/201616.4910/20/201614.45+12.49%
SHORT ^VIX10/12/201615.5310/20/201614.45+7.08%
SHORT ^VIX10/12/201615.5310/20/201614.45+7.08%
SHORT ^VIX10/12/201615.5310/24/201613.34+14.10%
SHORT ^VIX10/14/201616.4910/25/201612.93+21.49%


Kevin_in_GA
3,955 posts
msg #132228
Ignore Kevin_in_GA
10/29/2016 10:09:38 PM

a. Your first set of results/stats is entitled 'Fixed Trade Equity-Portfolio Size 5' but I believe you only ran the 20 strategies on a single symbol VIX, so what does Portfolio Size 5 refer to? If you read the first post carefully you will see that even though there are 20 systems that could signal, it is uncommon that more than 5 do so at any time. Hence, the best utilization of equity is with a Portfolio size of 5.

b. In the stats there is a phrase I'm not familiar with "Longest Flat period 48, could you explain what this measures? It is the longest period within the backtest where no trades were in play.

c. Further below in the section "And the absolute Number of Open...." ; you mention 20 slices and 5 units. Is a unit equivalent to a slice of investment funds? Yes, a unit is 1/5th of your trading equity for this approach.

d. In your #4 Vix long filter you use expressions pdi and mdi which I can't find defined anywhere in the usage guide, indicator list or the forum historical posts. Are they shortcuts for +DI and -DI? If so, where do I find a list of valid shortcuts/abbreviations? PDI and MDI are required to reference +DI and -DI within user-defined functions. As stated in the description:

stockfetcher
963 posts
msg #37388
8/10/2005 7:05:04 AM

Note:

When using the +DI or -DI inside set statements, you'll want to use PDI (for +DI) and MDI (for -DI).




nibor100
16 posts
msg #132233
Ignore nibor100
10/30/2016 3:03:31 PM

Kevin,

a. Your first set of results/stats is entitled 'Fixed Trade Equity-Portfolio Size 5' but I believe you only ran the 20 strategies on a single symbol VIX, so what does Portfolio Size 5 refer to?<< If you read the first post carefully you will see that even though there are 20 systems that could signal, it is uncommon that more than 5 do so at any time. Hence, the best utilization of equity is with a Portfolio size of 5.>>

Having read it carefully the 1st and 2nd time, I couldn’t reconcile the contrast between the purported results of a single symbol backtest having a title “Portfolio Size 5” when it appears from your original msg that the conclusion to use a portfolio size of 5 was made after the backtest was run; therefore my question.

b. Careful reading discovers a contrast between your stated first requirement of “Have a win % of 90% or higher” yet the first set of stats you provided only have an overall win rate of 86.7%. Did most of the individual strategies you picked have win rates of 90% and you had to pick a couple with less than 90% in order to have 10 long and 10 short strategies in your system?

c. Further below in the section "And the absolute Number of Open...." ; you mention 20 slices and 5 units. Is a unit equivalent to a slice of investment funds?<< Yes, a unit is 1/5th of your trading equity for this approach>>.

Ok I’m lost, did you mean to say 5 units is 1/4 th of your trading equity? (because in your original system description post you stated "so you make the best returns just trading five units whenever ANY filter triggers." ; however, if we trade 5 units of 1/5 th of our trading equity whenever any filter triggers we would only have a max of one trade open at any time).

d. Your out of sample test had a much lower win rate of 70.6%, leading me to wonder if you had first run the original test on the out of sample dates you would not have picked the same 20 strategies to follow. Especially since total gain for out of sample 10 year backtest was much much less than double the total gain of the 5 year backtest. The much longer flat period of 299 days seems to be a bit of a shocker considering the nature of the VIX and how many different trading signals are present in the 20 strategies.

e. I can't seem to be able to add bold to any text in my reply either typing in the forum text box directly or pasting in from MS Word. how are you bolding text in your responses?

Thanks,
Ed S.



Kevin_in_GA
3,955 posts
msg #132234
Ignore Kevin_in_GA
10/30/2016 4:31:19 PM

I don't want to turn this thread into a "how do you get XXX in Stratasearch?" discussion, so this will hopefully answer what you are asking and we can get back to the main topic.

a. Your first set of results/stats is entitled 'Fixed Trade Equity-Portfolio Size 5' but I believe you only ran the 20 strategies on a single symbol VIX, so what does Portfolio Size 5 refer to?<< If you read the first post carefully you will see that even though there are 20 systems that could signal, it is uncommon that more than 5 do so at any time. Hence, the best utilization of equity is with a Portfolio size of 5.>>

You can take multiple positions in the same ETF - portfolio size is the number of allowable open trades at any time. I looked at the max open trades (graph 2 in the first post) and it rarely exceeds 5. Therefor I advocate allocating your trading equity into 5 tradable "units" and trading each one based on the generation of new signals. If all 5 units are in play additional signals are ignored until a currently open trade is closed.

b. Careful reading discovers a contrast between your stated first requirement of “Have a win % of 90% or higher” yet the first set of stats you provided only have an overall win rate of 86.7%. Did most of the individual strategies you picked have win rates of 90% and you had to pick a couple with less than 90% in order to have 10 long and 10 short strategies in your system?

