Dump it Here

[MovingAverage]

A tale of two systems...importance of trading uncorrelated systems.

My weekly system was tracking well for the month of Oct, then yesterday happened. Result: all of Oct gains wiped out.



Meanwhile, it has pretty much been business as usual for my daily EOD system with yesterday not having a major impact. Result: Oct still looking good.



Difference between the two system: very different stock criterion to make the potential buy universe, very different entry methodologies, and different time frames.

Stay classy ASF.

 

[Warr87]

My daily EOD took a bunch of hits since Friday. half the months profits gone. Today it rebounded very well but still not near the peaks from a few weeks ago. The weekly is taking some major hits right now. The monthly is the one performing the best right now (though has lost a little over the last 2 weeks but still positive for the month).

mine are all probably relatively highly correlated. something i will need to work on.

 

[Skate]

Skate.

 

[ducati916]

Mr Skate,

Your chart of profit/loss:



Mirrors my VIX chart exactly:



Just check the dates/shape of the chart etc. So Aus. follows tick-for-tick the US. when trading a vol. based strategy.

There are 3 points of critical interest: 12 Oct. and 19/20 Oct. on the VIX and they correspond exactly with your profit chart. In fact you could draw a trend-line under your daily profit chart and exit when that trend is broken and the result would be +$17K in rescued profit.

Obviously in hindsight everything is clear and easy. Recognition of points 12 Oct and 19/20 Oct is the key. Clearly also 28 Oct (if that turns out to be the current bottom) or today 29 Oct (US dates) for re-entry on positions. I think I have the solution, which I have been testing over the last few days. I'll PM you later to see if it works for you.

jog on
duc

 

[Skate]

Mr Skate,

Your chart of profit/loss:

View attachment 113941

Mirrors my VIX chart exactly:

View attachment 113942

Just check the dates/shape of the chart etc. So Aus. follows tick-for-tick the US. when trading a vol. based strategy.

There are 3 points of critical interest: 12 Oct. and 19/20 Oct. on the VIX and they correspond exactly with your profit chart. In fact you could draw a trend-line under your daily profit chart and exit when that trend is broken and the result would be +$17K in rescued profit.

Obviously in hindsight everything is clear and easy. Recognition of points 12 Oct and 19/20 Oct is the key. Clearly also 28 Oct (if that turns out to be the current bottom) or today 29 Oct (US dates) for re-entry on positions. I think I have the solution, which I have been testing over the last few days. I'll PM you later to see if it works for you.

jog on
duc

Skate.

 

[Skate]

Skate.

 

[Skate]

Skate.

 

[Skate]

I think I have the solution, which I have been testing over the last few days. I'll PM you later to see if it works for you.

I'm sure the Duc won't mind if I give an update
@ducati916 was good to his word & contacted me with a tailored made solution to suit the "Daily VIX Strategy". Duc explained his solution in detail with additional confirmation graphs to further explain his exit methodology. The solution was coded & added to my original strategy with a slight improvement at first. The problem was that I added Duc's solution as an additional exit instead of the only "EXIT". It didn't take me long to work that one out.

Duc's solution worked a treat
Adding Duc's exit idea to my original strategy made a huge improvement by removing my looping exit & replacing it with the information that Duc had suggested.

I was blown away
Duc's raw idea worked a treat & I've been working on it all day - massaging the parameters trying to boost the results. To my surprise, Duc's original idea turned out to be the best fit for the strategy. (no fiddling was required at all)

Backtesting
I'm not a big fan of Amibrokers backtesting & without Norgate's "Platinum Subscription" the longer the backtest period the larger the inaccuracy of the results. I've used a short period as an example of the difference that it made.

Short backtest period
The backtest period for (1st July 2020 to 30th October 2020) should be accurate enough. I have circled some of the metrics in red that has impressed me.



Backtest
Amibroker's Backtest is one thing but using the signals from Duc's solution & recording them in "Share Trade Tracker" gives a 100% comparison between the Original Strategy versus Duc's modified exit - in the next few posts I'll supply the corresponding graphics.

Comparing "Share Trade Tracker" results
The "Share Trade Tracker" results of Duc's solution added to the strategy are impressive.

More to follow...

Skate.

 

[Skate]

=====================================================================================================================

Using Duc's modified exit



Skate.

 

[Skate]

=====================================================================================================================

Using Duc's modified exit



Thank you
I wish to publicly thank @ducati916 for his ongoing help & assistance in providing not only the idea for the "Daily VIX Strategy" but the solution to improve the profitability of the strategy. (caring is sharing)

Skate.

 

[Newt]

Sounds likes impressive work - kudos to both Skate and Duc.

Gosh this is a big slippery mountain we're climbing - just when you think you're closer to the top, turns out there's another 10,000 ft still hidden in the clouds....

Skate, do you feel these VIX parameters are mainly suited to your daily trading, or would there still be value incorporating into signals or index filter for Weekly timeframe trend following? I'm guessing that the extra degrees of "nimbleness" is very handy on Daily, but the edge is perhaps dulled on longer timeframes.

 

[ducati916]

Sounds likes impressive work - kudos to both Skate and Duc.

Gosh this is a big slippery mountain we're climbing - just when you think you're closer to the top, turns out there's another 10,000 ft still hidden in the clouds....

Skate, do you feel these VIX parameters are mainly suited to your daily trading, or would there still be value incorporating into signals or index filter for Weekly timeframe trend following? I'm guessing that the extra degrees of "nimbleness" is very handy on Daily, but the edge is perhaps dulled on longer timeframes.

Mr Newt,

As you can see from the charts (in the previous post which are weekly) posted by Mr Skate the improvement is in the weekly (and therefore included by definition will be the daily) profit drawdown.

However, Mr Skate's strategy is designed as a daily strategy. The exit is therefore a daily exit.

The beauty of mechanical trading is the on/off, black/white nature of it. It (mechanical trading) does away with excessive complexity and decision making, both of which I am convinced through many years of hit and miss subjective analysis, creates many of the issues traders encounter: ie. second guessing entries, exits; second guessing the importance of economic data, news stories covering the hot issue of the day (and there is always a hot issue of the day) and most deadly of all, slowly being seduced by the 'trend'.

Mechanical, if well designed and most importantly, executed, eliminates all of those issues for the trader.

So the new signal, is not a step up in complexity, rather a step down to (a logical) simplicity.


jog on
duc

 

[Skate]

Sounds likes impressive work - kudos to both Skate and Duc.

Duc deserves the kudos
The Action Strategy, the HappyCat Strategy, Duc’s Blue Bar Strategy & the Daily VIX Strategy have all been developed from information passed kindly to me by @ducati916

Skate, do you feel these VIX parameters are mainly suited to your daily trading, or would there still be value incorporating into signals or index filter for Weekly timeframe trend following? I'm guessing that the extra degrees of "nimbleness" is very handy on Daily, but the edge is perhaps dulled on longer timeframes.

@Newt, Duc's exit is finely tuned to the Daily VIX Strategy
I'm sure with additional work Duc's solution would translate over to a weekly strategy without issue "but not in its current form".

Where were the improvements?
Duc noticed the exit wasn't correlated with the volatility at the individual level. The entry signals are from the VIX. The exit signals were from a combination of (a) the VIX, (b) StaleStop & (c) the GTFO filter. The issue was to rid the strategy of all of my exit conditions I just described replacing it with a lower volatility band "ONLY". It's amazing how simple & clean Duc's solution was.

Duc could see the issue clearly
After it was explained to me in detail I could see the issue as well. Duc's solution was a two-fold solution as it not only improved the strategy performance but it also keeps me in the better trades a little longer. Previously I was exiting a good position too quickly.

Exits are now dependent on hitting or exceeding a volatility band
I'm hoping I haven't overstepped the mark as I make a habit of never talking about "private messages" but I'm sure I've given nothing away that can be replicated.

After the US elections
I was planning to trade the Daily VIX strategy after the US elections. Even in its original form, it's a handy strategy & "I was happy to trade it".

Duc's solution
The information Duc's has provided improved the Daily VIX Strategy immensely & it will now be the version I'll be trading shortly.

Skate.

 

[Newt]

Thanks both of you for the further detail. I'm becoming a firm believer that while its good to "soak up as much as possible", sometimes you can't make use of new information until your personal trading development and experience catches up - or you bump into someone else that questions long held assumptions and shows different ways of doing things.

Daily VIX, HappyCat etc all sound to be predicated on volatility changes as well as price trends. The pressure cooker of Coronavirus year has led to some revelations in my trend/momentum trading and index/GTFO filters, but yet to much make progress with (usefully) integrating volatility. Doesn't mean it won't happen when time and inspiration allows.....

 

[qldfrog]

Duc deserves the kudos
The Action Strategy, the HappyCat Strategy, Duc’s Blue Bar Strategy & the Daily VIX Strategy have all been developed from information passed kindly to me by @ducati916



@Newt, Duc's exit is finely tuned to the Daily VIX Strategy
I'm sure with additional work Duc's solution would translate over to a weekly strategy without issue "but not in its current form".

Where were the improvements?
Duc noticed the exit wasn't correlated with the volatility at the individual level. The entry signals are from the VIX. The exit signals were from a combination of (a) the VIX, (b) StaleStop & (c) the GTFO filter. The issue was to rid the strategy of all of my exit conditions I just described replacing it with a lower volatility band "ONLY". It's amazing how simple & clean Duc's solution was.

Duc could see the issue clearly
After it was explained to me in detail I could see the issue as well. Duc's solution was a two-fold solution as it not only improved the strategy performance but it also keeps me in the better trades a little longer. Previously I was exiting a good position too quickly.

Exits are now dependent on hitting or exceeding a volatility band
I'm hoping I haven't overstepped the mark as I make a habit of never talking about "private messages" but I'm sure I've given nothing away that can be replicated.

After the US elections
I was planning to trade the Daily VIX strategy after the US elections. Even in its original form, it's a handy strategy & "I was happy to trade it".

Duc's solution
The information Duc's has provided improved the Daily VIX Strategy immensely & it will now be the version I'll be trading shortly.

Skate.

Good morning,
A mammoth task ahead when i will finally have some free time to review these posts and check if and how they could improve my systems.

There is always, i feel, a problem of terminology and exact definition in these volatility related posts.
Vix as ^VIX - CBOE Volatility Index?
Vol as volume or volatility?
Individual vix..no such thing as standard so we mean volatility of an individual stock..
well:
Many ways to get a number there...i know, i tried several
So what are we talking about?
Volatility in price? in volume traded? a mix of? And which attached timeline?
We even have an asx volatility index..which i did not find very useful when i tried to use it a couple of months ago.
I am very interested in this domain as i have 4 live systems out of 5, plus Mr Skate happy cat ,which were initiated from the volatility talk from Mr Ducati and fully indepently developed.
But sometimes i am completely lost when reading some posts here.
We need a thesaurus....

 

[ducati916]

Good morning,
A mammoth task ahead when i will finally have some free time to review these posts and check if and how they could improve my systems.

There is always, i feel, a problem of terminology and exact definition in these volatility related posts.
Vix as ^VIX - CBOE Volatility Index?
Vol as volume or volatility?
Individual vix..no such thing as standard so we mean volatility of an individual stock..
well:
Many ways to get a number there...i know, i tried several
So what are we talking about?
Volatility in price? in volume traded? a mix of? And which attached timeline?
We even have an asx volatility index..which i did not find very useful when i tried to use it a couple of months ago.
I am very interested in this domain as i have 4 live systems out of 5, plus Mr Skate happy cat ,which were initiated from the volatility talk from Mr Ducati and fully indepently developed.
But sometimes i am completely lost when reading some posts here.
We need a thesaurus....

Mr Frog,

VIX as in VIX futures for the indices (SPX, DOW, NASDAQ):



Vol: as in implied volatility.

So essentially what I am looking at is an exit/entry based on historical levels of IV (10 years). You will find that Options traders will rank in quartiles, or by some other measure, levels or a ranking of IV of Options on individual stocks, which in part are based on the aggregate (VIX) IV levels.

As I trade 99% ETFs, the VIX is a robust entry/exit signal for my positions. Mr Skate's profit graph tracked the VIX VERY closely, hence an entry/exit predicated 100% on VIX was likely (proven now) to improve profitability.


jog on
duc

 

[Skate]

Individual vix..no such thing as standard so we mean volatility of an individual stock..

@qldfrog you have nailed it - exits are dependent on the volatility of individual stock.

Clarification
1. The entry signals are from an extended look-back period of the VIX.
2. The OLD (original) exit signals was a combination of (a) the VIX, (b) a momentum StaleStop & (c) a modified version of the GTFO filter.
3. The NEW exit is at the individual level - meaning each position exits when the close is less than or equal to a lower volatility band.
4. Good positions are held for longer as they are not dependent on a “momentum stale exit”.
5. The new exit depends only on “volatility “ at the individual level.
6. By not exiting until a position hits the lower volatility band it :
(a) reduces stock turns because of two reasons that have been explained to me that I won’t go into.
(B) The increase in profitability is because of a new volatility exit.

Volatility exit
Using a volatility exit gives stronger positions time to develop. Meaning, strong positions are held for a longer period because they are not subjected to a time dependent momentum exit.

Summary
Each position exits because of the volatility at the individual level. A position is not sold until it hits a lower volatility band. The average hold time has been extended from approx 14 days to 30 days.

Skate.

 

[debtfree]

Thank you @Skate and @ducati916 for your continued work in the VIX Strategy. The latest change is certainly impressive and makes a lot of sense, outstanding change so well done to you both.

Is there any chance of us having a peek at the lists of Buy Shares, Open Positions and Sold Trades that made these new results? Would love to view them on my charts.

Thank you both once again.

 

[Skate]

Thank you @Skate and @ducati916 for your continued work in the VIX Strategy. The latest change is certainly impressive and makes a lot of sense, outstanding change so well done to you both.

Is there any chance of us having a peek at the lists of Buy Shares, Open Positions and Sold Trades that made these new results? Would love to view them on my charts.

Thank you both once again.

@debtfree, i’ll post them with the Monday updates after the close.

Skate.

 

[qldfrog]

@qldfrog you have nailed it - exits are dependent on the volatility of individual stock.

Clarification
1. The entry signals are from an extended look-back period of the VIX.
2. The OLD (original) exit signals was a combination of (a) the VIX, (b) a momentum StaleStop & (c) a modified version of the GTFO filter.
3. The NEW exit is at the individual level - meaning each position exits when the close is less than or equal to a lower volatility band.
4. Good positions are held for longer as they are not dependent on a “momentum stale exit”.
5. The new exit depends only on “volatility “ at the individual level.
6. By not exiting until a position hits the lower volatility band it :
(a) reduces stock turns because of two reasons that have been explained to me that I won’t go into.
(B) The increase in profitability is because of a new volatility exit.

Volatility exit
Using a volatility exit gives stronger positions time to develop. Meaning, strong positions are held for a longer period because they are not subjected to a time dependent momentum exit.

Summary
Each position exits because of the volatility at the individual level. A position is not sold until it hits a lower volatility band. The average hold time has been extended from approx 14 days to 30 days.

Skate.

Clear and neat, thanks a lot