No, but that is what the results ended up as - I was a little surprised since each system should trade independently of any other and all should have 90% or greater win rates, but truth be told the overall performance is still exceptional at ~87% win rates.

c. Further below in the section "And the absolute Number of Open...." ; you mention 20 slices and 5 units. Is a unit equivalent to a slice of investment funds?<< Yes, a unit is 1/5th of your trading equity for this approach>>.

This was explained already, and again in answering your first question.

d. Your out of sample test had a much lower win rate of 70.6%, leading me to wonder if you had first run the original test on the out of sample dates you would not have picked the same 20 strategies to follow. Especially since total gain for out of sample 10 year backtest was much less than double the total gain of the 5 year backtest. The much longer flat period of 299 days seems to be a bit of a shocker considering the nature of the VIX and how many different trading signals are present in the 20 strategies.

It is rare that an OOS backest does as well as the in-sample dataset. The purpose is to see how the system does when faced with new data (it can be older data or new data as time goes on). The bottom line is that I wanted to be sure that recessionary periods (which were not included in the data used to construct the multi-system) were not detrimental. They were not, as a 70% per annum return indicates. If that concerns you, then I suggest not using the system.


e. I can't seem to be able to add bold to any text in my reply either typing in the forum text box directly or pasting in from MS Word. how are you bolding text in your responses?

HTML code can be written here to add effects to text. Use the following (I will write out certain symbols rather than type them since they would trigger text effect and not be seen in the post):

to turn on a text effect such as bold, italics, or underlining start the section with the "less than" symbol, then type in B, I, or U, then the "greater than" symbol. That starts the effect. To end it, type "less than", /B, /I, or /U, then the "greater than" symbol to end the effect.


nibor100
16 posts
msg #132248
Ignore nibor100
10/31/2016 2:09:19 PM

Kevin,
no reply expected,
Thanks for being patient with all of my questions regarding your system origination post, which being about 2 months old to you, is only 4 days old to me as a new StockFetcher user.

I definitely don't want a Stratasearch tutorial as I plan on doing a 30 day trial after I've used StockFetcher for a few weeks. I'm curious how it differs from Edgerater Pro.

This is the phrase from your original post that kept hanging me up on the units/slices questions:
"so you make the best returns just trading five units whenever ANY filter triggers"

Using a rough approximation of your absolute # of open positions column chart it looks like following the 5 unit portfolio allocations would have lowered the total possible trades in the first backtest results by about 10%, whether that's good or bad only time will tell going forward...

Ed S.







Kevin_in_GA
3,955 posts
msg #132256
Ignore Kevin_in_GA
10/31/2016 4:34:31 PM

... but doubled the size of each trade versus a portfolio of 10. If you are willing to concentrate your trades into a smaller portfolio of 1 or 2 you will achieve even higher returns but at a reduced Sharpe Ratio.

Kevin_in_GA
3,955 posts
msg #132258
Ignore Kevin_in_GA
10/31/2016 8:52:17 PM

SIGNALS FOR TUESDAY 11/01/16

SELL TO CLOSE THE SINGLE LONG POSITION AT THE OPEN TOMORROW.

OPEN TRADESDate Opened^VIX Opening ValueDate Closed^VIX Closing ValueNet Gain/Loss (%)
LONG ^VIX10/26/201613.66 SELL TO CLOSE17.06+24.09%
SHORT^VIX10/27/201614.37 - - - - -17.06-18.72%
SHORT^VIX10/28/201615.67 - - - - -17.06-8.87%
SHORT^VIX10/28/201615.67 - - - - -17.06-8.87%



CLOSED TRADESDate Opened^VIX Opening ValueDate Closed^VIX Closing ValueNet Gain/Loss (%)
SHORT^VIX08/29/201614.0709/08/201611.76 +16.42%
LONG ^VIX09/06/201612.5409/12/201620.13+61.66%
LONG ^VIX090/9/201612.4409/12/201620.13+60.37%
LONG ^VIX09/13/201615.9809/14/201617.63+10.32%
SHORT ^VIX09/12/201620.1309/19/201615.16+24.61%
SHORT ^VIX09/09/201612.5209/22/201613.41-7.28%
SHORT ^VIX09/12/201620.1309/22/201613.41+33.32%
SHORT ^VIX09/14/201617.6309/23/201612.02+31.74%
LONG ^VIX10/03/201613.7510/05/201613.54-1.67%
LONG ^VIX09/27/201613.3610/07/201613.50+0.90%
LONG ^VIX10/11/201613.7110/12/201615.53+12.96%
LONG ^VIX09/28/201612.9010/12/201615.53+20.05%
LONG ^VIX10/06/201613.1110/12/201615.53+18.13%
SHORT ^VIX10/12/201615.5310/19/201614.95+3.73%
SHORT ^VIX10/12/201615.5310/19/201614.95+3.73%
SHORT ^VIX10/14/201616.4910/20/201614.45+12.49%
SHORT ^VIX10/12/201615.5310/20/201614.45+7.08%
SHORT ^VIX10/12/201615.5310/20/201614.45+7.08%
SHORT ^VIX10/12/201615.5310/24/201613.34+14.10%
SHORT ^VIX10/14/201616.4910/25/201612.93+21.49%


StockFetcher Forums · Filter Exchange · A NEW ^VIX TRADING SYSTEM<< 1 ... 23 24 25 26 27 ... 40 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2016 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus